Patronage vs. ideology in Indian politics

ABSTRACT

How do ethnic, distributive and electoral politics interact in Indian democracy? This essay reviews recent monographs on the politics of the distribution of public and social goods that complicate the popular ethnic voting for patronage provision (EVPP) model. It then explores the disjuncture between micro-behavioural research in distributive politics and an influential account of the role of ideology in the Indian party system. The essay concludes with some thoughts about how ethnicity, distribution and ideology might be integrated in the study of Indian politics.

KEYWORDS:

• Patronage
• clientelism
• ethnicity
• party system
• ideology
• India

Much of the recent scholarship in Indian politics is located at the intersection of how Indians vote and why, what discretionary goods they might receive in return for their votes, and what incentives politicians might have to provide these goods, in order to maximise their chances of election and reelection. These core aspects of clientelistic exchange, or more crudely, vote-buying, have seeped deeply into how we understand Indian politics, and the politics of developing countries writ large. It is only natural, then, that we expect that elections in India – the spectacle of the world’s largest democracy at the polls – constitute the locus at which different actors meet, engage and transact over the terms of the selective distribution of goods and services. In other words, India’s distributive politics and its electoral politics are understood to be profoundly and inextricably intertwined. Further, they are thought to be woven together by a particular understanding of ethnicity (in India, caste) playing a key role in informing voters as to which parties they might vote for in order to maximise the receipt of patronage goods and services.

Yet if we separate the different components of this grand narrative of clientelism, examine them in isolation and then attempt to put them back together, we might find that the pieces do not easily fit together into a coherent and cohesive whole. There are several sources for this lack of coherence. Most notably, a new generation of scholars examining distributive politics, and particularly the regularised interaction between citizens and the state over the delivery of key non-programmatic public and social goods, have highlighted previously occluded actors and dynamics in the actual practices of public distribution. In Auerbach’s (2020) account of the public infrastructure politics of Indian slums, brokers act as agents and intermediaries between parties and poor communities, mobilising votes for politicians but also enabling access to key public goods for client communities. For Bussell (2018), elected politicians, as public officials, perform constituency service through solving problems in the provision of goods and services for their constituents, regardless of ethnic identification and partisan advantage. And Kruks-Wisner (2018) has emphasised how mobilised and empowered citizens can make successful claims on the state for the delivery of welfare. These nuances in distributive politics are welcome and indeed long overdue, but together, they complicate a simple model of vote-buying and thus upend heretofore settled relationships in the literature among ethnicity, patronage and electoral competition.

Classical accounts incorporated these objects of study through a traditional understanding of patronage: the performance of hierarchical submission and obligation in the distribution of resources among groups, operating through the social organisation of caste. This rather mechanical understanding of the relationship among caste, patronage and voting became increasingly out of step with the dynamism and competitiveness of Indian elections, especially at the state level, where parties representing subaltern groups have taken and held power. Using a powerful material-instrumental framework to resynthesize ethnic, distributive and electoral politics, Kanchan Chandra (2004) argued that Indian citizens with particular ethnic identities strategically vote for parties with the highest coethnic headcounts in order to receive patronage benefits. Subsequent scholars have used aspects of this logic to explain the success of the Hindu nationalist Bharatiya Janata Party (BJP) and the rise of regional parties in the 1990s (Thachil, 2014; Ziegfeld, 2016).

The new scholarship on distributive politics subverts the individual instrumentalism at the heart of this model, however. If brokers are responsible for both the distribution of benefits and the mobilisation of votes, then the characteristics of brokers and their networks, rather than the ethnic composition of party leadership, would determine how benefits translate into votes and vice-versa. If elected officials regardless of partisan affiliation undertake constituency service to make up for the deficiencies and inequities of public provision, then this represents a component of distributive politics that has nothing to say about the mechanisms of ethnicity and outcomes of electoral mobilisation. And if individual perspectives and experiences determine which citizens might make claims on the state and thus receive welfare goods, then claim-making outside the polling booth has far more impact than the casting of votes, based on ethnicity or otherwise, within it.

Together, this recent generation of scholarship has significantly eroded the theoretical foundations of the ethnicity-implicated relationship between distributive and electoral politics. Yet scholars have been wary of separating distributive and electoral politics entirely. Rather, many view Indian citizens as primarily voting on the basis of performance, especially in the delivery of goods and services, whether by parties, by candidates, or by brokers (Auerbach & Thachil, 2020; Bussell, 2018). This might be an eminently reasonable proposition, and indeed one deeply embedded in normative theories of electoral democracy. But it does not have a solid empirical base in India, where many fewer citizens receive discretionary welfare benefits than vote in elections.

Assuming that voting occurs on the basis of performance also tends to conceal ideology as a driver of electoral competition. There are certainly some reasons to think that ideological cleavages, at least as they are understood in the West, cannot characterise a political system as full of personalistic organisations and inconsistent policies as India’s. Yet ideology can represent an important set of explanations for understanding how Indians vote for (and against) parties, independent of the performance of candidates in the delivery of goods. It can also have traction on the more aggregate questions of electoral politics, such as changes in the party system over time, which microanalyses based on ethnic voting and patronage provision have some difficulty examining.

Chhibber and Verma (2018) present a novel ideological framework for understanding the Indian party system, as shaped by cleavages along the politics of statism and the politics of recognition. They argue that the nature of party competition, and its changes over time, are grounded in the expression of long-term social divisions over what the state should do or not do in relation to the economy and society. Their framework is important for understanding the recent consolidation of the party system under Narendra Modi’s BJP. While this ideological perspective on electoral politics in India has some lacuna, it represents an important alternative to either ethnicity and patronage or performance as explanations of how Indians vote and why, which can account for the dynamic nature and aggregate outcomes of electoral competition.

