Our spring issue, “Pandemic Politics,” is out now. Get a discounted subscription today!

Before Sweden Was Social-Democratic

Erik Bengtsson

Swedish social democracy is often thought of as somehow eternal — the fruit of a solidaristic national culture, or even its historic homogeneity. But Sweden used to be just as unequal as other European countries — and making social democracy “normal” took a fight against what was once considered traditional.

The park Kungsträdgården in Stockholm, Sweden, circa 1900. Library of Congress

Interview by
Giacomo Gabbuti
David Broder

It’s easy to think of Sweden as synonymous with social democracy. For much of the twentieth century ruled by left-wing governments, the Scandinavian country still today enjoys world-leading public services, high wages, and extensive worker representation.

A certain common sense likes to present today’s welfarism as an expression of a deep-rooted culture of community and egalitarianism. Yet all this had to be fought for. In fact, until the late nineteenth century Sweden suffered extreme disparities of wealth and power in line with other European countries at a similar level of development.

Based on a “classic” indicator of Marxist analysis — the division of national income between labor and capital — Erik Bengtsson’s work focuses on how these inequalities have changed over Swedish history. His reading shows the vital role of labor’s mobilizations from the 1920s onward, and in particular its success in shortening working hours, in laying the basis for social democracy.

Jacobin’s Giacomo Gabbuti and David Broder asked Bengtsson about his research and what the study of labor, inequality, and political economy can tell us about the bases of social change.


As you explain in a recent essay for Past & Present, it has been commonplace to see Sweden’s twentieth-century egalitarianism, its social-democratic “middle way” between capitalism and socialism, as the outcome of a long historical trajectory. This is part of a broader tendency to attribute modern political and social outcomes to long-term development and the “persistence” of historical settings. Indeed, you define this view as a “Swedish Sonderweg” — a reference to perhaps the most famous such “exceptionalist” theory, regarding Germany’s supposedly inevitable “special path” to Nazism. What is wrong with this idea, and what are the political consequences of this “national myth”?


It’s misleading because it directs attention from what actually happened — and a lot of interesting things happened in Swedish politics and society after 1870. This myth diverts real events into a narrative of historical continuity — a narrative which may seem plausible on its own terms, but does not fit with the facts. Politically, this narrative has played different roles over time. In the interwar era Social Democrats could, as the historian Åsa Linderborg has shown, use this myth to justify social-democratic policies as native and Swedish and not “dangerous” and foreign (German, Russian).

Today, this national narrative instead plays the opposite role. It represents the existence of organized Social Democracy as an unnecessary detail of Swedish history, thus implying that social-democratic policies are not needed today either. For in this view, the good outcomes associated with the Sonderweg story (equality, some sense of fairness and togetherness) are presented as an automatic product of national culture, simply “as we are Swedes.” This suggests that we can have just any economic and social policies (tax cuts, cuts to the welfare state, weakening trade unions etc.) without endangering the much-valued outcomes that are supposedly “Swedish” by definition.

This is, of course, a very useful narrative from a neoliberal point of view. It makes an honorary salute to a cherished tradition of solidaristic politics (talk is cheap!), combined with policies that work in the opposite direction. So, by misrepresenting Swedish history, the Sonderweg narrative can also be politically useful for contemporary purposes. The center-right government of 2006–14 showed serious interest in this kind of history, disconnecting the twentieth-century association between Swedishness and Social Democracy. It instead promoted a narrative of national peculiarities, emanating from a certain combination of individualism and solidarity dating back to the imagined “old peasant society” of yore.


The “prehistory” of social-democratic Sweden that you investigate in your recent projects is that of a poor, peripheral economy, with extremely high inequalities in wealth (or in the ownership of land and capital), and a very narrow franchise. Yet other historians have considered the peculiar strength of the Swedish peasant-farming class, and its inclusion in the political system, as the factor that prevented a German-style “authoritarian” solution to the weakness of the local bourgeoisie. You seem to reject this interpretation, whether in its more materialist, class-based version, or its more “culturalist” reinterpretations, based on the supposed existence of an “egalitarian” peasant ideology. Why? And how did these inequalities in capital and power translate into political conflict?


