CHAPTER 3

Värmland

Liberalism one step to the left

Citizen and society

Between 1870 and 1921 a total of twenty Second Chamber elections were held. By the end of the period universal suffrage had been introduced and the main elements of the Swedish party system and modern electioneering had fallen into place. Parts of the period are of lesser importance from our point of view. Regarding the election campaigns during the 1870s, these have justly been termed dull with respect to content as well as haphazard in management (Esaiasson 1990: 72–74). Neither did any regular party organizations come to the fore following the introduction of the new parliament. Somehow political life entered a stage of exhaustion once the bland outcomes of the 1865–66 reform became apparent. However, other elections, such as those held in the spring and autumn of 1887 and in 1893, 1896, 1911, and 1921, were the more important to the development of campaigning. These campaigns also illustrate the gradual shift from mobilization by means of proxy to organization by means of regular parties. Consequently, they have been chosen for the purposes of analysis.

The 1880s were not least important from the point of view of political association. The societal effects brought about by industrialization now became more tangible. Although Värmland lagged behind in terms of economic modernization, the period still remained one of critical transitions. It was a time of social and economic change, and of emerging grass-roots mobilization among an electorate struggling to embrace the positive effects of modernization and, at the same time, trying to preserve traditional ways of life. Whereas change for some posed a challenge to hallowed tradition and virtues, tradition could also be scolded for its attendant poverty and for its lack of justice and equality. Nowhere else were these contradictions more clearly put to words than in the works of Selma Lagerlöf and Gustaf Fröding. On the one hand, Lagerlöf, the first woman to become a Nobel laureate in literature, celebrated the refined, manorial culture of rural Värmland; something all but vanished by the late nineteenth century, since its economic foundations had altogether collapsed with the decline of regional iron-making. On the other hand, Fröding mocked the very same culture for the toil and back-breaking existence it imposed on the workers.1 And while the clergy continued to criticize the decline of church attendance, the moral failings of people, and the increasing lack of ancestral attachment to the soil among the peasantry,² what was commonly known as the ‘social issue’ meanwhile moved to the forefront of political discourse. There it mixed with old demands for political reform. These two forces mutually reinforced each other in the years to follow.

Through a mix of pragmatism, deliberate redefinition of critical concepts such as ‘citizenship’, and genuine and radical concern for social issues, influential groups of Swedish liberals did move in the direction of progressivism from the mid-1880s onwards, (Rönblom 1929; Vallinder 1984; Hedin 2002; Lundberg 2007). At the same time yet more groups of moderate liberals had begun to move closer to the conservative camp (Christensen 2009: 277). Ideological divisions gradually became more clearly defined. Among the former groups this development, as well as the attempts to coordinate the liberal movement along stricter lines, reflected the emergence of a new kind of leadership, represented by people such as Ernst Beckman, Emilia Broomé, Georg Halfred von Koch, and the future chairman of the Liberal Party, Karl Staaff. Indeed, when local liberals in Karlstad tried to summarize the role of the party in municipal matters in 1918, the keynote speaker, Dr Hugo Severin, depicted it as a movement by the average man and for the average man. Its focus was the struggle for improved housing, and the fight against drinking, ignorance, and general coarseness.³ As far as the liberals themselves were concerned, democratization and the struggle for improved social conditions were not to be left to the socialists to single-handedly benefit from politically. Also, for many liberals by 1900, the idea of deliberately linking social reform to state intervention had become more palatable as a solution to the problems caused by modernization. Often described in terms of a gradual turn away from unhampered ‘individualism’ to ‘collectivism’, this shift could in fact also be justified precisely from the point of view of individualism.

Although emerging forms of voluntary social cooperation differed compared to those thirty or forty years earlier, political development by the end of the nineteenth century did not result in a rebuttal of individualism among influential liberals. If anything, the history of liberalism demonstrated the paramount role of individual agents in combating outdated social and political institutions. This was, for instance, stressed at a meeting of the Liberal Debating Society in Karlstad, in March 1918.⁴ Duly considered contemporary events could also be described in terms of a victory for ‘political individualism’, viz. an acknowledgement of the capacity of the individual to make informed decisions with respect to the greater good of society. Consequently, this implied that the individual must now to a greater extent be prepared to accept binding decisions made on a collective, rather than solely individualistic basis, and, importantly, to accept the mediating role of the state between citizen and society. At the same time, however, the above meeting also emphasized that public duties on occasion had to take second place to duty ‘to kin and, particularly, family’.⁵ Corefamily values were, indeed, typical to the time, not least so among the middle classes. Yet the statement also illustrates–similarly to what had been the case with Erik Gustaf Geijer more than seventy years previously–how traditional conceptions of community continued to be a part of the liberal mindset.

To begin with, individualism and collectivism were not necessarily at loggerheads with each other, judged from a longer, historical perspective (Berggren & Trägårdh 2006: 9–11, 216–21). Interpreted from within the framework of the modern welfare state, Berggren and Trägårdh suggest that Geijer’s ideas may be considered simply as one of the many roots to what eventually became a specifically Swedish way of maximizing individual liberty, but without endangering the fundamentals of society. This was so since, at the same time as Geijer had spoken about the relationship between community, voluntary association, citizenship, and the state, he also made individual liberty the core of his model. Notions such as these filtered back into a conception according to which human freedom and dignity were ultimately guaranteed not by mutual relations and dependence on peers, but independence from traditional forms of community. The key to obtaining this freedom was the state and its welfare institutions. At the same time, however, it is obvious that such an ideological framework differs from that of Hume’s ideas on social cooperation. To Hume the basic dilemma seems ultimately not to have been that of the relationship between the individual and the state, but rather the conflict within the individual, between self-interest and altruism.

The conflict -ridden but also dynamic relation between individualism or, rather, different strands of individualism, and society was reflected in civic and political association on a grass-roots level by the late nineteenth century. As a confirmation of the dawning of a new society and new social identities, the popular movements had become firmly entrenched in many regions, such as Värmland. By gradually adopting progressivism, urban liberals, in principle, moved closer to nonconformism and teetotalism on the problem of individualism, although there were, of course, important exceptions to this pattern (see later in this chapter). For instance, in matters of faith, nonconformists–although stressing the personal relationship between the individual and God–still held the individual to be part of a greater community, that of the congregation; the secession from the state Church did not alter this fact. Similarly, as I have stressed, a more decisive turn from economic individualism in the direction of political individualism among liberals implied a greater emphasis on the collective dimension of citizenship, including traditional forms of community. To some extent a common ground for the understanding of the individual actor, and of his responsibilities as well as liberties, was laid.

Generally speaking, nonconformists were forerunners in creating a stronger sense of community by the help of egalitarian individualism and individual improvement, but a similar logic also applies in regard to teetotalism (Berggren & Trägårdh 2006: 212–15). In the latter case the weight of explicitly Christian motives differed between various types of teetotalist organizations. For instance, the IOGT was to become more and more secularized as time passed (Åberg 1975), but not the Blue Ribbon. Regardless of which, however, the temperance vows made by the individual members went hand in hand with accepting that the community represented by the lodges would monitor and police their behaviour with respect to alcohol. From the point of view of organization, established political factions were caught by surprise by this development of civic association and by what was for all purposes a new conception of partisanship.

Tariffs and party systems formation

By the late 1880s factionalism in parliament revolved around the transformation and subsequent break-up of the ‘Lantmannapartiet’. The split was a direct effect of the controversies brought out in full by the 1887 elections on the issue of protective duties on, primarily, imported agricultural produce and particularly grain, but also manufactured products. Signalling a definite end to deregulation and ideas of free trade, political opinion at that point also revealed the first signs of polarization along a class-based, left–right axis: among those against protective duties were a majority of the urban deputies in the Second Chamber, defending what they identified as the consumer interests of the urban working-class populations, as well as the majority of family small-scale farmers from the northern and northwestern parts of the country, including Värmland (areas which, in any event, did not usually produce grain for sale on the market). By contrast, opinion in favour of tariffs was strongest in the eastern and southern parts of Sweden, among major grain producers and industrialists. And, as it happened, the conflict over duties also involved the issue of extended political rights, similarly to what had occurred in Germany in the late 1870s during the so-called ‘Second Reichsgründung’ (chapter 4); indeed, Germany, where protective duties had been introduced in 1879, was used as an example by both sides to prove their point. It is significant that the increased public interest in the election spurred a dramatic rise in turnout–from circa 25 per cent previously to closer to 50 per cent (Esaiasson 1990: 84).

Emerging class-based politics and increased participation at the same time forced the established factions in parliament, including the liberals, to consider the problem of societal representation and social integration as an integral part of electoral mobilization. As indicated, this was a situation for which they were quite unprepared. Full mass party organizations, let alone on the regional level, did not yet exist. What did come to the fore, rather, were two nationwide umbrella organizations, for and against protectionism duties respectively, but in neither case with any formal ties to the parliamentary factions of the parliament. Local associations, for and against protective duties, also formed. Although these lacked any clearly defined political affiliations, the pro-free trade camp was, as in Värmland, inclined to the left, whereas protectionists more often included conservatives.

What appeared was, in the terminology of Gunther and Diamond (2001: 12–16; see also Daalder 2001), a compromise between traditional elite party structures and the principles of mass parties. Factions with no actual organization, or in any event a very narrow organizational base, and mobilization by force of vertical networks and personal allegiance to faction leaders, were supplemented by attempts at mass-mobilization. Due to the lack of formal organization, public rallies, a massive usage of pamphlets, and campaigning in the press became main forces in creating and shaping political opinion (on the 1887 campaign and its organizational foundations in general, see Esaiasson 1990: 84–85). On the regional level, in particular two newspapers would from now on constantly be at the head of political debate. Whereas Karlstads-tidningen, at this stage under the editorship of Anders Jeurling, served as a platform for the pro-free trade election campaign, Nya Wermlands-Tidningen became an advocate of conservative interests. In addition to this, the appearance of political proxies added a qualitatively speaking new ingredient to the campaign compared to earlier elections. Local liberals took the first steps towards using nonconformism and organized temperance as political agents.