In what follows, I will survey four important monographs and explore what they tell us about distributive politics, electoral politics and the intersections between the two. First, I will explore two generations of research at the intersection of ethnic, distributive and electoral politics in democratic India. Second, I will discuss three books that provide much nuance to distributive politics, but in so doing, complicate settled associations among ethnicity, patronage and electoral competition. Third, I will present the ideological alternative to the dominant patronage-based explanation of Indian politics. I conclude with some thoughts of the alternative ways in which the politics of ethnicity, distribution and electoral competition can be studied, separately and together.

Caste, patronage and elections in democratic India

India’s independence in 1947, and the adoption of a democratic constitution in 1950, was justly seen as a remarkable achievement, one that only became more remarkable as democratic competition was consolidated in one of the world’s poorest countries. As vigorous democratic participation and electoral competition became institutionalised, social scientists sought explanations for the persistence of India’s democracy. Weiner (1967) situated these outcomes in the institutional success of the Congress Party, which incorporated pre-existing traditional social elements, performing key roles in the process of intermediation between citizens and the state. Thus, the unique stature and effectiveness of Congress were rooted in continuing political activity that was associated with traditional pre-democratic norms: most notably, patronage from the state through the party to groups embedded in the social structure. Further, Rudolph and Rudolph (1967) argued that key elements of the social structure – caste and specifically caste-based associations – represented key bulwarks of, rather than hindrances to, democratic practice. Thus early characterisations of Indian democracy represented a traditional form of patronage that provided resources to groups and supported democracy, but could also reinforce group-based hierarchies.

Party organization and caste mobilization after congress decline

As Congress declined and new political parties, particularly those representing intermediate and subaltern groups, emerged and started to compete effectively, these earlier characterisations of Indian politics became less compelling. Upper-caste elites no longer completely controlled the state, and various communities increasingly looked to their own rising political leaders, from farmers to film stars, to shape policy and provide key goods and services. In contrast to earlier scholarship, Chhibber (1999) argued that, in the absence of formal associations and given the powerful influence of the state in Indian democracy, political parties themselves could construct politically salient social cleavages through unmediated relationships with client populations, and thus transform the party system at the state level. This argument explained critically important phenomena during this period, such as the collapse of the Congress Party in its Uttar Pradesh heartland in the 1990s, because its catch-all strategies, reliant on old patronage networks, were eclipsed by the formation of new coalitions based on ethnic and religious cleavages that polarised the electorate. Chhibber’s account placed party strategies and responses to government policy at the centre of electoral analysis and relegated patronage to a mediating rather than a primary explanatory factor in the shaping of partisan competition.

Broader trends in political science, however, have focused on more instrumental dimensions of ethnic identity and the straightforward exchange of votes for material rewards, transforming the traditional notion of patronage as supplicatory relationship to clientelistic exchange, or the mobilisation of votes by the selective provision of goods and services (Kitschelt & Wilkinson, 2007). Chandra (2004) incorporated these elements into a parsimonious framework that sought to explain the rise of the Dalit-led Bahujan Samaj Party (BSP) in Uttar Pradesh in the 1990s. Chandra began with a famous characterisation of India as a patronage democracy, a democracy in which ‘the state monopolizes access to jobs and services, and in which elected officials have discretion in the implementation of laws allocating the jobs and services at the disposal of the state’ (2004, p. 6). Given this monopolisation and this discretion, voters seeking access to publicly provided goods and services could conduct ethnic headcounts and vote for the party with the highest proportion of co-ethnics in its leadership ranks. Based on these assumptions, an ethnic party is capable of capturing office only ‘when it has competitive rules for intraparty advancement and when the size of the ethnic group(s) it seeks to mobilise exceeds the threshold of winning or leverage imposed by the electoral system’ (Chandra, 2004, p. 3). The book thus presents a model for individual voter preference that ties ethnic identity to clientelism.

Chandra’s book, heralding what I call the ethnic voting for patronage provision (EVPP) model, has been justly influential for bringing individual agency to the actions of individuals and operationalising caste not as a fixed hierarchical system but rather than as key (if ascriptive) identity that can represent the basis for instrumental mobilisation. The EVPP model has inspired other influential research on the changing electoral politics of India from the 1990s. Thachil (2014) argued that the Bharatiya Janata Party (BJP), as an elite party, was able to attract enough voters among subaltern communities through the provision of social services from affiliated civil society organisations, in addition to its ideological appeals to upper-caste communities, to achieve national power. Ziegfeld (2016) explained the rise of coalition governments at the state level in the ways that individual political elites with successful followings faced non-institutional incentives, including the mobilisation of clientelism, for the formation of new parties based on particular regional identities. Characterising this growing literature, Chauchard (2016, p. 254) stated that ‘ethnic parties win not because voters feel strongly about their identities but rather because the ethnic labels attached to parties constitute useful informational shortcuts – or heuristic devices – for patronage-seeking voters. The convergence towards this conclusion has been remarkable’.

EVPP and its discontents

Elegant as this framework is, we might question some of its assumptions. One of these relates to the idea that voters seek ‘psychic’ goods in addition to material ones (Chandra, 2004, p. 11). Yet the former are qualitatively different from the latter, both in terms of their creation and their distribution; they suggest a moral politics of assertiveness and dignity, often in the face of systematic discrimination and social domination, which is quite distinct from an instrumental one of patronage and clientelistic exchange.

