Parties,Civil Society and Democratic Deepening: Comparing India,Brazil and South Africa

Introduction

Indian democracy is in crisis (Varshney, 2022) but India is hardly alone. Not since the interwar period have we witnessed as much democratic regression across the globe. A burgeoning literature on the crisis of democracy has generally focused on OECD countries and has pointed to the economic effects of globalization and in particular how the increasing economic precarity of lower classes has fuelled support for right-wing populism. Yet as I have argued elsewhere (Heller, 2020), the sources of right-wing populism in post-colonial democracies are very different than in the North and the consequences are much more serious. In post-colonial democracies, instances of reaction are not just challenges to the liberal norms and institutions of democracy as in the OECD world, but also concerted efforts to control and even repress civil society and to sustain the power and influence of dominant class-led coalitions.

I develop this argument by comparing India to Brazil and South Africa, two other democracies that have recently experienced significant crises of democracy and that share similar structural and historical characteristics with India. Bolsonaro’s Presidency in Brazil (2018–2022) and the BJP in India (2014–) are illiberal regimes with clear authoritarian impulses. In South Africa, the ruling African National Congress (ANC), which has ruled continuously since the end of apartheid (1994), degenerated during Zuma’s Presidency (2009–2018) into a rent-seeking cabal that has seriously damaged democratic legitimacy and state capacity. For all the differences across these three cases of democratic regression, four shared developments are of deep concern. First, what were broadly respected democratic principles of separation of powers, including the independence of the judiciary, have come under sustained attack. Second, the parties in power have been openly hostile to liberal or progressive elements of civil society that have resisted regime domination. Third, there has been a significant rise in political violence, including state-abated vigilantism. And fourth, all three regimes (Zuma, Bolsonaro and the BJP) have built their support through the weaponization of social media and the propagation of disinformation, severely contaminating the public sphere. These crises are all the more alarming because all three, by the standards of post-colonial world, have been comparatively robust democracies built on the strength of historically broad and sustained democracy movements.

In all three, I argue that these democratic crises result from the formation of social coalitions that have been willing to subvert democratic institutions and practices in order to preserve or restore their social and economic privileges. The formation of these reactionary coalitions can be traced to a deep socio-structural factor as well as two inter-related but more contingent political dynamics. The deep socio-structural factor is that all three democracies were born of passive revolutions that preserved elite power and largely left intact the colonial developmental legacies of an unincorporated mass subproletariat. ² Stated as simply as possible, the inheritance of a colonial subproletariat has made the classic substantive challenge of democracy—achieving a degree of material incorporation of labouring classes in a private property economy—extremely difficult to resolve. ³ The fact that Brazil actually did make some progress on this front during the tenure of the Worker’s Party (PT—Partido dos Trabalhadores) (2002–2016) underscores that this legacy is tenacious, but not insurmountable. This deep structural problem has in turn been refracted through two key political dynamics that are critical to understanding the more immediate drivers of democratic crises. The first is the nature of dominant political parties or regimes and their relationship to nationalism. In most post-colonial democracies, electoral democracy was ushered in by broad-based but elite-dominated political formations that successfully claimed the nationalist mantle. These foundational nationalist parties or regimes—Vargas’s Estado Novo in Brazil, the Congress in India and the African National Congress (ANC) in South Africa—established liberal democracy to varying degrees and even provided some representation to sections of the popular classes, especially organized workers. But they also pre-empted the emergence of political formations that could have more effectively incorporated the masses. The break with this pattern emerges in Brazil with the electoral victory in 2002 of the Workers Party (PT—Partido dos Trabalhadores). Over the next decade and a half, PT governments were able to incorporate the subproletariat through a combination of institutional, welfare and labour reforms.

The difference between India and South Africa, on the one hand, and Brazil on the other, can be traced to the historical relationship between the party system and civil society and points to something of a paradox. The very success of mass nationalist parties in India and South Africa strengthened political society at the expense of popular (as opposed to bourgeois) civil society, which has in turn limited the impact that otherwise transformative social movements that sought to empower the subproletariat might have had on the political system. In Brazil, the failure of populist parties with reformist agendas in the 1960s led to military rule. The two-decade period of bureaucratic authoritarianism (O’Donnell, 1979) weakened the hold of political parties and nurtured the rise of a proactive, highly organized and transformative civil society with deep roots in the popular sectors. It is from this encompassing form of an active civil society that the PT emerged. Once in power, the PT instituted economic and social reforms that had significant redistributive effects and that for the first time in Brazil’s history truly incorporated the subproletariat.

In contrast, in India and South Africa, the prevalence of political society (the ANC in South Africa and the Congress and BJP in India) over popular civil society has sustained dominant social pacts. In South Africa, the ANC has become a party that has orchestrated an alliance of white capital and a rising black middle class that has come at the expense of pursuing the more transformative project of class and racial inclusion that drove the anti-apartheid movement. In India, the Congress and the BJP, for all their differences, built their electoral dominance on the strength of a social pact that links corporate and landed interests to a broadening middle class (first theorized by Bardhan (1983)), but much as in South Africa is predicated on state interventions that rather than incorporating the subproletariat are limited to ‘ensuring the subsistence of the vast population outside the circuits of capitalist accumulation’ (Chatterjee, 2020, p. 81).

