1. Why this symposium?
Why do left parties lose vote shares in times of economic crisis and hardship? Why do right-wing governments implement seemingly left-wing policies, such as labor market activation? Why do some countries react to increasing economic inequality by increasing welfare state expenditures and coverage, while other countries do not? Why is representation becoming more and more unequal? Why do workers vote for right-wing populist parties? And why do Social Democratic governments reduce welfare benefits in some countries but not in others?
These are some of the key questions dominating the politics of advanced capitalist democracies today. And—rightly so—they are also among the key questions that comparative politics research has been addressing over the past two decades. Many political science theories propose meaningful and important answers: voters are increasingly mobilized by identity politics, rather than distributive politics; parties in some countries converge programmatically, due to de-alignment and party cartelization; business interests and lobbies hijack political elites and reduce the quality of representation; and exogenous constraints—e.g. globalization—have reduced the room to maneuver for governments. There is certainly much validity and explanatory power in these theories. However, there is an additional and distinct approach to all of the questions above, which emphasizes socio-structural transformations in the labor market rather than elite politics or exogenous constraints: most of the processes above should also be understood with reference to increasing labor market inequality and its political implications.
We take rising labor market inequality to mean that employment, employment protection, as well as the social and economic rights attached to employment are distributed more unequally over time. Regular employment with full access to social and legal protection is increasingly concentrated in particular socio-economic groups, while other groups experience employment vulnerability of mounting scope and intensity. The precise terminology may vary: labor market inequality, labor market risk, dualization, or insider-outsider divides. All of these notions refer to a process that divides labor into more secure and more fragile employment. This divide takes a structural form in the labor market—the unequal distribution of risks across groups—but it can also take an institutional form when referring to the segmentation of labor markets by regulatory means (employment protection, wage coordination rules), or to the segmentation of welfare states through an increasing differentiation of social rights depending on employment status.
The relevance and explanatory power of labor market inequality for mass politics have not been fully acknowledged in comparative political science to date and this is the reason for our symposium. Labor market inequality affects political preferences and behavior, electoral politics, representation, and government strategies. It helps explain why parties cannot be responsive to the “working class” as a whole anymore, why the same activation policies can be both progressive or regressive in their distributive effects, why some parts of the society have almost lost their political voice, and why parts of the working class may perceive their interests best protected by populist parties.
In this introduction, we will explain these arguments in detail. The main claim and purpose of our symposium—and the reason we publish it in a generalist rather than specialized political science journal—are to make broader comparative politics research more aware of the crucial structural changes that labor markets have undergone in the advanced capitalist democracies of the OECD, and of the tremendous implications these changes have had for politics. While in comparative political economy, it is well established that labor market inequality has increased and that the working class cannot simply be conceptualized anymore as being homogenous in its risks and political demands, these insights have not been sufficiently integrated in wider political science research on mass politics, institutional change, and policy outcomes.
This introduction is meant to provide readers with an understanding of the origins and advances in research on labor market inequality and to convince readers of its relevance. It is structured along three key questions: where do we stand in research on labor market inequality and dualization? Why does it matter and to whom? And where do we go from here? The individual contributions then highlight the key debates in this research field and their implications for different research strands.
2. Where we stand: origins and advances in research on labor market inequality
This section provides a—non-exhaustive—account of the origins and the development of research on labor market inequality and dualization, in order to introduce some key concepts and notions, but also because these origins may account for the fact that mainstream political science has only weakly integrated this research in its theories. Indeed, the intellectual precursors of today’s research agenda on labor market divides can be separated into two broad streams: one in economic sociology and one in economics.
In economic sociology, research on labor market segmentation can be seen as the first approach that conceptualized the heterogeneity of labor (as a political class). This literature argued that structural characteristics of jobs and economic sectors create subgroups within the labor market, with far-reaching consequences for the welfare state and the electoral politics of advanced democracies (e.g. Berger and Piore Reference Berger and Piore1980; Goldthorpe Reference Goldthorpe1984). This research also brought about a rediscovery of potential conflicts of interest between employed people and people with a broken work history or only access to “atypical” forms of work contracts. Earlier welfare state studies conceptualized this distinction between workers and the poor in the analysis of the “dual welfare state” (Tussing Reference Tussing1975; Leibfried and Tennstedt Reference Leibfried and Tennstedt1985; Weir et al. Reference Weir, Orloff and Theda1988). The conceptual distinction of different segments of “labor”, however, rarely reached mainstream political science on mass politics.
