8.1 Introduction

Part II presented evidence for my theoretical argument based on multilevel comparative field research in northern India. The analysis showed that the historical divergence of bureaucratic norms has contributed to subnational differences in the implementation of primary schooling in four Hindi belt states. Part III extends the book’s argument to other empirical contexts and explores its implications for scholarship and policy. This chapter considers the plausibility of my argument beyond northern India. The four cases examined are the south Indian state of Kerala, along with China, Finland and France. These cases were chosen for theoretical reasons, to test different aspects of my theoretical argument in very different political economy contexts. I have a relied on a close reading of secondary literature examining the education system, bureaucracy and implementation processes in each setting.

Before proceeding with the analysis, some caveats are necessary. The abridged discussion of cases here is not intended to substitute for careful empirical investigation. I have no desire to extend my argument beyond what can be demonstrated empirically. Indeed, my objective in this book has been to signal the painstaking attention to research design and data collection that is needed to study bureaucratic norms and trace comparative their effects on implementation across different sociopolitical contexts. My aims here are therefore more modest. I seek to establish the plausibility of my theory outside northern India and to suggest directions for future research. At first glance, the inclusion of country-level cases may seem odd given the book’s focus on subnational variation across Indian states. Following Singh (Reference Singh2017), I intend to show how a theoretical framework developed through subnational comparative research can yield insights for comparative scholarship on national systems of education and social welfare provision, with an eye toward fostering dialogue between these approaches. To be sure, the discussion of country-level cases overlooks large, within-country variations and historical nuances. The literature that I reference accentuates the central tendencies of bureaucracy in each country, but pays less attention to subnational differences in bureaucratic norms.

This chapter’s discussion progresses as follows: I first examine Kerala, a recognized leader in education and health that has outperformed other Indian states, attaining social indicators that rival middle-income countries (Lieten Reference Lieten2002; Singh Reference Singh2011). Kerala has a history of social movements by subordinate castes and classes, along with a comparatively successful project of democratic decentralization (Heller Reference Heller1999; Isaac and Heller Reference Isaac, Heller, Fung and Wright2003). Less appreciated, I suggest, is how Kerala’s deliberative bureaucracy has helped strengthen community participation in public services. Next, I investigate China, a country that defies social scientific thinking on institutions and development. Scholars have investigated the Chinese state’s capacity for implementation through mechanisms of “adaptive governance” (Heilmann and Perry Reference Heilmann and Perry2011) and “directed improvisation” (Ang Reference Ang2016a). Building on their work, I argue that China offers theoretical lessons for how deliberative bureaucracy shapes social policy implementation in the absence of democracy. Third, I extend my argument to advanced economies by comparing the performance of school education in France and Finland, which have similar levels of public spending on education, but divergent outcomes. Finland’s school system has gained prominence among OECD countries for producing superior learning outcomes that are more equitably distributed across socioeconomic groups. France’s school system, meanwhile, produces learning outcomes closer to the OECD average, with relatively uneven outcomes between socioeconomic groups. Drawing on experimental governance and public administration scholarship (Sabel et al. Reference Sabel, Saxenian, Miettinen, Kristensen and Hautamäki2011; Peters Reference Peters2021), I suggest that deliberative and legalistic bureaucratic norms have generated distinct policy implementation patterns across these countries, contributing to the varied performance of their school systems.

8.2 Deliberative Bureaucracy with Social Movements: Revisiting the “Kerala Model” of Social Development

The southern coastal state of Kerala (pop. 33.4 million) has earned international acclaim for being a model of social development at a relatively low level of income. Many researchers have been struck by the differences in public service delivery and social welfare outcomes in Kerala and Hindi belt states (Drèze and Sen Reference Drèze and Sen2002; Singh Reference Singh2015). The comparison has helped accentuate the distinctiveness of the “Kerala model” of social development (Lieten Reference Lieten2002). Kerala has also had distinctive historical conditions and sociopolitical features that are not replicated in northern India. In 1951, Kerala’s literacy rate of 47 percent was already more than twice the all-India average of 18 percent. Kerala maintained its lead, and by the 2011 Census, recorded a literacy rate of 94 percent. The rural–urban divide in primary school access and enrollment was closed in the 1990s, long before the central government launched Sarva Shiksha Abhiyan (SSA). In 1999, more than 90 percent of rural Keralites lived in villages served by a primary school and 87 percent had access to middle school, against national averages of 80 percent and 45 percent, respectively (NFHS-2 2000). By then, Kerala’s enrollment rate for girls and boys (ages 6–17) was on par at 91 percent, compared to 61.4 percent for girls and 77.3 percent for boys at the national level. Kerala’s primary school system is found to produce superior learning outcomes, as 65.2 percent of rural fifth graders have a demonstrated ability to read and 45.9 percent can perform simple division, well above the national averages of 41.7 percent and 24.8 percent (ASER 2013).

To explain Kerala’s superior performance in mass education and other social programs, scholars have pointed to the historical importance of social movements by subordinate castes and classes, reinforced by leftist government and a robust civil society (Heller Reference Heller1999), along with the solidaristic ties of Malayali subnationalism (Singh Reference Singh2016). Missing from existing accounts is an appreciation for how bureaucratic norms in Kerala have evolved to support social policy implementation. Based on historical and secondary sources, I suggest that state-building processes in Kerala promoted deliberation between bureaucracy and societal actors, which has in turn supported Kerala’s superior educational performance. In contrast to the politics of elite cooperation in HP, deliberative governance in Kerala was forged through a combination of elite cooperation and societal conflict. By supporting local deliberation and citizen participation in service delivery, deliberative bureaucracy in Kerala has helped institutionalize conflict and steer it toward programmatic ends (Mansuri and Rao Reference Mansuri and Rao2013, 90–91).

At first glance, the superior performance of primary schooling in Kerala can be attributed to multiple factors. Kerala’s coastal geography and settlement patterns are more conducive to public service provision, allowing for maritime trade and financial remittances from migrants. Kerala is also among India’s most densely settled states (Bhat and Rajan Reference Bhat and Rajan1990). The 2011 Census shows that more than 92 percent of rural Keralites reside in villages with populations exceeding 10,000, compared to only 8.7 percent for the rest of India. Higher rural population density lessens administrative costs and raises political incentives for providing education, mitigating the urban bias in service provision. Also, the historical beneficence of Kerala’s political leadership encouraged more generous public spending. The princely states of Travancore and Cochin had socially progressive leadership supporting mass education. In 1817, Gowri Parvathi Bai, Queen of Travancore, declared, “The state should defray the entire cost of the education of its people,” a bold proclamation for any government at the time (Ramachandran Reference Ramachandran, Drèze and Sen1997, 268). Regrettably, the policy statements of progressive leaders were not always matched with attention to implementation (Sen Reference Sen, Wuyts, Mackintosh and Hewitt1992). Public funding for mass education reached notable levels by the nineteenth century’s close, but gender-and caste-based exclusion limited societal participation (Mathew Reference Mathew1999).

