Since the 2000s, interest in the middle classes, once heralded as the main forces of industrialization, urbanization and bureaucratization in Europe and the Americas, has experienced a renaissance, this time on a global scale. In his ‘tale of two middle classes’, Branko Milanovic (Reference Milanovic2014) has argued that recent globalization has produced ‘losers’ among the middle classes in the West, but ‘winners’ in the global South. Middle classes in the global South, including Africa, are seen by many as being on the rise (Kroeker et al. Reference Kroeker, O'Kane and Scharrer2018; Melber Reference Melber2016). They are believed to boost economic growth, promote desirable social dynamics and safeguard democracy. In short, they are regarded as modernizers who embody a positive vision of social mobility and a meritocratic social order.
Clearly, with regard to developments in the global South, the term ‘middle class’ has become increasingly attractive, both for political and market analysts, and as a term of self-description. Some scholars attribute this popularity to the global conjuncture of neoliberal politics, with its special emphasis on individual achievement and upward social mobility. For Don Kalb, the contemporary ‘discourse of the “global middle class” is … primarily an effort by global and local elites to intervene in ongoing class formations within globalizing capitalism’, aimed at neutralizing any possibility of popular protest (Reference Kalb and Nonini2014: 160). Others insist that we need more open questions about why these discourses are so meaningful to increasing groups of people around the world. As Göran Therborn has argued, the growing importance of the middle class may indeed imply ‘globalized middle-class consumerism’ and political acquiescence, but it can also point to instances of ‘middle-class political rebellion’ (Reference Therborn2012: 26–7).
As an intermediate category, the middle class defied, and continues to defy, neat class divisions. Discussions of the question of who belongs to this fuzzy category and whether the middle class is not, after all, a ‘myth’ (Maza Reference Maza2003), upheld by political discourse but not by social practice, are as old as the appearance of the term in the late eighteenth and early nineteenth centuries. However, most scholars agree that the middle class is not ‘a mere abstraction, a discourse, a metaphor, a rhetorical device’, but indeed a social formation ‘out there’. It is ‘a working social concept, a material experience, a political project, and a cultural practice’ (López Reference López, Weinstein, López and Weinstein2012: 20–1), although not all of these dimensions need to be present at all times or in all places.
But even if there is agreement on the social reality of the middle class(es), scholars have differed in their thinking about the degree to which the definition of the category should rely on the ‘natives’ point of view’, or rather on a set of ‘objective’ criteria such as income, wealth, educational attainments and occupational careers. Sociological and anthropological studies commonly emphasize that it is necessary to explore the material and symbolic struggles over class boundaries and analyse the middle classes’ dynamic and contested self-categorizations. Some have even suggested that ‘middle class’ should best be understood as an ‘aspirational category’ that is increasingly used because ‘middle-class membership becomes a powerful, life-altering goal for many of those poised on the margins’ (Heiman et al. Reference Heiman, Freeman and Liechty2012: 19). According to this perspective, the term ‘middle class’ implies a normative dimension, premised on aspirations of social mobility and certain ideals of professionalism, domestic life, and, more generally, respectability. However, if ‘middle class’ is to serve as an analytical term and enable historical and transnational comparison, some ‘hard’ criteria such as education, occupation and income will also have to enter the definition (Darbon and Toulabor Reference Darbon and Toulabor2011; Melber Reference Melber2013).
In any case, ‘middle class’ is at once a category ‘of social and political analysis’ and a category ‘of social and political practice’ (Brubaker and Cooper Reference Brubaker and Cooper2000) – two aspects that were, and continue to be, entangled in complex ways. Since the end of the eighteenth century, the term ‘middle class’, or the ‘middling sorts’, has become a catchword in European and American political discourse, well before scholars defined it in any systematic fashion. Once established as a conceptual tool of research, however, it began to take on an academic life of its own, with scholars also using it to describe people who did not themselves invoke this category for their self-description. But scholarly terms could also, and indeed did, feed back into folk understandings of social stratification. In particular, the recent global popularity of the term ‘middle class’ seems to be at least in part a result of the appropriation of academic categories by policymakers.
The following – necessarily highly selective – review of historical, sociological and anthropological research on the middle class starts from the assumption that it is important to explore struggles over class labels and examine the practices of distinction associated with these struggles as well as analyse the changing historical and contemporary structural conditions under which middle-class lives unfold.Footnote 1 The article hopes to contribute to the discussion on African middle classes by: (1) tracing the genealogy of theoretical perspectives on class; (2) sketching some observations on middle-class formation in the nineteenth and early twentieth centuries; and (3) outlining findings from research on middle classes in the global South, including India, China, Southeast Asia and Latin America. The article argues that future research on the dynamics of African social stratification has much to gain from a regionally and historically comparative perspective, but that there are also considerable conceptual challenges that scholars of the middle classes will have to confront.
Class theories
Although usually based on empirical findings from the global North, class theories have also informed, in one way or another, scholarship on the global middle classes. At the same time, research on the global South demonstrates that processes of social stratification – and their theoretical reflection – may be considerably complicated by, for instance, cross-cutting, multi-class networks of kinship, or ethnic and religious loyalties.
Some scholars have recently spotted ‘the return of class as an ever-more powerful determinant of inequality’ (Therborn Reference Therborn2012: 13). Not only the middle class but class in general seems to be back on the social-science research agenda.Footnote 2 In Rethinking Class, for instance, a volume discussing the history of research on social inequality in Great Britain, Rosemary Crompton and John Scott warn of the problematic consequences of the ‘cultural turn’ in the social sciences. They argue that the dominant focus of studies on the politics of identity and inequalities produced by gender, ethnicity or race carried ‘a very real danger of “winning cultural battles but losing the class war”’ (Reference Crompton, Scott, Devine, Savage, Scott and Crompton2005: 200). ‘[T]aking culture (too) seriously’, Crompton and Scott believe, risks obscuring enduring social inequality (ibid.: 191), and class therefore needs to be studied seriously.
Other contributors to Rethinking Class, such as Fiona Devine and Mike Savage, insist that the ‘cultural turn’ has been beneficial by overcoming the ‘straightjacket’ of the neo-Marxist ‘structure–consciousness–action’ formula (Reference Devine, Devine, Savage, Scott and Crompton2005: 21). They advocate Bourdieu's concepts of social fields and economic, cultural and social capital, as well as habitus, in order to better understand the symbolic work involved in the construction of social classes (ibid.: 14; Bourdieu Reference Bourdieu1984). Bourdieu does not posit that economic capital necessarily constitutes the most central dimension of social inequality, and thus has no problem in acknowledging ‘the complex and fractured nature of class-consciousness’ (Devine and Savage Reference Devine, Savage, Devine, Savage, Scott and Crompton2005: 15). That ‘people may not see themselves as members of classes’ does not make class unimportant, Devine and Savage claim, but is rather evidence of the ‘overarching power’ of dominant groups to structure the social field (ibid.: 15–16).Footnote 3
Unsurprisingly, none of the recent calls to revitalize class analysis, and certainly not scholars interested in the middle class, have argued for a straightforward return to Marxist class theory with its simple two-class model of capitalists versus proletarians. In his political writings, such as The Eighteenth Brumaire of Louis Bonaparte (Reference Marx1907 [1852]), Marx himself developed a much more refined analysis of social groups and their political interests. In his works on political economy, however, he defined classes exclusively with respect to the production process and the ownership of the means of production or of labour power. Rather than striving for a description of the social structure of nineteenth-century European societies, Marx's opposition of capitalists and workers was a model aimed at the future, expressing the expectation that the capitalist system would eventually develop towards a radical confrontation of these two central, antagonistic classes (see, for example, Marx Reference Marx1887 [1867]).