I started out doing this research quite convinced of the values of the model promoted by political sociologist Barrington Moore, where the strength of the independent peasantry more or less guaranteed an earlier and more gradual democratization than in other, “more feudal” countries. On the face of it, that story makes sense.

But what I found when I read up on the literature on suffrage, politics, and inequality was that Swedish democratization stalled in the nineteenth century. Around 1800, Swedish politics was certainly more inclusive than most other European countries — the existence of the peasant farmer estate meant that a larger share of the population had the right to take part in parliamentary elections than in Britain, for instance. But this didn’t translate (or transition) into early democratization. Rather, Sweden had a highly conservative reform in 1865–66 and then stagnation until a further extension of suffrage — which still was very undemocratic — in 1907.

My interpretation is that the decisive factor was not the existence of a strong farmer class per se, but rather the class alliances and coalitions that were built. Farmer radicalism in the 1840s to 1860s opened up room for alliances with semi-proletarian and proletarian groups, including over the issue of democratic suffrage reform, but rich farmers instead allied with the estate owners and dropped any interest in radical reform. This is quite different from the constitutional development of Denmark and Norway, for instance, where farmers contributed to more democratic reforms. So, I don’t think that a strong independent farmer class in itself guaranteed any particular sociopolitical development, whether on material or more cultural grounds. Rather, this change was more about the interactions between social and material forces and institutions.


As in other poor and unequal countries of the late nineteenth century, one working-class response was mass emigration. One-fifth of Swedes migrated between 1870 and 1910. According to your colleague Thor Berger and the Oxford sociologist Per Engzell, these “egalitarian” migrants in turn affected US patterns of inequality and mobility. However, contrary to theoretical insights, this did not translate in lower inequality in nineteenth-century Sweden. How would you characterize the relationship between emigration, inequality, and political struggles in Swedish history?


This is a very complex question. But I’m not too convinced by Thor and Per’s analysis — I feel that there are too many factors specific to the US context (land-tenure systems especially) that their analysis doesn’t control for. But if we do accept their argument that Scandinavians had more egalitarian values than others already in the late nineteenth century, then the puzzle that poses to my research is: why didn’t they then succeed in implementing more egalitarian economic and political institutions back home (in Sweden)?

Emigration surely had major effects on Swedish society itself. In a country with a small population, and a state that had since the 1700s had a major ambition to increase its population in order to grow stronger, large-scale emigration fostered a strong sense of crisis among elites. This became an impetus for social and labor-market reform. However, the connections are not completely clear. Wealth inequality worsened, and income inequality probably did too, in the last three decades of the nineteenth century. But I think that the emigration experience increased the legitimacy for reformist projects and strengthened the so-called Kathedersozialist current, promising “reformism from above” in Sweden (as distinct from change driven by a labor movement acting in its own interests).

The huge twenty-one-volume public commission on emigration of 1907 to 1913 became one of the first really ambitious social-scientific mappings of Swedish society and its deficiencies. The commission also strengthened the hands of those who wanted reform, by pointing to the many sources of popular discontent (the elitism and inequality of Swedish society, and such like). To convince people of the lower classes to stay in Sweden, elites would have to improve their conditions. According to recent research by Mounir Karadja and Erik Prawitz, emigration also directly improved workers’ bargaining power and labor movement strength locally, by making labor scarcer and dearer. So, in this sense, emigration strengthened reformism both “from above” and “from below.”


Following your narrative, it was only after World War I that inequality in Sweden fell to the low level we are used to. The debate on Thomas Piketty’s work, and even more that regarding Walter Scheidel’s The Great Leveller, led to an interpretation of both World Wars as a modern version of the “four horsemen of the apocalypse” (that is, catastrophes, such as mobilization warfare, transformative revolution, state failure, and lethal pandemics, that had the effect of radically lowering inequality for those who survived). Sweden was, however, a neutral country; other transformative, far less traumatic factors, seem to have driven this change, such as fiscal reforms, and most of all, the labor movements and its conquests (starting with the eight-hour working day, passed in 1920). What explains these developments? What factors explain the reformist, gradual way in which Swedish workers obtained these advancements — in contrast with either revolutionary Russia or the defeats imposed on socialists in countries like Italy?