Another oft-cited explanation is that Christian missionaries introduced English-medium education and opened formal schooling to women and lower castes, targeting marginalized groups (Tharakan Reference Tharakan1984; Sen Reference Sen, Wuyts, Mackintosh and Hewitt1992; Mathew Reference Mathew1999). Missionaries in Travancore, for example, set up boarding schools for girls and children of lower-caste slaves, despite protests by upper-caste landlords (Mathew Reference Mathew1999, 2814–2815). Although the direct impact of missionaries on mass education was diluted by their limited geographic presence, they were pivotal in bringing caste injustices to light, creating the ideological space for social reform movements (Singh Reference Singh2011, 284). The Christian message of equality, combined with fears of mass conversion, provoked religious competition from upper-caste Hindus. An historical study of social welfare in Travancore found that “the state was under considerable pressure from upper caste groups to open their own Hindu schools, for the missionary schools quite rightly were viewed as venues for the conversion of lower castes to Christianity” (Desai Reference Desai2005, 472).

A discussion of mass education in Kerala must give consideration to its legacy of bottom-up social movements challenging caste and class hierarchies, a defining feature of the state’s political development (Jeffrey Reference Jeffrey1975; Herring Reference Herring1983; Menon Reference Menon1994; Heller Reference Heller1999). Kerala’s elaborate caste system was one of the most oppressive ones in India. All castes below the Namboodiris, the highest-ranking Brahmin subcaste, were considered polluted and had restrictions imposed on their physical movement. The lowest castes suffered extreme humiliations; along with untouchability, these included “unapproachability,” which made the mere act of being seen by upper castes a punishable offense (Nossiter Reference Nossiter1982, 25–27). Slavery of Dalits was common as well in the colonial period, and it persisted in some forms, such as bonded labor, even after independence (Sivanandan Reference Sivanandan1976).

In the backdrop of caste oppression, mass movements challenging hierarchy and discrimination arose in the late nineteenth century. These movements raised political awareness among lower castes, who learned to articulate injustices to the state (Sen Reference Sen, Wuyts, Mackintosh and Hewitt1992, 263). An example is the Ezhavas, a lower caste that constitutes 22 percent of Kerala’s population. Denied entry into Hindu temples, the Ezhavas were discriminated against in schooling and employment. Upper-caste teachers and students boycotted classes to avoid making contact with Ezhava students (Mathew Reference Mathew1999, 2817). An Ezhava social movement arose in 1903 with the creation of the Sree Narayana Dharma Paripalan Yogam (SNDP), Kerala’s first caste association. SNDP promoted religious and secular education, while also working to secure Ezhavas access to Hindu temples (Jeffrey Reference Jeffrey1975). The movement stressed the importance of literacy and economic self-help. Ezhavas also learned to demand their rights from the state, establishing the link between access to formal schooling and social mobility.

If pre-independence social reform movements stimulated political demands from below, the communists in Kerala aggregated these demands into a coherent, class-based political movement. The world’s first democratically elected communist government, the Communist Party of India secured a majority of assembly seats in the 1956 state elections. More than simply collecting votes, Kerala communists operated as a “social movement party,” Heller writes, “occupying the trenches of civil society, building mass-based organizations, ratcheting up demands, and cultivating a noisy but effective politics of contention” (2000, 511). The CPM government led by E.M.S. Namboodiripad adopted a pro-poor policy platform, which included agrarian land and labor reforms. Kerala communists institutionalized societal demands in class-based terms and directed them toward the state, offsetting the destabilizing nature of class conflict. In addition to class movements, Singh (Reference Singh2016) demonstrates that a shared Malayali subnational identity motivated political elites to invest in social policies, while celebrating Kerala’s distinctive language and cultural traditions. For example, subnationalism was promoted by political elites during the Aikya Kerala (United Kerala) movement, which secured statehood for Kerala under the central government’s 1956 States Reorganisation Act. Malayali subnationalism, Singh argues, also triggered increased public spending on primary education and health. Faced with a common external enemy (non-Malayali Brahmins), political elites formed a cross-caste coalition, and in turn, were motivated by subnationalism to prioritize social welfare.

If state programming for primary education was a product of Kerala’s class-based mobilization and subnationalist politics, how services reached citizens effectively requires further explanation. Kerala’s bureaucracy is institutionally similar to those in other Indian states. Indian Administrative Service (IAS) officers occupying senior posts in Kerala’s administration are recruited from across India. Under service deployment rules, the vast majority of IAS officers belong to other states and are inculcated with a national outlook through training and promotion processes, which, by design, reduce the scope for subnational identification with a particular province. It is similarly unclear why class mobilization in Kerala led to participatory governance arrangements for primary education. The northeastern state of West Bengal, which experienced uninterrupted leftist government rule for over thirty years, has also been effective at poverty alleviation (Kohli Reference Kohli1987; Echeverri-Gent Reference Echeverri-Gent1992). However, its performance in primary education and health has remained abysmal (Mullen Reference Mullen2011). West Bengal’s Communist Party of India (Marxist) (CPM) used the bureaucratic machinery to cement its electoral dominance, stifling political competition and dissent from civil society (Desai Reference Desai2001). West Bengal also scores highly on the subnationalism scale, but the bonds of Bengali identity have not yielded superior education and health outcomes.

A fuller understanding of Kerala’s education performance requires an examination of the norms that emerged within the bureaucracy, and specifically, how state agencies learned to work with societal actors to provide services. As a first step, we must consider the historical state-building process, which shaped the nature of school governance in Kerala. A distinctive feature of state formation in Kerala, bureaucratic norms emerged in a setting in which non-state provision of social services was substantial. Historically, private provisioning comprised a larger part of Kerala’s education system. State programming was introduced only after religious authorities and lower-caste reform movements were well entrenched in the education sphere. Although the number of private schools was insufficient to guarantee universal coverage, the presence of multiple, competing non-state service providers allowed the state to experiment with mixed forms of public and private governance. In the princely state of Travancore, for example, a significant share of public expenditure on education went toward subsidies for private schools run by Christian missionaries and Hindu caste societies. As Tharakan (Reference Tharakan1984) observes, missionary education depended on state grants-in-aid to private schools. The Queen of Travancore’s 1817 proclamation that the state would bear the cost of education carried an important rejoinder that granted the state control over school management. In settings where rural education catered to different castes and religious denominations, centralized political control of schooling was highly controversial. Religious authorities were wary of administrative intrusion by the state in the affairs of schools. The state’s response to the concerns of private school bodies was remarkable. As an historian of Kerala has noted, the Travancore government took the decision not to interfere with existing school management structures; traditional headmasters were instead incorporated into the government school system (Jeffrey Reference Jeffrey1987). “By preserving the old schools,” Jeffrey observes, “the Travancore government retained the trust of elders willing to send their charges to institutions run by known and respected masters” (1987, 453). The incorporation of traditional schools in Travancore contrasted with the approach taken by the princely state of Cochin. In the 1890s, the government of Cochin adopted a strict regulatory stance toward the traditional school system, which fell into decline, contributing to a deterioration in literacy.