During the spring of 1887, free-traders and protectionists all over the country were rallying. For the first time, one and the same issue was discussed among voters across the country. Compared to earlier elections, campaigning became more detached from purely local interests. Tariffs had posed a troublesome issue since 1885, when protectionist interests had begun to consolidate in parliament. An intense but inconclusive debate followed and culminated in early 1887. In the First Chamber the issue fell through. Both sides in the Chamber agreed that duties were either to be introduced across the board, or not at all. The Second Chamber, however, was divided on the matter of tariffs on rye, with 111 deputies voting in favour and 101 against. Because of this situation, the king, Oscar II–himself positive towards free trade–decided to dissolve the Chamber and announce a re-election; in addition, regular elections were already scheduled for the autumn of the same year. Although opinion in Värmland as a whole was in favour of free trade (see, for instance, Rönblom 1929: 53), probably because of the large proportion of small-scale, family farming, both camps pushed their resources and arguments to the limit. From March onwards the newspapers filled with appeals enumerating the pros and cons of duties. For instance, Nya Wermlands-Tidningen, in an air of patriotism, criticized the free traders for reducing the whole matter to one of only duties on agricultural produce, when the issue, surely, concerned the entire national economy.6 Karlstads-tidningen, for its part, used the same logic, only in reverse; i.e. that duties would harm the economic foundations of the country. Indirectly the newspaper also touched on the issue of political reform by concluding that the Swedish workers and peasantry were, indeed, enlightened and clear-sighted enough to understand and manage the problem laid before them in the election.⁷

The role of the popular movements in this process becomes obvious when contemporary press coverage of the election campaign is examined. Material pertaining to the various organizations as such is scant for this particular period (and, when available, the minutes of the proceedings do not indicate any discussions on the matter). That being said, Karlstads-tidningen reported, in early April, that ‘several associations in Karlstad, such as the nonconformists, the Good Templars, and the members of the (non-socialist) workers’ association’ had met jointly to consider the forthcoming elections. According to the newspaper, they finally decided to support the liberal free trade candidate Gullbrand Elowson, a senior master from the secondary grammar school, for the constituency.⁸ Other preparations were made at the chapel of the Mission Covenant Church and, in Filipstad, the local society against protective duties; in this case, too, at the local branch of the Mission Covenant Church.⁹

Another novelty of the campaign was that candidates running for a seat now were more closely scrutinized in respect of their actual opinion on the matter at stake. The 1887 elections, to be sure, were still not about regular and competing party programmes, but focused solely on the one, critical issue of protective duties. For this reason, too, the individual qualities and, for want for a better word, charisma of the respective candidates probably became somewhat less important than the position taken on tariffs. An individual’s appearance and character were otherwise extremely important in a political landscape traditionally dominated by local elites, and in which the often small number of eligible voters were personally acquainted with the candidates. This was particularly the case in constituencies where direct rather than indirect elections were held and in which, as a rule, it was more important for the prospective deputy to attend regular meetings (Carlsson 1953: 35–39; Esaiasson 1990: 84; Gunther & Diamond 2001: 12–14).

Altogether this created a new situation for many of those who ran for a seat. The representation of Elowson in one of the local newspapers, Filipstads Stads- och Bergslags Tidning, gives a typical illustration of this. To begin with, a number of admirable features had been noted regarding Elowson’s person. Apart from being a man of learning and letters he also had

extensive experience from all walks of life … a demeanour characterized by calm and dignity whenever addressing an audience, as well as wit and wisdom; he expresses himself with great ease; he is a skilled negotiator and, last but not least, he is a just man of strong moral fibre.¹⁰

Whereas qualifications such as impeccable morals, a sense of fair play, and eloquence were hailed, these also, however, had to be supported by a thorough hands-on knowledge of the matter at hand, as well as the will and ability to argue and defend a specific course of political action. This, too, was a prominent feature of Elowson since he was, according to the newspaper, a warm and decided supporter of free trade.¹¹ At the same time Gullbrand Elowson’s candidature typified the flimsiness of factional affiliations at this time. To begin with, it was not uncommon for candidates to run as independents; something they would continue to do until the turn of the century. Furthermore many, like Elofsson, were prone to shift allegiance. He won a pro-free trade seat by a great majority in the spring elections of 1887, but only a few years earlier, in connection with the 1881 elections, had actually drifted towards the conservatives in Karlstad, before later siding with the left-wing camp (Moberg 1983: 126). This kind of background added fuel to the traditional smearing of the opposition candidates by the newspapers.

Together with other free traders, Elowson was targeted by the conservative Nya Wermlands-Tidningen. In hindsight, some of the critical remarks seem, in fact, quite to the point, considering the situation. Karlstads-tidningen, when rejecting the attacks, considered it unfair of Nya Wermlands-Tidningen to allude to those of Elowson’s qualities which, it presumed, would be viewed negatively by the ‘different social classes making up the free trade party’.¹² On the one hand, it could be argued that Nya Wermlands-Tidningen played the card of class conflict. On the other hand, the criticism clearly struck at the heart of what was, without doubt, a dilemma of liberal politics, i.e. precisely the difficulties in forming durable, cross-class alliances among the electorate. In the opinion of Nya Wermlands-Tidningen, the workers had simply been duped by free trade propaganda.¹³

The spring elections turned out to be a great success for the pro-free trade alliance in Värmland, however. Of eleven deputies elected, all but two were free traders.¹⁴ Part of the explanation was, as pointed out, that the issue of protective duties touched at the heart of everyday life. Another relevant aspect was that the problem of political representation and participation simmered more or less openly among the public during the campaign, and thus mobilized voters. Claiming to represent the rights of ‘the people’, therefore, rather than outright demands for ‘democracy’, became–as Lundberg (2007) has most recently demonstrated–a keynote of Swedish political modernization. Needless to say, the notion of ‘the people’ played an important part in German political discourse as well, but the important difference in Sweden was that this did not to the same extent as in Germany become infused with ethnic values and meanings (see, for instance, Stråth 1990a).¹⁵ But, always malleable as a concept, ‘the people’ could also be used in conservative rhetoric. When contemplating the outcome of the spring elections, Nya Wermlands-Tidningen queried whether the election had really allowed the government to take on board the opinion of the ‘the people’ on the matter. The answer had to be ‘No’, considering the conditions and, according to the newspaper, the aggressive campaigning of the free traders.¹⁶

The regular elections in September followed a similar pattern, although the result–initially promising from a free trade point of view–eventually ended in an anticlimax. Certainly, it was not unusual that election results in some constituencies were contested and so, too, was the case this time. From a free trade perspective the single most alarming of these incidents occurred in Stockholm, but its repercussions affected the whole country. It was discovered that one of the free trade candidates in the capital, Olof Larsson, had failed to pay local taxes and hence was not formally eligible to run for election. In its ruling the Supreme Court decided that all ballots including Larsson’s name and candidature would be declared invalid. In disappointment, Karlstads-tidningen concluded that, despite the Swedish people having so clearly shown their disapproval of protective duties, the election result would now be overthrown simply because of the negligence of a single citizen.¹⁷ Thus no fewer than twenty-two free trade deputies from Stockholm were replaced by protectionist candidates. Thus, too, the majority in parliament swung in favour of tariffs which, early in 1888, were introduced on a number of agricultural products and manufactured goods.

Nevertheless, the outcome must be seen as successful for the free traders as far as campaigning and mobilization of grass roots opinion are concerned. To begin with, an important step had been taken towards the ‘politicization of politics’, albeit without actual parties to aggregate and structure opinion among the voters. Effects on the established, intra-parliamentarian factions as a result of these changes were even more tangible. As pointed out, one of the outcomes was the split of the agrarians and the ‘Lantmannapartiet’ into two factions, one protectionist and the other pro-free trade (‘Nya lantmannapartiet’ and ‘Gamla lantmannapartiet’). Among the urban deputies, free traders similarly joined together to form a centre faction. Important to note is that the free trade faction of the ‘Lantmannapartiet’ and the left wing of the centre faction both contained elements which, together with the ‘Folkpartiet’ (People’s Party, 1895), were later to be joined in the Liberal Party. At the same time, however, the latter feature also indicated what would become a recurring dilemma of the liberal movement.

Although the People’s Party had originally been formed with the aim of reconciling urban and rural interests, the organizational roots of Swedish liberalism still meant that a cleavage between cities and the countryside was integrated with its electoral basis from the very outset. Ultimately this problem was part of the strategy chosen by the liberals in the 1880s. Although they had successfully managed to mobilize opinion when protective duties hit the political agenda, they had done so without any help of a coordinated, organizational framework. Considering the diverse, urban–rural composition of the liberal electorate, organization by means of proxy would therefore prove itself flawed in the long run. Confusion and, occasionally, dispute in connection with the nomination meetings would become a permanent ingredient during the following elections. At the same time, though, it is difficult to speak of these controversies in terms of clear-cut factional competition since, analogous to Panebianco’s model (1988), this definition necessarily implies a formal organizational arena and a formally defined institutional framework.

The Swedish party system was therefore still in the making, although the 1887 elections did send important signals to the body politic. While these elections by no means meant the introduction of parliamentarism, small steps had indeed been taken in this direction. The alignment of protective duties–including the social extensions of the issue–with the matter of suffrage demonstrated that programmes and doctrines were about to emerge at the expense of campaigning on single issues. Politics to an increasing extent became embedded in ideology; and importantly, this also implied a shift, however modest, away from the more general, utilitarian outlook of liberalism that had originally made it attractive to many farmers. In addition, precisely because of this, it became more and more apparent that government in the future would become contingent on the composition of parliament. Public opinion did not fail to make a point of this. And there were models to relate to. The breakthrough of parliamentarism in Norway a few years earlier had been met with great interest from the Swedish left, and it was a feature which, as in Värmland, received renewed interest in the press in the aftermath of the 1887 elections.¹⁸

The only problem was that the introduction of parliamentarism depended not only on the secession of monarchical power in relation to the executive, but also on the formation of durable factions and transparent, political agendas. At that stage the focus of opinion-making had shifted in the direction of political reform, but the controversy over tariffs was still alive. The rhetoric of ‘the people’ as well as a more collectivistic perspective on democratization clearly moved up to the front stage of political debate. Rhetoric drawing on the notion of ‘the people’ was frequently used by radical liberals and social democrats alike when they organized the franchise movement in 1890 and cooperated in mobilizing the two so-called ‘Folkriks-dagarna’ in Stockholm, in 1893 and 1896 (the word ‘Folkriksdag’ translates literally as ‘people’s parliament’, although ‘people’s assemblies’ is probably more accurate). These assemblies preceded the general elections held in the same years (for an extensive analysis of the Swedish franchise movement, see Lundberg 2007).

Mass-mobilization by means of proxy

When organized labour emerged in the 1890s, this confirmed that political cleavages in society had become more sharply drawn. Liberals and conservatives alike formed new support organizations, and compared to the 1880s there were also more tangible signs of mobilization at the rural grass-roots level. Particularly among the liberals, the possibility of an urban–rural alignment became of critical importance. On the one hand, opinion in the rural constituencies, among the landed peasantry, drifted towards a more conservative position on the matter of suffrage, since greater inclusion would threaten the delicate balance of power achieved through the 1865–66 reform of parliament. Universal suffrage, on the other hand, was, needless to say, demanded by the disenfranchised, and by urban radicals, progressive liberals, and socialists alike. At the core of this radicalization of liberalism were the formation of the franchise movement and the organization of people’s assemblies.

Although it is, of course, impossible to ‘measure’ precisely the impact of these factors in terms of ‘organizational efficacy’, it should be stressed that the franchise movement–which was not formally affiliated with the liberal faction in the Second Chamber–did serve as its campaigning platform in the 1893 as well as the 1896 general elections. Local liberal electoral assemblies (‘valmansföreningar’; similar to the German ‘Wählervereine’) had started to form in the major cities (in Stockholm, 1883, and in Gothenburg, 1892), whereas campaigning in the provinces was contingent on the efforts of the franchise movement and its organizational framework. Perhaps precisely because of the lack of formal ties to party, the franchise movement and the idea of summoning a ‘people’s assembly’ met with substantial interest. The issue was, after all, of general interest and, so it could be argued, transcended the traditional factionalism of politics.