Marx's class theory has been accused of economic reductionism as well as political teleology. However, in modified versions, Marx's approach to class still plays a role in some analyses of social stratification, and neo-Marxist approaches continue to regard the production process as the most important generator of social inequality.Footnote 4 However, they also try to account for the resilience, and growth, of intermediary groups such as managers, highly qualified employees or professionals who are neither capitalist nor proletarianized. In order to grasp the special position of such a middle class, Erik Olin Wright (Reference Wright1985), for instance, proposes a broader understanding of ownership rights, including the ownership of organizational rights (i.e. the right to decide about the use of the means of production), and of knowledge and qualifications. Members of the middle class thus own some important resources (although they do not hold the material means of production), which make them stakeholders in the capitalist system and thus undermine the confrontational relationship that Marx conjectured existed between the proletariat and the capitalist class. In Wright's view, the class position of managers, experts, supervisors and other members of the middle class is therefore contradictory, as they are both victims and privileged agents of exploitation. Other scholars, such as Thomas Piketty (Reference Piketty2017: especially 380–429), have emphasized the importance of the distribution and inheritance of wealth, not necessarily connected with the production process, for social stratification.
For most scholars of the middle class, however, Max Weber's writings have been more attractive than such neo-Marxist approaches. Weber (Reference Weber1968 [1922]: 302–7, 626–39) defined classes according to their position not in the production process but in the market, distinguishing ‘property’ from ‘commercial’ classes (Besitz- versus Erwerbsklassen), depending on whether people's dissimilar chances to procure goods, ‘gain a position in life’ and find ‘inner satisfaction’ are determined by differences in income from property (rents) or from the marketing of goods and skills (ibid.: 302).Footnote 5 ‘Middle classes’, including entrepreneurs, managers, peasants, artisans and state officials, can appear in both ‘property’ and ‘commercial’ classes. Most importantly, these economically defined categories only become ‘social classes’ when their members develop mechanisms of social closure and some sense of community: in short, class-consciousness. For Weber, economic classes do not necessarily evolve into social classes. If they do, membership in a social class is likely to be closely associated with belonging to a particular status group (Stand) that is defined by social esteem founded on a specific lifestyle, education, or hereditary or occupational prestige. However, people in the same economic class do not always belong to the same status group; conversely, one's economic position and class membership do not guarantee membership in a specific status group (ibid.: 935–8).
It is this idea of complex, non-deterministic combinations of economic and social classes with status groups that made Weber's approach a valuable starting point for researchers of social mobility and status incongruence in Europe and America (for example, Lipset and Bendix Reference Lipset and Bendix1959; Erikson and Goldthorpe Reference Erikson and Goldthorpe1992) as well as for scholars of middle-class formation in regions where race or caste played an important role in social stratification.Footnote 6 Jeremy Seekings (Reference Seekings2008), for instance, has shown how Weber's thoughts on class, status groups and ‘castes’ inspired research in the 1950s and 1960s on the emerging South African black middle class (e.g. Kuper Reference Kuper1965). Even recent attempts to open up class analysis for a nuanced study of social milieus – that is, configurations defined only in part by economic factors, but rather by shared normative orientations and lifestyles – are an outgrowth of Weber's non-deterministic approach to class and status groups.Footnote 7
Most contemporary scholars of social inequality would agree that class needs to be ‘considered as more than simply an occupational or economic category’, and that ‘symbolic and cultural forms’ play an important role in ‘class inequality’ (Lawler Reference Lawler2005: 804). However, scholars remain divided on the question regarding the respective weight of economic or cultural factors and what role ascriptive processes, based on race, ethnicity and gender, play in class membership and social stratification. Furthermore, there is a fundamental disagreement on whether class should be employed as an explanatory concept that assumes that class membership indeed produces social inequality, or as a descriptive concept, which simply categorizes members of society according to a set of criteria that may reflect, but not necessarily explain, social inequality. In German sociology, the first school would speak of Klassen, the second of Schichten, while in English-speaking scholarship both ideas are usually rendered by the term ‘class’. Scholars of the first approach proceed ‘by mapping individuals or families into mutually exclusive and exhaustive categories (i.e. classes)’. The implicit claim of the second approach is ‘that such dividing lines are largely the construction of overzealous sociologists, and that the underlying structure of modern stratification can, in fact, be more closely approximated with gradational measures of income, status, or prestige’ (Grusky Reference Grusky and Grusky1994: 17).
Such a gradational approach, however, relies more or less solely on the scholarly observer's decisions as to which criteria to privilege in the delimitation of classes (Schichten). It does not systematically explore the actors’ politics of difference and ‘boundary work’ (Lamont Reference Lamont1992) that go into the making of class distinctions. Most of the historical and anthropological studies of middle classes discussed below, therefore, go beyond gradational approaches. They tend to understand class as ‘something which is done … rather than a system into which we are slotted’, and as ‘dynamic … continually being re-made in the large- and small-scale processes of social life’ (Lawler Reference Lawler2005: 804, 797). Or, as Loïc Wacquant puts it, ‘class identities, practices and “lived experience” are not “afterthoughts” tacked onto pre-existing classes; they enter into the very making of these classes’ (Reference Lamont1992: 51). In this vein, many contemporary scholars of class would agree with Margaret Somers' appeal to ‘liberate the study of class action … from the constraints of the a priori independent variable’, namely the assumption that large-scale historical transformations, captured in master narratives of ‘the rise of capitalism’, ‘proletarianization’ or ‘modernization’, produce specific economic interests that become ‘the foundational causal factor’ of social action (Reference Lamont1992: 616–17). Instead, Somers suggests, researchers should engage with the multiple, and sometimes contradictory, individual and public narratives that social actors and institutions create in order to make sense of their lives and projects.Footnote 8 Somers developed this advice out of her critique of research on the English working class, but it reflects a perspective on class that has influenced many studies on the history of the European and American middle classes, to which I will now turn.
Middle classes in the global North in historical perspective
Research on the middle class in Africa – and the global South generally – can benefit immensely from broadening the historical perspective. Looking at processes of nineteenth-century middle-class formation in Europe and America allows us to discover surprising similarities with contemporary developments, but also to discern the latter's specificities. Furthermore, the current making of the middle classes is shaped by the past of this social formation. The concepts that both societal actors and scholars employ to make sense of these processes carry not only theoretical but also historical baggage.
Central insights into the formation of the European and American middle classes (classes moyennes, bourgeoisie, Bürgertum, etc.) have been developed by scholars of the ‘new social history’, an approach to historiography evolving since the 1960s that has aimed at studying the social dynamics, mentalities and everyday lives of ordinary people. E. P. Thompson's The Making of the English Working Class (Reference Thompson1980 [1963]), in particular, pioneered a new understanding of class as a ‘historical relationship’ of ‘real people … in a real context’: ‘class happens when some men, as a result of common experiences (inherited or shared), feel and articulate the identity of their interests as between themselves, and as against other men whose interests are different from (and usually opposed to) theirs’ (ibid.: 8–9). Unlike the working class studied by Thompson, however, the middle class comprised people with very different occupations and economic positions, including self-employed entrepreneurs, capitalists and merchants as well as employees, members of the free professions and public servants. The making of a self-conscious middle class therefore depended not on common ‘productive relations’ (ibid.: 9), but, as historians of the middle class have argued, on shared cultural values and a common lifestyle. The boundary work with which the new social formation set itself off from both the aristocracy and from the workers and the poor was crucial to the making of the middle classes. However, as Jürgen Kocka cautions, to ‘have common opponents and to share a culture defines those concerned only to a limited degree. In everything else, they differ: interests and experiences based in occupations and economic status, gender and region, religion and ethnicity. At any particular time, the middle class has been heterogeneous’ (Reference Kocka and Kaelble2004: 19).