Even more challenging is the concept of a patronage democracy. India, like most developing countries, is a society with both scarcity and hierarchy, and thus replete with elite connections, rents, rent-seeking, partiality and thus pervasive patrimonial relationships. Yet it is not clear exactly how this relates to democratic competition, that is, how much these elements are monopolised under the control of the state and placed completely under the gift of individual politicians to distribute. Many of the most valuable rents in India’s political economy – subsidised bank loans to favoured companies, admission to elite public and private schools and universities, employment in successful companies – are determined on the basis of murky connections over which most individual politicians have limited control. In fact, much of the most powerful operations of influence, in India and beyond, lie within elite networks at far remove from democratic politics, as Ashis Nandy noted at the Jaipur Literature Festival in an incident that remarkably led to his arrest (Outlook, 2013). At more quotidian levels, the state itself can be constrained from monopolising patronage by limited state capacity and subcontracted governance traditions (Naseemullah & Chhibber, 2018). Fortunately, a new generation of researchers have started to investigate the nature and impact of patronage relations on the ground in order to understand the politics of India’s welfare distribution on its own terms.

The distributive turn in Indian political research

The (mal)provision by the Indian state of key goods and services aimed at ameliorating the welfare of a poor population – from specific items like handpumps or gas cylinders to subsidised foodstuffs to infrastructure like paved roads and electrification to access to education and healthcare and employment guarantees – have been at the heart of academic and policy debates across an array of different disciplines. Partiality, discretion, exclusion and corruption in the distribution of these goods and services are prime targets of reform efforts from within government and among NGOs and international agencies. Academic and policy researchers regularly conduct monitoring and evaluation on welfare programmes as part of these efforts (see Banerjee & Duflo, 2011).

A number of political scientists working on India have established a successful research programme around distributive politics, with a primary focus on the analysis of the political incentives and constraints facing the agents of the state’s welfare and social service provision and how recipients might act politically to maximise the receipt of these benefits. In a mushrooming field, there has been a remarkable convergence around the parameters of this research. First, there is a (welcome) normative framing of its significance; distributivists characterise the Indian state’s welfare provision as woefully inadequate and a substantial barrier to the needs and aspirations of the poor, and are engaged in a collective enquiry to uncover the roots of these problems, in order to remove them. Second, most scholars locate explanations for this maldistribution in the dynamics of patronage and clientelism, or more broadly, examine the incentives of political actors to use targeting of goods and services for electoral advantage. But as a new generation of scholarship demonstrates, a full accounting of the politics behind public distribution, particularly at the local level, requires attention to a variety of actors and relational dynamics at some remove from the EVPP model. Several excellent recent monographs have focused on the constituency service of elected officials, the claim-making of empowered citizens and the activities of partisan brokers in slum communities.

Elected officials and constituency service

In an award-winning book that marshals different forms of evidence in the service of a narrative account of the distributive activities of high elected officials, Bussell (2018) argues that constituency service preoccupies much of the time and energy of Indian politicians holding public office at the state and national level. Bussell then frames constituency service as representing a powerful, hidden means of addressing deficiencies in the provision of publicly-provided goods and services. In representative democracies with defined geographic constituencies, there are powerful norms that allow constituents to engage with their representatives directly, and representatives can deploy their own influence and resources to solve problems in service delivery brought to them by individual constituents. Bussell shows that they tend to do so in a non-partisan, non-contingent manner, and further contends that it is tactically rational for politicians seeking reelection to conduct constituent service as a means of reaching swing voters. This sophisticated account highlights the importance of examining in detail how elected officials act in relation to public distribution, thus unravelling some of the assumed mechanisms of provision to targeted (ethnic) constituents for electoral mobilisation.

Bussell’s book has two virtues. First, it sheds light on an absolutely central, though often elided, figure in Indian distributive politics: the elected representative. Bussell deploys theories of representation, as well as considerable empirical evidence, to provide a corrective as to the meaning and purpose of these positions. She argues that assisting their constituents with problems negotiating the state and its delivery of goods and services represents ‘a predominant activity of many elected officials in India’ (Bussell, 2018, p. 4). Her fieldwork shadowing MLAs reveals the many, multifaceted problems with constituents’ access to public goods and services. The fact that elected officials in India spend a significant amount of their time performing these services – her surveys of legislators report that meeting with citizens accounts for between a fifth and a third of their working lives – reflects a dimension of their vocation that is orthogonal to clientelism (Bussell, 2018, pp. 52–53).

Second and relatedly, Bussell makes a powerful argument that disrupts the assumed operation of patronage democracy and the EVPP model. If the latter were a correct and comprehensive account of how politicians act, then we would see political actors performing services and directing resources only toward those individuals or communities who would plausibly vote for them; all other activities would be process-irrational. While it is undoubtedly the case that politicians (and brokers) arrange this targeting, particularly around election time, the sheer amount of time, energy and resources spent on constituency service acts as a form of revealed preference. Bussell additionally develops a theoretical framework that explicitly ties constituency service to a supplementary electoral strategy to appeal to persuadable voters, especially those excluded from partisan networks dominated by brokers and lower-level officials; she argues that the public performance of generosity and mediation from above allows legislators to build a personal base of support beyond partisanship (Bussell, 2018, ch. 4). At the very least, elected officials not available for constituency service might incur serious reputational costs for violating key norms of representation. In this sense, both the norms and the incentives for providing concrete public service are powerful across representative democracies, both rich and poor. These norms and incentives temper the purely clientelistic delivery of goods and services according to the EVPP model.