Discursively transformed by references to ‘the people’ and the ‘rights of the people’, the franchise could be launched as a universal interest. This universal interest bridged egoistic class interests and, therefore, made possible an appealing form of cross-class, non-partisan politics (cf. Lundberg 2007: 401–405; Lundberg does, however, also stress that the franchise movement gradually lost its grass roots attachment and became more elitist). That the franchise at the same time was a political issue was, however, beyond doubt. To a considerable extent it was class-related, and with differences in opinion corresponding to left/right ideological conviction. This situation added further to the increasing public interest in politics that had begun to accumulate in the previous decade, during the debates on protective duties.19

Certainly the idea of widening the movement by including more issues than just that of suffrage found support. Still other voices, such as Karlstads-tidningen, warned against what it saw as a perilous move, viz. creating a broadly based, left-wing organization on unnatural foundations.²⁰ The idea of a grand left-wing coalition was in part inspired by the Norwegian pattern, but a generally held suspicion among the Swedish liberals towards too close an alignment with the social democrats was a main reason why the proposal fell through (Rönblom 1929: 62–71). A coalition, therefore, was considered too risky by many liberals because of the ideological diversity and the internal divisions typical of the Swedish left. The franchise movement appeared largely as a liberal proxy, and it involved the popular movements as key actors, but still on the basis of only one single issue. In effect, if not by design, the result was a two-tiered structure. Värmland represented a prominent example of this pattern, and developments in the region illustrated both the advantages of the approach as well as its drawbacks.

To begin with, the formation of a national franchise movement in 1890 drew on the efforts previously made by a large and diverse number of local associations, united by the aim of promoting extended rights to vote. This process had been initiated by the formation of ‘Allmänna rösträttsföreningen’ (1886) in Stockholm, and ‘Östergötlands rösträttsförening’ (1887) (Lundberg 2007: 36–37). In Värmland, too, the formation of a regional franchise organization preceded organization at the national level: the former had, in fact, come into effect as early as March 1889, and by the end of the year no less than 32 local branches had appeared throughout the province.²¹ The annual report of the branches for 1892, i.e. as preparations for the first of the two people’s assemblies in Stockholm were in progress, proudly stated that the people of Värmland had earned themselves ‘a seat of honour among their comrades from other parts of the country’. According to the report no less than 17,000 people had petitioned in favour of the upcoming assembly. In no other region, the board noted, did such a large proportion of the population display open support for the movement. In addition no fewer than 18 new local branches had also formed during the previous year.²²

It is important to note that the franchise movement often grew strong organizations precisely in those areas where nonconformism and teetotalism had first struck root and, before these movements, where early nineteenth-century forms of association had made inroads. This was the case in, among other areas, southeastern Värmland, in the previously mentioned districts of Visnum, Väse, and Ölme (chapter 2). Considering nonconformism and teetotalism, the former of these popular movements traced its roots in the area back to at least the late 1850s, when a chapel, which later became part of the Mission Covenant Church, started in Väse, in 1859 (‘Väse Missionsförsamling’; this particular expression of nonconformism started out as low Church, but seceded in 1878). In Visnum a chapel had been opened in January 1871 (‘Visnums Missionsförsamling’). From the 1880s onwards the nonconformists were followed, first, by the Good Templars and, then, in a second wave, by the Blue Ribbon. In 1889–90 alone, five lodges of the IOGT were established. As for the Blue Ribbon, one of the very first local branches, the one in Ölme, was supposedly formed in 1892.²³

The franchise movement followed suit. While the Karlstad branch of the organization–the largest one–had 289 members by the end of 1889, the one in Visnum mustered a substantial number as well, viz. 121 members. During the following year, local branches were also established in Ölme and Väse.²⁴ These areas were by and large agricultural, although Visnum, south-east of Ölme and Väse, was also part of the Bergslagen mining region (Kåpe 2005: 73–76; Carlsson 1953:410 classified the district as dominated by small-scale family farming); in fact the ‘Björneborg’ IOGT lodge in Visnum seems to have related to a village with the same name, and where an iron works had been established as early as the 1660s (Kåpe 2005: 26; Furuskog 1924: 417). The deputies elected to parliament from these parts, such as Lars Anderson in Ölme and, between 1887 and 1908, Olof Andersson from Hasselbol, were farmers. The same was the case, for instance, with Olof Olson, who represented the neighbouring Mellansysslet judicial circuit in the period 1887–93. This reflected the requirements for owning property built into the electoral system. But at least Andersson and Olson were also closely connected to the franchise and popular movements. For instance, Olson was elected to the regional board of the former organization in 1889, and, later on, was a ranking member of the Blue Ribbon. Both he and Anderson were also active in the Mission Covenant Church.²⁵

Be that as it may, the interests of the well-to-do farmers and other, less fortunate groups among the rural population were diverse, not least on the issue of the franchise: examples such as these could certainly be taken as evidence of political modernization rooted in horizontal power relations and consensual bargaining. However, ‘Swedish model’ cross-class alliances of this kind were as much possible simply because a multitude of organizational arenas just happened to synchronize with one another at a particular point in time, i.e. the 1890s. Possible explanations for the subsequent successes of liberalism in areas such as the above may have included anything from peasant individualism and economic self-interest to radicalized grass roots demands for justice and equality. What temporarily brought them all together were, as Lundkvist (1977) has pointed out, the challenges brought on by modernization; a matter indeed open to many interpretations. An example of this was apparent at a Blue Ribbon meeting at Ransäter, Mellansysslet, in February 1900.There it was said that the immense societal changes that had taken place in the last twenty-five years had not only been beneficial. In the wake of progress had followed ‘arrogance, sloth, extravagance, banking credits [sic], infighting between individuals and entire peoples, drunkenness, counterfeiting and, consequently, suicidal tendencies’.²⁶

Needless to say, this scenario did not necessarily imply deep and heartfelt solidarity among the majority of the heterogeneous rural population itself. Neither did it imply a need for the creation of proper institutional mechanisms for securing organizational commitment and trust, let alone the formulation of uniform doctrines. Rather, the profound societal changes typical of the late nineteenth century, and with them an increasing polarization in the countryside, were in the long run detrimental to the kind of ad hoc organization depicted here.

With regard to the urban–rural dimension, the above problems were not least echoed in the different meanings read into modern individualism and citizenship by countryside grass-roots and city advocates of liberalism. In this case, too, the first signs of controversy appeared in connection with the 1890s elections. By the time of the formation of the franchise movement, the editorship of the Karlstads-tidningen had passed from Anders Jeurling to Mauritz Hellberg. Beginning in 1890, Hellberg gradually became the main contender of radical liberalism in Värmland and formally remained editor of the newspaper until 1939. Born in the province in 1859, the son of a land surveyor, Hellberg had studied at Uppsala. There he joined ‘Verdandi’, a radical organization of students formed in 1882 by Karl Staaff. Its aim had been to build a bridge between intellectual liberals and the emerging labour movement. For instance, Hjalmar Branting, leader and subsequent chairman of the Swedish social democrats, was one of its members. After his return to Värmland, Hellberg became editor of the Karlstads-tidningen; since his time at school he was also a close friend of Gustaf Fröding, who, from 1887 to 1894, periodically worked for the newspaper. Hellberg’s ideas and values had been firmly established during his years in Uppsala. Above all the idea of individuality appealed to him. To Hellberg individuality possessed a value in its own right, regardless of the ambition or achievements resulting from that individuality (Örnklint 1993: 12).

Unlike farmer deputies such as Olof Olson and Olof Andersson, and unlike other leaders of liberalism in the region, such as Anders Henrik Göthberg, or Carl Björling and August Lindh, who both emerged as leading functionaries around the turn of the century, Hellberg’s ideas on individuality caused problems. Although in principle in accordance with both nonconformism and teetotalism, in reality they brought him in conflict with precisely these increasingly influential groups. He was not an abstainer, nor was he, because of his view on matters of religion, credible among nonconformists (Kvick 1977: 216–20, in particular 220; Örnklint 1993: 23 characterizes Hellberg as an agnostic). Throughout his career he proved himself a formidable adversary when defending his ideas. With an instinct which in hindsight seems almost telepathic, Hellberg always displayed a capacity for choosing the approach best suited to the occasion, and the argument best suited to puncture his opponent and exploit the weakness of his arguments. Qualities such as these invariably created animosity, not only within the liberal camp. For instance, in 1911 the Church pessimistically noted that ‘political liberalism’, skilfully promoted in the radical spirit of the ‘Verdandi’ tradition, was probably stronger in Värmland than anywhere else in the country.²⁷

The first people’s assembly was held in March 1893, with Hellberg acting as one of the two vice-chairmen. The main line of action decided was, as previously indicated, against the formation of a more broadly based left-wing political organization, and a continued focus on the single issue of suffrage (although this strategy would, in effect, prove impossible to sustain once the election campaign had started). A manifesto on the results of the assembly was issued to the Swedish people, and a strategy was set for the general elections to parliament. ²⁸ The campaign was to be boosted by press coverage, and involved test elections as well as petitioning in order to provoke reform of the election laws. Certain novelties, compared to previous elections, were also included in the preparations, such as canvassing and the use of professional, out-of-town speakers (‘resetalare’). The 1890s were therefore characterized by a new intensity in terms of campaigning, which meant jettisoning the deeply rooted suspicion of organizational involvement in the elections (Esaiasson 1990: 86–91). This change was all the more important since it indicated new, external conditions for political organization and mobilization.

Yet within the franchise movement opinion was also divided on crucial issues, such as the notion of ‘citizenship’ and the inclusion of women in the assembly–an issue with important implications from the point of view of suffrage. According to the chairman of the 1893 assembly, Dr David Bergström, citizenship included both men and women, whereas others wanted women to be excluded from the proceedings. One female representative, however, was present. This was Emelie Rathou, subsequent founder of the Swedish branch of the Women’s Christian Temperance Union (WCTU), i.e. the White Ribbon (‘Vita bandet’). Using the Good Templars, among whom she was at the time active, as an example, Rathou pointed to the fact that it was precisely the cooperation between men and women that accounted for the success of that movement. By way of clarification, the assembly finally voted in favour of using the terms ‘men’ and ‘women’ so as to avoid any confusion on the issue.²⁹ Hence, despite the ambitions expressed at the March assembly, variation in opinion, as well as heterogeneity in terms of grass-roots level involvement, became characteristic of the election campaign.