Comparing Brazil, India and South Africa

Brazil, India and South Africa are among the most clear-cut cases of democratic consolidation in the developing world. In all three, the transition to democracy was driven by protracted, contentious and broad-based mobilization. This relationship between mobilization and effective consolidation finds empirical support in recent comparative work that shows that post-colonial democracies that were born of longer periods of mobilization are more durable, in large part because they have more built-up civil society structures (Kadivar, 2018; Kadivar et al., 2020). In Brazil, India and South Africa, the democratic regimes that resulted from protracted movements were clearly politically transformative. In India democracy has helped forge a nation from the most heterogeneous social fabric in the world and significantly eroded the social and economic power of traditional landed elites. In the post-Congress hegemony period (1977 onwards), the mobilizational base of Indian democracy has been deepened by the ‘second democratic upsurge,’ the broad-based entry of middle and lower caste groups into the electoral arena (Yadav, 2000). In South Africa, democratic politics and constitutional rule have managed a transition from white minority to black majority rule with minimal violence and brought to power a programmatic, left-of-centre encompassing party—the African National Congress—a party with deep and extensive roots in social movements. And in Brazil the transition to democracy based on universal suffrage in 1984, which came after a prolonged military dictatorship (1965–1984), eventually saw the Worker’s Party (PT) displace Brazil’s traditional political oligarchy and hold power for well over a decade (2003–2016).

What is most remarkable is that these realignments of power have been achieved against social backdrops of extreme inequality. South Africa and Brazil have long held the distinction of having the highest levels of income inequality in the world (as measured by the GINI) and if it were not for measurement problems (i.e., the massive size of the black economy), India would no doubt also be near the top. But even more distinctively, all three are marked by sharp and deep status inequalities: caste and religion in India and the most racially divided societies in the world in Brazil and South Africa. Patterns of social exclusion in all three are highly complex and diverse, but share a structural trait: as per Singer (2012) on Brazil, Chatterjee (2020) on India and Marais (1998) and Wolpe (1972) on South Africa, a dominant characteristic of the colonial legacies of all three is the presence of mass subproletariat, that is a workforce that remains outside of the legal and institutional structures of the formal capitalist economy. The subproletariat encompasses the labouring classes that are peripheral to the capitalist economy, representing a form of absolute surplus labour (Chatterjee, 2020, p. 79). With either very little property (small peasants) or no access to formal waged employment, their labour is casualized and highly precarious. In rural areas, this takes the form of subsistence farmers and landless labourers and in the urban sector, this takes the form of informal workers, including a growing population of unemployed or underemployed and highly peripheralized populations living in favelas (Brazil), township and informal settlements (South Africa) and slums or unauthorized settlements (India). In India the subproletariat is predominantly lower caste, Muslim and Adivasi. In Brazil and South Africa, it is black.

The classic comparative models of democratic deepening (Roberts 1998; Rueschemeyer et al., 1992) argue for going beyond the traditional institutionalist definition of democracy as competitive elections and the requisite civil and political freedoms to examining actual patterns of participation. This means in particular understanding the extent to which historically marginalized groups are politically effective, both in the electoral and the mobilizational/organizational sense. And this in turn raises the question of the degree to which democratic participation translates into substantive outcomes, most notably some degree of redistribution through transfers or taxes as well as through the expansion and socialization of public goods and services. On the participatory dimension, Brazil, India and South Africa have, comparatively speaking, faired quite well. A wide range of subaltern groups have mobilized and pressed claims on the polity and ‘pressure from below’ has been almost a constant in all three cases. That said, the relationship between civil and political society in India and South Africa is dramatically different than that in Brazil. This difference in turn helps explain why Brazilian democracy has been so much more substantively successful.

To be sure, both India and South Africa have made important advances in the past two decades. In India, sustained growth coupled with the rights-based welfare reforms of the UPA governments have seen both an expansion of the welfare state itself and a significant reduction in poverty. In South Africa, a range of government cash transfer programmes have made South Africa one of the more redistributive states among all middle-income countries (Ferguson, 2015; Seekings & Nattrass, 2005). Certainly, both have demonstrated real capacity for distributive reforms, but neither has made serious inroads into the structural problem of the subproletariat. In both cases, inequality has remained stubbornly high, and the conditions of those in the informal sector remain extraordinarily precarious. Most of India’s welfare reforms have focused on the rural subproletariat and despite the significant impact of NREGA in particular, landless labourers and marginal farmers remain at the margins of the capitalist economy. In India’s cities high growth rates have done little to remedy the marginality in work, housing and service access of the urban poor. Similarly in South Africa, the basic structural features of the apartheid economy remain largely intact. The black rural and urban populations that were not incorporated into the racial Fordism of apartheid industrialization remain largely unemployed or underemployed (Ferguson, 2015). And despite the significant capacity and resources of South Africa’s large cities, concerted efforts to ‘de-racialize’ the apartheid city (the most racially segregated cities in the world) have had limited effects.

In Brazil in contrast, the PT years produced what is arguably the deepest substantive impacts of any democracy in the 2000s. Throughout this period, there was a reduction in poverty, a reduction in inequality and a dramatic increase in social indicators. Some of this was made possible by the commodity export boom of this period, though Brazilian growth rates were well below India’s. But modest growth was translated into greater equity through determined policies of redistribution. Brazil’s conditional cash transfer programme, Bolsa Familia, not only contributed significantly to poverty reduction but also improved education and health outcomes. The PT governments raised and enforced multiple minimum wage increases, dramatically enhanced the collective bargaining capacity of workers, including sections of informal labour, and invested heavily in creating a universal healthcare system. These policies broke with the populist-patronage logic of directing public resources to targeted groups. They were instead part of a broader set of reforms in which welfare expansion was unwritten by a significant institutional overhaul of the state that included an expansion of bureaucratic capacities, the creation of a wide array of new policy instruments and in many sectors such as housing, social protection and community health, co-production with civil society partners (Marques, 2022). These transformations in basic state–society modes of engagement come into sharp focus at the municipal level. Beginning in the 1990s, Brazilian cities, having been significantly strengthened as developmental actors through the 1988 constitution, have made significant progress in upgrading slums, providing social housing, expanding public transport and universalizing access to basic services (Bradlow, 2021; Gibson, 2018). This new form of local state embeddedness was initiated in large party through participatory structures such as participatory budgeting and sectoral councils, which provided spaces through which civil society organizations and movements could play a direct role in shaping and implementing new policies (Baiocchi et al., 2008).