During the first half of the twentieth century, Kerala’s bureaucracy engaged in both cooperation and conflict with private school authorities. The assertion of state control in 1910 led to a drop in the number of private schools, while the subsequent relaxation of state control saw the swift reappearance of private schools (Lieten Reference Lieten1977, 6–7). Conflict was intense between the bureaucracy and Kerala’s church authorities, who competed for institutional oversight of schooling. By the late 1950s, Kerala had 9,500 primary and secondary schools, more than 60 percent of which were privately managed. As Jeffrey observes, “Nearly 70 per cent of corporate private managements were Christian (though schools run by individuals were overwhelmingly Hindu)” (1992, 155). Government primary schools in turn expanded in the 1960s, doubling in number. Nevertheless, by 1970–1971, the percentage of privately managed schools remained high, at 61.7 percent (Lieten Reference Lieten2002, 57). Given large public outlays on private schooling, the state kept close tabs on how church authorities ran the school system. The bureaucracy sought to instill uniform standards and professionalize teaching, thought it also needed buy-in from private schools.

Alongside the bureaucracy, the CPM sought to exert control over primary schooling, in part to secure ideological dominance over schooling curricula. Church authorities were suspicious of communist political interference. In 1957, Kerala’s communist-led government proposed a compromise education bill, which granted private management control of day-to-day administration while retaining state control over school inspections and curricular supervision, the latter responsibilities handed to the education department. The 1958 Kerala Education Act was adopted by the state legislature, albeit with an important amendment acknowledging the protection of minority (i.e. Christian and Muslim) rights under the Constitution. The Act’s provisions applied only to government schools and private-aided schools that received state financial support. An important stipulation was that private-aided schools were required to hire teachers from a government-approved roster. Teachers were paid directly by the government, a provision that aimed to protect teachers from unscrupulous school management. Non-aided private schools, those opting out of state financial support, continued to operate autonomously while having state recognition.

The Kerala Education Act bred conflict between the state and church authorities. Church leaders, who feared the spread of secular education, lodged protests against proposed revisions to textbooks. They were joined by members of the Kerala Provincial Congress Committee (KPCC), who argued that the Act was an attempt to spread communist ideology (Lieten Reference Lieten1977, 10). Controversy surrounding the Act may have contributed to the downfall of the CPM-led government in 1959. Nevertheless, the education policy withstood multiple changes of government. The bureaucracy continued to oversee private schools, which in practical terms meant that officials had to learn how to coordinate with religious authorities and community associations, whose interests at times conflicted with those of the state. These experiences of working with non-state agencies impacted bureaucracy’s orientation toward societal participation but also paved the way for reforms granting local communities greater voice in school management.

The bureaucracy’s catalytic role in encouraging societal participation in service delivery is equally apparent from Kerala’s more recent decentralization reforms. Following the 73rd and 74th constitutional amendments, Kerala has led India in establishing Panchayati Raj Institutions and devolving funds and substantive decision-making authority to village councils. Kerala’s panchayat reforms were spearheaded by the communist government in 1996, which came to power on the pledge to decentralize state authority.Footnote ¹ As Isaac and Heller (Reference Isaac, Heller, Fung and Wright2003) find, decentralization initiatives also depended critically on the bureaucracy, which worked with non-state agencies to strengthen societal involvement in state planning. As they argue, “The Kerala State Planning board formulated, designed and drove the Campaign for Democratic Decentralization. In doing so, the Board has relied on a stock of practical knowledge, ideas and experiences drawn from twenty-five years of local-level experiments conducted by NGOs” (Isaac and Heller Reference Isaac, Heller, Fung and Wright2003, 80). Studies also indicate that the Kerala Sastra Sahitya Parishad (KSSP) – the People’s Science Movement – helped the state integrate local knowledge for planning purposes (Isaac and Franke Reference Isaac and Franke2000; Véron Reference Véron2001). A grassroots organization created in the 1960s by retired teachers and other professionals, KSSP grew into a mass “people’s movement” for scientific thinking. The group supported the National Literacy Mission during the 1980s and 1990s and worked with local administration on the Total Literacy Campaign in Kerala’s Ernakulam district.Footnote ² These efforts earned Ernakulam the distinction of being India’s first “fully literate” district and a national model of state–society collaboration (Rao Reference Rao1993; Tharakan Reference Tharakan and Karlekar2004). In his analysis of Kerala’s literacy campaign, Tharakan notes, “In Ernakulam there remains an unquantifiable factor regarding the role played by the district administration and its personnel which went far beyond the call of duty. The attitudinal change that resulted in such participation has certainly persisted in the administration, leading to better results in the delivery of welfare and development programmes” (2004, 77). The bureaucracy’s ability to learn from local experiments helps account for how the Ernakulam experiment was later scaled up throughout Kerala.

Kerala’s bureaucracy institutionalized KSSP’s role with the implementation of panchayat reforms. As Heller notes, “The presence of a KSSP cell within the Planning Board has helped embed this state agency in social movement structures,” creating “new nonpartisan institutions of participation that build on civic associationalism” (2001, 154). Through its network of more than 6,000 volunteers, KSSP has provided technical support and assistance with fiscal decentralization.Footnote ³ The state bureaucracy undertook an elaborate training program, covering thousands of state and local officials, described as “one of the largest non-formal education programs ever undertaken in India” (Isaac and Heller Reference Isaac, Heller, Fung and Wright2003, 83). Consequently, Kerala’s panchayats are more likely to have functional village assemblies (gram sabhas) and draw high levels of citizen participation (Besley, Pande and Rao Reference Besley, Pande and Rao2007). Bureaucratic deliberation has thus made democratic decentralization comparatively deeper in Kerala.

This section revisited the Kerala model in light of my theoretical argument. The discussion pointed to the political process of state-building in a context of private schooling, religious pluralism and routinized conflict between government and religious authorities. The Kerala bureaucracy’s historical engagement with non-state agencies enabled joint management of schools. In a setting of social movements by subordinate groups, two-party competition and leftist government, the state bureaucracy’s policy experimentation during literacy campaigns, followed by decentralization in the 1990s, helped consolidate deliberative ties between local administration and civic bodies. Bureaucratic norms promoting inclusive deliberation, incorporating marginalized social groups in local governance, helped sustain the efforts of KSSP and other civil society groups. Deliberation between state and society has thus improved the quality of education services in Kerala and enhanced bureaucratic responsiveness to the needs of the least advantaged.

8.3 China: How Deliberative Bureaucracy Works in a Non-Democracy

In contrast to Kerala, the case of China raises questions as to whether and how deliberative bureaucracy works in the absence of democracy. Few countries have challenged mainstream social scientific thinking on institutions and development as much as China has. China’s embrace of global capitalism sans conventional rule of law and Weberian state institutions continues to puzzle scholars. Research comparing China and India has highlighted differences in the management and growth of their economies. China’s more open and dynamic market, steered by an organized party-state, has been contrasted with India’s fractious democratic federation and more rigid “License Raj” bureaucracy (Bardhan Reference Bardhan2010; Chibber Reference Chibber2003). When India gave up its autarkic, planned economy in the 1990s, China had already become a global economic powerhouse (Mukherji Reference Mukherji2014). However, the comparative human development performance of the two countries has not received due attention. China has gone far ahead of India in improving the quality of life for its people, arresting rural poverty, illiteracy and infant mortality (Sen Reference Sen2011).