Here, it is important to see that equalization was quite slow to start with. It seems to me that a lot of the research literature on the relationship between democratization (Sweden had its first parliamentary election with universal and equal suffrage in 1921) presupposes that this should have had an almost immediate effect on redistribution, or else no effect at all. I guess that this is due to the conventions of quantitative research.

But looking at the Swedish case and, indeed, other countries in the same period, it seems to me that even after democratization and with a well-organized labor movement, there was a jagged process of “policy learning” and experimentation. For instance, the estate tax — which, as Piketty’s story of twentieth-century equalization tells us, played a very important role in breaking with the rule of the rentiers — was proposed by the future minister of finance Ernst Wigforss in 1928, but only implemented at the levels he envisioned about twenty years later. (Back when he proposed it, the seemingly radical idea of an estate tax contributed to the red scare and an election victory for the center right!)

The Social Democrats governed for decades after 1932, but the reforms providing for universal social insurance — later identified as one of the core parts of the “Swedish model” — actually only came in the 1940s and 1950s. So, while the eight-hour working day and union militancy did matter already in the 1910s, political reforms were slower in the making.


A constant of your work has been the attention to the labor share — the “class” measure of the division of national income between labor and capital, workers and bosses. Long considered an “old” tool, in your historical work it is clear how this indicator reflects crucial leveling factors, such as the strength of unions, and power relations more broadly. What is the lesson for the modern debate on economic inequality? Does labor share — and therefore, trade unions — drive trends in income polarization?


This is a huge discussion. But I think that in the last five to ten years the debate has shifted, so that everyone now acknowledges that the capital–labor distribution of national income matters for inequality. Previously, say in the 1990s, when an implicit denial of the fact that we live in capitalist societies was part of civilized discourse (even in the social sciences), it seems that it was just assumed that all key inequalities in society are between different categories of wage- and salary-earners. It was a part of juste milieu thinking that classical capitalist polarization between capitalist and worker was a thing of the past in the “knowledge economy.” But whatever ideologists claim, it’s just a fact that capital incomes are an important part of the economy, and that they are very differently distributed than labor incomes are. Most people of working age have labor incomes, but very few people have capital incomes of any significance. So if the capital share grows, then inequality will increase, as the small elite who owns a lot of dividend- and interest-bearing assets benefit the most from rising capital income.

I think that there have been several important contributions to our understanding of this fact since Piketty’s book (and before that, since the important 2009 paper by Tony Atkinson, the late British economist and Labour activist among the first to draw attention on rising inequalities; his paper did a lot, I think, to call researchers’ attention to factor shares).

Daniel Waldenström and I had our paper on the historical connections between capital shares and inequality, which is mostly an empirical contribution. I think that in a sense something we have learned from this debate is that we live in a society which is much more “classical capitalist,” to use Branko Milanovic’s typology, where there are — in very simplified terms — capitalists and workers, rather than an imagined “new capitalism” where everyone has a little capital and does a little wage labor.

An economist called Ignacio Flores has a nice paper showing how using income surveys (like the often-used Luxembourg Income Survey) underestimates capital incomes and underestimates their importance for inequality. I think that this is the way the discussion — in research at least — is going now: we know that the distribution between capital and labor matters and that this has been misunderstood previously for both theoretical and methodological reasons. Now that we’ve got that misunderstanding out of the way, we can work on a more sophisticated understanding of the matter. Milanovic thinks that the solution to the problem of capital-driven inequality is equalization of capital ownership rather than the taxation of capital income which Piketty has argued for, and that’s an interesting debate — but I’m just happy that the insight is shared that we do in many ways live in a “classical capitalist” world. Building on that understanding, we can discuss what to do about it.


After World War II, and especially from the 1960s, Swedish social democracy radicalized, and impressed the world with some of the most ambitious social policies in world history, with radical transformative proposals such as the “Meidner Plan.” Recent news from the country, however, shows the far right winning their best results ever, and though Sweden remains socially more advanced than most Europe, 1970s social democracy looks increasingly like a paradise lost. Which factors do you think help explain these developments? Is the Swedish right also embracing the “national myth” of egalitarianism, and how do you think they will transform the institutions underlying Swedish exceptionalism?


