Introduction

In recent decades, a combination of political reforms and economic developments, including globalization and the ICT revolution, have led to rapid structural change from industry to services, to growing inequality, and to a new politics of status defense (Engler & Weisstanner, 2021; Rodrik, 2018). One striking manifestation of this new politics is the rise in rightwing populism, with the old middle and working classes turning against change and seeking to defend their economic positions and traditional lifestyles against real and perceived threats from trade, immigration, and increasing diversity (Gidron & Hall, 2017; Kurer 2020). This defense of status takes on discriminatory economic forms, such as tariffs against imports from developing countries and exclusion of immigrants from jobs and welfare benefits, as well as opposition to multiculturalism and the cultivation of nationalist and other in-group identities. A parallel, but no less striking, trend is the growing exclusiveness of many urban neighborhoods, which is rooted in the phenomenon whereby the well-educated sort into good school districts and adopt restrictive zoning and building codes that drive up housing costs and price out the working class (Glaeser et al., 2005; Einstein et al., 2019; Schuetz, 2022). While urban elites do not rely on overt discrimination, and in fact often express cosmopolitan worldviews and democratic values, their actions lead to opportunity hoarding and entrench inequality and status differentials. This new politics of status defense creates segregation, hierarchy, and a clash between parochial and cosmopolitan values. The aim of either “camp” is to defeat the other, but the consequence is to entrench both sides and erode common ground.

The emerging cleavage looks novel and intractable. Yet the coincidence of high uncertainty and growing inequality that we now associate with the ICT revolution is strikingly similar to the situation that gave rise to the modern welfare state in the early twentieth century. The economic driver of change then was the industrial revolution, causing major disruptions to traditional society, a transition from agriculture to industry, and rising inequality (as summarized in the Kuznets curve). It, too, created a conservative backlash, and arguably fueled fascism in parts of Europe (Inglehart, 1997, 90, 168–72), but many reformers at the time contended that the solution to profound uncertainty was not a return to the past status hierarchies but rather a comprehensive system of social insurance to accommodate and compensate for change while reducing status differentials (Baldwin, 1990; Dryzek & Goodin, 1986; Heclo, 1974). The normative principle that justified this solution, which Friedman (2020) calls probabilistic justice, finds application not only in social insurance narrowly construed as compensation for unemployment, illness, and old age, but also in the form of intergenerational mobility and opportunity, which equalize expectations over the long term (Iversen & Soskice, 2019). Social insurance is also a means to cope with international economic integration: Instead of closing off the economy to external disruption by erecting barriers to trade and capital mobility, countries can embrace an open international economy and the prosperity it promises by compensating the losers, an approach Ruggie (1982) called embedded liberalism.

This paper offers a theoretical framework for analyzing the contemporary political moment and the choices with which advanced democracies are now faced. In so doing, it makes a two-fold contribution to the recent literature on the political effects of technological and economic change, including the rise of populism and the future of the welfare state (Bonomi et al., 2021; Gidron & Hall, 2017; Margalit, 2019; Noury & Roland, 2020; Rathgeb & Busemeyer, 2022). The first is to show how the coincidence of high inequality and high uncertainty that characterizes the present moment can give rise to two very different political outcomes, one guided by a posture of status defense and the other by the principle of probabilistic justice. The former seeks to reduce uncertainty by reasserting barriers between groups and cultivating distinct in-group identities. The latter embraces uncertainty and encourages inter-group mobility based on broad notions of solidarity or shared fate. While there is currently profound pessimism about the future of western democracy, we argue that such pessimism is justified neither in the history of ideas nor in social history. The rise of the politics of status defense is real, but it is not new, and our political discourse and practice contain a powerful resource with which to challenge it, namely the principle of probabilistic justice.

The second, closely related contribution is to elucidate the normative foundations of the choice facing western democracies at this critical juncture, and to show how they are tied to broader conditions of social and economic change. Existing scholarship bifurcates into positive and normative perspectives: work in the positive tradition acknowledges that democratic politics is at a crossroad yet has no conception of the normative structure of the choices confronting us; work in political theory elucidates normative principles yet often lacks close engagement with the political cleavages of contemporary capitalism. While our aim is not to provide a thorough empirical defense of our claims here, we offer a descriptive typology of outcomes based on empirical observation (Gerring, 2012) and incorporate this set of inputs into a normative argument about our current impasse and possible avenues out of it (Ackerly et al., 2021; Schmidtz, 2017).

Uncertainty, status, and the two equilibria

The central claim of this paper is that uncertainty and economic stratification work together to either enhance or undermine the demand for solidaristic social insurance. Uncertainty refers here to the degree of destabilizing change taking place in a society, such as characterizes periods of rapid technological and economic transition. In general, uncertainty makes it difficult for individuals to accurately assess their own risks and to cope with those risks unaided. For example, under conditions of rapid technological change, job security declines and it is harder for individuals to predict whether their educational investments will lead to stable employment (Hacker, 2006).Footnote 1 Uncertainty thus describes both an objective set of circumstances and its subjective significance for decision-makers, a linkage that is justified by a long tradition in probability theory and economics of tying beliefs in situations of choice to observed states of the world (Arrow, 1958; Harsanyi, 1983).Footnote 2

Economic stratification refers here to the degree of material inequality in a society and the implied differentiation into distinct social strata. It refers not just to income levels but to life prospects more broadly, including access to economically well-connected social networks, good educational institutions, and economic opportunity. As with uncertainty, we argue that the objective aspects of inequality are closely linked to subjective constructions: Individuals in highly unequal societies who also lack opportunities for mobility will identify primarily with those in their narrow in-group rather than with the political community as a whole. This claim receives support from research in social psychology, which finds that when people see their society as rigidly stratified, they are more likely to relate to one another in terms of their social group and to pursue self-esteem and status through such affiliation (Tajfel & Turner, 1979; Wilkinson & Pickett, 2009). In our account, groups provide not only status, but also signals with which individuals estimate their risks and seek related protection. Correspondingly, we conjecture that in situations of high status differentiation and low uncertainty, when groups are firmly entrenched and their risks are relatively predictable, citizens will rely on narrow in-group affiliations as sources of both information and security. At the other extreme, if relative socioeconomic equality coincides with high uncertainty, then individuals will be more likely to seek social protection based on broad solidarities, because both their social identities and their sources of information will be less rigidly defined.