In nineteenth-century Europe, the middle classes were influential, but always very small. Their composition as well as the dominant target of their boundary work (those ‘above’ or those ‘below’) differed according to regional variations in the economic and political systems, and these changed over time. In Germany, for instance, the Bürgertum comprised not more than 5 per cent of the population in the early period, growing only towards the end of the nineteenth century (Kocka Reference Kocka and Kaelble2004: 16). It grew out of groups of townsmen (Bürger) of different local origins, developing into a post-corporate, urban, supra-local social formation whose members ‘shared a sense of social distance from the privileged aristocracy’. The Bürgertum stressed achievement and education, and propagated visions of an enlightened society (Kocka Reference Kocka and Kaelble2004: 17; also Budde Reference Budde2009: 5–14). In the course of the nineteenth century, distinctions against the nobility blurred while boundaries vis-à-vis the working class and urban poor as well as against the petite bourgeoisie (Kleinbürgertum) hardened (Budde Reference Budde2009: 92–107). The German middle class was characterized by a close association of Wirtschaftsbürgertum (economic middle class) and Bildungsbürgertum (educated or intellectual middle class), linked through multiple ties of marriage and shared sociability.
In England, by contrast, the economic middle class, comprising merchants, bankers, manufacturing entrepreneurs and industrial managers, was much stronger than in Germany, and its chances of entering the ranks of the nobility and gentry greater (Kocka Reference Kocka and Kaelble2004: 19–20, 22; Seed Reference Seed and Bush1992).Footnote 9 In nineteenth-century France, too, aristocratic and middle-class elements tended to blend, but the French middle class was strongly dominated by public servants, not the economic bourgeoisie (Maza Reference Maza2003). In Central and Eastern Europe, educated groups, in particular state officials, played an important role, and thus ideals of political autonomy and the image of economic power independent from the state did not feature as highly as among the English middle class. More generally, where the intellectual middle class dominated, the emphasis on qualifications and social mobility as well as claims to middle-class status through individual effort instead of property or family background were particularly pronounced; at the same time, these ideas supported exclusionary boundary lines vis-à-vis the petite bourgeoisie and the working class.
In North America, the frontier in America's West offered unique opportunities for social mobility; the civil war strengthened regional networks and loyalties and militated against the emergence of a national middle class; and, last but not least, issues of race and ethnicity haunted the creation of homogenized middle-class identities from the very beginning (Landry and Marsh Reference Landry and Marsh2011). As Burton J. Bledstein argues, the ‘line between working classes and the middle, both in neighborhoods and at work, was often fuzzy and in flux, tempered by ethnicity, nationality, and religion’ (Reference Bledstein, Bledstein and Johnston2001: 17). Entry into the middle class was not a one-time transition, necessarily embracing all spheres of life, but rather piecemeal, and ‘the multiplying interests of the middle classes, the mobile personal alliances, changing occupational fortunes, and shifting career affiliations dictated against a group consciousness or politics’ (ibid.: 23).
If there was anything common to these different constellations of the middle classes in Europe and America, then it was shared cultural values, centred on an emphasis on education and ‘respect for individual achievement … a positive attitude towards regular work, a propensity for rationality and emotional control and a fundamental striving for independence, either individually or through their associations and initiatives’ (Kocka Reference Kocka and Kaelble2004: 18). ‘At the heart of bourgeois culture was a specific ideal of family,’ writes Gunilla Budde (Reference Budde2009: 25, my translation), that was based on the separation of the private sphere from the workplace and public life, and on a gendered division of responsibilities that opposed the housewife at home to the male breadwinner in the outside world. In fact, however, gender roles were contested, and the division between the female domestic and the male public sphere was not as rigid as bourgeois ideology posited. Women often played a far more active role in their husbands’ firms than bourgeois discourse would admit (Davidoff and Hall Reference Davidoff and Hall1987). Furthermore, the model of love marriage was regularly supplanted by pragmatic considerations of strategically useful weddings.
With regard to lifestyle and patterns of consumption, ‘well-groomed modesty’ was to distinguish the middle-class home from the aristocratic love of ostentation (Budde Reference Budde2009: 83, my translation; see also Le Wita Reference Le Wita1994: 62–97). Claims to participate in the desirable habitus of middle-class ‘gentility’ found their material expression in an orderly but comfortably furnished domestic space (Young Reference Young2003). Modest but elegant clothing and an emphasis on bodily hygiene and respectful demeanour were further domains in which middle-class values were to become visible. Among members of the Swedish middle class, for instance, such ordinary practices helped in achieving cultural hegemony over a ‘declining peasantry’ and the ‘emerging working class’ (Frykman and Löfgren Reference Frykman and Löfgren1987: 265).
The middle class ‘molded its distinctive identity around domestic values and family practices’ (Ryan Reference Ryan1981: 15), but these values and practices were negotiated, exhibited and admired in a new urban public sphere created by salons, clubs and associations, festivities and cultural events such as theatre, concerts and art exhibitions, as well as by books, magazines and newspapers. These spaces and media of sociability provided mutual support and served as a ‘school of bourgeois habitus’ (Budde Reference Budde2009: 15, my translation). They embraced and integrated members of diverse occupations and professions, and thus promoted the development of a sense of middle-class commonality. At the same time, they were exclusive and reinforced gender as well as ethnic and racialized hierarchies.
Clubs and associations, as well as newspapers and magazines, offered a space in which new visions of the social order could be discussed and political aspirations articulated (Hausmann and Hettling Reference Hausmann, Hettling and Stearns2001). Beyond propagating the autonomy of civil society and freedoms in the public sphere, however, the middle classes never championed a common political project. Nevertheless, their participation in political debates promoted a liberal-progressive narrative of the productive role of the middle classes (Wahrman Reference Wahrman1995).
More generally, middle-class values and visions of social order were marked by an inherent tension. On the one hand, there was the claim to universality, centred on meritocratic ideals that posited the general accessibility of middle-class status through education and individual effort. On the other hand, there were practices of exclusivity that supported assertions of superiority and hegemonic control.
Kocka (Reference Kocka and Kaelble2004) has argued that central values of middle-class culture have spread throughout the twentieth century, but that the social formation as such has disappeared. The ever-increasing numbers of salaried instead of self-employed persons, he believes, had important political implications for the sense of autonomy from the state that earlier had been important in defining middle-classness. Other authors point to ‘increasing economic inequalities, downward mobility … and declining wages and productivity’ as well as growing difficulties ‘to translate education into good jobs’, which have undercut the ‘certainties of the middle class’ and eroded ‘the best tenets of the middle-class culture – tolerance, egalitarianism’ (Archer and Blau Reference Archer and Blau1993: 25, 35). Most scholars, however, suggest that, despite undeniable transformations, the middle class continues to constitute a distinct social category in the global North. They observe that middle-class men and women still uphold meritocratic ideals and draw boundaries against those ‘above’ and those ‘below’. And they examine how members of the middle class struggle for self-advancement by well-planned investments in education and lifestyle, and are anxious to transmit their status to the next generation (Groh-Samberg et al. Reference Groh-Samberg, Mau and Schimank2014; Mau Reference Mau2015). At the same time, these studies draw attention to the ‘fear of falling’ that Barbara Ehrenreich (Reference Ehrenreich1989) has so trenchantly diagnosed for the American professional middle class – the flipside of the aspirational project on which middle-classness is based.
Contemporary middle classes in the global South
The variety of constellations in which the middle classes have developed in the global North and the range of distinct national trajectories are considerable. They are reflected in different terminologies such as middle class, middling sorts, Bürgertum, Mittelschicht and bourgeoisie, which are not precise equivalents, and scholars of Europe and North America have often been hesitant with regard to the analytical value of the concept of middle class. Its purchase is even less self-evident for countries of the global South, with different histories of incorporation into the world economy and experiences of colonial rule, a late and often only partial industrialization, and great internal heterogeneity concerning rates of urbanization, access to formal education, and configurations of ethnicity and religion.