As the general elections approached, the movement sprang into action. Local committees were formed across the country to orchestrate the campaign along the lines decided by the national assembly. From this point onwards, though, different sources give somewhat different pictures of the outcome in Värmland. The regional board of the franchise movement recorded noticeable effects of the campaigning only in certain parts of the province, more specifically Mellansysslet, Södersysslet, Nyed-Älvdalen, Visnum, Väse, and Ölme. Furthermore, according to the board, only in regard to the election of Olof Andersson in the last three districts was the turnout successful.³⁰ In Mellansysslet, for instance, two moderate liberal candidates, both belonging to same nonconformist chapel of the Mission Covenant Church, competed with each other. The above-mentioned, pro-suffrage candidate Olof Olson finally suffered defeat against Anders Henrik Göthberg, who held a less radical position on the issue. Because of this, Göthberg in effect became a substitute candidate for the conservatives in the constituency, at the same time as he had widespread support among the many iron workers in the area.³¹

Regarding Mauritz Hellberg, finally, he as well had been personally involved in the campaign and initiated the organization of three new local branches of the franchise movement. The board did, however, conclude that, with respect to the actual elections, they had expected failure from the very outset and were thus not surprised by the meagre outcome. Probably the board had taken into account the specific social and organizational conditions in the region. For instance, financial problems–a recurrent feature of liberal political organization–had hampered work.³² By contrast, Karlstads-tidningen made a much less pessimistic assessment of the results, and warmly greeted the election of pro-free trade candidate Olof Andersson as the best outcome one could possibly have hoped for. Still, even this victory was marred by infighting, since the opponents had been able to recruit a nonconformist teetotaller to run against Andersson.³³

Map 4. The Väse, Ölme and Visnum judicial circuits.

The available sources, then, give a mixed picture of the political climate during the elections. As with the report in Karlstads-tidningen, they also shed some light on the ideological heterogeneity of the franchise movement and its complicated extensions into the various popular movements. What the comments reflect, more or less directly, were the difficulties involved in creating coherent standards not only for nomination procedures and issue structuring, but also that the franchise movement faced problems of societal integration and social representation (cf. Gunther & Diamond 2001). But regardless of the pessimistic opinions in the aftermath of the elections, the actual turnout did, however modestly, swing the composition of the Second Chamber somewhat to the left. Firstly, according to Carlsson (1953: 148), the number of left wing deputies increased from 73 to 76 of a total of 228. Secondly, on the regional level, interesting patterns come to the fore if we consider the various constituencies in a somewhat different light compared to the above. In agricultural constituencies such as Visnum, Väse, and Ölme, but also Fryksdal, there was a clear majority of left-wing votes; and so was, according to Carlsson’s estimates, the case also in regard to more industrialized rural areas, such as Mellansysslet and Älvdal-Nyed (Carlsson 1953: 410, 413). These included areas where the franchise movement, on the one hand, noted the effects of its campaigning but, on the other hand, detected only sparse tangible results in terms of turnout, such in the case of Mellansysslet and Göthberg’s subsequent victory there. Despite the fact that a moderate such as Göthberg succeeded in upsetting political opinion in Karlstad, this does not negate the overall impression that the 1893 elections, meant that an important step had been taken towards a more radicalized political climate in the region, albeit on the foundations of an occasionally extremely diverse social basis.

In fact, one of the most important outcomes triggered by the ‘external challenges’ (Panebianco 1988) of the 1893 elections was the formation in parliament of the People’s Party (1895–99). As a progressive faction it was a direct predecessor to ‘Liberala samlingspartiet’ (1900), which two years later was supplemented by ‘Frisinnade landsföreningen’ as its nationwide extension. This development was a befitting illustration of the extra-parliamentarian but at the same time intra-parliamentarian roots of Swedish liberalism, following Duverger’s description (1967 [1954]) of two, basically juxtaposed paths towards modern party systems.

As it happened, the new faction in parliament appeared just in time for the next, 1896, campaign. At that stage, though, it was also a disheartened franchise movement that went into action. Plans and strategies were roughly the same as in 1893 and the actual results similar to those of the previous election. The general level of turnout, however, dropped, and the press was less actively involved this time. Yet, in July 1896, Karlstads-tidningen remarked, in a fit of optimism, that the upcoming elections, after all, marked a turning point in Swedish politics. Among other things the increased polarization between left and right was viewed in positive terms, since this made it easier for the voters to distinguish more clearly between the various candidates. Consequently, among the results which the newspaper was looking forward to was the demise of ‘amalgamated’ factions, based on conflicting interests in parliament, and the establishing of an ‘influential and truly liberal party’ on the political stage.³⁴ Although considered too radical, and too left wing by some influential candidates–for instance, A. H. Göthberg this time also remained on the moderate side³⁵–the election and the formation of the People’s Party still became incremental to liberal organization.

With respect to the contributions made to party campaigning by the franchise movement, they did play an important part, but political development had also made clear, firstly, that liberalism and socialism did not, as some had hoped, share an organizational future together. Secondly, it was clear that liberalism had somehow to develop and consolidate its own grass-roots framework. In this process, the rank and file of the popular movements would, as within the franchise movement, become crucial. However, as the formation of the franchise movement and, not least, the 1893 campaign illustrate, in this they also became a topic of conflict between liberals and socialists and, particularly so in Värmland, a battleground for infighting among the liberals themselves. Nonconformism, to be sure, was for the time being a steadfast ally of the liberal movement. Organized teetotalism and, above all, the trade unions were another matter. That the trade unions opted for the social democrats was, perhaps, not surprising. However, in political terms organized teetotalism wavered between left-wing options, and in particular the Good Templars, if not the Blue Ribbon, became increasingly inclined towards the socialist camp (Edqvist 2001). Hence the lessons learned from political organization in the 1890s gave cause for alarm, as they demonstrated not only the merits but also the perils of organization by means of proxy. What, then, were the consequences of this strategy in the next stage, as ideologies and programmes started to make a more profound impression on the electorate?

The 1911 elections

With regard to nonconformism and teetotalism I have previously suggested the emergence of a new kind of modern and radical individualism, which, although in principle compatible with liberalism, was still more clearly orientated towards the collective compared to the traditional liberal mindset. Such an argument is compatible with the classic interpretation of the Swedish popular movements and their organizations. According to this line of argument they became important tools in the formation of new group identities and social integration in a rapidly changing society (Lundkvist 1977); viz. they performed functions which up to a certain point made them similar to political parties. Yet further similarities may be noted. Importantly, Lundkvist also stresses that the popular movements served as ‘training grounds’ for democracy, and for collective bargaining along democratic lines. This was because their formation required adaptation to more or less strict organizational standards, such as respect for majority rule. This pattern differed from early middle-class and liberal voluntary associations, where organizational standards were poorly developed. Another similarity to political parties proper should also be noted, especially considering their impact on the socio-economic composition of the Second Chamber in parliament: for example, by 1890 almost 15 per cent of the deputies were organized nonconformists (Lundkvist 1977: 175). In fact, compared to other regions, Värmland was quite unusual. Following the 1893 elections, no less than six of this province’s ten deputies belonged either to the Mission Covenant Church or ‘Evan-geliska Fosterlands-Stiftelsen’ (EFS, a low-church, non-secessionist forerunner of the former movement).36

Together with teetotalism, nonconformism rapidly became a critical factor in the formation of liberalism, not least by shaping public opinion out in the consituencies. For instance, in Värmland the Ansgarii-Posten newspaper (1887–1920) became important as a voice of nonconformist, regional radicalism. However, as previously pointed out, in many respects there were at the same time substantial differences in outlook between liberals such as Mauritz Hellberg and the representatives of the popular movements. Somewhat paradoxically, Hellberg’s concern for the grass roots of society went hand in hand with a fundamental lack of understanding of the very same groups, reflecting, among other things, his underestimation of nonconformism and teeototalism and their impact on local society.

Whereas Hellberg and his fellow urbanites would on occasion give historical lectures, such as on the topic of ‘Witches and witchcraft’, but also talks on subjects of more immediate interest, such as unemployment–in the Lecturing Society of Karlstad (‘Karlstads föreläsningsanstalt’)–it remains doubtful whether such efforts really engendered a genuine and permanent interest among the groups concerned.³⁷ Rather, for instance among members of the Good Templars, it had been noted, with some curiosity, that the movement failed to attract any serious attention from the middle classes and the intellectuals–as in a discussion held at an lodge meeting in the town of Arvika, western Värmland, in March 1893. There the participants observed that in the US or UK the educated classes, and not least women, were much more active in supporting the fight against liquor and drinking than their counterparts in Sweden. The discussants, however, also expressed the hope that teetotalism would attract a corps of doctors, clergy, and schoolteachers who could participate in its efforts more decidedly in the future.³⁸

Note that the discussants did not take into consideration the pre-1850 attempts made, most notably by the Swedish Temperance Society, to stem drinking, and which had originated in efforts by precisely the same professions as those mentioned in the above, and in particular the clergy. More than implying an outright negative attitude, the statement simply suggests that the rank and file of the popular movements were often quite unaware of their predecessors. In this particular aspect history did, indeed, not matter. Rather, if we look for traces of any contemporary, sequentialist interpretation of Swedish political modernization, such a reading can be found in the Temperance Society itself (chapter 2), or the Liberal Debating Society in Karlstad (see the introduction to this chapter). Ultimately it symbolized an attempt to redefine and adjust middle-class conceptions of individualism to the new demands brought about by rapid social and economic change. Many nonconformists, Good Templar activists, and Blue and White Ribbon members, though, did remain wary towards and uncertain of the ambitions of the liberals; something which, of course, only proves that the latter had considerable ground to cover before they could lay a firm foundation for party affiliation. That temporary alliances were possible had been shown in connection to the 1890s elections, but as the new century approached, not only did new such possibilities appear. Difficulties, too, mounted, as an effect of the marriage between two basically different organizational strategies–organization by means of proxy and mass-based organization–and the increasing political radicalism, not least of the IOGT.

When the Liberal Party finally appeared in 1900 and, two years later, its national organization (‘Frisinnade landsföreningen’), the latter set out on an ambitious attempt to organize liberals, or ‘frisinnade’, all around the country. These were ideas that, among others, Mauritz Hellberg had approved of since the heyday of the franchise movement and the 1893 and 1896 elections (see above). Above all, though, ‘Frisinnade landsföreningen’ was the creation of Karl Staaff, who was inspired by the British liberals. At the constituent meeting of the latter, held in Birmingham in 1877, the ambitions had been to create not only a basis for the national elections but, importantly, also a ‘liberal parliament’ outside parliament (Staaff 1917: 144–60; Kihlberg 1962: 259–63). This was certainly echoed in the emphasis Staaff repeatedly put on the importance of the ‘will of the people’ in relation to parliamentarism, such as in a speech in Dalarna during the 1911 campaign (Staaff 1918: 174–75). However, while this strategy meant recognition of the autonomy and activities of the individual citizens, it also ran the risk of creating tensions between the party convention, and the parliamentarian group of the party.

Therefore, among the many tasks left for the executive committee of ‘Frisinnade landsföreningen’ to grapple with, was coordinating itself with the Liberal Party; indeed, as late as 1904 some 80 out of a total of approximately 1,000 members of ‘landsföreningen’ (a number which, importantly, did not include the members of the various local branches around the country) had still refrained from joining the party.³⁹ Furthermore, it also had to encourage the many local level associations to integrate with the national organization, and stimulate liberal opinion-making in the regions,⁴⁰ or, in other words, instigate rules of social cooperation and organizational mechanisms securing commitment, monitoring, and accountability along the lines suggested by Ostrom (1990).