Some economists also investigated the relevance of intra-labor divides and they sharpened these conflicts of interests into the notions of “insiders” versus “outsiders.” Lindbeck and Snowers (Reference Lindbeck and Shower1988, Reference Lindbeck and Snower2001) used this idea to explain the prevalence of high unemployment in European labor markets. They argued that unionized workers demand excessive wages and shift the costs to more vulnerable segments in the labor force. Such an approach located the “implied” political mechanism accounting for unemployment not in a conflict of unions and employers reacting to exogenous structural pressure, but in a more political interpretation of trade union agency. This version of insider–outsider theory uses a form of interest group logic to explain labor market regulation and income policies but it falls short in accounting for the role and diversity of industrial relations in advanced economies (see Busemeyer and Kemmerling in this symposium). Another economist, Gilles Saint-Paul (Reference Saint-Paul1998), applied the notion of insiderness to the logic of general interest politics and voting: unionized voters agree on the introduction and maintenance of labor market institutions (employment protection, wage bargaining institutions, etc.) to enhance their wage bargaining power at the detriment of non-unionized, unemployed voters.
These ideas—though without the focus on rent-seeking dynamics that prevails in the economics literature—were picked up by political sociologists like Gosta Esping-Andersen (Reference Esping-Andersen, Kitschelt, Lange, Marks and Stephens1999), who discussed whether these labor market divides constitute a new form of class divide, or merely a symptom of eroding class structure. Along these lines, David Rueda’s work (Reference Rueda2005, Reference Rueda2007) goes farthest in applying the notions of insiders and outsider to political science. He relies on a simple dichotomy of individuals in regular versus atypical or unemployment to account for differences in the preferences of voters, and for differences in the programs of political parties who represent them. Rueda’s key concern was to understand whether, when and why employed people and their representatives (labor parties and unions) support measures that are beneficial to the unemployed, and what new political dilemmas labor market dualization creates for advocates of the “working class.” Rueda’s early work has resonated in the fields of comparative political economy and welfare state research, because it put on the table a research agenda, asking big, far-reaching questions. First, how do post-industrial labor market transformations affect individual-level policy preferences? This question brought back structural change to the agenda of welfare state analysis, and it linked it with a call for the development of a sound theoretical and empirical micro-foundation of post-industrial policy making. Second, how and by whom are changing preferences mobilized and represented? How do changing preferences transform party competition? These last questions went to the heart of changing partisan politics at a moment when most researchers in comparative political economy had understood that the traditional model of partisan politics had become inadequate, yet a new one was not yet in sight (Häusermann et al. Reference Häusermann, Picot and Geering2013). Third, what are the policy consequences of structural labor market change and changed party politics? This emerging literature reminded CPE and welfare scholars that the usual “dependent variables,” i.e. more versus less public spending or redistribution had become inadequate to understand the nature of post-industrial welfare politics. Finally, how does policy feed back into shaping risk allocation and politics? The topic of policy feedback had been brought up earlier by Pierson (Reference Pierson and Pierson2001), but Rueda’s piece very effectively laid the (sometimes paradoxical) effects of welfare state reform on the line by arguing that left-wing achievements themselves may over time undermine working class solidarity (see also Kemmerling Reference Kemmerling2014).
In sum, these were ambitious and—to many—provocative questions, which went to the heart of key topics in mainstream political science: the changing nature of class conflict, the changing role of political parties and the nature and effects of distributive policies. It is therefore not surprising that over the following decade emerged a quickly growing specialized literature on insider-outsider divides, dualization and labor market inequality more generally. This literature has achieved much progress mainly in two areas. On the one hand, it has provided ample and consistent evidence for the fact that the labor market position of individuals—labor market status, risk, vulnerability—does indeed matter for their specific social policy preferences (beyond the usual measures of socio-economic status). In other words, people do not just hold preference for more versus less welfare spending, depending on their income, but they do prefer specific social and labor market policies that cater to their labor market situation. On the other hand, the literature on dualization and insider–outsider divides has provided fruitful theory and evidence on policy feedback, i.e. on the ways in which policies and policy reforms may have dualizing effects.