By contrast, the outcome generated by a scenario of high uncertainty and high inequality is undetermined, because while solidaristic social insurance is a solution to uncertainty, high economic stratification may lead citizens to seek to entrench inequality by reducing mobility, and thus preserve their status. In short, people can either embrace uncertainty and compensate for it through social insurance, or they can try to reduce uncertainty by erecting barriers around their in-group and seeking to raise the relative status of that group, usually by distancing it from those below it. The latter tendency, we submit, is an important factor behind the recent rise of populism. While much has been written about this phenomenon in recent years, the analysis proposed here has an advantage over two other explanations that are prominent in the literature: those that stress the deteriorating economic situation of rightwing populist voters (Swank & Betz, 2003; Mewes & Man, 2012) and those that emphasize psychological factors such as identity, resentment, and the desire for recognition (Inglehart & Norris, 2016; Goodheart 2017; Fukuyama, 2018). The first group tends to overlook expressions of status defense among the “winners” in the new global economy, which we argue is a condition for the rise of populism among the excluded. The second group tends to downplay the centrality of material interests in leading to divisive politics at both ends of the socioeconomic spectrum. By contrast, the mechanism that we identify highlights the interaction between economic interests and the retreat to insular identities. The two are closely related because when stratification is high and uncertainty low, social networks and interactions become increasingly confined to the in-group, while affinity with the outgroup declines because the prospect of both upward and downward mobility declines.

Our conception is influenced by the seminal article by Gidron and Hall (2017), which shows that status concerns rooted in economic interests can animate working-class support for the populist right. They subsequently (2020) extended this argument to explain support for both radical right and radical left parties as a product of feelings of social marginalization. We agree regarding the salience of both economic and psychological factors, but we depart from their account in two important ways. First, whereas they emphasize insecurity in the old middle classes as the key driver of status defense, as do Fukuyama (2018) and Iversen and Soskice (2019, ch. 5), we understand the issue more broadly to include the well-off, well-educated middle classes in successful cities, who may engage in less reactive but no less consequential forms of status defense. Second, whereas they stress economic decline as the source of status anxiety and populist support, we argue that the threat of such decline can also explain support for class compromise and the solidaristic welfare state, via the principle of probabilistic justice. Because of this ambiguity, normative analysis and argumentation have a particularly important role to play in the eventual outcome. Our argument thus highlights the mutually reinforcing interaction between normative frames and socioeconomic conditions.

For the same reason, our account also addresses a longstanding question in the political economy model of social insurance. When inequality is high, people seek policies that will protect their position by reducing the risk of downward mobility. But when they experience a large decrease in their status, they are likely to want insurance. Standard models do not tell us when the insurance motive prevails over the status defense motive, or when on the contrary insurance is insufficient to protect against the threat of downward mobility. By considering the levels of both inequality and uncertainty, our theory reveals that at certain pivotal moments, the equilibrium is undetermined and therefore particularly ripe for normative intervention and political choice.

The argument proceeds as follows. We begin by outlining the principle of probabilistic justice and its expression in the solidaristic welfare state. In the remainder of the paper, we contrast this outcome with three other scenarios, each entailing a different combination of uncertainty and status differentiation. The first, which we denote the status defense equilibrium, refers to a state of high economic stratification and low uncertainty, in which social groups are clearly economically ranked and there is little room for socioeconomic mobility. We relate these two outcomes to the debate inspired by Robert Putnam’s concept of social capital, suggesting that probabilistic justice correlates with bridging capital, which crosses socio-economic and other divides, while status defense leads primarily to bonding capital. In this way, we argue, material conditions interact with perceptions of status and identity to shape how individuals define their communities of fate. The logic of our approach echoes a new political economy literature on culture, which argues for the possibility of multiple, usually two, equilibria of policies and dominant values (Bisin & Verdier, 2000; Tabellini, 2008; Besley & Persson, 2019). It is also consistent with Cathie Jo Martin’s new study of the divergent cultural trajectories of Britain and Denmark, which shows the deliberate use of cultural frames to create more or less inclusive public school systems (Martin, 2023).

The next scenario that we discuss is a situation of high economic stratification and high uncertainty, which results in the new politics of status anxiety. This is a politically volatile situation, the resolution of which is indeterminate and contested. The industrial revolution offers a historical precursor, which was eventually resolved in a solidaristic direction through the expansion of the welfare state. But such an outcome is not inevitable and is typically countered by a more reactionary alternative. In the current juncture, we identify two expressions of this alternative, one “defensive,” reflected in the turn to family, community, and nativism that characterizes right-wing populism; the other “assertive,” expressed in the various forms of local control that allow the economically successful to exclude outsiders from their neighborhoods and social networks. Despite their different political expressions, the two tend to reinforce each other and contrast with the principle of probabilistic justice. The final scenario, which we call contested solidarism, describes a situation of reduced uncertainty that may be followed by a period of growing economic stratification.

The key contributions of this analysis lie, first, in explaining the sources of our current malaise; and second, in mapping the normative foundations for the main political responses. By removing barriers to mobility from below and emphasizing shared perceptions of risk across the political community, it may be possible to imagine a return path to the solidaristic welfare state while also addressing the demand for security that underlies status defense. Our analysis thus highlights the dynamic interplay between material conditions and normative frames. Insofar as we find ourselves at a critical juncture, the principle that we choose to adopt and the policies enacted in its wake will either reduce inequality and pave the way to broad-based solidarities, or perpetuate a state of inequality, immobility, and status defense.