Bussell makes an additional claim, however: that constituency service can meaningfully compensate for the inequity, corruption and inefficiency in public distribution and delivery by providing an alternative means by which citizens can successfully make claims on the state. The primary difficulty here is one of scale: India has the largest district magnitudes of any democracy, with MLAs representing about 300,000 citizens, and MPs several times that many. An MLA, let alone an MP, working as hard as she possibly can in providing as many services as possible to her constituents, will only be able to engage individually with a small fraction of them. To provide support for the concrete impact of constituency service, Bussell (2018, p. 55) reports results from a citizen survey in Karnataka, and cites another by Kruks-Wisner (2018) in Rajasthan, that slightly more than 20 per cent of respondents have requested help from a politician in accessing a government service; a lower figure of 12 per cent would mean that MLAs would have to meet around 21 constituents, on average, every day of a five-year term. Accounts differ as to whether this is a realistic portrayal, and part of the ambiguity is whether these meetings are with the official herself or her agents, and whether such meetings are individual or collective. A prominent Hyderabad politician with a national following, and thus many external commitments, maintains a large constituency staff, who act on his behalf in addressing constituents’ concerns (Suneetha & Moid, 2019).

Moreover, it is also not clear the extent to which the complex problems of constituents can be fully solved by a short meeting with a public official. She might indeed order an underling to put in a call or write a letter to the relevant bureaucrat or office, but is unlikely to have the time or resources to follow-up on requests to be sure of its implementation on their own; in fact, Bussell reports that officials tend to ask the claimant to return if the problem is not resolved. In short, the time and attention of these high public officials is a far scarcer resource than most of the public goods in question.

Thus, while elected public officials might have the innate capacity and evince a willingness to intervene to right the wrongs of a deeply inequitable and corrupt public distribution system, they cannot do so effectively for that many people. It is therefore not clear how much concrete electoral benefit can be realised from the norm of constituency service.

Despite the clear willingness of elected politicians to help constituents regardless of ethnic identity or partisan leaning, it is also not clear how accessible they are. These offices might be in towns or cities dozens of kilometres away from the villages of aggrieved constituents. As Bussell notes, legislators and their agents are only present in the constituency on certain days of the week and weeks of the year, and are available to entertain constituents during particular times. Notwithstanding sanguine survey findings, there are few very poor Indians, especially living in rural areas, who could plausibly afford to take a day or more off work, forgoing daily wages, to travel and to wait for the possibility of an audience with a public official to intervene with the local welfare bureaucracy, much less be able to return in subsequent weeks if the problem is not resolved. This suggests that this avenue of recourse is most accessible to those with some time and resources at their disposal. Survey results from the book show that the most frequent requests for constituency services, apart from ration cards, involve building licenses and land records (Bussell, 2018, p. 60). These are not welfare goods for the poor, but rather key instruments for proprietary actors. While constituency service is indeed an alternative means of influencing service delivery, it is not viable for everyone seeking help in gaining access to welfare.

Yet Bussell’s book makes a vitally important statement about the activities of public officials, the supposed key actors in the supply of clientelism. If politicians themselves do not seem to be spending a significant amount of time directing service delivery solely to maximise their votes, then a crucial assumption of the EVPP model is at the very least diluted. Politicians seeking to achieve or remain in office and parties seeking power might try to still target resources to particular portions of the electorate particularly around elections, but these occur only every five years. Most of the rest of the time, the substance of distributive politics occurs in the interactions among elected officials, bureaucrats and citizens, thus straining any intimate, constant connection to electoral politics.

Activist-citizens and claim-making

If Bussell’s book focuses on political elites who might intervene to distribute welfare goods and services, Kruks-Wisner (2018) highlights ordinary citizens who act to ensure their receipt of these goods and services; it is an account of bottom-up distributive politics. As such, it engages with a complex landscape of negotiation, entreaty, activism and assertion between individuals, many from subaltern communities, and the local state in rural Rajasthan. It centres on the activities of everyday claim-making activities of the rural poor – ‘attending a meeting, filling out an application, visiting a government office, or approaching a local leader for assistance’ – in overcoming the deficiencies of public distribution and forcing otherwise disinterested bureaucrats to fulfil their responsibilities in distributing social welfare (Kruks-Wisner, 2018, p. 4).

Kruks-Wisner identifies claim-making as a key activity in distributive politics and highlights the individual characteristics of citizens in need of services that determine who feels empowered to make claims against the state. She argues (2018, p. 5) specifically that claim-making ‘is shaped, first, by the institutional terrain of the state (assessed in terms of its reach, visibility, and accessibility) and, second, by an individual’s social and spatial exposure to people and places beyond the immediate community and locality … ’. While the book focuses on both these demand-side and supply-side factors, it is rooted in an analysis of active citizenship, or the efforts of individuals to bridge the gap between what the state has promised in terms of welfare and what is delivered to citizens (Kruks-Wisner, 2018, pp. 8–11). Why do some individuals and not others make claims on the state? Kruks-Wisner argues (2018, p. 19) that ‘citizens who traverse local social and spatial boundaries will, all else equal, be more likely to make claims on the state, and will do so through a broader repertoire of practices, than those who remain more constrained’. These arguments follow social-psychological mechanisms of citizens in need of welfare, first framing these goods as entitlements and then discovering their own sense of efficacy in pursuing them.