Again, a very complex question! But starting from the Meidner Plan, it does seem that industrial relations and capital-labor relations polarized more in Sweden than in neighboring countries like Denmark and Norway in the 1970s and 1980s. The Swedish trade union movement was more radical than its colleagues in neighboring countries (and decreased the capital share more markedly). And it seems this also triggered a more radical, neoliberal reaction from the employers’ side. So that changes the context for party politics too.

As regards explanations of Social Democratic decline after that, I don’t have much to add to conventional accounts — there was a loss of self-confidence, a vicious cycle of unprincipled market liberal policies leading to the demobilization of party members and voters, leading to even less meaningful policy, leading to further demobilization. Specifically for the Swedish case, I also think it matters that the historical organizational wave of the labor movement had somehow run its course by the 1970s and 1980s.

In the late 1970s, electoral surveys still showed that middle-class people who had grown up in working-class households voted Social Democratic. There was a longstanding loyalty, there. But the children of these middle-class people have no such attachment to Social Democracy (and of course it has not attracted new members in the same powerful way as it once could).

Two sociologists, Jens Rydgren and Sara van der Meiden, recently wrote about “the end of Swedish exceptionalism,” here referring to the historic weakness of the extreme right. For a long time, we didn’t have a successful right-wing populist party. But since 2010 we have done so. As I interpret their paper, Rydgren and van der Meiden trace that development back to this crumbling of Social Democratic hegemony which I talked about above. For most young people in Sweden today, there is no obvious attraction to the Social Democratic Party, no reason to be loyal to it. The end of this hegemony, which had once meant that the party could count on hundreds of thousands of loyal voters (in Swedish called häströstare, “horse voters,” meaning that they would vote Social Democratic even if a horse was the party candidate), means that there are many more voters to compete for between the bourgeois bloc, the center-left, and the extreme right.


Your work and, indeed, your answers here, have provided insight into what is a much more complex society than many would imagine. But the usual counterargument to “Swedish” proposals — in countries like Italy, where both of us are based — invokes another kind of “exceptionalism.” For we often hear that Sweden is just “too small” and — with just one major city to win for the workers’ movement — culturally unique, to provide any kind of model. Those arguing for tighter immigration controls even contend that Sweden’s social solidarity is only possible because it is “ethnically homogeneous.” Does your debunking of the Swedish Sonderweg help answer this? And what does the history of labor and inequality in Sweden have to tell the wider world?


The one version of “Swedish exceptionalism” which I think that my research disproves is the notion that you have to somehow already be social-democratic (have a longstanding tradition of egalitarianism, etc.) to be able to pursue social-democratic policies today. When people dismiss the Swedish case as irrelevant because “Sweden was always like that” — I don’t believe that.

The arguments about smallness and ethnic homogeneity are not directly addressed by my research. I guess that what I show is that the connections between those factors and the outcomes of egalitarianism / Social Democracy aren’t really obvious. Sweden was a Lutheran country from the 1500s onwards, but the connection between the history of the state church and twentieth-century outcomes is very tenuous. Sweden was a small country with the ethnic makeup it long had, without becoming particularly equal or egalitarian. So, the connections between the structural factors and the sociopolitical outcomes are a lot weaker than arguments by exceptionalism presuppose.

It makes me think of a remark the historian David Cannadine made in 1987, during a conservative wave in British history-writing, when he said that “British historians today are mainly concerned to show that less happened, less dramatically, than was once thought.” I see the tendency to homogenize Swedish history and remove the conflicts, contradictions, and reversals as one instance of this kind of history-writing. Here, historians, working from a more or less conservative standpoint, aim to assert that not much changed in the past, and so the present cannot be changed either.

To demonstrate Sweden did not become egalitarian or social-democratic by virtue of some sort of ancient tradition, reinforces the counterpoint — that societies do change because of politics, because of conscious human action, and that in this sense, politics and organizing are indeed meaningful.