Considering organization by means of proxy, all this represented a break with tradition. Yet progress was slow. Regional differentiation in itself posed one hurdle, something the executive committee in Stockholm was also quite well aware of: from the outset of its formation, ‘Frisinnade landsföreningen’ had distributed a series of pamphlets entitled Political characters (‘Politiska fysionomier’). Navigating a delicate course between satire and parochialism, these pamphlets can be regarded as a plea to what the executive committee viewed as the regional political cultures and colourful activists typical of different parts of the country; admittedly, though, Värmland was not included among the regions approached.⁴¹ Also, as will become clear, liberal organization and electoral mobilization to a great extent remained contingent on the organizational networks of nonconformism and teetotalism. Although the immediate effects of this, such as in connection to the 1911 general elections, were far from clear, in hindsight it is, again, obvious that the drawbacks of such an approach outweighed the advantages. The kind of organization needed to balance grass roots and leadership was never fully implemented. The grass roots were very much left to their own devices.

In fact, the subtle differences signalled by the labels ‘liberal’ and ‘frisinnad’ capture something of the essence of what would rapidly become the most important dividing line within early twentieth-century liberalism. ‘Frisinnad’ literally means ‘broad-minded’, although any precise translation from Swedish to English is difficult. The expression had a corresponding notion in German, i.e. ‘freisinnig’, but in this case the context differed from what had become typical of Swedish ‘frisinnade’. Originally, liberals of all persuasions laid claim to being ‘broad-minded’; this had, indeed, been considered more radical compared to ‘liberal’ and was therefore, at the recommendation of Staaff, included in the name of the national organization of the party (Kihlberg 1962: 261). As time passed, though, ‘frisinnad’ increasingly came to designate a more or less distinct variety of liberalism, characterized by connections to rural society, and to nonconformism and teetotalism. Hence yet another feature was added to the conventional urban–rural division of Swedish liberalism (a feature which, importantly, was if not synonymous so in any event closely related to the intra-parliamentarian /extra-parliamentarian cleavage depicted by Duverger, 1967 [1954]; also cf. Vallinder 1984). The quality of emerging as a ‘soft party’ in Panebianco’s terminology (1988) thus spelled the roots of a conflict–over prohibition–that would eventually be the undoing of the party in 1922–23 (see below).

A similar development was typical of party formation in the provinces. In Värmland a regional office of ‘Frisinnade landsföreningen’ opened in June 1907 (‘Värmlands frisinnade länskommitté), although a permanent secretary, or ‘ombudsman’, was not appointed until 1916. Although local liberal assemblies had, in some instances, formed in advance, these remained more or less autonomous bodies until the formation of ‘Frisinnade landsföreningen’ and its regional network. Contrary to the situation in Schleswig-Holstein, therefore, liberal party organizations did not develop in parallel on the regional and national level at one and the same time; a circumstance which underlines both the difference in institutional settings between Sweden and Germany as well as the specificity of Schleswig-Holstein as a region.

In Värmland the regional office was supplemented by an executive committee and three subordinate organizations for the northern, western, and eastern constituencies of the province, as well as a number of local branches (the number of constituencies had been reduced from ten to three since the 1890s). This organization remained more or less intact until 1922, when ‘Värmlands frisinnade valkretsförbund’ was created. Whether this change resulted in any real centralization of the party organization is difficult to judge. The context does not, however, suggest this. Rather, the main impression of liberal organization remains that of improvisation and haphazard management; for instance, the executive committee itself had to be reinstituted as late as in 1918, as it had somehow fallen out of service.

Step by step, liberal organization nevertheless expanded during the first two decades of the twentieth century. Although not necessarily executed in any coordinated manner, this development reflected ambitions to cover all those aspects which were considered crucial to the life and activities of a political party. Quite naturally, an initial concern was to manage the campaigns, viz. the tasks of nomination and mobilization, mainly by rallying supporters to the cause.⁴² While interest aggregation and issue structuring became a concern at the central level, once a national organization had formed it still remained an important task of the regional and local assemblies to filter back the decisions made, and to communicate the party programme to the grass-roots level. But, as indicated, at the same time the executive committee in Stockholm recognized the difficulties involved with imposing centrally formulated directives. For practical purposes, the committee noted, it was possible only to make requests, more or less urgent, occasionally by telephone when possible.⁴³ Indirectly, therefore, functions of social representation and social integration, to follow Gunther and Diamond (2001), posed important and, most likely, unresolved issues from a national perspective. Despite its intentions, the allegedly universalistic liberal outlook was really a regionally entrenched outlook.

Precisely in regard to local and regional self-government, however, coordination of policies took place, though only gradually, and this left organizational imprints, at least on the regional party structures. For instance, in Värmland a specific branch of the organization was created for coordinating liberal policy-making in the Provincial Council. Furthermore, women were gradually included. Their inclusion began in 1909, with the formation of an autonomous organization for the promotion of women’s suffrage. Although formally non-partisan, this body was closely tied to the liberals, and the first elected chairperson of its executive committee was Gerda Hellberg, the wife of Mauritz Hellberg.⁴⁴ Eventually it evolved into a regional women’s branch of the party in 1919,⁴⁵ and in the new, 1922, organization several women were also elected to the board of the various local branches of ‘Frisinnade landsföreningen’.⁴⁶

When all was said and done, though, the core of the party was still its local branches. The national board had difficulties controlling these, not least when factionalizing on the local level continued to pose a problem. In Värmland, most notably the Mellansysslet area and the supporters of Anders Henrik Göthberg remained a regular source of irritation, at least from the point of view of Karlstads-tidningen. Göthberg’s position on the 1893 and 1896 elections are two examples; his manoeuvres in connection to the 1908 elections yet another. According to the newspaper, Göthberg had demanded that a conservative candidate should, against the rules of the party, be allowed to run in the test election organized by the Liberal Party, since prospective voters included people inclined to the right as well as the left.⁴⁷ Underpinning the conflict were, again, the tensions between radical and more moderate factions within the party, and between urban ‘liberals’ and rural ‘frisinnade’. However, not only the countryside but also the cities, such as Hellberg’s own Karlstad, were a matter of concern: as the 1906 elections to the provincial councils approached, Karlstad was, on the one hand, considered an easy win by the executive committee of ‘landsföreningen’. On the other hand, the victory was conditional since the committee felt uncomfortable with the candidate–Gullbrand Elowson–who, though part of the local establishment, would have ‘to be discarded’ and replaced.⁴⁸

Difficulties implementing discipline along the lines laid down by the party also originated in the reluctance of some prospective parliamentary deputies to embrace the very idea of centralized organization throughout the 1910s and the First World War. When Emil Rylander accepted his nomination as one of the liberal candidates in the northern constituency of Värmland in the 1917 elections, he did so reluctantly. He explained his doubts in terms of the need to choose between personal conviction and practical necessity (chapter 1). At a meeting held in Torsby, Rylander declared, firstly, that he had never fully accepted the programme of the Liberal Party since it was, in his opinion, much too imprecise. Secondly, however, there were also the effects of the new, proportional electoral system to consider, and with it the ‘inevitability’ of modern political parties. This was Rylander’s main thrust. Since the current situation forced candidates to join formally with parties, it also put ever-increasing pressure on them to declare their position on the political issues addressed to them. It was the decided opinion of Rylander that the latter was to the good of the legislative work in parliament, i.e. it provided an antidote to political renegades. However, the downside was, according to him, also obvious. The ideological compromises necessitated by formal affiliation to a party posed a new kind of moral dilemma for the individual deputies–in Rylander’s case more specifically in regard to prohibition, something which he personally advocated.⁴⁹

In light of the pressure brought to bear on the local liberal assemblies by formal organization or, perhaps, the institutionalized tension between local tradition and party, it is worthwhile to dwell at some length on the issue of what the average partisan in Värmland was like in the early twentieth century. The membership lists which were annually submitted by the local branches provide interesting information on this matter. Unfortunately the material is incomplete and, as far as the northern constituency is concerned, lacking entirely. In the two other constituencies, far from all branches bothered to file the requested information, and the main bulk of it pertains to the late 1910s and early 1920s. Though marred by inconsistencies, what remains still provides us with a skeleton key to the make-up of liberal partisanship shortly before the split of the party. Although the questionnaires differ somewhat in design between years, returned data includes not only the total number of members, but also the proportion of women as well as nonconformists and teetotallers among them. Usually the regional branch also asked for a breakdown with respect to occupation. (Particularly the latter data is sketchy, though, but in cases where it does exist it indicates that a majority of the members were, in fact, farmers.)⁵⁰

In 1920 the western districts reported a total of 865 members spread among 26 local organizations.⁵¹ In the east, though, a mere ten branches reported a total of 373 members by 1922.⁵² This certainly does not correspond to the image of these areas, such as the Visnum, Väse, and Ölme districts, as traditionally simmering with civic and political association of all kinds. However, the data was compiled only a year before the conflict on prohibition in the Liberal Party culminated, and therefore in all likelihood reflects what had at this stage become a general problem to the party–an increasing number of members were defecting because of the conflict (Johansson 1980: 60). And in which direction the secessionists chose to direct their faltering loyalties we can only guess. As for those remaining we should, however, note that evidence, although circumstantial, does indicate affiliations and networks similar to those of parliamentary deputies such as Anders Henrik Göthberg, Olof Olson, and Olof Anderson thirty years previously.

Firstly, the overwhelming majority of members belonged to organizations based in the countryside. Despite the fact that the local branch in Karlstad (included among the eastern districts) had the largest number of members–151 (1922 figures)–the majority of members were still organized in rural areas: 76.4 per cent in western Värmland (1920 figures), and 59.5 per cent in eastern Värmland (1922 figures).⁵³ Secondly, women were in the minority, but in eastern Värmland 12.3 per cent of all party members in these areas were women, mainly because of the relatively speaking large number of females in the Karlstad branch.⁵⁴ Thirdly, in western Värmland 7.1 per cent of the party members were nonconformists whereas 13.2 per cent were teetotallers; in the eastern districts the proportion was 7.8 per cent and 29.0 per cent respectively (as previously indicated, the former were usually affiliated with the Mission Covenant Church).⁵⁵

Table 3. Membership of ‘Frisinnade landsföreningen’, c. 1920. Proportion of members in urban and rural districts (per cent).

District	Urban districts	Rural districts
Western Värmland	23.6	76.4
Eastern Värmland	40.5	59.5

Sources: Föreningsrapporter till valkretsförbundet 1920. Vol. 7, Inkomna handlingar från lokalavdelningar 1918–21. Västra Värmlands frisinnade valkretsförbund; Föreningsrapporter till valkretsförbundet 1922. Vol. 8, Inkomna handlingar från lokalavdelningar 1919–26. Östra Värmlands frisinnade valkretsförbund. Folkrörelsernas arkiv för Värmland, Arkivcentrum Karlstad.