Passive Revolutions, Subproletariats and Arrested Democracies

The Gramscian concept of passive revolution has gotten a lot of traction in India. In contrast to the classic democratic bourgeois revolutions of Europe that marked a decisive defeat of feudal elites, post-colonial democracies broke with colonialism without breaking with their inherited class structures. Riley and Desai have defined passive revolution as one ‘in which a mass political party rather than the state promotes economic development and national integration while leaving the pre-existing social and political order largely undisturbed’ (2007, pp. 815–816). Passive revolutions are paradoxical, as Riley and Desai argue, because they are conservative revolutions that employ a revolutionary model of political mobilization. Dominant groups preserve their socio-economic power, but confronted with worker or peasant mobilization accept the need for a change in the political order. The mass-based nationalist party is the vehicle of change, but does so while managing and containing lower class demands. Formal and even participatory democracy emerges but with limited substantive change. With minor qualifications, this pattern also holds for South Africa and Brazil, though in the latter case the ascendancy of the PT which is a mass-based redistributive party, the legacy of the passive revolution has been partially ruptured.

India is unique in the democratic world for having moved directly to universal suffrage at the time of independence but political parties, in particular the Congress, were monopolized by upper caste/class elites and the popular classes/castes never organized on their own terms (Frankel & Rao, 1989), with some notable sub-national exceptions. Aside from the Emergency (1975–1977), the electoral dominance of the Congress as the party of national liberation underwrote democratic stability. But the combination of a stagnant economy and little or no social reform triggered various forms of social mobilization that challenged Congress domination (Corbridge & Harriss, 2000). By the 1980s new political competitors were emerging, expressing both regional and lower caste aspirations. In what Yadav (2000) has famously called the ‘second democratic uprising,’ lower castes and in particular other backward castes (OBCs) started to mobilize on their own terms, supporting politicians from their own castes. The regional thrust of these pressures from below pre-empted the emergence of a national level party like the PT, but nonetheless did destabilize the traditional dominant pact. The emergence of the BJP as a significant electoral force at precisely this time has been widely interpreted as an ‘elite revolt’ and specifically an upper caste response to mobilization from below (Corbridge & Harriss, 2000). When the Congress returned to power in 2004 it was a shadow of its former self, more an assemblage of opportunistic rent-seekers and assorted political scions, than a party with a programme. A powerful faction of the party’s leadership however was close to key actors in civil society, which itself had increasingly coalesced around demands for rights-based social reforms. This faction pushed through a remarkable set of rights-based laws that included the right to information, as well as legislation and policies designed to universalize access to education and food. Most significantly, the passage and implementation of National Rural Employment Guarantee Act (NREGA) has directly confronted the problem of the rural subproletariat. A large body of research has demonstrated that NREGA has pushed up rural wages (Jenkins & Manor, 2017) and that in some parts of India it has clearly disrupted traditional relations of labour domination and local upper caste dominance (Veeraraghavan, 2021). There is also increasing evidence that political parties in India now recognize the limits of traditional clientelism and are increasingly receptive to programmatic or post-clientelistic policies (Jenkins & Manor, 2017). Yet, the transformative possibilities of these otherwise quite significant reforms have been limited in two key respects. Though the rights-based nature of these reforms has broken in principle with the group and patronage-based logic of traditional welfare schemes (Chatterjee, 2020), implementation has been processed and hence refracted through state and local-level institutions and political configurations. There is little doubt that in some cases, organization from below (rural labour unions or various accountability movements) has allowed the reforms to take root and begin to transform local power relations (Adhikari, in press; Veeraraghavan, 2021; Sharan, 2021). But such cases stand against the norm, one in which weak local level state capacity and local power configurations have blunted the effects of the reforms. Second, in material terms, the UPA reforms would have to be sustained and deepened over a much longer period and at a much higher level of public investment to achieve the economic incorporation of the rural subproletariat. Similar policies would also have to be extended to urban sectors, which were not the focus of the UPA reforms. And this is now that much less likely because of the limited political effect of the reforms. Thus, in contrast to the PT which was able to capitalize on its early welfare reform measures in the first Lula government (2002–2006) by securing the electoral support of the precarious workforce that benefited the most from Bolsa Familia and minimum wage increases, the UPA completely failed to translate welfare reforms into electoral support. This failure reflects the organizational and ideological disarray of the Congress, and more broadly the fact that the Congress has limited links to civil society, and in particular no real ties to popular organizations.