Doubts whether class concepts are appropriate for the analysis of social stratification and social mobility were particularly pronounced with regard to Africa. Jean Copans, for instance, argued that class theory was ‘an imported product in the hands of specialists in the import and export [of] ideas’; it was even a ‘luxury product’ of limited analytical and political usefulness because it had been appropriated neither by the broader population nor by African ‘organic intellectuals’ (Reference Copans1985: 36). The problem of how to conceptualize social stratification in Africa was intensely debated in the 1970s and 1980s. Many scholars insisted that even contemporary, and certainly colonial and precolonial, African societies could not be understood in terms of social classes. Jack Goody (Reference Goody1971), for instance, claimed that consistent patterns of out-marriage (instead of social endogamy, as in Europe and Asia) continued to militate against the formation of a class system. Others laid more emphasis on ethnic belonging and vertical ties of patronage as factors that held back the development of ‘a bureaucratic bourgeoisie’ (Miller Reference Miller1974: 526) and its counterpart, a lower class of peasants and workers. Some researchers, by contrast, observed an increasing closure of the formerly open educational systems in sub-Saharan Africa and predicted ‘a movement from achievement to ascription’ that would ‘reinforce the boundaries between what could be legitimately regarded as “social classes”’ (Foster Reference Foster1980: 231, 222). Marxist scholars such as Alpheus Manghezi (Reference Manghezi1976) argued that the term ‘elite’, fashionable in analyses of social stratification in Africa, concealed the class character of African societies and thus served to entrench power and privilege. Similarly, Irving L. Markovitz (Reference Markovitz1977; Reference Markovitz and Markovitz1987) argued that even precolonial Africa – and contemporary societies even more so – was characterized by the existence of classes (see also Katz Reference Katz1980; Sklar Reference Sklar1979). With the exception of Gavin Kitching's study (1980) on the Kenyan ‘petite bourgeoisie’, however, these debates did not produce in-depth empirical research on specific upper or middle classes, and scholars eventually shifted their focus to ethnicity, patron–client relations, and the role of the state in the accumulation of power and wealth. Since the 2000s, however, ‘class is back on the agenda of discussions of Africa’, and, more specifically, ‘the apparent emergence of the middle is generating interest’, as Claire Mercer observes (Reference Mercer2014: 228).
Dominique Darbon and Coumi Toulabour (Reference Darbon and Toulabor2011) suggest that there seem to be three reasons why the term ‘middle class’ in particular is currently taking hold among broader strata of the African population. First, the recent economic growth has created a broader middle-income segment. Second, the increasing globalization of categories of social-structure analysis through institutions such as the World Bank introduced this category into African policy discourses. And third, the self-designation as middle-class is attractive since it contains the promise of upward mobility and implicitly presents economic advantage as being based on achievement and thus as morally legitimate. Precisely because of the increasing globalization of discussions on the middle class, among scholars as well as among social actors, a closer look at some studies of Asian and South American middle classes can yield useful insights for research on Africa's ‘middle’.Footnote 10
Some authors, including the historian Ricardo López, criticize the ‘universalization of Euro-American modernity and … a teleological narrative of an intrinsically Western modernity’ (Reference Behrends and Lentz2012: 10). In order to transcend narratives of the histories of non-European middle classes as ‘deviations, failure, or, at best, close emulations’ of hegemonic models, López argues, it is necessary to study the ‘historical formation of the middle class in a comparative, connective and transnational framework’ (ibid.: 12, 11). Such a transnational history of middle-class formation, Mrinalini Sinha claims, should trace ‘the story of modernity as the effects of a globally connected history’. Furthermore, it should explore ‘the repertoire of ideas and institutions that form part of the common stock on which middle classes in different parts of the world draw freely’ as well as ‘the more self-conscious “anxiety of influence” that shadows these formations’ (Sinha Reference Sinha, López and Weinstein2012: 389–90). The parallels revealed by middle-class formation in nineteenth-century Euro-America and in the contemporary global South are indeed remarkable. In what follows, I will first briefly discuss some of these striking continuities and repercussions: namely, the centrality of boundary work; the role of education, ‘intelligent’ work and meritocratic values; new ideals of domesticity and gender relations; practices of consumption; and middle-class sociability and political engagement. I will then point to three novel aspects of middle-class formation that various scholars have explored: the role of the state in making and breaking the middle classes; the relations between middle-class status and ethnicity or race; and the transnational dimension of middle-class formation.Footnote 11
Middle-class boundary work
Virtually all historical and anthropological studies on the global middle classes agree that economistic definitions of the middle class through its location in the occupational structure or its income and expenditure do not suffice. Most scholars emphasize that it is necessary to analyse the middle classes’ dynamic, shifting and contested relational self-categorizations as being different from ‘upper’ and ‘lower’ classes – in terms of education and occupation, but also of lifestyle, values and political aspirations. Interestingly, the criteria along which class boundaries are drawn can shift significantly over time. In Latin America, for instance, when the continent entered into an era of ‘neoliberal globalization’ since the late 1980s, middle classes defined themselves ‘less by education and stable employment in the public sector and more by … income and possessions’ as well as ‘rampant consumerism’. However, these ‘shifting standards’ are not agreed upon by everybody; rather, there is ‘competition among diverse groups for recognition as the class's standard-bearers’ (Parker Reference Parker, Parker and Walker2013: 12–13). With unprecedented economic opportunities in India, ever larger groups are able to engage in middle-class lifestyles and consumption patterns, and ‘values such as material austerity, earlier important to urban middle-class identity, do not carry the same weight today’. As in Latin America, the ‘inclusive moment’ has prompted intense contestations over ‘what makes middleclassness, and who can claim such a status’ (Donner and De Neve Reference Donner, De Neve and Donner2011: 4–5).
An important aspect of the middle class's boundary work is the reference groups that people invoke when defining their social location. In a survey of the social structure and use of class labels in South Africa's largest township, Soweto, Mosa Phadi and Claire Ceruti (Reference Phadi and Ceruti2011) have found that 66 per cent of their respondents were comfortable with placing themselves in the ‘middle class’, although many of them would not qualify as middle-class by any hard socio-economic standards. All interviewees saw ‘affordability’, ‘ability to consume’ and ‘self-sufficiency’ as central characteristics of middle-classness. However, they judged their own capacities in these respects in comparison to different reference groups, ranging from a narrow spectrum of even poorer neighbours to a much larger social world of well-to-do South Africans or even groups of cosmopolitan professionals. Such lack of agreement on social location seems to be particularly pronounced in societies such as South Africa that are characterized by separated spheres of life and consumption with little interaction or overlap (Southall Reference Southall2016). However, the coexistence of different reference groups is characteristic for many societies in the global South. Furthermore, frames of reference and concomitant self-designations may shift over time (Behrends and Lentz Reference Behrends and Lentz2012).
Education and work
The middle classes’ boundary work is shaped by changing historical configurations of educational opportunities, occupational structure, levels of income and distribution of wealth, state legislation concerning inheritance, social security, and so forth. There have been long-term shifts in the economic bases and composition of the middle classes, often analysed in terms of the partial displacement of the ‘old’ by emerging ‘new middle classes’ (Heiman et al. Reference Heiman, Freeman and Liechty2012: 12–15). However, what precisely is regarded as ‘old’ and ‘new’ can vary, depending on the country and period under study. In the early 1960s, for instance, Japan's ‘old’ middle class comprised mainly independent businessmen and landowners, while the ‘new’ middle class consisted of bureaucrats and white-collar employees of large corporations (Vogel Reference Vogel1963). Latin America's ‘old’ middle class, by contrast, had ‘deep roots in the state’, encompassing ‘occupational groups closely tied to state-supported industries and mushrooming public bureaucracies’; since the 1980s, when state-led development models eroded and ‘neoliberal globalization’ set in, this old middle class was eventually sidelined by a ‘“new” middle class of self-employed entrepreneurs large and small, particularly in importing–exporting, computing, telecommunications, and services’ (Parker Reference Parker, Parker and Walker2013: 12–13). Recent research on Asian middle classes, too, generally regards ‘salaried bureaucrats and professionals’ as well as ‘employees in private firms’ as the ‘old middle classes’, while the ‘new middle classes’ that emerge in the ‘post-liberalization period’ comprise various categories of ‘newly rich’ with a much broader range of occupations and often fewer educational credentials (Donner and De Neve Reference Donner, De Neve and Donner2011: 3–4).