A. The principle of probabilistic justice

Social insurance has long been a defining policy of the advanced welfare state, and for good reason: It is an unusual case of political philosophy, mathematics, and social science converging on a common understanding with both explanatory and normative power. The logic supporting social insurance holds that people who are in a “good state”—healthy, employed, earning a reasonable wage, etc.—will seek protection against the risk of falling into a “bad” state such as illness, unemployment, or plain bad luck. Economic models typically assume that individuals have concave utility functions, which implies risk aversion and a demand for insurance to smooth income across “good” and “bad” states. Combine this insight with the common assumption that markets for private insurance often fail due to incomplete information and adverse selection, and much of the postwar welfare state can be attributed to popular demand for insurance, as has indeed been argued by some of the most prominent scholars of social policy history (Baldwin, 1990; Barr, 2001; Esping-Andersen, 1990; Heclo, 1974). In addition, social insurance may promote economic prosperity by allowing countries to benefit from trade while compensating losers (Katzenstein, 1985; Rodrik, 1998; Ruggie, 1982); by encouraging economic risk-taking (Sinn, 1998); by incentivizing workers to invest in specific skillsets (Estevez-Abe et al., 2001; Iversen & Soskice, 2019); and by promoting flexible labor markets by insuring against unemployment instead of protecting existing jobs, a policy approach known as flexicurity (de la Porte & Jacobsson, 2012). According to the OECD’s measure of net social spending, publicly provided social insurance accounts for between one quarter and one third of GDP in advanced democracies (Elkjaer & Iversen, 2022).

The principle of probabilistic justice (Friedman, 2020) provides a central normative justification for social insurance. This principle states that the idea of equality to which distributive claims should correspond is equality in the face of chance, expressed mathematically as a probability value or expectation. Those who are equally likely to encounter a given eventuality, and contribute appropriately to the shared resource pool, should receive an equal distribution of resources if and when that eventuality comes to pass. Probabilistic justice is similar to what others have referred to as “chance solidarity,” or the sharing of responsibility for misfortune that characterizes all insurance pools (Lehtonen & Liukko, 2015; Thiery & Van Schoubroeck, 2006). Unlike chance solidarity, however, probabilistic justice allows for differing assessments of risk and therefore raises important puzzles about whom to include in the shared risk pool and on what terms.

The intellectual and political history of the welfare state shows that this principle was highly influential in the development of mutual and then social insurance (Ewald, 1986; Jones 2005). At its core, it represents an understanding of shared fate among those who pledge to reciprocally protect one another against adverse outcomes. At the same time, because it rests on assessments of probability, it has an expansive and unifying potential that more traditional forms of identity—such as those derived from ethnicity, class, or religion—do not.

The solidaristic welfare states that proliferated in the period following World War II expressed a particular interpretation of probabilistic justice, which sought to extend the perception of equal vulnerability beyond limited pools of risk-prone workers to the polity as a whole, providing comprehensive social protection from birth through old age (Baldwin, 1990; Beveridge, 1942). At the root of probabilistic justice thus understood is a fundamental equality in the human condition: Everyone can fall ill; everyone needs care when old; and everyone is susceptible to losing their income or earning capacity, or falling into poverty should truly bad luck strike. Therefore, as we employ the term here, it rests on a situation of socioeconomic equality and high uncertainty, which encourages broad risk pooling based on hazards widely perceived as common. Such insurance aspires to align prudential calculations of self-interest and personal responsibility, which are closely linked to traditional liberal ideals, with more collectivist notions of our responsibility to one another as members of a larger community (Ewald 1986) . While elements of status preservation do persist in some modern social insurance schemes, as discussed below, on the whole the post-war solidaristic welfare state entailed a significant expansion relative to previous programs, aimed at providing both equal opportunity and security for all (Dryzek & Goodin, 1986; Goodin, 2003). In this respect, it exemplifies our understanding of probabilistic justice as an ideal type.

  • i. Individual support for social insurance: two approaches.

There is a puzzle here, however, which any normative account of insurance must address. For the individual, the rationale for participating in a mutual insurance arrangement, as with any contract involving an uncertain outcome, rests on the expected utility of the wager, which reflects both the probability and amount of the relevant loss (Bernoulli, 1954). In a given insurance pool, expected contributions and expected payouts (plus operating costs and normal profits) equal one another. Since people are risk-averse, they enjoy a net gain in utility, or “consumer surplus,” even if their payments exceed their strict mathematical expectations. In this very fundamental sense, social insurance is welfare-improving and hence efficient.Footnote 3

The puzzle arises insofar as the relevant probabilities are based on the observed frequency of an event within a given reference class. Belonging to a group whose members face similar exposure provides a reason for individuals to support a mutual insurance scheme that will protect them against the risk on more or less equal terms. Yet observed frequencies, while valuable for the statistician who has the convenience of working from a given group, are of little use to individuals seeking to determine which group they belong to and how much insurance to obtain (Hájek, 2007). This is true even if people are told that insurance must be provided through a single national scheme, since they may not support such a scheme if they believe their risk is much lower than the national average, and they will in any event be left with the question of how much insurance they need.

In theorizing the social decision procedure, there are two plausible responses to this quandary. For our purposes, the best known is the Rawlsian social contract, in which distributive principles are decided behind a “veil of ignorance” that denies individuals all knowledge of probabilities pertaining to their future lives (Rawls, 1971, 154–55, 172; 1974). In decision theory, this situation is known as decision-making under uncertainty, and it is not clear how people make choices in this case. Rawls invokes the so-called maximin criterion, according to which individuals maximize their worst-possible outcome in the different feasible states of the world. Although this account does not tell us what level of insurance individuals would choose, and although Rawls denies that participants are inherently risk-averse, there is an insurance-like character to the principles of justice that result, insofar as parties choose to protect themselves against the worst-possible eventualities (Rawls, 1971, 176; Hare, 1975). Rawls’s principles of justice thus provide a rationale for redistribution that does not depend on individual estimates of risk. But this very achievement is arguably also a weakness of his theory, since by dispensing with probabilities in the social decision procedure, it precludes the empirically grounded solidarities of self-identified risk groups that have historically been a major driver of welfare policy (Baldwin, 1990; Esping-Andersen, 1990; Huber & Stephens, 2001). Rawls was right to worry that any agreement made with knowledge of personal probabilities could reflect existing power differentials (Rawls, 1974). Nevertheless, by distancing his theory of justice from the lived experience of shared vulnerability, and in particular from the empirically-based perceptions of common hazards that lead citizens to demand mutual protection, he also ruled out some of the most significant sources of solidaristic social policy.