The comparison between India and China in the primary education arena is particularly instructive. In the 1950s, the countries had comparable levels of education, with India’s literacy rate of 18 percent slightly below China’s rate of 20 percent. However, China’s education expansion happened at a much faster pace in the next three decades. By 1980, India’s literacy rate of 40 percent was surpassed by China’s rate of 65.5 percent (World Bank 2020). China achieved universal enrollment in primary schooling in the 1970s, while it took India until the early 2000s. China also made significant progress in tackling education inequalities. In 1981, the ratio of rural to urban literacy in India was 54 (per 100), rising to 61 in 1991. Over that period, China’s ratio climbed from 76 to 90 (Drèze and Loh Reference Drèze and Loh1995). Similarly, the ratio of female to male literacy in China was 78 (per 100) in 1991; in India it was only 61. By 2018, China’s gender gap in literacy vanished and India’s ratio improved to 85 (World Bank Group 2021). China is now at the cusp of having full adult literacy. Meanwhile, India’s road to catching up remains a long one.

China’s superior education performance may seem puzzling given its authoritarian political system. Leading theories of political economy expect democratization to encourage redistribution in favor of the poor (Meltzer and Richard Reference Meltzer and Richard1981; Boix Reference Boix2003; Acemoglu and Robinson Reference Acemoglu and Robinson2006). Numerous studies associate democratic regimes and political competition with increased public spending on primary schooling, as well as greater enrollment and other positive outcomes (Brown Reference Brown1999; Lake and Baum Reference Lake and Baum2001; Brown and Hunter Reference Brown and Hunter2004; Stasavage Reference Stasavage2005; Ansell Reference Ansell2010).Footnote ⁴ On the other hand, historical examples abound of non-democratic regimes that provided mass education, from Meiji Japan to modern Turkey. A study of 109 countries over a 200-year period finds that democracy produces little or no effect on primary school enrollment, and that in most cases, state provision of primary schooling was introduced prior to democratization (Paglayan Reference Paglayan2021). There can be multiple motivations for non-democracies to promote mass education. Not all of these motivations are related to human capital development, such as extending state control over the political beliefs and ideologies held by citizens (Pritchett Reference Pritchett2003). Setting aside the various political purposes for non-democracies to provide primary education, questions remain as to how these policies get implemented.

There are many potential factors behind China’s superior implementation of primary schooling, some continuing since the Maoist period, if not earlier in history. An underappreciated factor, I suggest, is the presence of deliberative bureaucracy in China. Political philosophers associate deliberation with democratic forms of government and substantive citizen participation in decision-making (Cohen Reference Cohen, Hamlin and Pettit1989). China may appear, then, an unlikely place for deliberative bureaucracy to flourish. Yet, deliberation has historical importance within China’s party-state.Footnote ⁵ Central and provincial elites in the Chinese Communist Party (CCP) exert authority and top-down hierarchical control over governmental decisions. However, they have also been open to persuasion and have responded to policy ideas and experiences from below, including from local deliberative fora. He and Warren (Reference He and Warren2011) theorize the concept of “authoritarian deliberation” and trace its lineage within China’s political development. In authoritarian deliberation, “powers of decision are concentrated, but power holders enable communicative contexts that generate influence (responsiveness to claims and reasons) among the participants. Power holders are influenced in their decisions by the reasons generated by communication among participants and/or by the legitimacy of the process of reason-giving” (He and Warren Reference He and Warren2011, 274). Functionalist accounts suggest that deliberative mechanisms provide state officials information about public sentiments, facilitating negotiation with society, as well as legitimacy for policy decisions. Deliberation encourages “willing compliance” among citizens for state policies, lessening the costs of authoritarian control (He and Warren Reference He and Warren2011, 280).Footnote ⁶ Deliberation with societal groups during public service delivery has also enabled the Chinese state to make inroads into civil society.Footnote ⁷

Deliberative politics was institutionalized in China in the 1980s with the enactment of decentralization, village elections and other participatory reforms.Footnote ⁸ However, deliberative practices have an older pedigree in Chinese political thought and were later adopted by Mao, who emphasized “learning from the people through direct engagement with their conditions and struggles” (He and Warren Reference He and Warren2011, 276).Footnote ⁹ Along with “deliberation,” the concepts of “pragmatism,” “experimentalism” and “adaptation” appear in the scholarly lexicon of Chinese governance. Heilmann and Perry (Reference Heilmann and Perry2011) explore the Chinese party-state’s capacity for “adaptive governance,” an ability to experiment and learn from unfamiliar and changing circumstances, which they argue has facilitated economic and social development.Footnote ¹⁰ The experimental approach to policy development in China has been associated with Mao’s embrace of pragmatism. “In line with his practice-based epistemology,” Heilmann notes, “Mao held that policy implementation, not policy debate, provided the crucial device for learning and innovation” (2008a, 8).Footnote ¹¹

Decentralized administration sets the stage for local policy experiments, which, importantly, are guided from above by national and provincial agencies (Heilmann Reference Heilmann2008b; Ang Reference Ang2016a). Despite its unitary political system, China is administratively among the most decentralized countries in the world (Landry Reference Landry2008). City, county and township governments are responsible for funding and implementing public services. Along with formal decentralization, informal bureaucratic norms promote experimentation by local officials. As Heilmann elaborates: “Most experimental efforts are set off by local policymakers … Encouragement and protection extended by senior leaders to local experimenters, a mechanism of informal ‘policy hedging,’ is a major determinant of pioneering behavior on the local level” (2008b, 9). Rather than top-down political control, senior officials guide lower-level functionaries and enable them to take risks.

The significance of informal bureaucratic norms in the Chinese state is drawn out in a compelling fashion by Ang (Reference Ang2016a), who demonstrates how state officials have improvised with society to address practical challenges of development. Ang advances a theoretical framework of “directive improvisation,” a set of behavioral norms motivating entrepreneurial behavior by bureaucrats on a range of development problems. Improvisation, she argues, is not idiosyncratic or spontaneous, but directed by senior officials, who articulate policy mandates in ways that balance national goals against subnational differences. In addition to the “hardware” of a decentralized authoritarian state, Ang emphasizes the institutional “software” facilitating adaptation, namely, administrative guidance and flexibility, as well as the rewarding of successful policy adaptation. To illustrate, she describes the puzzling behavior of Mr. Zhou, the principal of a public elementary school (Ang Reference Ang2016a, 103–104). In a display of entrepreneurship, Mr. Zhou upgraded the school’s infrastructure into a state-of-the-art facility, drawing on schoolteachers’ savings and a bank loan secured against their property. His unorthodox behavior broke with administrative rules on fundraising, but helped solve a practical problem, contradicting stereotypes of the shirking bureaucrat. In a related example, Paine (Reference Paine, Lieberthal and Lampton1992) describes how local bureaucracy in China helped address the urban housing shortage, which had negatively impacted teachers. In the mid-1980s, 32.3 percent of elementary and secondary schoolteachers lacked access to housing, creating irregularities in the teaching force. Education officials cooperated with other public and private agencies to address the matter. In one locale, school administrators negotiated a pragmatic solution with a nearby factory, agreeing to admit a section of workers’ children in exchange for the construction of new housing.