However, a certain level of educational credentials seems to be regarded as necessary in order to legitimately claim middle-class status even by the ‘new rich’ (Pinches Reference Pinches and Pinches1999) who demonstrate middle-classness mainly through their participation in ‘new public consumer cultures’ (Donner and De Neve Reference Donner, De Neve and Donner2011: 4). Education and the respectable work to which it gives access serve to legitimate one's social position as earned through one's own effort. At the same time, middle-class discourses regard education not only ‘as key to an individual career and personal success in life’, but also ‘as the means to build a moral community of virtuous and responsible citizens’ (Fumanti Reference Fumanti2006: 97). Moreover, many members of the middle class believe in education as the universal remedy for poverty and marginalization. And, indeed, in much of the global South, education has been – and still is – the most important tangible instrument of upward mobility.Footnote 12 In spite of the increasing inflation of educational credentials (see, for example, Foster Reference Foster1980; Parker Reference Parker, Parker and Walker2013: 14–15) and increasing rates of unemployment among school and university graduates, education is still regarded as an investment that does, by and large, improve one's access to better-paid employment. Many studies show how middle-class as well as aspiring middle-class parents therefore devote much money and energy to supporting their children's education (Berry Reference Berry1985; Donner Reference Donner2008; Abelmann Reference Abelmann2003).
Just as central as education for middle-class formation and reproduction is work. ‘Nothing seems more middle class than the fact of having a steady well-paying job,’ assert Abhijit Banerjee and Esther Duflo in their comparative analysis of household surveys from thirteen African, Asian and Latin American countries. Of great importance is ‘the sense of control over the future’, they argue, ‘that one gets from knowing that there will be an income every month – and not just the income itself’. It is these well-founded hopes for the future that allow ‘the middle class to focus on building their own careers and those of their children’ (Reference Banerjee and Duflo2008: 26). Work also stands at the centre of middle-class images of dignity and worthiness. However, middle-class work is of a particular kind. The middle class often regards itself as ‘the thinking class: freed from the dulling effects of manual labor but not corrupted by a leisured existence’ (Weinstein Reference Weinstein, López and Weinstein2012: 109).
It is interesting to note, however, that upward social mobility is often aided by the combination of work in different sectors, and by income derived from various types of activity. Kenya's ‘petite bourgeoisie’ of public servants, for instance, often used their salaries to invest in commercial ventures or cash-crop production (Kitching Reference Kitching1980). Similarly, South African black middle-class households benefited from access to, and property rights over, land, enabling them to invest in the education of the next generation and to endure economic and political risks as well as to exploit new opportunities (Mabandla Reference Mabandla2013). In any case, the creation of property, beyond salaried income that is often vulnerable to economic and political crises, is important for transmitting middle-class status to the next generation. As Roger Southall (Reference Southall2018b) argues, much more fine-grained research on such processes is needed.
Family life and gender relations
Globally circulating ideas of romantic love, family planning and responsible parenthood, and of privileging the nuclear family over extended kin networks, have become part of middle-class imaginaries worldwide and have fuelled intergenerational conflicts about gender roles, marriage choices, child-rearing practices and responsibilities of old-age care. However, actual middle-class practices with regard to modern marriage, gender relations and family life are rather diverse.
A number of older studies have shown that, contrary to modernist aspirations to the nuclear family, three-generation households and multiple material as well as symbolic obligations towards extended kin networks are the rule rather than the exception, even among upper-middle-class groups. Where this has been the traditional norm, payment of bride price and practices of polygyny often continue (Harrell-Bond Reference Harrell-Bond1975; Stichter Reference Stichter, Stichter and Parpart1988). Loyalties towards the matrilineage may conflict with the conjugal family's preference for virilocality and patrilineal inheritance (Oppong Reference Oppong1981). Multi-class households with sometimes numerous poorer relatives sustained by, and partly working for, educated middle-class couples can be found in many places. Of course, much depends on the institutional framework. Where the state does not regularly offer facilities of childcare, support during periods of unemployment and illness, or old-age care, extended family networks are of vital importance for individual survival and well-being. At the same time, however, changes do take place – sometimes not immediately visible – and middle-class couples do place emphasis on having fewer children, and often at a later age; they do find ways to circumvent or alleviate some of the extended family's material expectations towards them; and they invest heavily in the upbringing and education of their own offspring (Noll Reference Noll2017).
The complex negotiation of traditional family norms and modernist aspirations continues to be an important theme in more recent studies of middle-class families. Erdmute Alber (Reference Alber2014: 263–316), for instance, analyses how West African urban middle-class families and their rural relatives draw on well-established patterns of child fostering in order to support educational ambitions and strategies of upward social mobility. Henrike Donner (Reference Donner2008) shows that Kolkata's rather traditionalist Bengali-speaking middle-class families tend to live in smaller household units, but social obligations such as (female) care for next of kin and members of the extended family remain in place. And Jun Zhang and Peidong Sun (Reference Zhang, Sun, Davis and Friedman2014) explore how parental matchmaking practices for grown-up daughters resurface in the context of urban middle-class families in Shanghai.
Research on changing gender roles has revealed interesting differences between continents. As Michael West's study of black middle-class families in colonial Zimbabwe shows, there have been repeated attempts by European missionaries and colonial authorities to instil ‘the bourgeois domestic ideal’ (Reference West2002: 73), particularly among educated Africans in the colonies. The latter often accepted the idea that being ‘good wives’ meant self-fulfilment through ‘homemaking’ and heeding ‘the sacred calling of home and hearth’ (ibid.: 98). However, despite such patriarchal ideologies, salaried employment outside the house tends to be the norm for African educated women, just like work on the farm or as traders has been, and continues to be, widespread practice among lower-class women (e.g. Dinan Reference Dinan1977; Behrends Reference Behrends2002; Okeke-Ihejirika Reference Okeke-Ihejirika2004). Young professional women in Nairobi, for instance, self-confidently challenge gender stereotypes and claim a right to work in well-paid, respected jobs as well as to engage in egalitarian partnerships with men (Spronk Reference Spronk2012). In South Korea, India and other Asian countries, on the other hand, middle-class status seems to be more closely connected with the ideal of ‘full-time housewives’. The new ‘ideology of egalitarian marriage’ has not prevented a ‘new form of patriarchal family’ from taking root that associates the woman with the private sphere at home, and the man with the workplace and the public sphere (Kim Reference Kim1993: 73; see also Abelmann Reference Abelmann2003; Donner Reference Donner2008).
Beyond the often hidden, everyday negotiations in the domestic sphere, middle-class family relations become highly visible affairs during certain lifecycle celebrations. For urban middle-class families in India, for instance, the traditional marriage ceremony is a public site for exhibiting their wealth and their ability to reconcile a cosmopolitan lifestyle with respect for cultural ‘heritage’ (Brosius Reference Brosius2010: 272–306). In Africa, it is often funerals rather than weddings that are made to serve as grand public statements of wealth and middle-class status (see, for example, Smith Reference Smith2004; Lentz Reference Lentz, Rosenthal and Bogner2009). However, here, too, middle-class families increasingly turn wedding ceremonies into occasions of conspicuous consumption and demonstrations of familiarity with a modern, cosmopolitan lifestyle (Pauli Reference Pauli and Etim2016).
Taken together, these studies on middle-class marriage, family life and domestic arrangements show that, ‘while the construction of the new middle class as a cultural ideal is relatively homogenous, the outcomes of the processes leading to new middle-class lifestyles are often ambiguous and surprising’ (Donner Reference Donner2008: 182).