A key value of this approach is its incisive understanding that welfare must be actively demanded from an often-unresponsive state apparatus as a feature of active social citizenship; in this sense, distribution on the ground is not technocratic or administrative but deeply political. She provides an account of the state’s institutional terrain in which claim-making is likely to occur. The breadth of welfare programmes but also unevenness in their implementation are key institutional features, common to contexts like Rajasthan, that drive citizens to make claims. And, importantly for distributive politics, Kruks-Wisner argues that claim-making is efficacious; those who make claims receive more from the state.

The book’s analytic emphasis on ‘micro-determinants of citizen action at the level of the individual’ and the related practices and repertoires of claims-making tends not to emphasize contentious and movement politics as a central component of group claim- making (Kruks-Wisner, 2018, p. 25). But this is a large part of the politics of the poor and marginalised, who can act more safely and effectively in concert, such as village communities gheraoing a collector’s office demanding relief to farmers marching on the national capital against neoliberal reform. Ahuja (2020) argues that collective contentious action among Dalit activists has been crucially important for the provision of goods and services for Dalit communities, and is moreover a vital means for the community’s empowerment and assertion of collective citizenship against systems of social domination that go much further than the state’s neglect.

Important for the purposes of this essay, however, Kruks-Wisner’s monograph challenges any intimate link between the politics of public distribution and partisan electoral competition. Claim-making suggests ‘a degree of room to maneuver beyond partisan structures – even in settings that remain distinctly ‘non-programmatic’ given a lack of regularised and rule-bound allocation’ (Kruks-Wisner, 2018, p. 199). Her research indicates that parties are not central to claim-making – they are one of a large collection of mediating actors – and that there is little evidence of any quid-pro-quo relationships of services for votes. The book, when engaging with representative politics, does so at the level of the gram panchayat, an often (nominally) non-partisan village-level body that has complicated relationships to higher-order politics. More broadly, she notes that citizens needing services face ‘a claim-making environment appears broader and less static than most clientelistic readings of citizen-state relations would suggest’ and details multiple pathways into a porous state structure, including NGOs, social movements and local associations in addition to parties and brokers, for citizens seeking public welfare (Kruks-Wisner, 2018, p. 201). These findings are significant; unlike Bussell and Auerbach, Kruks- Wisner set out to explain patterns of distribution independent of its links to electoral competition, and her findings are consistent with a view that disrupts the EVPP model.

Brokers and slum communities

In a monograph explicitly linking electoral mobilisation and the provision of public goods and services in the slums of Jaipur and Bhopal, Auerbach (2020) focuses on a key actor: the broker. He asks ‘why are some vulnerable communities able to demand and secure development from the state while others fail?’ (Auerbach, 2020, p. 3). His deeply intuitive yet compelling answer is that slum settlements differ in their political organisation; some have a larger number of brokers associated with different partisan organisations, and as a result, they are better able to extract key public services from governments at various levels in return for electoral mobilisation, whereas broker-poor settlements lack access to the state and the parties that control it at different levels. This is, in turn, a function of the size and ethnic diversity of settlements; larger slums increase the incentives of parties to invest in establishing a political presence, and more ethnically diverse ones fragment local leadership, as slum residents prefer to seek services from co-ethnic leaders and brokers. As a result, larger and more diverse slums with denser networks of party-affiliated brokers enjoy more group-based public goods in the non-rival, non-excludable sense, from paved roads and street lines to water and sewage connections to municipal waste removal and government medical camps.

The figure of the broker in Auerbach’s research lies squarely at the intersection of distributive and electoral politics. Yet contrary to extant accounts of broker-driven clientelism (for example, Stokes et al., 2013), brokers are not simply the foot-soldiers of political machines, but rather individuals who hold dual roles as community leaders and party workers. As problem-solvers, representatives and advocates for a client population within the slum, they engage with the state on their behalf, yet as operatives for political parties, they remain responsible for mobilising voters for municipal, state and national elections, as well as organising rallies and associated activities. Brokers themselves establish themselves as slum leaders first, emerging from settlements through a local reputation for solving individual problems and community-organising, and are only then incorporated into partisan structures when party officials recognise their growing influence (Auerbach, 2020, chapter 4).

The book also presents a detailed picture of the internal organisational structure of the Congress and the BJP in Bhopal and Jaipur and explains the place of successful brokers in these party organisations; if successful, they can be appointed to sought-after ward-level committees in addition to cells and wings, such as for women and young people. To achieve and maintain these positions and even climb up hierarchies, partisan brokers compete intensely with one another for a large following by providing services to clients. Auerbach argues that this serves as a monitoring and sanctioning mechanism used by parties from above and communities from below to keep brokers honest and efficacious (2020, pp. 68–73).

There are some lingering questions about the scope of activities that partisan brokers must undertake within the context of the political economy of rent-seeking in Indian cities. Partisan actors not only mobilise votes and (perhaps grudgingly) provide services but also engage with the seamy underbelly of urban politics, replete with contractors and construction companies, teamsters, illicit finance, mafia and criminal networks and groups that engage in violence (Berenschot, 2011; Vaishnav, 2017). Sinha and Wyatt (2019) record that more than a quarter of MPs after the 2019 elections have continuing careers in business, and brokers seeking advancement might well seek to get involved in the lucrative but extra-institutional activities pursued by such ‘entrepreneurial politicians’. While slum leaders may be very rooted in their communities and preoccupied with the concerns of their citizens, connections to these wider machine dynamics, particularly as they ascend rungs in the party hierarchy and seek to gain materially from politics, would only be natural.