This reading of Rawls is similar in spirit to that of commentators in the political “realist” tradition, broadly defined, who argue that in treating political philosophy as a form of applied moral theory, Rawls downplayed the core problems of politics as such (Larmore, 2018; Williams, 2005). Our critique echoes this one in suggesting that Rawls’s theory is insufficiently attuned to the political conditions of social solidarity, including the need for a legitimation story that appeals to actual citizens of welfare states (Hall 2015). Insofar as longstanding arguments for social insurance invoke probabilistic reasoning of one kind or another, a theory that rules out such reasoning will have limited normative force in contemporary political life.

A second approach to the quandary of individual support for social insurance is Harsanyi’s “objectivist” interpretation of Bayesian probability theory, which links directly to modern theories of belief formation (Harsanyi, 1983). In Harsanyi’s conceptualization, people base their initial expectations on the best information available (their “priors”) and then update this information as they acquire new knowledge (“signals”) through their daily encounters with the real world. Harsanyi argued that the prior would ordinarily be the population mean, which can be observed fairly easily as the national unemployment rate, the share of people who reach retirement age, the share who fall ill, and so on. Over time, people may acquire more precise information that is more relevant to their own situation through their workplaces, social networks, or membership in organizations, such as unions, that are charged with protecting their economic interests.Footnote 4 This approach is similar to that articulated by Dworkin (2000), with the crucial difference that it allows for an intertemporal dimension in the updating of priors. As a result, it offers a more realistic account of individual reasoning in the face of chance. It also forces us to confront the political challenge of sustaining social insurance in a way that Dworkin’s theory, like Rawls’s, does not.

  • ii. Prior probabilities, path dependency, and solidarity.

How does the second approach relate to the principle of probabilistic justice? It clearly captures the prudential aspect of that principle: Individuals support social insurance because they see it as advancing their own interests. Yet it does not fully capture the dynamic that allows such interests to align with equality and the idea that we are responsible to each other as members of a larger community.

A crucial insight of a large body of research in political science on “path-dependency” and “policy feedback” is that the initial structure of social protection shapes social policy preferences in ways that reinforce that structure (Campbell, 2003; Esping-Andersen, 1990; Pierson, 2000). In our model, we can conceive of such feedback effects in terms of Harsanyi’s priors and the logic of Bayesian updating. If the pool is initially broadly defined and people’s priors conform to a common risk, then the path of subsequent updating will be different than if the pool is narrowly defined with risks differentiated across multiple pools.

Consider two scenarios that both start from the same underlying (but unobserved) distribution of risk. The first is that people conceive of the risk pool broadly and adopt similar priors, which produce redistribution in the sense that those at (objectively) higher risk will draw more on the common insurance than those with lower risk. This means that in the face of adverse life events, the unfortunate will be better able to maintain their current lifestyle, stay in their current neighborhood, send their children to the same schools, and access the same social networks and labor markets as before. This will preserve opportunities for current and future generations and thereby help maintain a more compressed distribution of risks, especially when conceived broadly to include intergenerational mobility. In turn, when people update their risks, the distribution of these will not deviate too much from the common prior, in turn entrenching support for broad social insurance. This line of argument has been used to explain how it is possible to sustain a universal welfare state (Korpi & Palme, 1998; Rothstein, 1998).

In the second scenario, people link their priors to observable differences in income, social status, and other group markers. The insurance system will then tend to reinforce differences by reducing risk pooling and making it less likely that people will mix across class and group boundaries, leading to a self-fulfilling confirmation of the divergent priors. As we discuss further on, this may be exacerbated by deliberate actions to cut off upward mobility from below, what we call localism. The consequence is that existing inequalities become amplified thanks to lower mobility and a further segmentation of risks.

The second scenario corresponds to what we have called the principle of status defense, and we discuss it in more detail in the next section. The principle of probabilistic justice corresponds to the first scenario. What gives this normative principle its valence is that it accords with ethical intuitions about our common vulnerability as human beings and thus supports a vision of equality across otherwise diverse groups (Freeden, 2003). What makes it politically realistic is that it appeals to personal interest and can be sustained by prudential calculation, and furthermore taps into a longstanding narrative in liberal democracies about the legitimacy and desirability of welfare state institutions (Dryzek & Goodin, 1986). Finally, as our analysis shows, what renders it vital for liberal democratic politics is the ambiguity about which path is chosen, meaning that the solidaristic outcome depends on our political choices.

Our two scenarios are meant to elucidate mechanisms that can lead to the different outcomes in Table 1. Yet status defense and probabilistic justice are never pure states of the world. In Esping-Andersen’s (1990) influential typology of welfare states, for example, social insurance in what he calls the “Conservative” or “Christian Democratic” type is tied to broad occupational groups, which has a status-preserving effect. In fact, however, because Christian democratic welfare states support “generous safety nets for people in all income groups,” Bradley et al. (2003) finds that they are far more redistributive than Liberal welfare states such as the US (p. 223). As our later discussion of mutual aid societies underscores, the insurance principle can be applied to more or less broadly defined risk pools. Broadly defined pools approximate our “probabilistic justice” ideal type; narrowly defined ones the “status defense” ideal type. Despite institutional feedback mechanisms, these boundaries are always contested and subject to redefinition, as implied by Mahoney and Thelen’s (2010) influential theory of institutional change.

Table 1 Status, uncertainty, and the politics of social protection
Full size table

The probabilistic justice path appears most likely under two conditions: (i) when people are in fact fairly similar in terms of socioeconomic position, including their access to opportunities and their ability to cope with risk; and (ii) when uncertainty is fairly high, so that it is difficult for individuals to distinguish their own situation from that of others. The first case corresponds to a state in which the signals people receive tend to be fairly similar to each other. The second case corresponds to a situation involving imprecise signals, which force people to base their judgments more on broad averages (the “priors”) than on individualized probability estimates. This combination of conditions is captured by the lower left-hand cell in Table 1.

B. The principle of status defense

Today, of course, most wealthy democracies are far from reflecting this ideal type. In several countries, particularly the United States, unequal housing and education policies and the absence of a robust safety net have combined with rising inequality to create sharp and stable boundaries between communities and social classes. Such status differentiation makes it less likely for people to identify their personal probabilities with the national average and support broad-based social insurance. If uncertainty is low and status differentiation high (the top right-hand cell in Table 1), visible inequalities will lead people to adopt very different estimates of their own life prospects. As a result, if social groups remain closed off from one another, the demand for risk-sharing arrangements across groups will be low.