The figures still require some further comment. To begin with, we should particularly note that only a minority of all liberal voters were actual party members. In the country as a whole the ratio was approximately 17.5 per cent at the time of the 1911 general elections. ⁵⁶ Figures for the local branches are, as I have pointed out, impossible to calculate; although outside Stockholm, Värmland had the highest number of members directly joined to the national level organization by 1910: 155 out of a total of 2,127 members.⁵⁷ In any event, the Swedish pattern almost certainly indicates a more successful case of grass-roots organization compared to Schleswig-Holstein. In this region the liberals managed to organize only 5.0–10.0 per cent of their supporters (left-liberals and national liberals respectively, Schultz Hansen 2003: 473). At the same time party organization on the grass-roots level still lagged behind compared to the social democrats, who organized roughly one-third (32.0 per cent) of their voters in the 1911 elections.⁵⁸

Finally, we do not know the precise extent of overlapping membership in the case of nonconformist and teetotalist liberals, although more generally this combination of affiliations was far from uncommon, and in particular with regard to the Mission Covenant Church, and the Blue Ribbon. But, as a reflection of the omnipresent and unresolved tension between different strands of political individualism, the conflict within the party in some cases seems to have spread to these organizations as well. For instance, among the former, Göthberg’s own congregation at Munkfors (Ransäter) simply split in the autumn of 1923,⁵⁹ only a few months after the division of the party into two factions. In even other respects some of the alignments between civic and political associations had most probably started to dissolve long before the concluding battle on prohibition. Anders Henrik Göthberg, one of those who had originally answered the call to gather in Stockholm for the constituent meeting of the Liberal Party in 1900,⁶⁰ had left the IOGT as early as 1892 (Svenska folkrörelser, I, 1936: 450): as opposed to the Blue Ribbon, the IOGT had become both more radicalized in political terms and more secularized by the end of the nineteenth century. Among some of the former, though, affiliation with the Liberal Party and, at the same time, teetotalism and nonconformism remained, as for ombudsman Carl Björling, a viable option at least up to the 1923 split (Svenska folkrörelser, I, 1936: 278).

As the September 1911 elections moved closer, the new regional organization of the party was put to a critical test. An important innovation was also that the electoral system had now been reformed from a majoritarian to a proportional system. Portending Emil Rylander’s shrewd observation some years later, this meant that party organizations, however weakly developed, henceforth became more important to campaigning activities. Indeed, immediately following the reform of the electoral system, the efforts on the part of ‘Frisinnade landsföreningen’ to strengthen their grass-roots organization had surged, and included renewed activism and the formation of new, local branches.⁶¹ Among other things, this also meant that the 1911 campaign was initiated earlier–in June–compared to previous elections. Preparations for the campaign on the regional level started as early as January, though. For example, the liberals in eastern Värmland then stressed the necessity of making the new citizens ‘aware’ of the liberal credo. Since the conservatives had ‘unlimited’ financial resources at their disposal, and since the social democrats could rely on their organizational skill and equally substantial funds, it was necessary for the liberals to make a vigorous effort not to fall behind. Regardless of how realistic the latter’s analysis of the competing parties was, the executive committee pleaded with its voters to raise money to be used for the election campaign.⁶²

Although, technically speaking, it was still possible to run in the election without a party label, candidates seldom seem to have used this opportunity (Esaiasson 1990: 110–13, including n. 17, 134–35). Rather, with the 1911 elections, campaigning entered a more ‘modern’ phase and in terms of issues focused on military expenditure, constitutional reform, socio-economic reform (including the agricultural sector), and the alcohol issue. The extent to which the different parties were successful in adapting to the new conditions did, however, vary. For instance, the social democrats started using motor cars to transport speakers between rallies; a move that was replicated by the conservatives. In particular the latter, though, still suffered from the lack of a centralized coordination of the campaign. The same was the case with the liberals, despite the efforts previously made by the executive committee in Stockholm (Esaiasson 1990). Particularly in the countryside, the absence of efficient communications also remained an impediment to campaigning. As late as the mid-1930s, the executive committee in Stockholm proposed, among other things, using motorboats for campaigning; predominantly in the coastal regions, one assumes, but probably in the extensive inland lake districts as well.⁶³ Some confusion, finally, occurred because liberals and social democrats entered the election in the form of cartels in fifteen constituencies, including four parishes in the western constituency of Värmland (Trankil, Silbodal, Skillingmark, and Järnskog),⁶⁴ thus provoking conservatives such as Albert Petersson to complain about ‘improper forms of campaigning’ (see introduction).

It appears, however, that the liberals devoted some time and energy to ‘improper campaigning’ among themselves, too. The tensions between, on the one hand, the Liberal Party and the urban liberals, and on the other hand ‘Frisinnade landsföreningen’ and its supporters, surfaced at an early stage of the preparations. They included the role of the nonconformist faction in the party. In March the executive committee in Stockholm pondered an official letter–one couched in very careful words, however–from the local branch in Kristinehamn. Unfortunately, the original document has not survived, but it is clear that in this town on the eastern border of Värmland some dispute had arisen regarding ‘the attitude among some of the liberal newspapers towards the nonconformists’. The executive committee probably decided that the matter was too sensitive, since it simply filed the complaint.⁶⁵ One important reason for this might have been that, considering the upcoming elections, it felt impelled to be guided by organizational necessity rather than ideology–this because, as Esiasson (1990: 111) points out, the popular movements still had well-developed organizational networks for campaigning and, in that capacity, continued to exercise considerable influence on the behaviour of the political parties.

Among the liberals the lion’s share of campaigning did, in a strictly formal sense, fall on the local branches. Tasks included the recruitment of campaign workers and preparation for the nomination of candidates (Ransäter). Meetings were held at which, as in the town of Filipstad, the issue of women’s suffrage was discussed.⁶⁶ At Munkfors attempts were made to invite Karl Staaff himself to hold a rally, since it seemed likely that he would anyway visit Karlstad during the campaign. That the branch experienced some difficulties in attracting speakers, though, is clear from the proceedings.⁶⁷ In effect, however, the outcome of the campaign depended heavily on the activities of both the nonconformists and the teetotallers. Similarly to earlier campaigns, the venues of these movements were used to hold liberal campaign meetings.⁶⁸ One important distinction compared to earlier elections, though, was that the difference between civic and political association had become more blurred.

Not only the IOGT but also the Blue Ribbon had thus become more overtly politicized. The IOGT had already undergone a radicalization, and its loyalties were not unanimously orientated towards the liberals (Edquist 2001). And they coordinated the election preparations behind locked doors, as in Arvika where the Good Templars decided to organize a closed, member-only deliberation on the elections.⁶⁹ In the Blue Ribbon similar opinions were voiced. Carl Björling, who would become a regional ombudsman for ‘Frisinnade landsföreningen’ in 1919 (Svenska folkrörelser, I, 1936: 278), was chairman of the Högerud section of the Blue Ribbon when some among his sixty-odd members suggested nothing less than an entirely new ‘Christian and political temperance party’ in order to advance their cause. The keeper of the minutes concluded, somewhat wearily, that the issue was best resolved by latching onto the established political parties.⁷⁰ It was as if the pattern typical of the 1890s elections had been replayed, only in reverse, with the popular movements permeating the Liberal Party, and pressing for commitment, rather than the other way around.

When election day in September was approaching, Karlstads-tidningen nevertheless conveyed optimism, considering the intense campaign.⁷¹ And the actual results did, indeed, once more confirm the dominant position of liberalism in the province. Although the elections were an all-out success for the social democrats on the national level, the liberals in Värmland received a 46.1 per cent result compared to the national average of 40.2 per cent (Table 1, chapter 1). By comparison the social democrats won 29.5 per cent of the votes in Värmland, which, although a very substantial increase, was not above the national average for the party–unlike the case of the liberals.⁷² Karlstads-tidningen commented on the results under the headline ‘The People’s Verdict’ (‘Folkets dom’).⁷³ However, in light of the noticeable advances made by the social democrats, this very verdict could also be read as a warning to the liberals of what might be in store for them in future elections should they fail to ride the rising tide of popular radicalism.

Difficult as it is to explain electoral behaviour, it probably had less to do with organizational factors at this stage. It is possible that voting liberal had in fact become something of a political–cultural tradition, despite the difficulties and factionalism within the party. This might not have been the case among the workers, who, rather, tended to move towards the social democrats. However, it almost certainly applied to the farmers, a crucial group in the rural districts. On the one hand, the problem–that of preserving agriculture and, not least, small-scale family farming in a rapidly industrializing society–reflected the traditional urban–rural split among the Swedish liberals. On the other hand, this was an issue which, from the point of view of the individual voter, was not necessarily enmeshed with the never-ending controversy on prohibition and, hence, the tangled web of alignments between various forms of civic and political forms of association. As much as these two strands of organization had mutually reinforced each other since the late nineteenth century, analogous to what Tocqueville had suggested (2003 [1840]), the result was at the same time a stark illustration of the perils of political self-interest.

Therefore, an important factor on the regional level was presumably the, relatively speaking, greater homogeneity among the peasantry in Värmland compared, for instance, to the eastern and southern parts of the country, or Schleswig-Holstein, where manorial economies were still prominent at the turn of the century (chapter 4). This is not to say that the rural population had not differentiated along the lines suggested in the above (see also Olausson 2004), but only that this feature was probably less extreme compared to many other regions. Indeed, Carlsson (1953: 407–15), who studied national voting patterns in the general elections up to 1902, including the socio-economic composition of the electorate, identified no constituencies in the region which were dominated by large-scale farming. Rönblom (1929: 174) might well have exaggerated when he claimed that rural workers to a great extent opted for liberalism by the turn of the century. What remains a fact, though, is that from 1900 onwards the established political parties, the liberals included, increasingly tried to attract these workers as well as the farmers, not least the small-scale family farmers. Focusing on the liberals, the interest taken in agriculture and its modernization became part of their new progressivism–for example, when the executive committee of ‘Frisinnade landsföreningen’ started its preparations for the 1911 elections, it also discussed the matter of creating a ‘non-biased organization for the rural workers’.⁷⁴ Although the incorporation of agrarianism with ‘non-biased’, viz. cross-class progressivism, might in the long run have failed to attract the workers and, rather, driven them towards the social democrats, it is nonetheless likely to have appealed to the practical, or ‘utilitarian’ needs of the farmers.