This in turn opened the electoral space for the BJP’s traditional urban middle classes to align with elements of an emerging neo-middle class, including upwardly mobile OBCs and Dalits, forging an ethno-nationalist coalition predicated on Muslim exclusion. In other words, the Congress’s failure to consolidate a redistributive class coalition paved the way for a reactionary regime which now threatens the very foundations of Indian democracy. This counterfactual is not simply hypothetical. It finds support in the sub-national successes of Kerala (Heller, 1999) and Tamil Nadu (Kalaiyarasan & Vijayabaskar, 2021), which respectively built left social democratic and Dravidian populist welfare regimes on the basis of lower class/caste mobilization. Even during the BJP’s tenure in power, the possibility of a politics of regional mass incorporation remain significant, as witnessed by the electoral success of the of Trinamool Congress in West Bengal and the Biju Janata Dal in Orissa. In both cases, concerted efforts by the BJP to win at the state level on a communalist platform were repelled by regional parties that banked support for their state-level models of welfarism. In other words, it is precisely where state-level political alignments have had some success in incorporating popular classes that the BJP has failed to make electoral headway. It is telling that since returning to power in 2019 the BJP has assiduously repurposed the incipient welfare state from a right-based logic of distribution in which citizens receive entitlements to a neo-patrimonial logic in which beneficiaries (labhartis) are rewarded by the largess of the state, which is itself identified with the person of Modi (Aiyar, 2022). The BJP has also shifted the UPA logic of associating welfare with active citizen participation and democratization of local institutions (e.g., social audits) as pioneered by movements such as Mazdoor Kisan Shakti Sangathan (MKSS) to reliance on IT systems to distribute benefits directly from the centre. This centralization and neo-patrimonialization of welfare has only further eroded the possibility that demand-making from below could congeal into a form of subproletarian political mobilization. This is especially evident in the BJP governments more recent efforts to challenge the role of state governments in providing welfare, deriding state programmes as ‘freebies.’

Though its democratic history is shorter, the rise of the ANC as a mass-based nationalist party has obvious parallels to the INC. The apartheid regime that emerged in the immediate post-WWII period was first and foremost a regime of colonial racial exclusion in which a white minority monopolized political rights and economic power. The white minority itself was built on a class compromise of British capital and an English-speaking middle class aligned, after a bitter civil war, with the Dutch-descendant Afrikaners, most of whom were working class or land-owning farmers (Marx, 1998). The white minority competed for power and resolved internal conflicts through a competitive electoral system that excluded non-whites. Enforcing apartheid and its systemic exclusions required building significant state capacity.

The ANC emerged as a powerful mass movement in the 1970s, deeply rooted in the union movement, black churches and township associations (civics). As the vehicle of national liberation, the ANC has much the same mass nationalist party organizational structure and underlying social coalition as the INC in its heyday. By the 1980s, it had successfully unified the majority black community (which included the apartheid era racial categories of Africans, Coloured’s and Asians) across all classes in opposition to colonial–apartheid white minority rule. If ever there was a case of post-colonial democratic transition that was favourably endowed to sustain democratic deepening, then surely it was South Africa. The ANC had broad and encompassing support, reaching even into liberal elements of the white middle class. The ANC itself was disciplined, organized, ideologically leftist and supported by a deep reservoir of trained and experienced activists and led by highly charismatic leaders. And the ANC inherited one of the highest capacity states in the post-colonial world, including one of the most effective tax mobilization regimes in the global south (Lieberman, 2003)

If the ANC’s transformative capacity was very quickly dissipated, the reasons can be found in both external constraints and in internal political reconfigurations. During the transition period, determined resistance from propertied whites and very real concerns about capital flight in a period of global neo-liberal ascendancy saw the ANC leadership abandon its social democratic ambitions in favour of national unity and market accommodation. Despite significant push back from unions and organized civil society, an increasingly top-heavy and insulated ANC leadership that had developed close ties to technocratic elites both globally and nationally, quickly pivoted from a transformative agenda that was built explicitly around political and economic incorporation to a growth-led agenda that completely abandoned any pretence of redistributive reforms.

After nearly three decades of rule, the ANC has suffered a dramatic erosion of legitimacy and a steady loss of its vote share over the past decade, but continues to command electoral majorities at the national level. The drivers and effects of ANC hegemony have four distinct dimensions. First, it has systematically insulated itself from the civil society formations that were the basis of its anti-apartheid mobilization. With the exception of unions representing organized workers, all of the ANCs civil society partners, from the township associations that were at the leading edge of the anti-apartheid struggle to the vast array of NGOs and associations that were the backbone of the United Democratic Front, have been marginalized and even demonized by the ANC as anti-nationalist. A wide range of participatory structures that were part of the foundational commitments of the new democracy were almost immediately dismantled and the ANC, especially during the period of Mbeki’s rule, embraced increasingly technocratic and centralized modes of policymaking and implementation. Second, the ANC has carefully cultivated a new elite alliance of white capital and an emerging black middle class. It did so first by foregoing the redistributive demands that were at the heart of the anti-apartheid movement, including nationalization of national resource sectors and land reforms. Instead, the state promoted the de-racialization of private sector leadership and stock ownership through affirmative action policies and stock diversification. These ‘black economic empowerment’ policies only benefitted a small but growing black middle class, and especially those with close ties to the ruling party. These diversification strategies saw the rise of a mega rich class of black capitalists, that included many of the leading figures of the ANC. Third, the ANC maintained significant distributive policies, most notably various welfare grants that have provided baseline social protection to the poor black majority who continue to suffer from extremely high unemployment. Fourth, the ANC built a vast network of patronage, both through direct personal corruption and also the highly institutionalized delivery of public contracts to national and local party officials. During the Zuma’s presidency, this network congealed into what might well be the most prolific rent-extracting regime seen in recent democratic history and was indeed described by the public prosecutor and subsequent commissions as a case of ‘state capture.’ At its peak, Zuma’s tight circle of relatives and close political allies and the industrious Gupta family that had become his preferred business partners were buying up public assets at discounted rates, off shoring vast sums skimmed from public revenues and gaining direct control of state apparatuses. The Zondo commission that investigated state capture after Zuma’s fall found that the cabal had secured control over most of the state’s major public enterprises. The commission went on to conclude that a small group of state and non-state actors ‘conspired systematically … to redirect resources from the state for their own gain’ and that this was ‘facilitated by a deliberate effort to exploit or weaken key state institutions and public entities.’ It added that these efforts were not only about personal financial gain but also geared ‘to shaping and gaining control of the political order … in a manner that was necessarily opaque and intrinsically unconstitutional’ (Public Affairs Research Institute, 2022, p. 3).