These examples from the education sector point to general features of the Chinese bureaucracy. Problem-oriented deliberation by local administration (city, county and township) is supported by central and provincial governments. Consider China’s comparatively successful implementation of agricultural reforms.Footnote ¹² In the 1970s, most agricultural households belonged to communes, whose output was based on predetermined prices and production targets. To meet their targets, some communes gave households individual farm plots and let them retain output beyond predetermined quotas. After selling the required output at fixed prices to the state, any residual output could be sold at market prices, allowing households to earn profits from their labor. The household responsibility system, as it came to be known, created marginal incentives for households to produce more, enabling subsequent market reforms (Lin Reference Lin1988). Notably, the household responsibility system arose organically through local experimentation by communes. Successful experiments were identified by local leadership, emulated elsewhere and later endorsed by the central leadership. Xu describes how “proto-types” of the household responsibility system emerged in the 1970s (2011, 1111–1113). Land contracting experiments were attempted by local officials in Anhui and Sichuan, with support from provincial governors. The management of political risks associated with experimentation is noteworthy in this case, since land reforms were illegal and incurred heavy penalties. After experiments were validated in the 1980s, the central government extended these reforms nationwide. Similar bureaucratic processes supported the growth of rural industry through township and village enterprises, the creation of a “manager responsibility system” to reform state-owned enterprises, and the introduction of special economic zones to attract foreign investment and trade. These reforms were not conceived in Beijing and imposed all over China. Instead, they emerged as subnational initiatives and were tested in geographic pockets before being scaled up countrywide.

The bureaucratic mechanisms that Ang and other scholars describe have an affinity with the flexible, problem-based orientation of deliberative bureaucracy theorized in this book. The cases of Himachal Pradesh (HP) (Chapter 5) and Mahila Samakhya in Uttar Pradesh (UP) (Chapter 7) brought out the importance of deliberation within India’s public agencies. In this respect, however, Chinese bureaucracy far exceeds its Indian counterpart. A bureaucratic culture of deliberation in China is sustained with support from central and provincial governments. By contrast, central government agencies in India have exhibited a “culture of rule-following,” which, as Chhibber (2002) observes, limits their capacity to undertake complex development tasks and coordinate with society. Along with differences in bureaucratic culture, formal administrative structures vary decidedly between the two countries. In comparison to China, administrative decentralization in India has been much shallower. The Indian state is highly centralized with respect to fiscal responsibilities and social policy implementation, barring a few subnational exceptions, such as Kerala. At the subnational level, legalistic bureaucracy in states like UP has prevented education officials from pursuing practical responses to citizen grievances. As we saw in Chapter 4, local education officials in UP who attempted to innovate experienced institutional setbacks and had little support from senior officials. In that sense, the Indian state operates more like a conventional “rule-of-law” system, which, as Heilmann notes, “virtually rules out discretionary and experimental administrative measures” (2008b, 9). Successful policy experiments in India are less likely to receive encouragement from state leadership, weakening the prospects for institutional learning and scaling up of local innovations. It is not surprising, then, that policymakers in India refer to its governance landscape as a “graveyard of pilots.”Footnote ¹³

This brief foray into China’s social development suggests that deliberative bureaucracy’s promise is not reserved for democracies. This is not to deny the value of democracy and the political freedoms it affords. If bureaucratic deliberation is well institutionalized in China, the constraints posed by its authoritarian political system are considerable. In particular, the mechanism of societal feedback during implementation is weakened when citizens cannot freely organize and voice collective demands on the state. In India, the political dynamics unleashed by democratic elections, associational freedoms and protection of civil liberties are more conducive to societal involvement in policy implementation. Even in underperforming states like UP, democratic processes have increased the scope for women and lower castes to mobilize collectively. By contrast, China’s authoritarian system limits deliberative bureaucracy’s ability to fully incorporate societal feedback, especially when it comes from underprivileged groups who lack preferential access to the party-state.Footnote ¹⁴ China’s authoritarian restrictions, particularly state suppression of public criticism, can lead to major policy mistakes, as witnessed in the Great Famine (Drèze and Sen Reference Drèze and Sen1989). Moreover, authoritarianism can weaken society’s ability to exercise voice collectively on an everyday basis. Tsai’s (Reference Tsai2007) study of rural public goods provision shows, for example, that while village temple groups can hold local officials accountable, village church groups lack the same opportunity since the Chinese party-state closely regulates churches, prohibiting local officials from participating in them. The limits of authoritarian deliberation are perhaps further revealed in the recent period of market-oriented reforms, which have brought unmatched economic prosperity but with heightened inequality and diminished progress in basic health and education.Footnote ¹⁵ As Bardhan (Reference Bardhan2010, 104–116) observes, the collapse of local public finances and growing privatization of public services has widened China’s class divisions and regional disparities. Likewise, Ang notes, the Chinese bureaucracy’s profit orientation “has exacerbated unequal access to essential services and provoked public resentment” (Ang Reference Ang2016b, 14). Thus, while deliberative bureaucracy in China has facilitated extraordinary gains in economic and social development, how the political system addresses these mounting concerns is yet to be seen. This brief foray into China’s development suggests that the concept of deliberative bureaucracy has analytical reach beyond democracies such as India, and merits further attention in comparative research on the state and service delivery.

8.4 Finland and France: Varieties of Bureaucracy in Advanced Economies

This section considers the book’s argument in the context of advanced economies through a comparison of school education in Finland and France. The theory connecting bureaucratic norms to differences in policy implementation pertains primarily to weak institutional environments, where the gap between formal rules and bureaucratic behavior is often substantial (Levitsky and Murillo Reference Levitsky and Murillo2009). In advanced economies, states tend to have greater legal and fiscal capacity, improving the likelihood for public services to be well implemented (Besley and Persson Reference Besley and Persson2014). Citizens in these countries will also tend to have better material conditions and more knowledge of their rights, raising the demand for quality services. Yet, even wealthy countries encounter hurdles when implementing education services, as evidenced by the OECD’s Programme for International Student Assessment (PISA). A well-recognized international assessment for education, PISA measures the aptitude of fifteen-year-old students in reading, mathematics and science. Learning improvements in OECD countries have been marginal since PISA’s first round in 2000, despite major public outlays for education.Footnote ¹⁶ In the 2018 PISA round, 24 percent of students from OECD countries scored below the minimum proficiency level in mathematics and 12 percent did so in reading (Crato Reference Crato2021). Against these disappointing results, there is significant variation across OECD countries in average PISA test scores as well as disparities in the achievement gap between different socioeconomic groups. Finland is a leading performer on PISA, having comparatively superior learning outcomes across students of different socioeconomic categories. France, meanwhile, is a middling performer among OECD countries, with relatively high inequality in learning outcomes across students.