Certainly, the sharpened focus on agriculture had already stimulated the landed interests and peasantry alike to create their own interest organizations. This movement had begun with the formation of ‘Svenska agrarföreningen’ (1895), which was fiercely protectionist in outlook, and directly inspired by the German ‘Bund der Landwirte’ (Carlsson 1953: 162–63). Later on the farmers also created their own political parties: ‘Bondeförbundet’ (1914) and ‘Jordbrukarnas riksförbund’ (1915; the two parties amalgamated in 1921 under the former name). In general terms, therefore, the liberals again seemed trapped in their traditional, uneasy position in-between political chairs. However, if we extend the perspective somewhat, beyond the 1911 elections, it does at the same time become clear that ‘Bondeförbundet’, which was more clearly orientated towards the smaller farmers–a crucial category in Värmland–had great difficulties establishing itself in the province. In the 1917 elections the party received virtually no votes at all in Värmland (it formed too late to participate in the 1914 elections). In the 1921 elections the result was only half the national average, and only in the 1924 elections did ‘Bondeförbundet’ reach 10.8 per cent of the votes in the province.⁷⁵ Recently split in two factions, the ‘frisinnade’ still fared much better compared to the country as a whole and, after the social democrats, became the second largest party in the region.⁷⁶

In general terms the 1911 elections were instrumental to the formation of the modern, Swedish party system. From a liberal point of view, though, the elections had also highlighted the complicated relationship between the national leadership of the Liberal Party/’Frisinnade landsföreningen’ and its grass roots; a problem to a considerable extent related to its organization by means of proxy. Part of this problem, too, was the differentiation between urban liberalism and rural ‘frisinne’. But as much as this differentiation signalled different outlooks, and different approaches to citizenship and the challenge of modernity, it also reflected a generational shift within the liberal movement (insofar as this description is valid at all after 1911). In brief, around 1910, old-style liberal candidates of the nineteenth century mould, moderates and radicals alike, had become a dying breed, something which went hand in hand with the increasingly important role of ideology, of programmes, and of mass-based organization in Swedish political life.

In Karlstad, Gullbrand Elowson, who had joined the liberal camp during the 1880s, had departed in 1908. By the end of his carreer considered expendable by the executive committee in Stockholm (see above), he had been succeeded in 1906 by Axel Schotte, a high-ranking official in the provincial administration, and a more up-to-date liberal by the standards of the executive committee. Mauritz Hellberg, of course, had involved himself in the 1911 campaign, and he did so in part on the strength of his own seat in the prestigious First Chamber, which he had finally been able to secure the previous year. More stalwart than Elowson, Hellberg was nevertheless also at odds with social and political developments. On the threshold of a new era, his intellectualism and particular views on individuality had made him increasingly detached from the grass roots of the party. Originally considered too radical to fit in with mainstream liberalism, it was significant that he became a deputy only at a late stage in life. In a 1917 letter from his old friend, professor Johan Bergman, the latter was regretful about Hellberg’s late entry into parliament, but at the same time tried to strike a jesting tone: Sweden had, according to him, indeed become a ‘country for old men’.⁷⁷ Instead, liberal leadership on the regional level had gradually moved into the hands of younger people, such as Carl Björling and August Lindh, who had both been brought up in environments ideologically close to the popular movements and a more collectivistic conception of political individualism. These were among the men who were destined to play prominent roles in the ‘frisinnade’ faction when the Liberal Party finally split.⁷⁸

The winding road to destruction

During the First World War and the outbreak of the revolution in Russia, food rationing and shortages led to social unrest and political crisis in Sweden. In March 1917, at the same time as the first phase of the revolution culminated and the Tsar abdicated, one conservative cabinet replaced another after fierce opposition from socialists and liberals. On the domestic scene the social democrats had gathered strength during the war, but nonetheless split in May 1917; one faction remained reformist whereas the other, the left-wing faction of the old party, included the core of what later on (1921) became the Swedish Communist Party. As the 1917 general elections approached, demands for the introduction of universal suffrage increased. The elections resulted in victory for the social democrats, but also in a recovery for the liberals, who had declined in the previous, 1914, elections. Consequently, the social democrats and the liberals joined forces and formed a coalition government.

The latter were in for some nasty surprises, though. As the 1920 elections had turned out to be a political parenthesis, the 1921 elections, viz. the first in which universal suffrage for men and women was applied, became the more important. To the liberals–although the Karlstads-tidningen among other newspapers tried to refute this –the immediate threat was that of being cut to pieces in the struggle between the social democrats and the conservatives; thus, the ‘brute clash of interests between class-based parties [simply] was a reflection of the [recent] war between nations’.79 In reality, the onslaught came from within the ranks of their own party.

The main competition for voters came from the social democrats. But the tool with which to match this challenge–the party organization–was slowly crumbling. For one thing, the matter of prohibition continued to put a strain on the efforts to coordinate policies. Teetotalism had culminated in terms of organized members around 1910 (Lundkvist 1977), but within ‘Frisinnade landsföreningen’ itself, teetotallers together with nonconformists continued as a formidable and, as it would prove, subversive lobby. Furthermore, the liberals had suffered from yet other internal conflicts since 1914, one of which was on defence issues and military expenditure. In connection to the elections that year, a right-wing faction actually left and eventually joined the Conservative Party. Finally, Karl Staaff had died in October 1915. With his replacement by Nils Edén, a professor from Uppsala and leader of the liberal deputies in parliament, the party entered a severe leadership crisis. Staaff had guided the party wisely and paid more than lip service to its grass roots in an attempt to maintain a balance (Johansson 1980: 22–25). Although skilful, Edén lacked these qualities. He failed to keep together a movement and an organization which, to follow Rönblom (1929: 96–97), more than anything else was a cross-class coalition of ideas.

Compared to Staaff, Edén was more elitist, in the sense that he was much less prone to let the party conventions and, thus, the grass roots, decide on matters pertaining to the orientation of the party; importantly, following such a line also meant tying the deputies in parliament to the decisions made by the conventions. The latter, Edén had argued in the spring of 1916, simply lacked the necessary insight and knowledge to decide on complicated political matters (in this case specifically on military expenditure) (Johansson 1980: 24–25). When put to it, Edéns perspective in effect implied a centralization of ‘Frisinnade landsföreningen’, which, whether by necessity, considering its organizational pedigree, or by creed, had been open to diversity in local and regional opinion. Edén’s strategy could–perhaps–have posed a harsh but nevertheless useful remedy for the tensions between ‘liberals’ and ‘frisinnade’. Under current conditions, however, it created an organizational crisis which brought out this conflict in full. His resistance towards accepting the leading role of the party conventions thus seemed to confirm the problem of party alignment touched upon by Emil Rylander during the 1917 elections.

By the time of the 1921 general elections, these competing ideas of party came out into the open. The hardliners among the ‘frisinnade’, led by editor Carl G. Ekman, mobilized their teetotalist supporters and demanded prohibition. In this they also found support among the social democrats, who by this time had become well entrenched in the organizational network of the teetotallers, and in particular the Good Templars. The moderate minority led by Edén favoured restrictions and control, but only based on existing legislation. Their approach was therefore not entirely dissimilar to the strategies used by the ‘Deutscher Verein’ (see chapter 2) in Germany. Prohibitionists and moderates were, however, unanimous on the need for a referendum to decide the extent of public support for prohibition. The conflict culminated at the party convention in June 1920, although a compromise was eventually negotiated. It was, as Johansson (1980: 51) has stressed, to great extent a conflict over the right of the majority to set the party programme. Since the moderates insisted on the right of the minority to veto any too far-reaching proposals on prohibition–a position that was reflected in the final compromise–in effect the majority rule of the convention was overruled.

The bitterest struggles took place among the Stockholm liberals, but the conflict inflicted damage on the organization across the country. In Karlstad the local branch of ‘Frisinnade landsföreningen’ was divided, but at a meeting held in August several moderates claimed that the prohibitionists tried to impose a ‘much too narrow-minded view’ on the matter. Even though the prohibitionists were in the majority, they did not have the right to suppress the party minority’s views.⁸⁰ Opinion in the countryside was likewise divided, but more generally tended towards the protectionists, i.e. the ‘frisinnade’. This reflected the close connections that still existed between the party and organized teetotalism and nonconformism.

Old organizational networks were once again activated in the 1921 elections. For example, as early as March the nonconformists of the Mission Covenant Church closed ranks on the matter. Formally non-partisan in political matters, all congregations in Värmland decided, firstly, that any member who for any reason possessed ration coupons for alcohol (alcohol rationing had been introduced nationwide in the previous year) should immediately return them to the proper authorities. Moreover, the members in Väse decided to sign a mass petition in favour of temporary and all-out prohibition.⁸¹ Secondly, it seems that in some parts of Värmland the teetotalist organizations of the Good Templars and the Blue Ribbon had by this time tried to join forces by forming joint committees, such as in the Ölme district.⁸² As the elections approached, campaigning in favour of prohibition was extended. On 13 August, the IOGT lodge Heijkensköld in Älvsbacka, west of Filipstad, decided to hold an open session on the issue, prompted by the upcoming elections. The meeting was to be organized in cooperation with other lodges in the area.⁸³ And at an extraordinary meeting three weeks later, in September, the lodge ‘Björneborg’ of the IOGT in Visnum pondered the position on prohibition among the candidates from the various parties. The keynote speaker was Gustaf Svensson, a prominent Good Templar member and social democrat from Stockholm.⁸⁴ The double-edged political role of the popular movements had become fully apparent.

On the one hand, all of these organizations had an intensely mobilizing effect on the electorate. On the other hand, their party affiliations were more than ever before uncertain. For every successful move that could be noted by, for instance, the teeotallers, there was a corresponding defeat of the idea of a synchronized election campaign among the liberals. These difficulties had long been reflected in the brief reports–and briefer they became as the elections approached–from the local branches. As early as the previous year, Adolf Carlsson, head of ‘Frisinnade landsföreningen’ at Värmskog, western Värmland, succinctly pointed out to the regional office that it was ‘not easy to keep a local branch here’.⁸⁵ We can note two things in relation to this. To begin with, social cooperation is never easily achieved. Even in small-scale associations of the kind typical to the early nineteenth century or, as here, small, local-level party branches, successful cooperation requires repeated and continuous experience within the framework of the organization and its ideological values and beliefs. It remains a statistical fact that this pertains particularly to the formation of mutual or horizontal trust between citizens (Åberg & Sandberg 2003: 116–117).⁸⁶ Such conditions were no longer in place as the activities of the local branches of ‘Frisinnade landsföreningen’ dwindled to a minimum, or simply ceased.

Furthermore, the tensions among the liberals on the issue of prohibition suggest more deeply running ideological controversies, which ultimately pertained to competing conceptions of political individualism. Urban liberals had, indeed, moved closer to the popular movements in their views on citizenship as an effect of the turn towards progressivism, but obviously not close enough. These were differences that dated back to at least the 1890s and the endeavours of the suffrage movement, and later contributed to the increasing divide between liberals and social democrats. To some extent the same kind of differentiation also characterized the struggle between the liberal party conventions and the parliamentarian group around Edén. In a sense, therefore, the upheaval most notable among the teetotallers was in actual fact an expression of a more deeply rooted disagreement on fundamental matters of civic and political association and political modernization. Hence, when viewed from the perspectives of ‘Frisinnade landsföreningen’, the situation looked ominous when the 1921 elections began.