If the trajectory of democratic deepening in Brazil has proven more successful, the pathway was marked by much greater disjunctures than in India and South Africa. In the post-World War II period, democratic rule has been routinely disrupted by military interventions instigated by elites in response to popular mobilization. Brazil cycled from a period of elite-dominated electoral democracy with a very limited franchise and punctuated by military coups (1945 to1964) to continuous authoritarian rule by the military (1964–1984) to holding its first direct presidential elections based on universal franchise in 1985. The immediate post-war period was highly unstable because the balance of power was a delicate stalemate between popular forces, especially unionized industrial workers, and entrenched elites that included conservative rural landholders. In a period when the Brazilian Community Party was very active, populist Presidents (Vargas and Goulart) had to toe a very careful line. Every time they drifted to the left the military stepped in (Skidmore, 2007). Vargas built his support among the urban working class by building Latin America’s most corporatist state which on the one hand provided significant social benefits to unionized workers and on the other hand demobilized more radical elements by making union leadership and benefits dependent on the state. But incorporation was highly segmented. Vargas never directly challenged the absolute control that conservative rural elites had over the rural poor. In 1954, when Vargas started to more directly seek working-class support for his fledging labour party, he was forced out of power (specifically driven to suicide) by the military with the support of the bourgeoisie and the middle class. A decade later, his ideological successor Goulart was formally deposed in a coup, this time to be succeeded by a military dictatorship that would endure two decades. Goulart had directly threatened the dominant social block by pushing not only for urban reforms to address the problem of informal housing but by proposing rural land reform in response to a recent wave of agrarian mobilization. As Kowarick and Bonduki note, ‘The culmination of the process in the 1964 coup was at once the climax and total defeat of a wide variety of popular mobilizations’ (1994, p.131).

The period of military rule was one of accelerated growth and industrialization, but one that saw inequality rise to historical heights. Politically it was marked by two key dynamics. First, the military preserved the form, if not the substance, of electoral democracy. It sponsored two political parties—a government party and an opposition party—that competed in regular elections. Second, though the regime was repressive, it was much more tempered than its Southern Cone counterparts (Stepan, 1988, p. 69) and allowed for a fairly high degree of associational activity. While radical students were targeted (especially when they sponsored urban guerrilla movements) the church and other civil society organizations were given a degree of latitude. During the first half of the regime, labour was successfully tamed through traditional corporatist structures but in the second half when the military initiated a political opening (the arbertura), labour militancy skyrocketed. In the absence of more effective political incorporation, the capacity for mobilization remained high.

An emergent democracy movement in the 1980s brought together large swaths of the middle class increasingly disillusioned by the lack of political freedoms, a working class that was organized and increasingly determined to push up wages that had stagnated even as the economy had boomed and various rights-based movements organized around housing, urban transport, health and gender issues. This broad-based coalition of the middle class and the popular sectors brought the authoritarian cycle to an end in 1985, ushering in not only a return to electoral democracy but also a sustained period of democratic deepening (Baiocch et al., 2008). The 1988 constitution massively expanded social rights and institutionalized participatory democracy. The two Presidential terms of Fernando Enrique Cardoso (1995–2002) stabilized the economy, consolidated democratic institutions and introduced some modest social welfare reforms. Cardoso’s party—the PSDB (Brazilian social democratic party) —was centrist and technocratic, rooted in the liberal business community and the middle class, with no organizational ties to the popular classes. During this period, the Workers Party (Partido dos Trabalhadores or PT) emerged as the main opposition party, a programmatic party firmly rooted in the labour movement but with broad ties to a diverse and densely organized democratic civil society. Luiz Inácio da Silva’s (known as ‘Lula’) election in 2002 marked the beginning of a 14-year period of uninterrupted rule by the PT during which the welfare state was significantly expanded. This expansion had four distinct dimensions. First, increases in minimum wages (rising 75% between 2003 and 2013) and direct transfer programmes such as Bolsa Familia fuelled a dramatic rise in working and lower middle-class incomes. Between 2002 and 2014, the bottom incomes of the bottom 70% grew faster than the top three deciles (Gethin & Morgan, 2018). Extreme poverty was cut in half between 2003 and 2013 (Lero, 2019). Second, with new constitutional provisions for decentralization and devolution of resources, municipal governance improved dramatically (Heller, 2019). The result was a significant improvement in the quality of basic urban services, especially in urban peripheries. Third, concerted efforts to formalize informal labour sectors such as domestic workers and sugar cane workers pushed up wages but also gave large segments of the labouring class new collective bargaining rights and some social protection. Fourth, this period was marked by a dramatic expansion of public education and the universalization of basic health care. This included aggressive affirmative action policies that significantly desegregated higher education. Overall, expansion of the welfare state not only socialized basic opportunities but also resulted in a social decompression of institutional spaces. Spaces of traditional upper middle class and largely white privilege, including political institutions from the centre to the local, universities, shopping malls, public spaces and urban neighbourhoods became more inclusive and more diverse. In sum, mass electoral incorporation translated into social incorporation and a degree of improved economic incorporation, though inequality remained comparatively high. But just as dramatic and essential to explaining the reaction (i.e., Bolsonaro’s election in 2018) was the challenge to ‘social authoritarianism’ (Dagnino’s term). Not only was the PT period marked by very real and substantive gains in social opportunity, but the language of rights saturated the political culture, fuelling a wide range of movements throughout civil society, including around race, gender and sexuality. The creation of new political subjects (to borrow from the title from Paschel’s book on anti-racism movements) directly confronted the legacies of social authoritarianism and Brazil’s rigid status hierarchies (Paschel, 2016).