What explains the variation in education services and associated learning outcomes between the two countries? In what follows, I suggest that differences in bureaucratic norms help account for some of the variation in learning, especially among the least advantaged student cohorts. It should be noted at the outset that student achievement is determined by a multitude of factors, some operating outside of the school system. Student socioeconomic background–conditioned by factors such as parental education and work status, household income, family immigration status–explains to a large extent the individual and cross-national variation in learning (Hanushek and Woessmann Reference Hanushek, Woessmann, Hanushek, Welch, Machin and Woessmann2011). Yet, the magnitude of the effect of socioeconomic factors on learning varies between countries, suggesting that schooling systems have an important influence on student achievement. As Woessmann writes, “The considerable differences in student achievement across countries are systematically related to differences in the organization and governance of school systems” (2016, 3). Quantitative indicators of schooling provision, such as public expenditure per student and class sizes, are important up to a threshold, but have less of an impact on learning outcomes than one might expect (Schleicher Reference Schleicher2018, 20). Education researchers are therefore paying more attention to “nonresource institutional features of school systems,” such as teacher training and incentives, but also school governance, monitoring and accountability systems (Hanushek and Woessmann Reference Hanushek, Woessmann, Hanushek, Welch, Machin and Woessmann2011, 138).

Before proceeding, it is important to validate the comparison between France and Finland. The Varieties of Capitalism literature has shown how different institutional configurations for the economy and welfare state produce varying patterns of skill development (Hall and Soskice Reference Hall and Soskice2001; Thelen Reference Thelen2004; Iversen and Stephens Reference Iversen and Stephens2008; Busemeyer Reference Busemeyer2014). Their many differences apart, France and Finland have similar levels of public spending on education – 5.4 percent and 5.9 of GDP, respectively – and social spending overall is also comparable (Busemeyer Reference Busemeyer2014, 125–126). France’s welfare state lies near the cluster of Scandinavian countries, where education is regarded as “a complementary addition to social policy” (Busemeyer Reference Busemeyer2014, 134). State commitment to education in France diverges from the approach taken by liberal market economies like the United Kingdom, where education is largely privatized. And yet, inequalities in student achievement are considerable in France, and may be even worse than in the United Kingdom (Duru-Bellat and Suchaut Reference Duru-Bellat and Suchaut2005; Doyle Reference Doyle2021). There are, at the same time, important differences in the formal institutions governing school education in Finland and France. While recognizing these differences, I suggest that bureaucratic norms may also contribute to the variations in education delivery and learning outcomes between the two countries.

8.5 Conclusion

In this chapter I have placed the book’s arguments in comparative perspective, examining how legalistic and deliberative bureaucratic norms operate and shape public service delivery in cases outside northern India. The investigation of Kerala’s comparative achievements in education revealed that deliberative bureaucracy is most responsive to citizens where social movements and political competition help sustain democratic articulation of subaltern demands (Sandbrook et al. Reference Sandbrook, Edelman, Heller and Teichman2007). The case also suggested how state-building processes can influence contemporary service delivery, emphasizing bureaucracy’s historical engagement with non-state education providers, an underappreciated attribute of social development in Kerala. Bureaucracy in Kerala learned how to work with religious authorities, private schools and civil society organizations that provided social services, establishing mechanisms for the joint management of schools and co-governance of public goods more broadly. These observations point to the need for further research investigating how, in the process of state-building, bureaucracies learn to govern public services and forge relationships with non-state service providers (Cammett and MacLean Reference Cammett and MacLean2014; Ansell and Lindvall 2021).

In contrast to Kerala, the Chinese case explored how deliberative bureaucracy functions in a non-democratic context. Mechanisms of policy adaptation in China’s education system (and other policy domains) revealed unanticipated parallels with India, highlighting the importance of deliberation by higher-level agencies to support local experimentation. The provisional observation, that bureaucratic commitment to deliberation has facilitated the implementation of public services in China, in some ways more effectively than it has in India, merits further study. The comparison of school education in Finland and France pointed to the advantages of deliberation for improving the education of underprivileged students in advanced economies. The case of Finland suggested how deliberative governance supports customized education services, encouraging cross-sectoral coordination to identify and respond to individual needs, thereby enabling students from different social backgrounds and learning levels to succeed. Meanwhile, the case of France showed how legalistic bureaucracy encourages uniformity in the implementation of education services, but at the same time leads to inequalities in achievement between students from different geographies and social classes. While the findings are tentative and require deeper investigation, the comparative analysis of Finland and France suggests that informal bureaucratic norms matter for the implementation of high quality education, even where formal institutions are strong.

9.1 Introduction

This book was animated by the puzzling differences in the implementation of primary education in northern India. Subject to the same national policies, constitutional-legal framework, administrative structures and electoral institutions, it sought to explain why and how bureaucracies in some cases have delivered primary education more effectively than others. The theory and evidence presented underscores the importance of bureaucratic norms–unwritten rules that influence the behavior of state officials and their relations with citizens. Through a multilevel comparative study, I have shown that bureaucratic norms vary systematically across north Indian states and help account for variation in service delivery. The findings stress the advantages of deliberative bureaucracy, which encourages a problem-based orientation, over legalistic rule-following bureaucracy. Deliberative bureaucratic norms were shown to enable state officials to undertake complex tasks, coordinate with society and adapt policies to local needs, yielding higher quality education services. These findings are impressive considering that rural north India is not where one would anticipate bureaucracy to function well in the first instance.

In highlighting the impact of bureaucratic norms on implementation, I should stress that my argument also recognizes the significance of other factors. Policy implementation is a causally complex process. The research design took alternative explanations into account, including geography, colonial land institutions and the agrarian economic structure. Chapter 2 also took note of prominent political explanations, operating at different levels of analysis, that could influence implementation. My findings at the state level accord with research demonstrating that two-party competition in India is more conducive for social service provision than fragmented, multiparty systems (Chhibber and Nooruddin Reference Chhibber and Nooruddin2004; Nooruddin and Simmons Reference Nooruddin and Simmons2015; Thachil and Teitelbaum Reference Thachil and Teitelbaum2015). However, in contrast to prior work, which shows how electoral incentives influence levels of social spending, I suggest that party systems can also operate through administrative mechanisms, influencing how officials plan and utilize public resources. Furthermore, research at the village level investigated social capital’s bearing on community monitoring of schools (Krishna Reference Krishna2003). Sociocultural norms and inequalities were observed to influence collective action by marginalized groups. Nevertheless, my fieldwork revealed that bureaucratic norms can shape the nature of village collective action over time, leading to differences in the coproduction of services and local accountability of schools.

In this concluding chapter, I review my findings and contributions to the fields of comparative politics, development and public administration. The theoretical framework, anchored in bureaucratic norms, brings institutionalist perspectives on the state and social policy together with insights from literatures on street-level bureaucracy and local collective action. The conceptual interweaving of meso-level institutional processes with the micro-politics of service delivery yields a new understanding of bureaucracy and its connection to human development. Along with scholarly contributions, I explore policy implications, particularly for reform efforts to enhance primary education and other frontline public services in developing countries.