Regarding the first strand of prolific research, focused on individual-level analyses of labor market risks and preferences, there have been various contributions on the ways of conceptualizing and measuring individuals’ positions in the labor market and on where to draw the line of “dualization.” In this respect, the literature on labor market inequality (as other contributions before it) has moved away from Meltzer–Richard political economy approaches in which the only relationship that mattered was the one between present income and redistribution. Rueda relied on the type of labor market contract and others followed this operationalization (see the contributions by Marx/Picot and by Vlandas in this symposium). Other researchers have modeled individual labor market heterogeneity differently: Iversen and Soskice, for instance, focused on skill levels (Reference Iversen and Soskice2015) or on the transferability of (specific) assets (Reference Iversen and Soskice2001). In other words, employed people’s attitudes to redistribution depend on the transferability of skills from one job or occupation to the other. Rehm (Reference Rehm2009, Reference Rehm2016), as well as Schwander and Häusermann (Reference Schwander and Häusermann2013) proposed measuring labor market risk in the form of frequencies, i.e. rates of unemployment or atypical employment in different occupational categories. They use such risk measures to explain variation of individual preferences and of aggregate outcomes such as redistributive policies.
Rueda and Stegmueller (Reference Rueda and StegmuellerForthcoming) propose a way, based on the labor economics literature on life-cycle profiles, of estimating the present value of an individual’s expected future income as the result of the interplay between age and experience and argue that many of the existing approaches in comparative political economy (emphasizing concepts as diverse as risk or mobility) are best understood both theoretically and empirically by being integrated into this model of expected life income. Despite these differences in operationalization, all these contributions agree that labor market differences provide strong political dividing lines in the electorate.
The second line of cumulative scholarly progress in this field refers to studies on institutions and institutional change, more precisely studies on the determinants and effects of dualizing institutions. These studies take top-down perspective, and conceptualize dualization as an—at least partly—institutionally created phenomenon related to macro politics. Hence, scholars have often used institutionalist arguments to explain (aggregate) dualization and segmentation of labor markets (Esping-Andersen Reference Esping-Andersen, Kitschelt, Lange, Marks and Stephens1999, Rueda Reference Rueda2007, Kemmerling Reference Kemmerling2009, Palier/Thelen Reference Palier and Thelen2010, Emmenegger et al. Reference Emmenegger, Häusermann, Palier and Seeleib-Kaiser2012). Dualization, in this literature, means the regulatory, institutionalized differentiation of labor market participants into two segmented pools: one with a relatively high level of socio-economic security, good protection, and material compensation, and one with a more fragile character, lacking in protection and compensation. In this perspective, it is first and foremost a particular type of institutional configuration and interest group representation—mainly found in continental European welfare states—which lends itself to driving wedges between different types of workers (Emmenegger et al. Reference Emmenegger, Häusermann, Palier and Seeleib-Kaiser2012; see also Busemeyer and Kemmerling in this symposium). Technological change is another factor affecting labor market vulnerability, and its effects for individuals whose occupation is susceptible to automation are not clearly correlated to income or education, but rather to the type of tasks. See, for example, Thewissen and Rueda (Reference Thewissen and Rueda2019).
The main ongoing debates in this literature are still to be found in these two areas. Scholars argue about the appropriate way to conceptualize and measure labor market inequality, and they discuss its political consequences for policy preferences, as well as the distributive effects of macro institutions, i.e. social and labor market policies. However, the implications of this research to key questions of comparative politics loom much larger and have not been sufficiently explored yet.
3. Why it matters: the relevance of labor market inequality to research on mass politics, institutional change, and policy outcomes
We contend that it is crucial for research on labor market inequality to move beyond the analysis of labor market trends and their immediate micro-level consequences. Labor market inequality and dualization need to be integrated into mainstream theories of mass politics, institutional change and policy outcomes.