In the ideal typical representation of this scenario, people live in segregated neighborhoods, send their kids to different schools, and intermingle mainly with “their own” in separate social networks. As a result, they find it difficult to imagine themselves as part of a wider community, let alone an encompassing national risk pool. Even if Harsanyi’s shared prior were the starting point for most people in this setting, their divergent lives would produce vastly different estimates of their risks, as well as their possibilities for upward mobility, over time, reflecting the experience of prosperity and widespread opportunity among some, and the experience of hardship and stagnation among others. This is a situation of extreme socioeconomic sorting and limited mobility, which undermines support for national-level social insurance while facilitating the ability of private insurers to offer differentiated products based on segmented risk pools. This phenomenon will in turn increase inequality as those at higher risk are charged higher rates or left under- or uninsured (Iversen & Rehm, 2022).

If flexicurity is a natural extension of the principle of probabilistic justice, we may say—a little tongue-in-cheek—that high status segmentation is a system of inflexisecurity. Just like the typical voter in a low-differentiation regime, the typical middle-class voter in a highly differentiated society seeks guarantees of material security. Yet in the latter case, these guarantees take the form of policies that defend existing class standings in a hierarchical system. Such policies erect barriers to mobility from below, making it harder for outsiders to enter better neighborhoods, to send their kids to better schools, and to join more empowering social networks. This reinforces status differentials even as it simultaneously reassures the stability of the position of each class.

To better elucidate the micro-logic of this phenomenon, it is useful to distinguish between the lower and upper middle classes. Both have reasons to support policies that open opportunities for socio-economic advancement among their own, but in highly inegalitarian settings this may only be realistic for those at the upper end. A main driver is access to good school districts, which offer children a path to higher education and hence careers in the expanding knowledge economy. This in turn drives up the costs of housing and shuts out lower-income households. The hyper-meritocracy of American elite education, in which access to top universities is highly rationed, perpetuates status differentials through prohibitive costs and skewed admissions criteria that favor families able to afford private tutors and career coaches. The dynamic is particularly pronounced in a system like the American one, where school financing is mostly local, but it is increasingly also observed in more centralized public systems in which the highly-educated self-sort into specific districts and create barriers to entry through higher housing costs and restrictive zoning policies (Gingrich & Ansell, 2014). For the middle and upper middle classes, who can afford the price of admission, the system is consistent with a belief in meritocracy, but de facto it restricts mobility and often leads to opportunity hoarding as well as growing segmentation in insurance and credit markets.

For the lower middle classes, where the lack of upward mobility shuts down aspirations for a better life, status defense is likely to take on more overt forms of discrimination, including the cultivation of identities that underscore the superiority of the ingroup (Lamont & Molnár, 2002; Tajfel & Turner, 1979). An extreme example is the American South during the era of Jim Crow, but the practice of “redlining,” whereby people of color are unable to purchase homes in white middle-class neighborhoods because of discriminatory mortgage-lending and related practices, persisted long after the end of formal segregation (Aaronson et al. 2021).

Support for such discriminatory defenses of ingroup status are likely to be more common in the middle of the distribution, where resources and opportunities are limited yet mobility from below is seen as a threat. Häusermann et al. (2021), for example, find that middle-class voters in Western European countries who feel shut out of opportunities for upward mobility are much more likely to support right populist parties than middle-class voters who perceive good opportunities for upward mobility. A similar logic can sometimes inform political action even at the socioeconomic apex. For example, in a recent book on India, Suryanarayan (2021) argues that the upper caste of Brahmans used their dominant position in the state bureaucracy over tax collection and the education system to block lower castes from expanding primary education and integrating schools and neighborhoods. They did so by undermining the collection of land taxes, even though Brahmans were rarely large landowners, which shows that their purpose was not to enhance Brahman wealth, but rather to defend the status of their caste by “hollowing out” the fiscal state and preempting the expansion of public goods. While such state capture is not feasible in advanced democracies, this is an extreme example that prejudice and discrimination against outsiders is not just a working-class phenomenon.

A key mechanism in all of these cases is sorting on the basis of socioeconomic status, in particular education and income, which for the well-off in advanced democracies is often accomplished by zoning, building codes, and other local regulations favoring insiders. When inequality is high, sorting increases, and this in turn bifurcates opportunities for acquiring a good education, buying homes in desirable neighborhoods, and becoming members of social networks with well-connected people. Economists have correspondingly found a strong negative relationship between inequality and intergenerational mobility, known as the Great Gatsby Curve. It is this combination of high status differentiation and low mobility that turns distributive politics into a game of status preservation rather than encouraging support for broad-based social insurance.

C. Probabilistic justice and status defense through the lens of social capital

The distinction between probabilistic justice and status defense relates to an older debate in political science inspired by Robert Putnam’s concept of social capital, which helps to elucidate the logic of our argument. Building on de Tocqueville, Putnam (1992) proposed that widespread participation in dense networks of civic engagement, from neighborhood associations to choral societies, taught people mutual respect and a sense of community, which in turn translated into a “generalized trust” that contributed to good governance, high investment in public goods, and a well-functioning system of social insurance.

Sheri Berman (1997) challenged Putnam’s account with reference to the pervasive participation of middle-class Germans in civic associations in the late 19th and early twentieth centuries, many of which proved fertile recruiting grounds for the NSDAP in the fateful breakdown of the Weimar Republic (see Satyanath et al., 2017 for supporting evidence). In that case, civic associations were a source of differentiation and exclusion rather than broad-based trust and solidarity (see also Levi, 1996). In later work, Putnam acknowledged the issue by drawing a distinction between “bonding” and “bridging” capital (Putnam, 2000). Bonding social capital reflects close-knit and largely homogenous groups that develop highly trusting relationships among themselves but view outsiders with suspicion, disdain, and sometimes outright hostility. Bridging social capital, by contrast, is drawn from the original concept of generalized trust and refers to connections made across socioeconomic and related divides.

We see a parallel between Putnam’s two varieties of social capital and our own distinction between low and high status differentiation. Conditions of high inequality and low mobility will lead groups to develop mostly bonding capital, while states of low inequality and high mobility will be more conducive to bridging capital. Objective economic conditions and the subjective sense of status are thus closely connected, shaping who individuals perceive as their equals and how they define their communities of fate.