Middle-class practices of consumption
Like lifecycle rituals situated at the interface between the domestic and the public sphere, but of a more durable and solid character, are the diverse forms of middle-class housing. They are also a particular form of conspicuous consumption that has captured the interest of many scholars of the global middle classes. On the one hand, domestic architecture has a pragmatic dimension. It expresses as well as defines the separation of the domestic world from the sphere of work and prefigures family arrangements, such as the size of the household, the gendered division of labour, or the place of children in family life. On the other hand, the aesthetic dimension of houses publicly demonstrates inter- and intra-class distinctions with regard to prosperity and the owners’ imaginaries of tradition and modernity. Middle-class architecture in Tanzania, for instance, exhibits different styles that range from more traditionalist forms of the ‘respectable house’ to ‘the global aspirational house’, which incorporates architectural details from European or Arab traditions (Mercer Reference Mercer2014). Among New Delhi's middle class, insertion into gated communities is part of real-estate developers’ projects to reshape the urban landscape, with modern condominiums, luxurious shopping malls, entertainment centres and religious leisure sites. Middle-class Delhians aspire to live in a ‘world-class city’ (Brosius Reference Brosius2010: 40) that combines the amenities of a modern global city with carefully selected, beautified elements of India's Hindu traditions.
More generally, scholars exploring the ‘new’ middle classes have often focused on consumption practices as a central field in which a middle-class politics of distinction can best be observed. Consumption is explored as an arena where global images of a desirable life are circulated and, in complex negotiations, domesticated by local norms of respectability. Members of the Nepali middle class, for instance, are ‘pioneers’ of ‘a new space of cultural “betweenness” – between high and low, global and local, new and old, “tradition” and “modernity”’ (Liechty Reference Liechty2003: 25). Similarly, young professionals in Nairobi seek to avoid any semblance of uncritically copying American or British ways of life and insist on creating a modern but distinctly African lifestyle that is respectful of ‘tradition’ – but, of course, a tradition that is reimagined and renegotiated (Spronk Reference Spronk2012).
Some scholars see the prioritization of consumption practices in middle-class boundary work as typical of the era of neoliberal globalization (Jaffrelot and van der Veer Reference Jaffrelot and van der Veer2008; Pinches Reference Pinches and Pinches1999: 32). Others warn that consumption is ‘not the sole interpretive and experiential framework of … middle-class subjects’ (Donner and De Neve Reference Donner, De Neve and Donner2011: 9). It is important not to lose sight of the hard economic realities of consumption. As David Parker observed on the ‘new’ middle class in Latin America: ‘a class defined by lifestyle … was more than ever at the mercy of the ups and downs of the market’ (Reference Parker, Parker and Walker2013: 14). In Brazil in the 1990s, middle-class households were obliged to develop shopping and stockpiling strategies that could mitigate the effects of inflation rates of 2,000 per cent and more (O'Dougherty Reference O'Dougherty2002: 22). In South Africa, the ‘upwardly mobile and fast-growing African middle class’ has been drawn into massive levels of indebtedness (James Reference James2014). The ‘black diamonds’, as they are sometimes jokingly called, borrow considerable sums of money, not so much in order to acquire ‘flashy goods and branded clothes’, but to invest in higher education, marriage and bridewealth, and funerals (James Reference James2012: 20). The (re)distributional effects of borrowing and lending have increasingly become a policy concern, not only in South Africa, but in many other countries as well (see, for example, Walker Reference Walker2013 on Mexico).
Middle-class sociability and politics
The multiple challenges and anxieties of middle-class lives can congeal into collective concerns and inspire political debate and activities. But this has an important condition: namely, the emergence of spaces of communication and sociability beyond the household. The media obviously play an important role in creating shared middle-class agendas. Just as important for the construction of the middle class as a self-aware as well as a self-confident social formation and potential political force, however, is face-to-face communication. It may occur in informal circles of friends who regularly meet in private homes or in public places as well as in more formal clubs or voluntary associations. Some of the studies of middle-class consumption practices discussed above explore urban youth cultures and examine how new norms, lifestyles and political visions are negotiated among peers and friends (Liechty Reference Liechty2003; Nisbett Reference Nisbett2007; Spronk Reference Spronk2012). Sociability of married middle-class men and women seems to take place more in clubs and voluntary associations which help to create middle-class networks that can be useful in many ways, including individual professional advancement. Moreover, if they are part of an international organization such as the Rotary Club, these networks may reach well beyond national borders (Errington and Gewertz Reference Errington and Gewertz1997). Furthermore, middle-class voluntary associations usually have not only mutual support on their agenda, but also charity schemes and development projects for the ‘grass roots’ that may help to justify middle-class privilege in the face of redistributive expectations of poorer relatives (Lentz Reference Lentz1995). At the same time, associational activities may develop into political projects that challenge ruling elites and suggest reforms towards a more inclusive and egalitarian order based on achievement and merit rather than ‘undeserved’ privilege. In any case, voluntary associations provide an effective instrument for members of the middle class to ‘sort themselves out’ (Muncy Reference Muncy, López and Weinstein2012: 379) and to assert a collective understanding of their place in society with respect to other social groups as well as vis-à-vis the state.
In colonial Zimbabwe, informal sociability and a rich associational life laid the foundations of what became an increasingly militant nationalist movement (West Reference West2002; on South Africa, see Kuper Reference Kuper1965: 309–87). Propagating the ‘foundational creed of “equal rights for all civilized men”’, the educated black activists first hoped to improve their situation within a framework of ‘racial partnership’. However, the ‘failure of [their] historic quest for nonracial meritocracy’ eventually led the Zimbabwean African middle class ‘to the conclusion that the advancement of its collective interests required taking power from white settlers’ (West Reference West2002: 239). What had begun as an interracial middle-class reform project thus developed into black popular nationalism that downplayed distinctions of class.
In Latin America, the middle classes’ potential to create a strong civil society and oppose military dictatorships has been at the centre of attention of both (foreign) policymakers and scholars after the Second World War. Many set great hopes on the middle classes’ capacity to reshape their fragmented societies by championing moderate social and political reforms that could contain communist mobilization. In the 1970s, however, middle-class reformism abated and middle-class politicians instead turned to the armed forces to defend their interests against an increasingly radical working class ‘clamoring for the very same rights and privileges that the middle class has so recently and precariously won’ (Parker Reference Parker, Parker and Walker2013: 9).
‘The fundamental precariousness of their privilege’, resulting in a ubiquitous ‘fear of losing class status’ in times of economic instability, seems to be the most powerful motivation of middle-class citizens to take to the streets and demand government policies protecting their investments in education, housing and employment (Walker Reference Walker2013: 14). As Göran Therborn has put it: ‘[T]here is nothing inherently democratic about the middle class … It is “situationally” (opportunistically) democratic – or anti-democratic’ (Reference Therborn2012: 18). At the same time, just as scholars have showed for the nineteenth century, the political uses of the term ‘middle class’ in contemporary public discourse is often ‘infused with heavy doses of mythology and moralizing’ (Parker Reference Parker, Parker and Walker2013: 16). This, in turn, plays an important role in amalgamating the heterogeneous ‘middle sectors’ into a self-conscious, albeit politically heterogeneous, social formation.
New themes: the role of the state, ascriptive differences and the transnational dimension of middle-class formation
While these fields of enquiry resonate with findings from research on European and American middle-class formation in the nineteenth and early twentieth centuries, scholarship on middle classes in the global South has also discerned several novel themes: (1) the role of the state in the making and breaking of the middle classes; (2) the importance of regional, ethnic and religious differences in the formation of a national middle class; and (3) the transnational dimension of middle-class trajectories.
First, with regard to relations between the state and the contemporary middle classes in the global South, most researchers would concur that even though neoliberal policies have provoked ‘a certain disengagement of the middle classes from the project of the developmentalist state … a majority of middle-class households still depend – directly or indirectly – on services and opportunities provided by the public sector’ (Donner and De Neve Reference Donner, De Neve and Donner2011: 15). The public sector was, and often continues to be, one of the major employers of upwardly mobile educated men and women. However, even where middle-class careers depend on state employment, members of the middle class may still develop their own political ideas that can be critical of the regime in power. Political independence, so to speak, is not necessarily premised on economic autonomy.