In this section, we want to illustrate the relevance of labor market inequality to three broad sub-fields of comparative politics research: the demand side of electoral mass politics, the supply side of partisan politics, and, finally, the determinants of social and economic outcomes. For each field, we discuss key questions and how labor market inequality and dualization speak to them. This illustration is necessarily selective, but it should highlight—beyond the examples discussed—why we would remain blind to relevant explanations, or even misinterpret political developments, if we overlooked the key changes that have happened in post-industrial labor markets.
3.1. The demand-side of mass electoral politics
The study of voters, their preferences and their electoral choices, is a key subject in comparative politics and the concept of class has always—implicitly or explicitly—structured key assumptions that characterize this field. Class voting theories propose that the working class votes for parties of the Left because of its socio-economically disadvantaged position and, ultimately, its preferences for high levels of state intervention and income redistribution. This is a theoretical foundation that—indirectly or directly—informs a wide range of studies on electoral choice, representation, and party system development. It is also a theory whose explanatory power is strongly affected by labor market dualization. Indeed, a wide literature has observed a decline in traditional class voting, as working-class status has become a weak predictor of individual vote choices for the Left (Evans Reference Evans1999, Elff Reference Elff2007). Why is that so? More specifically: why do Left parties—Social Democrats in particular—lose working class votes?
Labor market dualization provides an answer to this question that has largely been overlooked in the dominant contributions on changing mass politics. So far, there are two main –interrelated—explanations for the decline in traditional class voting: identity-politics and programmatic changes in parties’ positions. The first explanation refers to the increased saliency of socio-cultural programmatic issues and identity-politics in Western mass politics (e.g. Kriesi et al. Reference Kriesi, Grande, Lachat, Dolezal, Bornschier and Frey2008, Kitschelt and Rehm Reference Kitschelt, Rehm, Beramendi, Häusermann, Kitschelt and Kriesi2015, Häusermann and Kriesi Reference Häusermann, Kriesi, Beramendi, Häusermann, Kitschelt and Kriesi2015): workers have stopped voting for the Left because they care more about issues of recognition (immigration, cultural liberalism) than redistribution. The second explanation refers to changes in the political programs of Left parties towards more centrist or more conservative positions (e.g. Karreth et al. Reference Karreth, Polk and Allen2013; Evans and Tilley Reference Evans and Tilley2017): as left-wing parties have stopped representing the interests of the working class, workers in turn have started to withdraw electoral support from the Left.
Both of these arguments provide possible reasons for Left party decline, and they certainly capture important and real dynamics. However, research on labor market inequality and dualization provides a third possible factor: the “working class” is not homogenous anymore. Instead, different parts of the working class are exposed to very different forms and levels of labor market vulnerability: While parts of the working class are exposed to increasingly precarious employment conditions in terms of income and rights, others are in regulated employment, fully covered by social protection, and may fear losing precisely this status and its associated rights. These two groups do not have convergent preferences and policy demands. Hence, left parties have become increasingly unable to mobilize them jointly. How exactly this divide within the working class is structured differs among countries, and current research still strives to find the most adequate ways of drawing the line between insiders and outsiders (see the contributions by Marx/Picot and Vlandas in this symposium), but the basic structural development is fairly unequivocal: in the postindustrial knowledge economy, the working class has become split into a more vulnerable and a still secure part.