9.2 Bringing Bureaucracy Back In

Bureaucracy does not often inspire praise. Researchers have cast bureaucracy as “inefficient, lazy, incompetent, wasteful, inflexible, unaccountable, and inhumane” (Olsen Reference Olsen2008, 14).Footnote ¹ These negative assessments motivate drives for reform, such as New Public Management, which calls for private sector-style contracts and incentives (Dunleavy and Hood Reference Dunleavy and Hood1994). Bureaucracy is also found wanting from the perspective of democratic accountability, as noted by calls for collaborative and network-based governance (Ansell and Gash Reference Ansell and Gash2007; Bevir Reference Bevir2010). The literature on distributive politics in developing countries is no less pessimistic about bureaucracy’s constructive potential. Bureaucrats are portrayed as cogs in the wheels of clientelism (Kitschelt and Wilkinson Reference Kitschelt and Wilkinson2007). State agencies that deliver social services are seen as egregious sites of patronage and political interference (Brun and Diamond Reference Brun and Diamond2014). Under conditions of weak state capacity, mechanisms of democratic competition are found to yield “visible” gains in school enrollment, but little improvement in the provision of schooling inputs, services or learning outcomes (Harding and Stasavage Reference Harding and Stasavage2014). Along these lines, Indian democracy is weighed down by a bureaucracy that is deemed to be captured and corrupt, incapable of executing its core functions (Wade Reference Wade1982; Iyer and Mani Reference Iyer and Mani2011).

Nevertheless, as I have shown, bureaucracy is not everywhere the same in India. This book departs from the pessimistic view of public agencies in developing countries, bringing to the fore the remarkable differences in bureaucratic performance within the Hindi belt, a region associated with clientelism and patronage politics (Chandra Reference Chandra2004; Keefer and Khemani Reference Keefer and Khemani2004). Delivery of education services poses a particularly difficult test for bureaucracy given chronic political neglect of primary education in this region, as well as the absence of leftist government and subaltern movements demanding social rights (Heller Reference Heller1999; Ahuja Reference Ahuja2019). During twenty-eight months of field research, I observed local administrators and teachers travel great distances to ensure that primary education reached villages. Even in rural Uttar Pradesh, where identity politics and ethnic targeting are thought to divert public resources away from programmatic ends, bureaucrats committed to legalism used their discretion to direct resources to primary schools in need. It is all too easy to overlook the effort of bureaucrats laboring quietly from a sense of duty. Instead, my approach builds on scholarship exploring the possibilities for bureaucracy to function well amid institutional weaknesses and political constraints (Uphoff Reference Uphoff1994; Grindle Reference Grindle1997; Tendler Reference Tendler1997; Chand Reference Chand2006).

There are, of course, many instances where bureaucracy falls short. In India and elsewhere, citizen encounters with the state can end in frustration. Such cases, considered throughout this book, often get framed as examples of bureaucratic shirking and corruption (Klitgaard Reference Klitgaard1988). Yet, deficiencies in service delivery can occur even when officials abide by policy rules. Collapsing all such cases as instances of corruption depends on an overly stylized view of implementation and bureaucratic discretion, which distorts empirical reality more than it clarifies it. Discretion is deemed to be problematic because, theoretically, it provides rent-seeking opportunities. However, discretion is a necessary feature of public service work and, in some cases, it is preferable to administrative control (Tendler and Freedheim Reference Tendler and Freedheim1994; Piore Reference Piore2011; Honig Reference Honig2018). That is not to say that bureaucrats always apply discretion in public-spirited ways; they can and do succumb to the temptations of rent-seeking. It is inappropriate, however, to treat discretion as synonymous with rent-seeking. Nor is it obvious that more and tighter rules are a bulwark against misuse of discretion.Footnote ² As Gupta (2012) shows, conscientious adherence to rules and procedures in India can generate the opposite effect, leading officials to apply “red tape” and deny citizens essential services. Instead of assuming that discretion is harmful, my framework of legalistic and deliberative bureaucratic norms helps decipher alternative ways in which discretion can be applied in various administrative tasks.

The pessimism surrounding bureaucracy is in part due to the paucity of field-based investigations of public agencies in developing countries. In India, few empirical studies catalog the internal dynamics of the state and its ordinary relations with citizens. I have attempted to help fill this void. My analysis of the striking variation in policy implementation by education bureaucracies in India joins recent studies of state capacity in developing countries (Piore and Schrank Reference Piore and Schrank2008; Schrank 2009; Canales Reference Canales2014; Amengual Reference Amengual2016; Ang Reference Ang2016a; Williams Reference Williams2017; Bersch Reference Bersch2019; Dasgupta and Kapur Reference Kapur2020; McDonnell Reference McDonnell2020). Whereas others have emphasized the role of policy entrepreneurs, administrative resources and formal structures of the state, my argument underscores the stark differences in the norms that guide state officials on how to understand their duties. As I have shown, bureaucracies in India with identical formal administrative structures have developed very different informal norms that shape their internal operations and external ties with society. There is more than one bureaucratic ethos.

The advantages of integrating norms into the analysis of bureaucracy can be gleaned from the vibrant debates on state capacity in the United States. On the one hand, public agencies in the United States have historically been depicted as weak and politicized (Novak Reference Novak2008). Formal bureaucratic structures are shown to bend to the political will of elected officials and interest groups (Moe Reference Moe, Chubb and Peterson1989). At the same time, studies identify vast differences in the informal norms and culture of public agencies in the United States (Kaufman Reference Kaufman1967; Wilson Reference Wilson1978; DiIulio 1994). These differences in norms can have political implications. As Carpenter (Reference Carpenter2001) demonstrates, executive agencies like the US Postal Service used their discretionary authority to build an organizational culture of performance. They created network ties with societal groups and earned a reputation for innovative work, which over time allowed them to loosen the shackles of political control and secure their autonomy. These debates suggest that bureaucracy is not merely an instrument for political ends, but an institution bound by norms, with political authority in its own right.

9.3 Deliberation and the Developmental State

This book contributes to comparative scholarship on the state and inclusive development. The developmental state literature has demonstrated how bureaucracy has facilitated industrial development and economic growth in East Asia (Haggard Reference Haggard2018). Inclusive development, however, calls for administrative mechanisms that support the expansion of human capabilities (Sen Reference Sen1999). There is a need, therefore, to reorient our understanding of the developmental state toward the provision of public services, in particular mass education and health (Evans and Heller Reference Evans, Heller, Leibfried, Evelyne, Lange, Jonah, Nullmeier and Stephens2015; Evans, Huber, and Stephens Reference Evans, Huber, Stephens, Centeno, Kohli, Yashar and Mistree2017). Accordingly, my argument has cast the spotlight on quotidian bureaucracies that coproduce primary education with citizens (Ostrom Reference Ostrom1996). Coproduction of services is a cognitively difficult, transaction intensive and politically challenging work, presenting “wicked problems” for the state (Andrews, Pritchett and Woolcock Reference Andrews, Pritchett and Woolcock2017; Peters Reference Peters2017). Beyond the technical details of policy design, the inclusive development agenda warrants a fundamental rethinking of bureaucracy and its relationship with society, and marginalized citizens in particular, who are asked to participate in coproduction.