There has been a lively debate in Brazil on just how transformative these reforms have been. There is a broad consensus that inequality has come down to some extent, that poverty declined significantly and the social protection has expanded very significantly. Whether or not the reforms of this period taken together add up to structural break with the dualistic economy that reproduced the subproletariat remains an open question, especially since the transformative momentum was seriously stalled during Bolsonaro’s presidency (2018–2022). At a structural level, Brazil’s economy continues to rely heavily on commodity exports and foreign capital. But in two key respects there has been a clear break with the populist-patronage model of the restricted democracy period. The first has been the significant growth of the social state and its surface area. Social protection and redistributive programmes are rights-based and highly institutionalized, including at the local level. While Bolsonaro launched a concerted attack against the social state, especially by cutting funding to health care and social housing, he actually expanded Bolsa Familia in response to the COVID crisis. He also had little success in reversing the basic laws and governance structures of the social state, and it is clear that Lula’s government will be able to revitalize these quite quickly. The second break is political. As Singer (2012) has shown, it is very clear that during the three full terms that the PT was in power it build a very strong, solid and reliable support based among the subproletariat. When first elected in 2002, Lula’s electoral base was very much in the middle class and unionized working class and regionally concentrated in the more economically developed and much whiter Southern half of the country. By the time he was re-elected in 2006, his support base had shifted to the much poorer and more rural Northeast, where Bolsa Familia has not coincidentally had its biggest impact. The 2022 Presidential elections point to a consolidation of this pattern. Lula was returned to power by a coalition of the poor and the progressive middle class, especially women. The only income group that Lula won was the lowest, that is those at two or fewer minimum wages. He easily won in the predominantly black states (and the historical centre of the slavery-plantation economy) of the northeast. Though his margin of victory was only 1.7%, he won the women’s vote by 16%. And in Sao Paulo, he won handily in the periphery where favelas are concentrated and in the most progressive middle-class neighbourhoods. In sum, while poverty and marginalization in Brazil remain significant, the PT has successfully built out an impressive rights-based welfare state. And in a country in which politics has long been monopolized by a mostly white oligarchy, the PT has built a powerful political force composed of the poor, who are predominantly black, and those who support social and human rights.

I can now briefly summarize these respective trajectories of democratic deepening. The timelines differ, but the basic sequenced phases are similar. A first period of restricted democracy in the post-war period was dominated by a nationalist party in the Indian case, and state enforced exclusion in Brazil and South Africa. In all three cases propertied classes and their middle-class allies dominated with the support of segments of the popular classes co-opted through corporatism or patronage. The dominant pacts in turn produced patterns of development characterized by a disarticulated economy that was highly exclusionary. In a second phase, pressures from below led to basic regime reform in Brazil and South Africa, and to a period in India where lower class/castes interests began to find more autonomous forms of political expression. In South Africa, this phase saw the consolidation of the ANC as a mass-based, nationalist party, that much like the INC in the Nehruvian period secured an alliance between propertied interests and a rising middle class. The ANCs hegemony has diffused the transformative thrust of the anti-apartheid movement and effectively stabilized a highly dualized labour market structure in which a very significant portion of the black popular classes remains completely redundant to the capitalist economy. The ANC’s increasing insulation as a political force has led almost inexorably to state capture and a crisis of democratic legitimacy. Quite similarly, pressure from below in India led to significant welfare reforms during the UPA period. But with its internal fragmentation, organizational weaknesses and lack of ties to popular civil society, the Congress failed to capitalize electorally on these reforms. This failure to build a stable lower class coalition set the stage for the BJP’s electoral success in forging a reactionary coalition built on Hindu majoritarianism and the defense of elite privileges (Heller 2000). In Brazil, higher levels of popular mobilization and contestation during the period of restricted democracy induced greater and more frequent forms of elite reaction. The authoritarian period produced accelerated economic growth and urbanization, but increasing inequality. Limited political incorporation fed an oppositional civil society, which eventually congealed into the electoral support base for an ascendant PT. Once in power, the PT expanded the welfare state and built a solid electoral coalition of the poor and middle-class progressives. Though displaced by Bolsonaro’s coalition, which in class and status group terms closely mirrors the BJP’s social base, Lula’s base has proven highly resilient and returned him to power in 2022.

The Balance of Political and Civil Society

Any discussion of the history of democracy in Brazil, South Africa and India must begin with a foundational paradox. On the one hand, these are societies marked by extremes of categorically based and segmented inequality: slavery and race in Brazil, apartheid in South Africa and caste/religion in India. These have produced what Dagnino (1998) in the Brazilian context has called ‘social authoritarianisms,’ deep-seated inequalities of income and property, as well as of cultural and social capital that permeate social practices and govern social interactions.

On the other hand, and flying in the face of Tocquevillian assumptions that democracy can only take root where there is broad-based associational equality, in all three countries democratization was driven by mass-based social movements. The foundational democratic movements in all three—the Indian National Congress, the ANC/United Democratic Front and the 1980s civil society alliance in Brazil—were encompassing, self-consciously cutting across class, race and caste, and sustained high levels of mass mobilization. They not only pushed for and secured representative democracy but also produced rights-based discourses and constitutions that challenged social inequality and domination. In India and South Africa, however, civil mobilization was ultimately subordinated to the political primacy of a nationalist, mass-based political parties. In contrast, in Brazil, democratic civil society emerged in direct opposition to parties that had been sponsored by the authoritarian regime.