Hence, irrespectively of whether we draw the line between insiders and outsiders on the basis of work contract (Rueda Reference Rueda2005), occupational class (Gingrich and Häusermann Reference Gingrich and Häusermann2015, Beramendi et al. Reference Beramendi, Häusermann, Kitschelt and Kriesi2015), skill-level (Iversen and Soskice Reference Iversen and Soskice2015) or type of task (Thewissen and Rueda Reference Thewissen and Rueda2019), the implication is similar: different parts of the working class have increasingly diverged in their economic situation and in their political demands. It has become more difficult for political parties to mobilize “the working class” with a coherent redistributive program. A few studies have started to conceptualize and test the implications of this divide for electoral politics: Marx (Reference Marx2014, Reference Marx2015) shows that temporary workers differ in their electoral choices from workers in regular contracts; Iversen and Soskice (Reference Iversen and Soskice2015) show that labor market outsiders have largely lost electoral relevance in both majoritarian countries and continental European party systems, meaning that dominant party coalitions do not encompass their interests anymore; Häusermann (this symposium) suggests that the working class vote for the populist right is actually a “labor market insider” vote; Rueda (Reference Rueda2005, Reference Rueda2007) has long argued that Social Democrats face a dilemma of representation, because they have difficulties serving the interests of insiders and outsiders at the same time. Lindvall and Rueda (Reference Lindvall and Rueda2013) have tested this dilemma most explicitly in the context of Sweden, a country usually characterized as having a fairly universal welfare state and unsegmented labor market (i.e. a least likely case). But even in Sweden, as they show, Social Democrats experienced the electoral dilemma of labor market dualization first hand: When they advocated the interests of labor market outsiders in the mid-1990s (strengthening unemployment relief and active labor market support), they lost votes among the skilled insider voters. However, when they focused their electoral program more clearly on the issues relevant for skilled working class and middle-class interests (such as childcare support), they won back the votes of labor market insiders while losing support among the outsiders. The strategic dilemma ensues because the policies that are supportive of outsiders directly harm the interests of insiders, by creating labor market competition or implying higher taxes.
The theoretical implication of all these contributions is very clear: dualization of the labor market explains a loss of vote shares by mainstream Left parties that has less to do with rival conflict lines (identity-politics, nationalism, immigration) or with a “neo-liberal” right-shift of the Social Democrats. Rather, different segments of workers want different policies and parties are facing hard decisions and trade-offs.
In a wider perspective, the increasing divide within the working class affects the way in which we should conceptualize the analysis of present-day mass politics. Aligning voter preferences on a single economic dimension—pitching more income redistribution against less redistribution—and deriving electoral preferences from this positioning does not do justice to the demands different voter groups have. Research on mass electoral politics needs to acknowledge the distinct preferences of different groups among lower class voters, especially in countries that are affected by strong labor market segmentation.
3.2. The supply-side of partisan politics
Labor market dualization affects voters and their electoral preferences, but it also affects partisan politics, i.e. what political parties advocate, who they form coalitions with and what policies they implement when in government.
Two main mechanisms account for the fact that dualization changes partisan politics: the first mechanism refers to changing voter preferences and voting behavior, as illustrated in the section above (parties may face trade-offs in the representation of voters’ interests). The second mechanism is the one we want to illustrate here: depending on the labor market structure in a country, the same policies do not have the same distributive effects. In other words, one policy can be egalitarian in one country and stratifying in the other—“left” in one country and “right” in the other. We illustrate this argument with reference to two recent studies that have made this point very clearly. Gingrich and Ansell (Reference Gingrich, Ansell, Beramendi, Häusermann, Kitschelt and Kriesi2015) show that investing public money in labor market activation policies (training, employment incentives) is a left-wing policy in dualized labor markets with high levels of employment protection for insiders, but a right-wing policy in countries with flexibly and weakly regulated labor markets. The reason for these diametrically opposed consequences lies in the fact that the distributive effects of the same policy are conditioned by labor market dualization. In dualized labor markets, insiders are well protected through wage regulation and employment protection, and training systems tend to be controlled by trade unions. Therefore, expanding activation policies would not undermine the position of insiders, but would rather sustain high wages throughout the economy. However, in flexible, weakly regulated labor markets, bringing more outsiders in the labor market tends to increase wage dispersion and competition. Consequently, at low levels of labor market dualization, left parties are not associated with higher training spending then right parties, while they are in countries with high employment protection levels. Hence, when a policy (increasing spending on training and activation) is expanded in two different countries, its distributive effects (help workers and cost employers versus help employers and cost workers) may be the opposite. Since labor market dualization conditions the distributive effects of policies (and expanding the welfare state is not necessarily egalitarian), we should not expect all Left parties to advocate welfare state expansion.