But on its own terms, this represents a nuanced and authoritative account of the mechanics of electoral mobilisation and patronage on the ground in slum settlements, highlighting brokers as the lynchpin that connects these two forms of political activity. Auerbach, perhaps more than Bussell and Kruks-Wisner, provides solid empirical support to claims about clientelism in general, and the EVPP model in particular, because brokers, if not the parties they work for, tend to engage with both coethnic mobilisation and public goods distribution. Yet the book also emphasises conceptual distinctions that are awkward for any straightforward account of targeting discretionary resources only for electoral purposes. First, following Kruks-Wisner, the book demonstrates the significant forms of political agency on the part of slum residents in constructing and controlling political networks of varying strengths and thus engaging with the state. These mechanisms clash with a framing of clientelism as top-down and brokers as perfect agents for the targeting of goods and services on behalf of parties.

Second, Auerbach highlights important analytical differences in the timing and scale of the distributive and the electoral aspects of brokers’ responsibilities:

the slow grind of these efforts [of routine forms of claim-making] – mobilizing residents, writing petitions and navigating state institutions – differs from the episodic, top-down acts of vote-buying. This is not to say that election time party politics do not inform the allocation of public services. Instead, it is to argue that many of the most pressing problems that face slum residents do not wait for elections to present themselves (Auerbach, 2020, p. 25).

This frames distributive politics as constituting a discrete set of activities, by a particular set of actors, to claim, argue, contend or compete for public goods and services. These politics unfold at very local levels at a constant, quotidian rhythm, and as a result, have tenuous and contingent connections to the grander but more episodic dramas of electoral mobilisation.

Distributive politics and electoral politics?

The focus on the micro-dynamics and internal mechanisms of distribution thus complicates assumed theories of the impact of patronage provision on electoral outcomes, and the role of caste in providing informational cues to allow citizens to vote in order to maximise their receipt of such patronage. The books discussed above survey multiple ways of accessing the distributive aspects of the state, but they identify more sustained action at the local level than the casting of a ballot towards a party with high proportion of co-ethics; if one urgently needs a ration-card or a handpump, the polling booth is not the first place one would look.

This challenges the putative importance of electoral mobilisation on distributive politics. So should it also complicate the presumed importance of patronage distribution on electoral outcomes? There are reasons to think it might. Scholars are starting to bring some of the insights of research on public distribution to bear on questions of electoral competition. A wide-ranging review of recent literature, coauthored by Auerbach and Bussell as well as ten others, highlights how emerging research undermines some of the key tropes of the EVPP model (Auerbach et al., 2021). They suggest that distributive politics are not dominated by partisanship and that any quid-pro-quo of discretionary goods for votes has little empirical support even in elections, leading ‘recent studies of Indian politics [to] pivot away from studying episodic forms of ‘vote-buying’ toward more quotidian – and arguably more substantively important – forms of distributive politics that guide public service delivery’ (Auerbach et al., 2021, p. 13). A focus on distribution in the specific context of electoral campaigns signals a significant retreat from the EVPP model, which does not assume such scope conditions. The paper also argues that there is weak empirical support for instrumental voting primarily on the basis of ethnicity, with political operatives seeking to establish cross-ethnic support and intra-ethnic political preferences varying substantially across time and space (Auerbach et al., 2021, pp. 21–23). These together seriously erode two foundational assumptions of the EVPP model.

This paper and the broader scholarship on distributive politics tends to assume that distributive politics continues to have electoral relevance, however. To account for this, distributivists have returned to a commonsensical notion of the efficacy of politicians in performing key services for their constituents. This suggests that voters would reward legislators who are more successful in arranging the provision of publicly-provided goods. Bussell’s (2018) conceptualisation of constituency service is especially invested in this notion.

Yet there are some reasons to be cautious about alternative explanations of voting based on performance as a key driver of the vote. First, much of the research in support of efficacy has been conducted at the lowest levels of government in India, where partisanship may be more muted, officials live among their constituents and monitoring of their efforts and related outcomes is thus possible (Auerbach & Thachil, 2020; Dunning & Nilekani, 2013). Research set in urban areas, particularly among slum populations, might also perhaps overemphasise the extent to which public goods distribution is transactional and easily monitored. Context matters; while these sites of research are ideal cases for adjudicating the influence of clientelism and efficacy, much as with studies of urban political machines in municipal governments anywhere, they have limited external validity.

Second, despite more sanguine assessments in the books above, there is substantial evidence that the vast majority of citizens simply do not receive politically salient welfare goods and services; exclusion from public distribution is the norm rather than the exception. Some of the most universal welfare schemes only reached a small proportion of potential beneficiaries, with no category of benefits reaching more than 21 per cent of the voters surveyed in the 2014 National Election Survey (Chhibber & Verma, 2018, p. 113). In a household survey in rural West Bengal, Bardhan et al. (2009, pp. 51–53) found slightly higher figures for survey respondents in terms of non-excludable benefits like paved roads and water connections, at 32 and 23 per cent respectively, but much lower figures for excludable ones, with ration cards provided to only 12 per cent of households and only 3 per cent were Integrated Rural Development Programme (IRDP) beneficiaries. This stands in stark contrast to turnout – 67 per cent in the 2014 and 2019 parliamentary elections, not less than 50 per cent in recent assembly elections – indicating that many more people cast ballots than receive discretionary benefits from which to form a conclusion of the relative performance of candidates and parties. To be sure, researchers have argued that citizens might cast their ballots based on the government’s performance not just in selective distribution but also the health of the overall economy and responses to disasters (Cole et al., 2012; Jensenius & Suryanarayan, 2021). But even these broader notions of performance still represent fundamentally non-ideological ways of voting, on which more below.