The rise of the twentieth century welfare state, from this perspective, was mostly an expression of bridging capital, with the associated demand for shared public goods and social insurance. As Polanyi, Esping-Andersen, and other prominent scholars of European economic and social history have argued, the largescale expansion of capitalist markets produced massive demand for social insurance, or what Polanyi (1944) called a “counter-movement” and Esping-Andersen “decommodification.” Starting in the late-nineteenth century, insecure groups increasingly turned to the state to provide accident compensation, pensions, and other policies that would protect them against economic risks (Baldwin, 1990). These programs laid the political and institutional foundations for the emergence of the solidaristic welfare state after World War II, reflecting the initiatives of cross-class coalitions of citizens who experienced similar vulnerabilities and were willing to share the burdens of mutual support.

Before these developments, there were attempts at “private” social insurance through mutual aid societies (MAS). Although they often advocated lofty ideals of solidarity and equality, MAS quickly faced a double bind that ultimately doomed them. Seeking inclusiveness, they attracted sicker, older, and more vulnerable workers, which in turn triggered an exodus of younger, healthier, and more employable members, either into private plans or into more upscale MAS. Most MAS therefore ended up as more or less exclusive clubs, often organized around ethnicity, religion, residency, or occupation, and in nearly all cases excluding unskilled laborers, the infirm, and those unable to qualify for benefits through long qualifying periods of contributions. Such restrictions reduced the relevance of MAS as urbanization, industrialization, and democratization multiplied risks and demands for more generous and encompassing social insurance (de Swaan, 1988).

Historians paint a complex picture regarding whether MAS and other community-based groups promoted bonding or bridging social capital (Cordery, 2003; Ismay, 2018). Yet insofar as the welfare state is associated with generalized trust rather than localist attachments (Rothstein & Stolle, 2003), it differs significantly from the original MAS model. Conservatives who look nostalgically to MAS as expressions of fraternal solidarity (Murray, 2013) are therefore correct that such institutions and the modern welfare state are based on distinct and even competing normative logics.

This discussion also relates to a new literature in political economy on the role of civic culture for good governance. Work by Bisin and Verdier (2000), Tabellini (2008), and Besley and Persson (2019) show formally that is possible to have self-reinforcing combinations of values and government policies that are either supportive of a “generalized morality,” according to which most people behave cooperatively towards others, or a “limited morality” that reserves cooperation for ingroup members, whether the family, clan, ethnic group, or local community. Banfield’s (1958) concept of “immoral familism” is an extreme example of the latter, in which insular culture (here in rural southern Italy) goes hand in hand with fear, disdain, and other hostile emotions toward outsiders. In our context, public policies also matter because targeting, restrictive local zoning regulations, and symbolic allegiance to exclusive identities reinforce reliance on the in-group, whereas policies that promote the public good, mobility, and spatial integration increase the payoff from holding cooperative, civic values. Borrowing a term from evolutionary anthropology, good policies increase the “relative cultural fitness” of holding civic values, which in turn increases the effectiveness of good policies. We highlight this idea in the economics of culture because we, too, emphasize the possibility of mutually reinforcing feedback loops between values or attitudes on one hand, including psychological states, social identities, and moral commitments, and public policies on the other.

Our distinction between probabilistic justice and status defense resonates with the abstract economic notion of cultural equilibria. Yet while the new political economy of culture offers convincing accounts of how distinct value orientations can be sustained over time, it has nothing to say about how different equilibria emerge or change, and in particular the role of normative principles at these critical junctures. This omission is highlighted in Persson and Tabellini’s (2021) authoritative review of the new economics of culture: “normative questions are absolutely vital.” Like other political economists, however, they leave them unanswered.

D. The new politics of status anxiety

We have argued that it is possible to envision two relatively stable political and economic ideal types. The first operates under conditions of relatively low inequality and high uncertainty, and tends to reproduce these conditions through generous social insurance and class mobility. The second arises under conditions of high inequality and low uncertainty and tends to reinforce high status differentiation because, in the absence of opportunities to improve their life prospects, members of segregated groups seek to protect what they have against competition from outsiders.

Both of these ideal types have been challenged by major structural shocks, notably the industrial and ICT revolutions. Such shocks create the conditions for high uncertainly and high inequality (captured by the bottom right-hand cell in Table 1). This is the situation of many affluent democracies today, where a confluence of technological, demographic, political, and other factors has led to a sense of extreme vulnerability for many. At the same time, disparities between “winners” and “losers” in the new global economy have become increasingly pronounced. The combination of rapid change and growing inequality has given rise to what we call the new politics of status anxiety. This unstable situation is perhaps best characterized as a critical juncture, and we believe it may be resolved in the direction of either the social insurance state (probabilistic justice in Table 1) or more extensive and permanent status differentiation (status defense in Table 1). The outcome will depend not only on material factors but also on normative commitments and on the dispositions, identities, and communities of fate those commitments encourage or undermine.

The new politics of status anxiety has two major expressions, one on the populist right and another among the affluent elite. We refer to these as defensive and assertive localism, respectively. While their policy expressions differ radically, both reflect a state of affairs in which life chances are seen as highly unequal yet uncertainty generates powerful fears about the prospect of downward mobility. The higher the level of differentiation and the fewer the opportunities to move up the socioeconomic ladder, the more individuals will have to fear from a loss of status.

At the same time, it is important to understand that this retreat to localism is not inevitable, since the rise in inequality may also be a reason to strengthen social protections and reassert a politics of opportunity. The key is that high levels of uncertainty and inequality can induce either a reactionary defense of privilege or a progressive reassertion of solidarity and shared fate.

We begin with the right-wing expression of status anxiety, which represents a new variation on the theme. The combination of high uncertainty and high inequality has led many voters to support policies they believe will protect their status against the instabilities caused by globalization, population migration, and technological change. Such policies set out to bolster national state sovereignty, limit immigration, and restrict welfare benefits to “insiders,” to the exclusion of ethnic and cultural minorities (Burgoon et al., 2019; Gidron & Hall, 2017; Ketola & Nordensvard, 2018; Pettigrew, 2017; Sainsbury, 2012; Schain et al., 2002). In keeping with the predictions above, then, status anxiety driven by high uncertainty and barriers to mobility—and in particular, low opportunity for economic betterment among the lower middle classes—will tend to generate support for populist platforms that promise voters a form of insider protection in the face of destabilizing change. If upward mobility is blocked, the logic goes, one should seek to at least prevent downward mobility by undercutting the opportunities of those below. Such socioeconomic encapsulation encourages group members to look inward and to seek status affirmation within their community.