In several respects these elections were unique. They had been triggered by the 1918–21 constitutional reforms which, to follow Kahan (2003), marked the passing from a politics of ‘limited suffrage’ to that of democracy. For one thing, this meant an enormous increase in eligible voters, mainly by the inclusion of women in the political system. Also, although social and economic issues played an important role as an effect of the post-war crisis, no single issue dominated the campaign if we consider the emerging party system as a whole (Esiasson 1990: 127). This feature confirmed that doctrines and programmes had become more important at the expense of single issues. Still, precisely the single issue of prohibition was rapidly tearing up the liberal front, this at a time when competing parties devoted their efforts elsewhere. If not sooner, so in any event when the nomination procedure was completed in early August, it was clear that factionalism would obstruct the elections: the liberals in Värmland entered the elections with no less than three, separate lists: one rural list, one for the towns, and one for anti-prohibition candidates, with A. Zachrisson from Karlstad as primary candidate.⁸⁷

Not surprisingly, the most distinguishing feature of the elections on the national level was the victory of the Social Democratic Party, which attained a national average of 36.2 per cent. Another prominent feature was the overwhelming defeat of the Liberal Party, with a result of 18.7 per cent. As in previous elections the liberals, however, still fared much better in Värmland with a 29.5 per cent result but, importantly, simply because of the overwhelming support from the rural and, in effect, prohibition list.⁸⁸ Clearly, this showed how political events had overtaken the position held by ‘classic’ liberals, now among the moderate camp, such as most notably Mauritz Hellberg. Indeed, the rhetoric of Karlstads-tidningen as the first election results were made public serves as an pertinent illustration both of this and the fact that the ideological leadership of the party and its members now stood very far from one another. Rather than dwelling on the issue of prohibition and intra-party factionalism–obviously too embarrassing a topic to be discussed–the newspaper identified the lack of ‘class-related prejudice’ as a main source of defeat. It stressed that, contrary to the socialists and the agrarians (‘Bondeförbundet’), the Liberal Party fomented no such ideas to profit politically from.⁸⁹ This line of argument, viz. rooted in a late nineteenth-century middle-class morality of cross-class cooperation, became a smokescreen intended to conceal the bitter factionalism within the party.

If anything, the prohibitionists now gathered strength from the results in the belief that the demands for a ban on liquor would prevail, not least since a date for a referendum on the issue was set in the aftermath of the elections: August 1922. Any hopes the prohibitionists had for a ban on alcohol, though, proved to be wishful thinking considering the outcome. Värmland was among those regions where the majority voted in favour of prohibition, but in the country as a whole a narrow majority of 50.7 per cent voted against the idea (Johansson 1980: 74–77). It is certainly true, as Johansson (1980) argued, that internal divisions had hampered the Liberal Party from the outset, and that these were at the core of the conflict. However, drinking as a social issue had been part of liberalism and its predecessors since the very beginning–including among the intra-parliamentarian factions in the second half of the nineteenth century, but also, even more importantly, among the non-political, civic associations outside parliament which had once provided the base for the extra-parliamentary branch of the movement.

Both phalanxes now recovered and redeployed for the final showdown. It unfolded in conjunction with the May 1923 party convention in Stockholm. The proceedings have been dealt with in detail by previous research (see Johansson 1980: 193–253), but to outline them briefly, opinions were divided on how the results of the referendum should be interpreted in relation to the party programme and, consequently, how strongly the demands for prohibition should be formulated. More important from our point of view were the consequences, in particular on the regional level. As an effect of the conflict, the moderates left the party, although many delegates warned against the consequences, among them Mauritz Hellberg. Would not the effect of a split marginalize the liberals similarly to what, according to him, had proved to be the case with the agrarians? Hellberg instead favoured the idea of a united party and, in this respect, also quoted the women’s and youth organizations of the party as examples to be followed.⁹⁰

Alas, a split was virtually impossible to avoid at this stage. Following the negotiations on 27 May, some 70 delegates immediately gathered–at 10 p.m.–to elect a working committee that would prepare the formation of a new Liberal Party. From now on, then, the Swedish liberals would be divided between, on one side, ‘Frisinnade folkpartiet’, containing the majority of the old ‘Liberala partiet/ Frisinnade landsföreningen’, and, on the other side, ‘Sveriges liberala parti’; the latter was formally established on 7 October, although the process was not fully completed until the following year. The split also meant that Nils Edén in effect retired from political life, although he retained his seat as deputy to parliament until 1924. While Carl G. Ekman took the chair of ‘Frisinnade folkpartiet’, a lawyer from Stockholm, Eliel Löfgren, became the chairman of ‘Sveriges liberala parti’. Mauritz Hellberg sided with the latter faction.

If anything, the 1923–24 events gave final proof of the deficiencies of weakly developed party organizations cum organization by means of proxy. Not surpisingly, professor Nils Alexandersson, who had joined in the efforts to organize the new ‘Sveriges liberala parti’, emphasized the need for a coherent programme. In his view it was ‘unsound’ for political association to be based on privately negotiated opinions alone. Rather, the ‘spine and backbone of a party must consist of a mutual and carefully considered doctrine’.⁹¹ Though the former option had been feasible, to some extent, in the nineteenth century, it was an untenable position in the era of democratic mass politics and modern party systems. However, when considered in hindsight, ideology as such was only one of two problems faced by the new party. An equally pressing concern was that of the organization. As secessionists, the liberals had been forced to leave the ‘frisinnade’ in possession of the established organization of the old ‘Frisinnade landsföreningen’. Of course the new party rallied to put together a roster of reliable people who could mobilize prospective members, but as the central registers for 1923–24 suggest, it met with considerable difficulties in doing so. Värmland was no exception. ⁹² On the regional level, no actual material has been recovered for ‘Sveriges liberala parti’.

By contrast, the ‘frisinnade’, which continued to operate through the organizational framework of the old party, suffered from the split to a much lesser extent. This was amply demonstrated by the first post-split election results of 1924. Between them, ‘Frisinnade folkpartiet’ and ‘Sveriges liberala parti’ yielded a better result in Värmland compared to the national average (21.3 per cent as compared to 16.9 per cent). However, the absolute majority of votes were in favour of the ‘frisinnade’ and only a small minority of voters aligned themselves with the new Liberal Party. Out of a total of 1,578 votes cast for ‘Sveriges liberala parti’ (2.8 per cent of the overall results in the elections), 531 of these were from cities and towns, whereas 12,116 of 14,176 ‘frisinnade’ votes were cast in the countryside.⁹³ From the late 1920s things went from bad to worse. For instance, in the 1932 elections the liberals received a result of 1.9 per cent (1.8 per cent in Värmland), and the ‘frisinnade’ 9.8 per cent (13.0 per cent in Värmland). Simultaneously the social democrats consolidated as the largest party in parliament.⁹⁴

Something had to be done, and it came in the shape of a reunion between the two factions in 1934. The liberals still considered themselves a left wing option;⁹⁵ and to ‘strengthen and stabilize’ the organization quite naturally became the main task of the new party–‘Folkpartiet’ (the People’s Party; distinct from the 1895 intra-parliamentary faction bearing the same name).⁹⁶ About this time, however, liberalism as an ideology had lost political momentum; something that would trouble it for some considerable time. Rather, in parliament the liberals would now increasingly become reduced to what Lindström and Wörlund (1988: 273) have called ‘the politics of unholy alliances’, i.e. politics more concerned with how to squeeze the most from its parliamentary position than politics guided by ideological conviction.

A transitory phenomenon

Ironically, the formation of a modern party system, and the increased role of political organization, programmes, and doctrines, spelled the demise of liberalism as an ideological movement. The principle of organizations by means of proxy had clearly proven to be a transitory phenomenon at the same time as it tended to shape indeterminately many a liberal’s view on how to mobilize grass-roots support. This is all the more ironical since liberalism and its ideals of civic and political association had been the major radical alternative in nineteenth-century society. In the main, liberalism certainly moved further to the left during the crucial years around the turn of the century, the only problem being that many of its supporters moved on once the party decided to stop. By the 1930s liberalism had been surpassed by the social democrats and lost its prerogative to identify, aggregate, and structure the major political issues in society. Yet this development was at the same time part of a more profound modernization of politics. Once universal suffrage was introduced everybody, except the most truculent of conservatives and communists, subscribed to those basic civic values which had originally nurtured liberalism. These no longer provided enough of a platform from which to form and sustain a government, to follow Gunther and Diamond (2001).

There is one final point to made with respect to this, viz. the development of liberalism in relation to the idea of a specifically ‘Swedish model’ for political modernization and democratization. While it is beyond doubt that the organizational weaknesses of liberalism and, looked at from the other end of the spectrum, the organizational skills of the Social Democratic Party, played important roles in the changes of political culture, other aspects of the problem make less sense, at least judging by the case of Värmland. To begin with, the ‘organizational skills’ of social democracy, however formidable, had yet to be cemented in a more long-term strategy. As most notably Schüllerqvist (1992) points out, the 1920s remained a period of probing different options in this respect. From that point of view it would not, perhaps, have been entirely impossible for the liberals to regain political momentum. Had it not been for their internal weakness, there was nothing inevitable about their defeat. Related to this, too, is the matter of the composition of the liberal electorate.

Since the assumption of a ‘Swedish model’ rests heavily on the relative independence of the peasantry as well as the political feebleness of the middle classes and, therefore, implicitly that of liberalism (Therborn 1989), it is worth stressing that the self-reliant peasantry were, indeed, the body and soul of the Liberal Party: As late as the early 1920s, nearly eight out of ten liberal voters in the country were rural dwellers, and close to nine out of ten in Värmland (chapter 1). Even though far from all of them were farmers, of course, the latter still had a strong voice in the party organization, both as passive members and as activists. There was certainly nothing feeble about the farmers as such. What one finds lacking in the 1850–1920 period were coherent organizational and ideological mechanisms that could bridge the urban-rural divide and, later on, the related separation of ‘liberals’ and ‘frisinnade’. Where such cross-cultural alignments were possible, such as in Denmark (see Thomas 1988), liberalism continued as a vital political force. Where they proved unfeasible, as in Sweden, liberalism failed.

Adding to this were the increasing doubts among many farmers about the true value of liberalism from the point of view of what they came to consider as agricultural necessities, such as state support and protection. It is this perspective that adherents of the ‘Swedish model’ have made into one of their most important points: the fact that the peasantry finally discarded what had become a much weakened middle-class liberalism and, instead, opted for alliance with the social democrats, in 1933, in exchange for protectionist legislation, has very often been interpreted as a victory for consensual bargaining. Yet, while the ‘Swedish model’ adherents have a point, not only the weaknesses of the middle classes should be considered an important factor, but also the inability of the farmers to extend their perspective and look beyond their ‘class interests’, rather than their contribution to consensual bargaining as the linchpin of parliamentary politics. If anything, it was the class interests of the farmers that provided the rationale for their alliance with the social democrats, not political universalism. At the same time as their position may be described as being closer to the economic individualism of early liberalism, their apparent reluctance, at least in Värmland, to immediately shift their support to agrarian organizations and parties also suggests that liberal, cross-class politics were, after all, possible up to a point. Whether by force of culture, tradition, or the socio-economic composition of the peasantry, events in Värmland were not an historical inevitability. They suggest the possibility of alternative paths in the development of grass-roots democratization.