Apart from the question of the role that state employment, the free professions or entrepreneurship play in the ongoing formation of middle classes in the global South, their economic and social trajectories were, and continue to be, significantly affected by governmental policies with respect to taxes, inheritance laws, credit, social services, and so forth. The educational system in particular, which is still one of the central sites that (re)produces social status and class membership, is usually regarded as a major responsibility of the state; the costs of higher education, for instance, have an immediate effect on middle-class parents’ capacity to lay the foundations of their children's middle-class careers. Members of the middle class thus have vested interests in how, by whom and to whose benefit the state offers its services. The question of middle-class ‘capture’ of the state has been debated with regard to the future of the welfare state in the global North since the 1980s (Bertram Reference Bertram1988) and is now beginning to make its way into scholarly discussions of the global middle classes. The concept of ‘capture’ can mean that state-provided benefits are disproportionally beneficial to, and appropriated by, members of the middle class. But it can also imply the argument that the very policies concerning education, social security and other welfare provisions are strongly influenced by middle-class interests. However, it seems that there are not many examples as yet of straightforward ‘state capture’ by the middle class. On the contrary, at least for a number of African as well as Southeast Asian states, one would rather have to speak of authoritarianism, patrimonial rule and state capture by the ‘elites’ and upper classes, not the middle classes (Welsh Reference Welsh, Murphy and Welsh2008: 57–60; Melber Reference Melber2013).
Some scholars, however, regard ‘state privileging of middle-class subjects’ as a strategic project of ‘booster states’, predicated on and sustaining the ‘capitalist logic’ that aims at ‘class containment’ (Heiman et al. Reference Heiman, Freeman and Liechty2012: 18, 15). Contemporary state policies in the global South, Heiman et al. believe, shift ‘the desires of marginalized groups away from liberatory politics … and toward relatively depoliticized aspirations for middle-class goods and lifestyles’ (ibid.: 19, emphasis in the original). Such an almost conspiracy-theory interpretation, however, does not prevent these anthropologists from also conceding that, in some cases, ‘middle-class politics may represent progressive political agency’ (ibid.: 26). In any case, as these different strands of discussion and competing arguments show, the middle classes’ relationship with the state is becoming an important field of enquiry.
Second, questions of race and ethnicity have rarely been addressed by research on nineteenth-century middle-class formations, but studies on their role with regard to contemporary middle classes in the global South suggest that ethnic (and linguistic) homogeneity may have been an important precondition for the formation of nineteenth-century middle-class culture within the borders of emerging nation states. In the global South, ethno-regional origins and race have often determined access to educational institutions and thus access to more lucrative employment and other kinds of economic capital. Class status thus regularly implies, or is even built on, ascriptive differences. In Latin America, for instance, the middle class is predominantly white (e.g. O'Dougherty Reference O'Dougherty2002; Garguin Reference Garguin, López and Weinstein2012). In South Africa under apartheid, the black middle class's position in the social hierarchy was mitigated by the colour bar (see, for example, Kuper Reference Kuper1965; Seekings Reference Seekings2008), and even the current middle-class status of black corporate managers, for instance, can be significantly influenced by their working-class backgrounds in poorer black neighbourhoods (Modisha Reference Modisha2007).
Case studies on India in particular have explored how caste and religion are articulated with the emergence of the middle class(es). Delhi's traditional Muslim elites, for instance, transformed in the late nineteenth century into a modernizing bourgeoisie that translated Western knowledge into Indian vernaculars and held fast to the religious rhetoric of a reformed Islam (Pernau Reference Pernau2008). Indian Hindus converted Brahman caste status into middle-class lifestyles (Fuller and Narasimhan Reference Fuller and Narasimhan2014). New constructions of ‘Oriental essentialism’, and particularly Confucian ideology, have fed into the self-fashioning of Southeast and East Asian middle classes (Pinches Reference Pinches and Pinches1999). And the predominantly Christian middle class in Aleppo (Syria) in the early twentieth century negotiated its social standing around notions of ‘modernity’ that created distance from the ruling Sunni Muslims, but also aimed at refuting ‘the political and cultural hegemony of the West’ (Watenpaugh Reference Watenpaugh2006: 5).
Mary Kay Vaughan (Reference Vaughan, López and Weinstein2012) has interpreted such ‘traditional’ cultural identifications as expressions of the deep anxiety that ‘modernity’ breeds everywhere. As Vaughan argues, middle-class anxieties result from the experience of a rapid pace of societal transformations, and ultimately ‘derive from the sense that modernity cannot entirely meet our emotional needs’. Consequently, ‘the sense of loss and lack of fulfilment … create[s] nostalgia’ and ‘will fortify – at least for some time – traditional ways of doing or thinking things that enable survival and mobility in a rapidly changing world’ (ibid.: 229). Not all scholars would attribute the persistence of apparently older distinctions and memberships to the existential challenges of modernity that may engender a growing nostalgia for a (reimagined) past. Some would argue that continued, and even revitalized, membership in cross-class institutions such as kin networks, caste, religious communities or ethnic associations remain relevant to middle-class lives for rather pragmatic reasons. These include the precariousness of state-provided social security and welfare and the redistributive pressure that non-middle-class kin exert on their middle-class relatives. However, all scholars would agree that it is necessary to examine more closely how class is articulated with these other identifications and memberships.
Third, and finally, these aspects are closely linked to the question of how transnational, national and local spaces of middle-class formation and reproduction intersect. What seems clear is that globally circulating discourses on and concepts of middle-classness play a much more important role in middle-class formation today than in the nineteenth and early twentieth centuries. Furthermore, through global migration and migrants’ remittances to their home families, transnationally generated income has come to play a crucial role in the making of national middle classes. At the same time, middle-class boundary work may transcend national borders, and those ‘above’, ‘in the middle’ or ‘below’ may no longer necessarily reside in the same locality or country. Mark Liechty has proposed the concept of ‘interjacency’ to capture these complex intersections between space and class. He argues that the ‘current global order produces an unprecedented level of class spatialization in which the interproductive interjacency of class communities occurs less and less within localities or even nations and more and more across national borders and even hemispheres’ (Reference Liechty, Heiman, Freeman and Liechty2012: 276). However, such global class geographies may well coexist with more local forms of middle-class boundary work. Or, as Donner and De Neve put it: ‘While middle-classness is increasingly shaped by transnational movements and interdependencies, the homogenizing tendencies are counterbalanced by the simultaneous promotion of regional and communal distinctiveness, often most strongly mobilized by the very communities that have become global players’ (Reference Donner, De Neve and Donner2011: 17). Indeed, the nation state still seems to constitute the most important framework within which people, including members of the middle class, negotiate their own place in society and imagine a desirable, just social order. Nevertheless, it is imperative to pay attention to the transnational dimensions of these negotiations and imaginations.
Conclusion: challenges for future research on African middle classes
In this article I have engaged with theoretical approaches to studying (middle) classes and have reviewed literature on the history of the middle classes in Europe and America as well as the contemporary development of middle classes in the global South. I have discussed the challenges of conceptualizing the middle class, a particularly fluid social formation. All theories of class have to address the tension between the observers’ analytical categories and the social actors’ self-categorizations. However, this issue seems particularly acute in the case of the middle class for which ‘boundary work’ is often regarded as a defining characteristic. The middle class's ‘boundary work’ was indeed one of the central themes discussed in the empirical research that I reviewed, and there were striking parallels between the history of the middle class in nineteenth-century Europe and America and recent developments in the global South. Further similarities concerned meritocratic values, perspectives on education and work, concern with new forms of domesticity and respectable consumption, investment in new forms of leisure, and engagement in social associations as well as, sometimes, political activities. At the same time, research on the contemporary global middle classes has explored themes that did not feature prominently in studies of earlier periods: namely, the central role of the state in the making and breaking of the middle classes, the continued importance of regional, ethnic and religious loyalties, and the transnational dimension of class formation.
What, then, are the lessons of this review of existing studies for future research on African middle classes? I will first make some suggestions on how the concept of class may be further developed in order to enhance fruitful empirical research. Second, I ask what specific challenges social dynamics and life worlds in Africa may hold for using middle-class concepts that have been developed for rather different historical configurations.