In India, the principle and especially the norm of political and civil rights born of the independence movement have been sustained by a wide range of social movements and have even been the basis of transformative regimes in Kerala and Tamil Nadu. Ahuja (2019) has shown that where anti-caste social movements predated Independence and the consolidation of parties, post-independent politics have been more inclusive. Congress Party hegemony itself was challenged in the 1980s, with the emergence of new political formations driven by regional and lower-caste reform movements. In terms of social gains, a wide range of movement and civil society organizations played a critical role in shaping the expansion of rights-based welfare policies during the Congress-led United Progressive Front government. With the ascendancy of the BJP, the balance has however shifted decisively in favour of state power. The BJP has systematically politicized government institutions, centralizing power in the Prime Minister’s office. The suppression of democratic civil society has become widespread with NGOs harassed, human rights activists persecuted and a wide range of civil society institutions, from media to universities, subjected to state control. All states are wary of independent and critical expressions of civil society, but when a state openly talks of being at war with civil society, condones vigilantism, routinely denounces critics and entire communities as anti-national and insists that duties prevail over rights (Varshney, 2022, p. 116), then the downward spiral into authoritarianism is clear.

The imbalance in South Africa may not be as threatening to democracy. The ANC has lost some of its lustre as the party of national liberation and has seen its vote share erode significantly in the past decade. But it has also built powerful networks of patronage ties, so powerful in fact that a small faction could capture the state under Zuma. Yet, throughout the ANC’s rule, a wide array of movements from local civics to ‘single-issue’ campaigns and HIV/Aids movements deployed a range of ‘in-system’ and ‘extra-institutional’ tactics to press both rights-based demands (HIV treatment) and counter-hegemonic challenges (such as opposition to neoliberalism) on the state (Ballard et al., 2006). More recently, new movements have explicitly challenged the ANCs hegemony (Von Holdt & Naidoo, 2019). In the medium term, the political dominance of the ANC seems assured. But the source of its ideological hegemony—its claim to represent the ‘national democratic revolution’—sets a high bar. For large numbers of South Africans, the promise of a just and inclusive society continues to inflect the meaning of politics with a transformative thrust that by definition leaves much to be redeemed. As Von Holdt and Naidoo (2019) have argued, recent movements, including community protests, revolts of mining unions and student movements, have challenged the ANC’s monopoly claim to post-apartheid transformation, reinvigorated discursive claims that the ANC has suppressed (socialism, participation) and developed new organizational forms.

The general point here is that in both India and South Africa, subordinate groups, despite pervasive social exclusions, have used the political space created by democratic institutions to make public claims. Thus, it is both possible to argue that democratic power in both countries continues to be concentrated in the hands of elites and intermediaries, while at the same time recognizing that civil society continues to be a vibrant arena for democratic claim-making and participatory practices (Jayal, 2011; Mosoetsa, 2005; Von Holdt & Naidoo, 2019). If civil society in South Africa has been containerized, it has not been defanged. Indeed, it was a coalition of judicial actors, media, NGOs and academic research centres that ultimately exposed Zuma’s state capture and triggered his downfall. Elements of Zuma’s cabal still have a strong presence in the ANC and the state, and his successor, Cyril Ramaphosa, has made limited progress in reforming the state. But civil society has remained relentless in its efforts to expose corruption and the abuses of one-party domination. In India, democratic civil society is under siege and state actions have increasingly tilted the terrain of associational life is favour of communalist and illiberal forces. That said, India’s extraordinary pluralism, the accumulated experience of progressive civil society actors (e.g., the MKSS, SEWA) and maybe most importantly federalism, provide some hope of democratic sustenance. It is important to note that at a time when political opposition in most of India, and especially at the national level has become alarmingly ineffectual, the only effective resistance to the BJP’s hegemonic project has come from social movements, most notably the anti-CAA and the farmers movement (Behl, 2022).

From this brief overview then, it becomes clear that the democratic deficit in India and South Africa lies neither in civil society per se nor in the formal character of the state. The state in both cases is a democratic one and, social inequalities notwithstanding, subordinate groups have organized in civil society. The intractable problem has been the vertical axis of democracy. Despite the conditions of highly consolidated democracies with legally guaranteed rights, citizens from subordinate groups find it difficult to engage the state effectively. There are two interrelated problems here. First, the institutional surface area of the state remains limited, especially when it comes to local government. In India, local governments, including those of its vast megacities, have limited powers of autonomous governance when compared with most democracies. ⁴ As a result, Indian citizens face significant constraints in exercising their rights at the local level (Heller et al., forthcoming). South African cities have greater resources and greater autonomy, but ANC dominance at the local level has made for insulated and patronage-driven local politics, including well-documented cases of local state capture (Hart, 2014; Olver, 2017). Popular discontent with the limits of local democracy has fuelled almost continuous waves of ‘service delivery’ protests over the past two decades (Atkinson, 2007). Second, in both democracies, political parties not only monopolize the channels of influence but also exert near-absolute power in setting the agenda to determine which issues, claims and even identities enter the political domain. As a result, the public sphere is shaped primarily by forms of influence that flow directly from political or economic power (parties, lobbies, powerful brokers) rather than from the communicative practices of civil society actors. It is in this sense that I argue that the problem of democratic deepening lies less in the electoral institutions of democracy or the party system (which are dramatically different) than in the political practices and channels that link civil society to the state (which are similar).

This point is reinforced by Brazil. Political scientists have long pointed out that Brazil’s fractious party system is one of the most dysfunctional of any democracy, dominated as it is by small parties that are little more than vehicles for regional bosses, hereditary oligarchs or narrow interests such as the agricultural business lobby. No party in Brazil has ever established itself as a dominant party and none has ever ruled without multiple coalition partners. The paradox of Brazil’s democracy is that an authoritarian regime helped nurture a diverse, rights-based civil society, deeply rooted in the popular sectors. The terms of the transition to democracy in 1985 tilted the balance form political society to civil society and indeed produced a constitution that deeply reflected the emancipatory demands of civil society.