In the case of right-wing populism, then, our claim is that key aspects of status defense can be explained in terms of a desire for a restored sense of protection and self-determination among those exposed to the costs of economic change but largely excluded from its benefits (Fukuyama, 2018; Gidron & Hall, 2017). These voters are drawn to a model of politics that cultivates bonding capital, including a turn inward to the national community and an appeal to traditional hierarchies to shore up the sources of security and self-respect. For example, Diana Mutz has found that a perceived threat to social status played a salient role in motivating less-educated voters who supported Donald Trump in the 2016 US presidential election (Mutz, 2018). Crucially, however, our explanation does not prioritize identity over economic interests, but rather indicates that the two are closely linked. Right populism in both the US and Europe, for example, has been linked the displacements of manufacturing employment caused by the so-called “China shock” (Autor et al., 2013; Colantone & Stanig, 2018), but in reviewing the evidence, Rodrik (2018) concludes that the political effects are strongly mediated by inward-looking, nativist cultural frames.

We call this phenomenon, and the disposition that motivates it, the politics of defensive localism. It can be seen to emanate from elements of the conservative political tradition, although in themselves these elements need not produce pathological outcomes. Contemporary conservatives frequently invoke Edmund Burke’s view that social attachments begin in the “subdivision” or “little platoon,” from which “we proceed toward a love to our country and to mankind” (Burke, 1987, 41; Miller, 2017). Such “subordinate partiality” is “a sort of elemental training to those higher and more large regards” that attach citizens to the nation as a whole (Burke, 1987, 173). In the twentieth century, this idea was given a new valence in Russell Kirk’s The Conservative Mind and Robert Nisbet’s The Quest for Community, which held that local, particular attachments are the source of identity and distinction, preventing individuals from getting lost within a homogeneous mass (Kirk, 1953; Nisbet, 1953). Notwithstanding important differences between Burke’s meaning and that of later conservative thinkers, this disposition on the whole represents an appeal to older, more traditional forms of association in times of rapid change and disruption (Woods, 1999).

The politics of defensive localism thus looks to exclusive attachments as the sources of security, belonging, and meaning (Goodhart, 2017, 2020). Moreover, insofar it seeks to counteract the atomizing and homogenizing tendencies of the globalist order, this phenomenon is anti-liberal or “postliberal” in its normative orientation (Deneen, 2018). This feature, in turn, relates to its more pathological manifestations, in particular the right-wing populism that has become increasingly prominent over the past several decades. Francis Fukuyama has argued that populist voters are motivated more by feelings of resentment over their loss of dignity and recognition than by economic concerns (Fukuyama, 2018). While it is clearly true that status anxiety reflects psychological factors that cannot be reduced to material interests, we agree with Gidron and Hall (2017) that what matters is the interaction between these explanations. By understanding the new status anxiety as seeking a renewed self-determination and sense of belonging in the face of uncertainty and entrenched inequality, both results of the transition to a new knowledge economy, we offer a fuller account of its genesis and character.

In addition to reflecting a demand for security and recognition, the appeal to particularistic attachments and bonding capital can be seen as a turn away from the methodological individualism of the inclusive welfare state (Deneen, 2018). Social insurance as a technology of governance tends to regard the insured in abstract terms, as equal and interchangeable (for the purposes of any given policy) with others in her risk category (Ewald, 1986; Walters, 2000). This is not to say that all models of social insurance dis-embed individuals from their social contexts. Nevertheless, because the principle of probabilistic justice requires relations of equality among relatively diverse and, at times, geographically remote individuals, the type of solidarity that it entails is more diffuse and less personal than the “thick” ties of family, association, and community.

Localism is not the exclusive province of lower-income groups or right-wing nationalists, however. Wealthy elites exhibit some of the same fencing-off tendencies as supporters of right-populist movements. This is another major oversight in much of the current literature on populism. Among elites, these tendencies manifest less in seeking to exclude ethnic minorities and protect local jobs than in hoarding opportunities to live in desirable areas, attend prestigious schools, and enter powerful social networks (Hansen & Toft, 2021). Precisely because policies designed to strengthen the local community tend to make it more desirable and therefore more exclusive, local self-determination and democracy are complements to this impulse. Nevertheless, such “assertive localism,” as we call it, is no less toxic from the perspective of building inclusive national communities.

This second expression of status defense in contemporary politics thus emerges from the cosmopolitan outlook of the “creative classes” (Florida, 2002) in the successful cities. The well-educated and secure winners of the new economy tend to be welcoming to outsiders and to promote inclusive, democratic norms (Kitschelt, 1992). Indeed, this may be construed as the main cleavage of the knowledge economy: cosmopolitanism against parochialism. But we caution against such a view because it overlooks the extent to which upscale groups contribute to segmentation by engaging in assertive localism. NIMBY’ism is an obvious example, but a large recent literature on housing documents how local zoning rules, restrictive building codes, and environmental reviews, combined with market-driven sorting mechanisms, create exclusive enclaves of well-functioning local communities with good schools and participatory democracy (Einstein et al., 2019; Glaeser et al., 2005; Schuetz, 2022). These communities have prohibitive costs of entry because of high and rising housing prices, just as market-based segmentation of insurance and credit markets shut out lower-income, higher-risk types without the need for overtly discriminatory policies (Iversen & Rehm, 2022). Tolerance is cheap when the barrier to entry is structural, whereas the lower middle classes often rely on overt discrimination to defend their own status.

Our explanation for the contemporary politics of status anxiety thus stresses the interaction between material interests and psychological factors, or values broadly understood. Focusing on such interactions further underscores that the politics of localism is only one possible outcome of the current juncture. Existing literature, even when emphasizing both interests and values, has largely ignored the principle of probabilistic justice, which played a central role in the theory and development of the solidaristic welfare state. This work therefore neglects the motivational and even path-determining potential of normative principles in such an unsettled moment.