First, class is a comprehensive and relational concept, and most class theories would assume that the entire society is divided into classes. To speak of a middle class therefore entails positing the existence of other classes, such as upper, lower or working class. Classes and class boundaries, however, do ‘not exist ready-made in reality’ but ‘must be constituted through material and symbolic struggles waged simultaneously over class and between classes’ (Wacquant Reference Wacquant, McNall, Levine and Fantasia1992: 57). Furthermore, despite the continuing debate on the precise weight of economic and non-economic factors, the concept of class always invokes economic standing, social status and similar not-innate distinctions. Finally, membership in classes is usually thought of as relatively permanent, comprising not only individuals but the entire family, or at least the household, and, incidences of individual mobility notwithstanding, as being transmitted to the next generation.
Methodologically, the concept of class can be employed in three different ways. First, class can be used as a taxonomic or analytical concept, ‘from an external vantage point to describe the make-up of society and to analyse social process’ (Goldstein Reference Goldstein2005: 89). Second, class may be understood as a bundle of social practices that may or may not be conscious but that create and mark differences between social groups: in other words, an embodied and experiential sense of class. Third, the concept of class can imply conscious self-identification and become part of a ‘social imaginary’ (Castoriadis Reference Castoriadis1987); class then works as a discursive construction and political language that certain groups in society use in order to define their social position and interests and further their projects of social order (Furbank Reference Furbank1985; Somers Reference Somers1992).
As the literature reviewed in this article has shown, these three understandings may, or may not, be combined. Some scholars would question whether it makes sense to speak of middle classes if only the first dimension is present, and not the other two. I would argue in favour of a fairly wide understanding of middle class(es) as social formations that can embrace a broad variety of socio-economic situations and lifestyles. At the same time, I agree that some element of active boundary work is necessary in order to transform a socio-economic category of people into a (middle) class. As some scholars claim, middle-classness can be premised on aspirations of (future) social mobility and certain ideals of professionalism, domestic life and, more generally, respectability. Research on the middle classes thus should explore why and how such normative discourses become an attractive means of positioning oneself. At the same time, however, if ‘middle class’ is to serve as an analytical term that enables historical and transnational comparison, some of the ‘hard’ criteria regarding education, occupation and income will also have to enter the definition.
The heterogeneity of the social strata grouped by scholars or the actors themselves under the umbrella term ‘middle class’ is impressive. This constitutes a challenge for any classical groupist and totalizing understanding of class. Members of the middle class may share some supposedly distinctive values and practices with non-middle-class groups. They exhibit such a range of different lifestyles that the idea of shared class membership may appear spurious. I would still argue that we should retain the concept of middle class, not least precisely because it embraces such a broad variety of socio-economic situations and lifestyles, and because it forces us to examine the ways in which it has become such a powerful resource in social and political practice. Rogers Brubaker's (Reference Brubaker2004) critique of the naive reproduction of ethnic ‘groupism’ by social scientists can fruitfully be extended to social classes. Instead of thinking of classes as large social groups to which one belongs (or not), it seems more useful to explore the social dynamics in terms of ‘doing being middle-class’, as bundles of discourses and performances that can, in specific situations or even over longer periods, be mobilized or disregarded, emphasized or de-emphasized.
Second, there are a number of challenges that seem specific to research on African middle classes. These challenges result, not least, from unresolved tensions that are intrinsic to class concepts in general, but become particularly obvious when these are employed in the analysis of African societies. However, this should not be understood to mean that these tensions are entirely specific to Africa – they may be found in many other areas of the global South and they may be more or less relevant in different African countries. It is more a matter of degree than of essential difference. As mentioned above, class is conventionally understood as an enduring membership that embraces all aspects of life. When looking at typical biographies of educated and well-to-do Africans, however, one may ask whether belonging to the middle class necessarily involves all domains of life and the entire life course. Could not someone be ‘middle-class’ for most of his working life in an urban environment, but behave as a traditionalist elder, appreciating rather different values and styles of consumption after retirement to his home region, or even during his regular visits to the natal village? Similarly, could he or she not participate in rather different circles of sociability even when living in the city, by, for instance, belonging to an employees’ union, a posh tennis club, a Catholic lodge and a popular ethnic association or family foundation at the same time? How do such multiple loyalties and different registers of behaviour affect class membership and styles of middle-classness?
Similar questions arise from the relationship of class and family relations. Conventionally, social classes are considered as comprising not individuals, but households and families. But what does this mean, for instance, for African first-generation middle-class men who married less educated or illiterate village women? How large a section of the family needs to be included to allow individuals or couples to lead middle-class lifestyles? How does solidarity within extended kin networks that cross-cut class boundaries affect middle-class lives, and what does this mean for social mobility? How do state regulations affect the intergenerational transmission of social status and resources, and how does this interact with the transfer of resources and liabilities between (kin) related persons and households? Particularly among recent social climbers, we often find ‘multi-class’ households and families. Furthermore, class positions may change in the course of a lifetime, depending on demographic, economic and other contingent developments, and on the mitigating effects of social security systems. Research on the formation and reproduction of middle classes thus needs to pay attention to both individual biographical, family and historical time (Hareven Reference Hareven1977) and their complex interrelations.
Finally, moving from the individual and his or her family to the wider social formation, both historical and contemporary African examples remind us that class formation may have important subnational as well as transnational dimensions. Most African countries are characterized by pronounced regional and ethnic heterogeneity. This has important consequences for the rhythm of social change, which may occur unevenly across a country, opening specific windows of opportunity or leading to disadvantage at different times in different regions. National and regional criteria of what constitutes middle-class status may therefore differ. To what extent are regional trajectories of social stratification eventually synchronized, and can we observe the formation of comprehensive national middle classes? In what ways do subnational memberships, organized around locality, region, ethnicity or religion, defy, intersect with or drive the emergence of national middle classes? How and to what degree have the different segments of national middle classes developed a shared culture (ideals of family life, standards of worthiness and respectability, patterns of consumption) and a common public sphere?
With regard to the transnational dimension, we may ask what role the involvement in global flows of resources and images plays for middle-class formation. To what extent, for instance, can educational achievements and material resources acquired abroad be converted into social status ‘back home’, and how do they impact on social inequality and class membership within a nation state? What reference groups for class status become relevant in transnational family networks, or among highly mobile professionals participating in global social circles? In what ways are the aspirations of African middle classes inspired by global discourses on middle-classness and imaginaries of ‘good’ middle-class lives?
As Donner and De Neve have suggested, it is productive to ‘substitute the question of what constitutes the middle class … in favour of an exploration of when, where, how and why being middle-class becomes an option, a possibility, desirable or a problem’ (Reference Donner, De Neve and Donner2011: 8). A more open perspective, like this one, on how and why people ‘do being middle-class’ will provide an important corrective to all-too-reified and deterministic theorizations of class. At the same time, we should integrate this into more rigorous analytical conceptions of class. In any case, ‘middle class’ is a multidimensional concept that refers to a socio-economic category, a cultural world and a political discourse. It remains to be seen what further research on African middle classes can contribute both to the knowledge of social stratification and its dynamics on the ground and to the theoretical instruments of scholarly analysis. These challenges also bear innovative potential as they encourage researchers to develop more nuanced understandings of middle-class formation and social stratification in complex societies in general.
Acknowledgements
Large parts of this article were previously published in a chapter in Henning Melber's edited volume The Rise of Africa's Middle Class (Zed Books, Reference Melber2016). Thanks to Zed Books for their permission to reuse these parts in this article. I also wish to thank the Stellenbosch Institute for Advanced Study (STIAS), where I enjoyed a fellowship in the spring of 2019 as part of a group of scholars working on African middle classes. The time at STIAS allowed me to engage with the recent literature on African middle classes and edit the final version of this article. The first extended period of comprehensive engagement with the existing literature on historical and global middle classes was facilitated by a fellowship at the International Research Center ‘Work and Human Life Cycle in Global History’ at Humboldt University, Berlin. Thanks to all colleagues at Re:Work, STIAS and elsewhere who offered valuable comments.