When in power, the PT, itself a product of civil society, claimed a patented model of governing (o modo petista de governor) that included a procedural commitment to ‘incorporating and even institutionalising popular participation in decision-making’ (Hochstetler, 2004, p. 8). A varied literature has documented not only the capacity of civil society actors to make demands and push policies, but also clear instances of civil society projecting itself into the state to shape state intervention (Bradlow, 2021; Gibson, 2018; Rich, 2019). Civil society pressures have resulted in the institutionalization of a wide range of participatory structures and the strengthening of local democratic government (Baiocchi et al., 2008; Wampler, 2015). This blurring of the traditional boundaries between state and civil society has, moreover, had substantive effects. One cannot account for changes in Brazil’s health sector, the formalization of labour markets, environmental reforms, the democratization of urban governance and the expansion of social housing and basic services, without reference to the role of movements (Kerstenetzky, 2014; Seidman, 2010). Finally, these movements have engaged with the state while preserving their autonomy. Even during the PT’s 13 years in power, social movements have openly criticized the government’s economic policies and continued to engage in contentious actions (Hochstetler, 2004, p. 21). The mass street protests in the run-up to the 2014 FIFA World Cup were evidence that despite more than a decade of having the PT in power, civil society had hardly relinquished any of its autonomy. The protests decried the corruption and wasteful expenditures associated with the showpiece, and explicitly critiqued political society. But they did so by affirming support for the public sector by demanding an expansion of health and education expenditures.

Given such power and status shifts in a deeply hierarchical society, the Bolsonaro backlash in 2018, with its sharp class, racial and gender inflections, is hardly surprising and underscores just how transformative democratic deepening has been (Heller, 2000). Bolsonaro’s election is certainly a pointed reminder of the vagaries of representative democracy, especially in a democracy with a notoriously weak and fragmented party system. But this should not distract us from recognizing the clear gains that Brazil had made in terms of participation and state capacity and, in particular, the strengthening of rights culture and local government. The resilience of civil society and a democratized state was in full evidence during the pandemic. Even as Bolsonaro refused to engage the crisis, civil society organizations and local officials mounted effective public health campaigns and successfully lobbied the National Congress for significant welfare measures. More broadly, even as Bolsonaro assailed democratic institutions, unleashed vigilantes and para-militaries, liberalized access to guns, polluted the public sphere with hate speech (discurso de ódio) and demonized human rights activists, civil society organizations and the institutions held fast. The judiciary by and large held his government to account. Though new right-wing media emerged as vehicles of Bolsonaro’s disinformation, the mainstream media was highly critical. In sharp contrast to India, foundations, universities and NGOs remained independent and indeed combative, actively pushing back on Bolsonaro’s policy agenda, especially in education, environmental protection and preserving social protections.

The presidential elections in October 2022 saw Bolsonaro loose to Lula, but by a much tighter margin than predicted. The reservoir of support for reaction in Brazil remains deep, an alliance of right-wing business interests, the agro-business sector, rising evangelicals and the neo-middle class (Heller, 2000). But Lula was able to forge a winning coalition of the poor and progressive elements of the middle class by emphasising the urgency of re-invigorating democratic institutions and expanding the welfare state. More than anything else, he was rewarded for his previous government’s success in building support among the subproletariat.

Conclusion

The comparative literature on democracy has been dominated by global north paradigms and cases and has all too often dismissed post-colonial democracies as shallow and highly fragile. Such analyses are superficial at best and driven by teleological modernization assumptions at worse. Brazil, India and South Africa’s democracies were born of extended periods of mass mobilization that were politically transformative. Formal democracies created spaces for subaltern mobilization. Born as they were in the crucible of deep, categorical inequalities, Brazil, India and South Africa’s democracies have cycled through periods of elite domination and revitalized popular sovereignty. Each has recently experienced crises of democracy, driven primarily by forms of elite reaction (India and Brazil) or entrenchment (South Africa). Taken together, these three cases underscore that elite subversion of democracy is neither an inevitable outcome of some macro process of neoliberal globalization nor the inevitable result of the assumed institutional weakness of the state in the global South, but rather a problem best understood through the lenses of political sociology, and more specifically through a comparative historical analysis of trajectories of democratic deepening.

The Brazilian case demonstrates that social movements and civil society can have transformative effects. At the other end of the spectrum, India points to the extreme dangers of illiberalism that can result from a hegemonic political party asserting itself over and against civil society. South Africa lies somewhere in between. The ANC has been able to contain but not completely dominate civil society. Zuma’s capture of the state represented a period of extreme imbalance that was exposed and pulled back by a resurgent civil society.

At the deepest level of causation, the crises that these democracies have faced stem from the unresolved problem of incorporating the subproletariat. As long as large segments of the population remain socially and economically excluded, the problem of the political incorporation of the masses will continue to challenge the stability of the democratic order. The tensions that arise are played out most immediately in the balance between political parties and civil societies. Where that balance favours parties, the outcome tends towards supporting pacts of domination based on co-optation and patronage, rather than incorporation. Moreover, a weakened or contained civil society is less likely to be able to push back against the authoritarian impulses that have become ever more pronounced in the current age of reaction. When the balance is more favourable to active civil societies, parties are more likely to seek genuine accommodations with the popular classes, or what Gramsci called a ‘concrete co-ordination of class interests.’ (Burawoy, 2003) Though such accommodations may not mark a complete rupture with the legacy of passive revolutions, they clearly represent an advance in democratic deepening.