Third, Indian elections exhibit extraordinary levels of anti-incumbency. In most elections, less than half of incumbents win re-election (Verniers, 2020; Wyatt, 2013, p. 37). Further, candidate deselection is common; in recent Bihar assembly elections, slightly over half of sitting MLAs ran for a second term (Nissa et al., 2020). This level of volatility sits uneasily with the notion that legislators are rewarded for their performance either in terms of individual provision or the health of the economy as a whole.

A broader analytical problem of the distributive-electoral nexus is that it occurs at very low levels of analysis, often focusing on individual political actors or slices of the electorate in particular locations to explore micro-dynamics of electoral campaigns. This implicit redefinition of what studying electoral politics means excludes important aggregate outcomes such as which parties win elections, either alone or in coalitions, and, crucially, the nature of party competition at state and national levels. In other words, there is a significant disconnect in the objects of inquiry between research at the local level on distributive politics and the traditional emphasis on aggregate electoral outcomes in party systems. The former is singularly unsuited for understanding dramatic shifts in the nature of party competition over time, and in particular, the consolidation of a majoritarian party system under Modi’s BJP, after decades of fragmentation.

Bringing ideology back in?

Ideological sorting has long competed with ‘checkbook’ models of voting and the managerial credibility of parties in the study of party systems in developed democracies. Foundational theories of party systems argue that the instigation of fundamental social cleavages – between business and labour, but also between cities and the countryside, or church and state – give rise to parties that represent different sides of these cleavages; these provide signalling policy agendas that allow citizens to vote according to their ideological proclivities (Downs, 1957; Lipset & Rokkan, 1967). Electoral rules are also important; single-member plurality systems incentivize competition between two parties at the constituency level (Duverger, 1954).

Scholars of Indian politics have long been chary about incorporating ideology into their analysis of party politics, however, because parties are widely seen as organisationally weak and ideologically incoherent vehicles for personalist elites to gain or retain power. Ideological cleavages in particular seem less salient because India became a democracy without going through the process of modernisation and industrialisation that constructed the core social cleavages in western Europe (Varshney, 1998). This suspicion has, if anything, grown as the party system fragmented in the 1990s. Chandra (2004, chapter 2) essentially argued that ethnicity stood in for ideological issue positions and policies as signals for voters seeking to maximise their material interests.

Yet the EVPP model has been somewhat less capable of explaining the volatility of the party system, because electoral outcomes shift more often and more dramatically than the rise of parties catering for underrepresented ethnic groups, let alone caste identities themselves. It has even more limited capacities for explaining the remarkable success of Modi’s BJP in fashioning a much more consolidated party system, reversing the general trends toward fragmentation, even though it might lack some of the features of previous dominant party systems in India (Ziegfeld, 2020). Clientelism itself connotes a stable, steady-state system that could only be disrupted by exogenous shocks, often leading to greater fragmentation; it has less facility to accommodate endogenous change towards system convergence.

Chhibber and Verma (2018) address these questions by making an affirmative case for the importance of ideological cleavages in explaining party system dynamics in India. Crucially, they argue that in India, cleavages do not straightforwardly resemble the left-right spectrum of class-based competition in western Europe, but rather ideologically divide the electorate over the politics of statism and the politics of recognition, or ‘the extent to which the state should dominate society, regulate social norms, and redistribute private property … and how the state should accommodate the needs of various marginalized groups and protect minority rights from assertive majoritarian tendencies’ (Chhibber & Verma, 2018, p. 2). They show that even parties (and their activists and voters) that are putatively quite similar – such as the DMK and the AIADMK in Tamil Nadu – do in fact align and compete along these cleavages.

Through this framework, Chhibber and Verma provide an explanatory account of the decline of Congress and the rise of the BJP under Modi. They do so by exploring how Congress attempted to straddle cleavages while increasingly relying on the distribution of patronage, thus allowing oppositions at the centre and the states to consolidate animating ideological positions. Under Modi, the BJP has constructed a powerful electoral coalition with greater ideological consistency against state intervention in society and the recognition of particular social groups. Thus, the victory of the BJP in 2014 (and in the 2019 elections, after the book’s publication) was a case of partisan realignment through ideological mobilisation.

Innovative as it is, the Chhibber and Verma framework is underspecified in a number of respects. First, the inductive nature of this bidimensional cleavage arises from a series of questions in India’s National Election Survey, a long questionnaire that naturally cannot represent salience or strength of conviction. This is inevitable with large-scale survey work, and is usefully supplemented by survey experiments, but the book lacks other complementary approaches, such as interviews and focus groups, that might have been used to explore the salience of these categories and their roots in individual worldviews. Second, the historical provenance of these cleavages might be debated; Chhibber and Verma situate them in the dynamics of state formation around the framing of the Constitution, though reflecting older traditions in ancient and medieval Indian political thought. But other historical perspectives might highlight disagreements within the nationalist movement in the early twentieth century, at a time when the colonial state increased intervention in economy and society and expanded institutions of participation (Erdman, 1967; Naseemullah, 2017). Finally, the state can create ideological cleavages based on its coercive as well as its infrastructural avatar, with different partisan implications; Malik (2020) persuasively argues that the persistence of support for the Muttahida Quami Movement (MQM) among the Mohajir community in Karachi, despite deeply underwhelming social provision, is due to feelings of defiant solidarity arising from their discriminatory treatment by the security services. These dynamics may be operable in Indian contexts of systemic state repression and violence as well, from Kashmir to the regions of eastern and central India impacted by the Maoist insurgency.