Probabilistic justice requires a readiness to imagine oneself as facing similar odds as others in the national community, whether in the context of particular risks or in terms of life prospects more broadly (Freeden, 2003). This dispositional stance toward uncertainty manifests in a willingness to cooperate reciprocally with diverse others in providing the means of security for all. It is very difficult to achieve inclusive solutions, however, if the educated middle classes weaponize local democracy to exclude others from their communities. Solidarity requires democracy to assert itself at higher levels of aggregation, at the state or national levels.

It is important to stress that while the dangers of rising inequality turning into a politics of status defense are real, and while the US has been moving dangerously towards such an outcome, this is not a common pattern across advanced democracies. Market inequality has increased everywhere and expressions of localism, notably the rise of right populism, have become more vocal. Yet governments in Western Europe have also significantly raised spending on public goods and social insurance. The after-tax and -transfer incomes of the bottom and middle thirds of the distribution have consequently largely kept up with overall economic growth (Elkjaer & Iversen, 2022). Support for the welfare state has been maintained at high levels even as it has become more contested (Gingrich & Häusermann, 2015; Häusermann et al., 2022). The logic of probabilistic justice, in other words, continues to play a major role in politics.

E. Contested solidarism

Before concluding, we briefly note a fourth logical outcome, in which status differentiation and uncertainty are both low (the top left-hand cell in Table 1). While it is true that low status differentiation means that the population is already relatively homogeneous socioeconomically, any reduction in uncertainty has the potential to lock in existing inequalities, particularly following a period of dynamic change. One could fear that this is another path to a more status-differentiated system (Iversen & Rehm, 2022). By contrast, the principle of probabilistic justice builds on a widespread sense of shared fate, which is rooted in uncertainty and continued social mobility.

Conclusion

The question of whether our current moment will lead the way to even greater status anxiety and political division or to the return of solidarity is a decisive one for the future of liberal democracies. We have argued that high levels of uncertainty and status differentiation encourage pathological forms of status defense, but that because of the indeterminacy in our current moment, the normative principle we choose can help shape material outcomes in a meaningful way. Indeed, the emergence of the modern welfare state offers an example of how conditions of high inequality and uncertainty can lead to an embrace of broad solidarities rather than a retreat to narrow ones. While contemporary debates have largely focused on the threat that populism poses to liberal democracy, we show that considering inequality and uncertainty together allows us to perceive the current moment as more malleable than these pessimistic accounts suggest.

This observation in turn relates to an ambiguity in the insurance model that our analysis helps to resolve. In situations of inequality, people may seek policies that will protect their position by reducing the risk of downward mobility and blocking upward movement from below. Yet when they do experience a decrease in status, they are likely to want more insurance, especially when the loss is large. It is not obvious, therefore, which principle will dominate in situations of high uncertainty. When inequality is low, status is less of a concern and social insurance may be enough to preserve standing after an adverse outcome. When inequality is high, however, it may be impossible to prevent a status loss through insurance alone, since the drop tends to be life-changing. Individuals’ focus may therefore turn to preventing such movement in the first place, even if they recognize that they will need insurance as well. Where one principle takes over from the other involves ambiguity and is therefore subject to ideational mobilization and political choice. This is precisely why it is so important to have a clear understanding of the normative foundations and significance of the different policy responses.

As we have argued, the solidaristic welfare state rests on a sense of shared fate across socioeconomic, geographical, religious, and other differences. Returning to Harsanyi’s approach, this means that individuals treat the population average as their own prior probability value and, despite subsequent updating, see themselves as more or less equally likely as everyone else to incur a particular fate. Interpreting this condition quite broadly, we might say that such solidarity is more likely when individuals regard themselves as justifiably sharing, on roughly equal terms, not only in the downsides of uncertainty and change, but also in their upsides. Iversen and Soskice (2019) refer to such individuals as “aspirational voters,” and there is evidence that they are much less likely to vote for right populist parties (see Häusermann et al., 2021). Equalizing opportunities for social integration and economic betterment may therefore be one way of transitioning from a situation of high uncertainty and status differentiation to one of probabilistic justice. If mobility is taken as a given, insurance models have the unambiguous implication that when inequality across states of the world rises, risk-averse individuals will want to increase spending on insurance. This response therefore emphasizes traditional redistributive social insurance policies, in addition to public investment in education and housing, which would help bridge the divisions that separate neighborhoods and social networks.

A complementary response is to cultivate bridging capital in the context of particular risks. This would call for a discursive framing that highlights voters’ shared vulnerabilities, an approach that is likely to be particularly fruitful with regard to intergenerational risks. In the case of defensive localists, an emphasis on extending probabilistic justice across generations could speak to the drive to protect “one’s own.” In the case of assertive localists, a focus on widely shared vulnerabilities could be a reminder that their personal interests do extend beyond their narrow enclaves to the political community as a whole. According to the OECD, while parents’ income and social class influence those of their children, it is still true that two-thirds of individuals in OECD countries have a different social class than their parents (OECD, 2018, ch. 4). Class is not destiny.

This leads to a further conclusion about the relationship between status anxiety and the welfare state. Just as policy choices about housing regulations, school financing, and the public–private insurance mix can generate and entrench stratification, alternative policy choices in these domains can promote integration and help equalize life chances. As a normative matter, then, the answer to status anxiety lies as much in our institutions as it does in our identities. By reframing the political debate to focus on opportunity and mobility, ideals long central to welfare economics but more recently sidelined in political science, we can begin to replace localist identities with solidaristic ones, bonding social capital with bridging. In short, we need not choose between material interests and the demands of identity, belonging, and self-determination. The same institutions can respond to both.

The challenge for liberal democratic politics in an era of high uncertainty is to identify ways to promote probabilistic justice and with it a more equal distribution of both the upsides and the downsides of change. This means not only reducing social gaps so that those who encounter misfortune do not have as far to fall, but also ensuring that there are sufficient opportunities for social integration in a variety of domains and scales. In this respect, perhaps the new status anxiety can be a force for positive change. Insofar as it compels us to confront the ways in which stratifying policies and welfare state retrenchment underlie contemporary pathologies, it may also prompt us to turn uncertainty into a source of opportunity and solidarity instead.