A second telling example should further illustrate this insight. Both Fernandez-Albertos and Manzano (Reference Fernández-Albertos and Manzano2016) and Beramendi and Rehm (Reference Beramendi and Rehm2015) start from the puzzle that preferences for welfare state expansion are correlated with income to very different extents in different countries. Fernandez-Albertos and Manzano, e.g. show that people in precarious socio-economic conditions in Spain are much less supportive for welfare state expansion than British people in the same situation. To solve this puzzle, both recent contributions argue that the relationship between the socioeconomic attributes of individuals and their views of the welfare state is conditional on the structure of the labor market (Fernandez-Albertos and Manzano Reference Fernández-Albertos and Manzano2016), or more particularly on the progressivity of tax and transfer policies in a country (Beramendi and Rehm Reference Beramendi and Rehm2015). In dualized labor markets, the welfare state (which has itself contributed to this dualization by creating different categories of beneficiaries) is to a large extent regressive in nature: social benefits are tied to regular employment, which is the prerogative of labor market insiders. Hence, expanding pension insurance or unemployment benefits may hardly benefit labor market outsiders at all.
This conditioning effect of labor market dualization on the distributive effects of policies has very important implications for how we conceptualize partisan politics. The simple expectation that left parties want more welfare spending and right parties want less may be simply wrong. Hence, research that explains government choices, party programs or party competition needs to pay attention to the distributive effects of policies in a particular context.
3.3. Institutional change and policy outcomes
Given that research on labor market divides is often directly related to research on the welfare state, it is perhaps surprising that we argue that its full implications are still understated. However, we see three major areas where the implications of the literature on labor market divides can still contribute significantly to our understanding of political and socio-economic outcomes: research on the nature and evolution of the welfare state, research on inequality, and research on the politics of well-being and satisfaction.
Let us begin with research on welfare state change. Labor market dualization is relevant for our understanding of welfare state politics, because it affects actors, issues, and power relations in two ways: labor market transformations structure the dynamics of politics and they affect reform outputs. Indeed, current welfare state politics take place in a context of austerity and budgetary constraint. For a long time, the literature on welfare state politics assumed that institutional legacies and path dependency account for this changed context (Pierson Reference Pierson and Pierson2001, Hacker Reference Hacker2005), since mature welfare states have over time created an imbalance between benefits and funding. Research on labor market inequality, however, sheds more light on the structural changes in the labor markets, which create new risks that are distributed unequally across countries. These new risks (unemployment, working poverty, precarious employment, lacking contributory record) compete with traditional demands on the welfare state (Bonoli Reference Bonoli2005, Reference Bonoli2013). Thereby, they shape the welfare reform agenda, but also its political dynamics: as the distribution of risks is not aligned with the distribution of income and other socio-economic resources (e.g. Häusermann et al. Reference Häusermann, Kurer and Schwander2014; Schwander in this symposium), the preferences and power of particular risk groups are not equal to the demands usually conceptualized in power resources approaches.
A good illustration of a fruitful exchange between these two dynamics—welfare state and labor market transformations—is Philipp Rehm’s work (Reference Rehm2016) on the effects of historical risk shifts on welfare state dynamics. He shows that economic shocks affect the distribution of risks, with far-reaching consequences for welfare politics: while the Great Depression aligned risk- and income-driven demands for redistribution (i.e. similar social groups were affected by the two forms of hardship), asymmetric shocks in more recent times have driven the two forms of demands apart, creating conflicting claims to the welfare state. Thereby, they have increased the political pressure on existing welfare state arrangements.
Perhaps more importantly, the political consequences of new labor market divides go beyond the narrow realm of welfare state politics. To begin with, tax systems redistribute not only income, but also job opportunities (Kemmerling Reference Kemmerling2009), thereby affecting labor market risks directly and also affecting the distribution of power and power resources in distributive politics. This implies that tax politics is deeply affected by insider-outsider differences. It is not hard to extrapolate these examples to other policy areas such as foreign economic policies or immigration (e.g. Emmenegger et al. Reference Emmenegger, Häusermann, Palier and Seeleib-Kaiser2012).