But ideological explanations such as that of Chhibber and Verma provide a theoretical and empirical grounding to a long and storied tradition of analysing Indian party systems, which focus on the patterns of aggregate outcomes across thousands of constituencies at the parliamentary and state assembly level (Chhibber & Kollman, 2004; Kothari, 1964; Yadav & Palshikar, 2003). They do not easily engage with micro-level electoral accounts at the level of the gram panchayat or the municipal ward, the sites where much of the contemporary research on the electoral implications of distributive politics are located. Nuances at the very local level are lost when analysis occurs at the level of the party system, but equally, aggregate patterns of the party system are invisible when focusing on politics at its most local. Both may be important scales of enquiry, but it is possible that research into the granular dynamics of distributive politics – brokers and claim-making, especially – may at once be important, but teach us very little about electoral politics. It is only clientelistic models such as EVPP that force the two to engage with one another, and this link has been subject to more rigorous empirical scrutiny by the current generation of scholarship in distributive politics, as well as alternative party system accounts.

Conclusion

One of social science’s deep mysteries is what occurs in the mind of a voter as she approaches the polling booth, waits in line, presses a button and has her finger stained with ink. We are left with inferences arising from various longstanding theories of political psychology and behaviour, which we attempt to test by means of the aggregate analysis of outcomes, pre- and post-election surveys and field experiments, interviews and the like. The theoretical frameworks of Indian politics were first bound up in the narrative of modernisation, but then instrumentalist frameworks became dominant, emphasising the material benefits that citizens might derive in exchange for their votes. This consolidated a theory of clientelism that brought elections and the distribution of public goods and welfare together, and highlighted ethnicity as a key informational mechanism that could facilitate this transaction, or what I have called the EVPP model. Recent scholarship in distributive politics has undercut the foundations of this model, and Chhibber and Verma have introduced an alternative paradigm that challenges any intimate link among ethnic, electoral and distributive politics.

Can ethnicity, patronage and ideology be reconciled as causal factors in the study of Indian politics? After all, it is undeniable that Indian society is saturated with patronage, that elections involve the significant if spasmodic distribution of goods, and that caste- based mobilisation is a powerful force in Indian elections, particularly in the populous states of north India. Chhibber and Verma suggest one possibility of integrating ideology, ethnicity and patronage: that voters might vote according to ethnic and/or clientelistic patterns when their partisan options are ideologically weak or incoherent, but that coherent and assertive ideological positions trump ethnic and clientelistic appeals, particularly in contexts of realignment. But what really stands in the way of a reconciliation is a framing of identity as purely instrumental and material distribution as purely transactional and utility-maximising. This leaves out the multiple ways that Indians might understand their material lives, which is deeply implicated in their ethnic identities, in political terms.

This alternative socioeconomic framing of political action, including but not limited to the vote, is not alien to the literature: Auerbach et al. (2021, pp. 24–25) call for more attention to be paid to the intersection of class and ethnicity and Ahuja and Chhibber (2012) have powerfully argued that different socioeconomic groups understand the meaning of the vote differently, with the poor casting ballots as a means of asserting their formal equality as citizens to an otherwise deeply inequitable state, rather than for concrete benefits. But these insights in electoral contexts have not yet been fully combined with a deep appreciation of the structures of inequity, exclusion and social domination that pervade India’s political economy, which go far beyond public distribution of goods by the state, and which are felt most keenly by subaltern groups. For Indians – especially the poor – there are particular political meanings attached to caste identity and to their understanding of how resources flow through the state and the broader economy to benefit the wealthy and well-connected; these meanings inform political action, from contentious politics to the vote. Ethnicity, still-powerful hierarchical structures of wealth and power and the flow of material resources from the state, together with the coordinating mechanisms of ideology, thus work together to form electoral choices and outcomes, but just not as obviously in the individual, transactional manner that is predicted by theories of clientelism.

Where we go from here in research terms depends a lot on how crucial we think it is for distributive and electoral politics to speak to one another directly. One eminently plausible approach would be to separate the two entirely as objects of enquiry. Distributive politics is at its best when examining the social and political factors behind systems of public distribution at very specific locations, be it slum communities, rural villages or constituency offices. That specificity is laudable, particularly in uncovering mechanisms, interests and incentives that might be of help in designing policy reforms. But it does hazard generalizability, which is crucial when we seek explanations for the outcomes other than the specific micro-mechanisms of electoral mobilisation by brokers and constituency candidates. Further, distributive politics portray a landscape filled with so many different actors and channels of entry to the state; parties and elected officials are frequently minor players standing in the background, at least until elections, and privileging these players risks underemphasising the most important aspects of the politics behind distribution. Electoral politics, by contrast, cannot solely operate at the local level because they are concerned with aggregate patterns across space and over time; here, patronage must compete on a level empirical playing field with other causal factors, from electoral institutions to party governance to ideological cleavages to exogenous events, in explaining how votes translate into seats and how seats might translate into party systems.

Disclosure statement

No potential conflict of interest was reported by the author(s).