Moreover, the usefulness of dualization as a theoretical concept should not be limited to “mature welfare states.” Rueda, Wibbels and Altamirano (Reference Rueda, Wibbels, Altamirano, Beramendi, Häusermann, Kitschelt and Kriesi2015) argue that dualized labor markets are a worldwide phenomenon. In many non-industrialized countries, workers are divided between those with formal contracts that come with valuable benefits and extensive labor market protections and informal workers who work under contingent contracts or no contracts at all. This latter group receives few or no labor market protections and lower or no levels of social benefits. They are the world’s labor market outsiders. In the OECD context, the process of dualization has been linked to a number of political and economic processes: increasing competition in manufacturing, the rise of the service sector, the decline of unionization, political choices by Left governments, etc. Echoes of these arguments are present in work on developing countries, where dualization is closely linked to the informal sector and has received a lot of attention from economists, sociologists and, more recently, political scientists (e.g. Berens Reference Berens2015). Indeed, a long tradition of models in development economics emphasize the stark income and productivity gaps inherent in “dual economies” and the uneven growth that characterizes broad swaths of the developing world (Rosenstein-Rodan Reference Rosenstein-Rodan1943; Ray 2010). Rueda, Wibbels and Altamirano propose that there is considerable room for unifying the study of welfare states and social protection across the developing and developed world around a set of factors emphasized in the dualization literature (labor market risks, the competitiveness of markets, interest organization and representation, and partisan politics). And they provide evidence that dualization is a powerful explanatory factor when looking at the development of labor markets and employment protection legislation both in OECD and non-OECD countries.
Turning now to the second area, research on labor market dualization also has implications for the rekindled interest in economic inequality. Thomas Piketty (Reference Piketty2014), to mention a prominent example, focuses on the problem of top incomes “running away” from middle or median incomes in advanced economies. This is certainly correct and could, in part, be explained by the enormous lobbying power of the very rich (Bartels Reference Bartels2009) and owners of freely moving capital (Hacker and Pierson Reference Hacker and Pierson2010). However, what is often missing in the picture is why poorer voters do not mobilize more strongly against this lobbying. To explain this lack of mobilization, scholars often cite “diversion” of mobilization through identity-politics (e.g. religion or cultural authoritarianism, see e.g. Kriesi et al Reference Kriesi, Grande, Lachat, Dolezal, Bornschier and Frey2008; de la O and Rodden Reference De la O and Rodden2008; Alesina et al Reference Alesina, Glaeser and Sacerdote2001). However, research on dualization and labor market divides adds an alternative story: difficulties in mobilization may stem from labor market dynamics, and from the structural problems they pose for “organizing labor” in the 21st century. Labor market outsiders often lack the resources to mobilize (Rovny and Rovny Reference Rovny and Rovny2017) and political actors may be torn between contradictory incentives regarding their organization (see the trade-offs mentioned above). Such an approach would also explain why it is much more the structure than the size of inequality that matters to explain the size of the welfare state and redistribution on the macro level (Lupu and Pontusson Reference Lupu and Pontusson2011). Understanding the segmented nature of political support at the lower end of the income scale and the role of labor market divides in this debate is therefore of vital importance to understand the prevalence of inequality in modern societies.
Finally, research on labor market inequality also reminds us that the consequences of job security and labor market participation go far beyond the realms of income and economic inequality (e.g. Brand Reference Brand2015). This research is directly relevant to what might be called the politics of well-being in (partially) post-material societies. In the political economy literature on labor market divides, labor market risks ultimately only matter in the form of contingent losses in income and material well-being. However, the sociological and psychological literature has long shown that the importance of having a (specific kind of) job goes beyond guaranteeing a minimum income (Jahoda et al. Reference Jahoda, Lazarsfeld and Zeisel1933). Non-employment, unemployment, or atypical employment means more than losses in income and economic security. It means a total or partial lack of social inclusion in society. It means a substantive loss of meaning and often psychological trauma. Research on labor market divides is particularly well suited to elicit the consequences of such a perspective (e.g. Marx and Picot in this symposium). It may illustrate, for instance, why often well-insured, well-protected parts of the society may be afraid that despite compensation mechanisms such as unemployment benefits, they would lose social status and experience relative deprivation once being unemployed.
4. Where we go: ongoing debates in research on labor market divides and the structure of this symposium
This symposium aims at speaking to an audience broader than the circle of scholars working on labor market divides and dualization. Through the above discussion, our goal was to illustrate various ways in which labor market inequality affects, transforms and conditions politics in the post-industrial capitalist economies. Beyond increases in wage inequality or welfare state retrenchment, the structural transformation of labor markets is one of the deepest and most consequential changes in our societies, driven by broad macro societal processes of deindustrialization, globalization, or changing gender roles. In the near future, structural processes like technological change and automation will most likely even exacerbate these trends. An analysis of politics—both mass and elite politics—in this context needs to be aware of the tremendous socio-structural developments that shape needs, actors and power relations.
While this introduction focuses on motivating the need for and purpose of the present symposium, the various contributions that follow are more specific, and they dive deeper in order to introduce interested readers to the main debates in this research area. We identify two areas that are particularly relevant: (1) the changing nature of labor market divides and how to conceptualize and measure them and (2) the political consequences of labor market divides on the micro-level and the macro level.
The first three contributions focus on (1). This is the area where—in the wake of the general micro-level turn of comparative political economy—research on labor market inequality and dualization has invested most and progressed furthest. However, the literature is now at a point where at least three main ways of conceptualizing and operationalizing labor market vulnerability co-exist, each with their respective definition of insiders/outsiders and labor market vulnerability. For scholars referring to labor market dualization and its impact, it is crucial to be aware of these approaches and their respective theoretical and empirical implications. The line between the more vulnerable and the more secure parts of the workforce can be drawn based on one of these three criteria: labor market status (being unemployed or in irregular employment versus holding a regular employment contract), labor market risk (the probability of becoming unemployed or insecure), or subjective insecurity with regard to employment. These three contributions to the symposium review the different conceptualizations and spell out how the choice of a particular operationalization may depend on the research question at hand. Paul Marx and Georg Picot review the different empirical approaches to measure “outsiderness” and discuss their relation to different theoretical expectations. They also explain in which research areas subjective measures should play a stronger role in the analysis. Tim Vlandas illustrates the complementarities of different approaches with a stylized analysis of data on public opinion. He shows that both labor market status and occupational unemployment risk do have explanatory power for individual-level preferences, but they do not measure the same forms of vulnerability and they reflect different aspects of labor market inequality. Finally, Hanna Schwander addresses two key debates in the literature on labor market dualization, namely the relationship between class and labor market vulnerability, and the importance of family status and gender patterns in labor market stratification. Schwander shows that labor market vulnerability has spread also in the middle class. She then unpacks the large group of high-skilled workers to show that both young and female workers are strongly affected by labor market risk. Thereby, she raises the question of labor market vulnerability and household structures: do mixed households of insiders and outsiders demobilize political divides and conflict? Her answer is negative: even though labor market risk is strongly gendered, the household provides a safety net to only a small minority of women.
This brings us to the second area, the political consequences of labor market divides, which are the topic of the subsequent three contributions. Most studies to date focus on the micro-level, while questions of mobilization, preference aggregation, and representation at the macro level remain implicit. Busemeyer and Kemmerling start this discussion by highlighting the difference between the macro and the micro level. While differences in regulation and coverage of social security systems very often result in a binary dualization, especially with hindsight of recent reforms in many continental European countries, the micro level question of labor market stratification gives rise to more complex forms of political contestation. Keeping the two levels conceptually apart is also helpful to understand the long-term trajectory of dualization, which can be seen, in part, as a reaction to more general processes of liberalization. While Busemeyer and Kemmerling discuss the consequences of the macro for the micro level discussion, Silja Häusermann, in contrast, focuses on bottom-up approaches, and the electoral channel of mobilization and representation in particular. She argues that dualization is highly relevant to explain forms of current political cleavage formation. Her contribution suggests how we could explain the rise of right-wing populist parties in Europe with reference to dynamics of transforming labor market inequality. She argues that it is not the precarious outsiders—i.e. the losers of globalization and deindustrialization—who mount the populist challenge, but rather insiders who feel threatened by changes in labor markets.
Philipp Rehm, finally, combines the macro and the micro levels by looking at feedback effects of new forms of regulation and social policy on political mobilization. He extrapolates this argument analytically into the future, anticipating the implications of dualization for distributive politics in the longer run. He argues that high-skilled individuals with significant levels of wealth might have no interest any longer in participating in pooled, publicly regulated social protection schemes. As risk (and income) inequality will increase, inheritance and wealth might well overshadow insider-outsider or traditional partisan conflicts and lead to a new kind of dualization, with the privileged opting out and the less privileged being confined to restricted public pools of solidarity.