I push the developmental state literature forward by offering a new theory centered on differences in bureaucratic norms that advances our understanding of bureaucracy’s role in inclusive development. Chapter 2 theorized the ideal types of bureaucracy–legalistic and deliberative– and delineated the mechanisms whereby they generate differences in policy implementation. Subnational comparative field research in four north Indian states (Chapters 4–7) provided evidence for the material impacts of bureaucratic norms on the delivery of primary education. These divergent bureaucratic norms were shown to spark very different orientations among officials, influencing how they understand their policy mandates, apply discretion and encourage societal input. The empirical findings suggest that deliberation is a central component of developmental states oriented toward human capability expansion.Footnote ³ Although legalistic bureaucracy was shown to succeed at codified tasks, such as enrollment and the provision of schooling inputs, deliberative bureaucracy performed far better on complex tasks of school monitoring and teaching, yielding higher quality education services. Deliberative spaces for collective problem-solving by senior and frontline administrators, schoolteachers and parents enabled “thick” information exchange and coordination, as demonstrated through the case of Himachal Pradesh, allowing the state to adapt education policies to meet diverse local needs.

In arguing for the significance of bureaucratic norms, I do not discount the importance of formal Weberian institutions. Bureaucratic autonomy, corporate coherence and commitment to impartial rules are important restraints against corruption, predation and other pathologies (Evans and Rauch Reference Evans and Rauch1999; Rothstein Reference Rothstein2011). Rule-following is encouraged by esprit de corps, a corporate culture supported by meritocratic selection, training and promotion (Levi and Sherman Reference Levi, Sherman and Clague1997, 332–333). Where formal state institutions are comparatively weak and politicized, creating a bureaucratic ethos of rule-following remains a first-order challenge, as McDonnell (Reference McDonnell2020) observes in her study of public institutions in Ghana. McDonnell demonstrates that “pockets of effectiveness” within the state depend on bureaucratic subcultures that build commitment to administrative rules and boundaries, countering political pressures. Seen from this angle, the presence of legalistic bureaucracy in parts of northern India is a noteworthy achievement. Nevertheless, when it comes to the challenges of inclusive development, bureaucratic commitment to rules is not enough. To implement quality services, bureaucracies need to solve complex problems and adapt to local needs, which is best achieved when bureaucratic norms encourage robust deliberation.

My book’s arguments about bureaucracy and inclusive development are based on evidence from India’s primary education sector. However, the theoretical framework of bureaucratic norms offers insights for other policy arenas as well. Ever since Gunnar Myrdal’s (Reference Myrdal1968) depiction of India as a “soft state,” scholars have cataloged its bureaucracy’s uneven performance across policy functions. The Indian state has been variously described as “weak-strong” (Rudolph and Rudolph Reference Rudolph and Rudolph1987), “failed developmental” (Herring Reference Herring and Woo-Cumings1999), “Sisyphean” (Kapur and Mukhopadhyay 2007) and “flailing” (Pritchett Reference Pritchett2009). These formulations suggest that Indian bureaucracy has the potential to achieve development goals by virtue of its semi-Weberian design, but it falls short of its theoretical promise owing to societal demands, resource constraints and particularistic politics. Pritchett’s (Reference Pritchett2009) notion of a “flailing” state adds another dimension to the assessment, highlighting the disconnect between the “brains” of Indian bureaucracy, the elite agencies that formulate policy, and its “limbs,” the frontline agencies tasked with implementation. There are likely many reasons for this disconnect. My argument demonstrates that institutional barriers to deliberation are one critical source of dysfunction, while the social distance that legalistic bureaucracy creates between state planners and lower-level functionaries is another.Footnote ⁴ These barriers are reinforced by legalism’s penchant for upholding hierarchical divisions and written documentation as a mode of internal communication, which can disempower frontline workers and divert their energies away from solving citizens’ practical needs.Footnote ⁵

Equally, it may be that the elite agencies constituting the “brain” of the Indian state are not as well functioning as they appear. Chibber (Reference Chibber2002; 2003) has demonstrated that a bureaucratic culture of rule-following in elite economic institutions, such as the Planning Commission, stifled interagency coordination and weakened the execution of industrial policy in the decades following independence. Despite having capable officers from the Indian Administrative Service, who drew up sophisticated five-year plans, Chibber finds that an excessive reliance on formal rules and procedures undermined policy implementation. On the other hand, there are also moments where bureaucratic norms support interdepartmental coordination. Indian bureaucracy is not always flailing. As Kapur (Reference Kapur2020) observes, the periodic administering of elections in India brings disparate arms of the state together on a massive scale, with intensive coordination among the police and other agencies. He notes puzzling discrepancies in state performance as well, such as the remarkable strides India has made in eradicating polio, even as public health systems languish. More research is needed to identify the causes of uneven bureaucratic performance in India.Footnote ⁶ Bureaucratic norms, I submit, merit consideration alongside other plausible factors.

These examples from outside the primary education domain also cast doubt on the proposition that rational-legal norms are always the surest path to effective state performance. Perhaps the most striking example of divergence from rational-legal norms comes from China. The Chinese bureaucracy’s contributions to economic and social development are bewildering from the vantage point of Weberian bureaucracy. Alternative models of public administration have been proposed, theorizing norms that vary widely from the Weberian rational-legal ideal (Rothstein Reference Rothstein2015; Ang Reference Ang2016b; Heilmann and Perry Reference Heilmann and Perry2011). Explored in Chapter 8, Ang’s (Reference Ang2016a) account of “directed improvisation” underscores the importance of bureaucratic flexibility and local adaptation for poverty reduction. Her argument resonates with my conception of deliberative bureaucracy, albeit with the difference that my framework also highlights how different bureaucratic norms activate alternative pathways for societal feedback. The emphasis on feedback also brings out the advantages of democratic arrangements, especially the freedom to associate and voice dissent. The brief examination of China in Chapter 8 does not consider subnational differences in bureaucratic norms, the study of which would enrich our understanding of Chinese bureaucracy.

My argument linking state action to societal feedback during implementation also advances scholarship on embeddedness, the state–society linkages associated with high performing developmental states (Evans Reference Evans1995). Instead of industrial capitalists, however, inclusive development requires bureaucracy to connect with mass publics. Chapter 2 theorized societal feedback as a dynamic process of interaction between citizens and local bureaucracies. When citizens seek public services, they learn how local agencies function, and, gradually, form expectations and collective strategies. Legalistic bureaucracy embeds itself through officially recognized channels, such as Village Education Committees (VECs), encouraging citizens to monitor services and voice demands through systems for grievance redressal. However, these channels impose administrative burdens on citizens, which limit the state’s ability to “see” and respond adequately to societal demands (Scott Reference Scott1998). Meanwhile, the problem-based orientation of deliberative bureaucracy motivates state officials to discuss policy problems with societal groups, including informal associations, enlarging the spaces for marginalized citizens to monitor services and communicate demands.

The difficulties of promoting substantive citizen participation in public service delivery are well recognized, especially in contexts of social inequality (Fox Reference Fox2007; Houtzager and Acharya Reference Houtzager and Acharya2011; Heller and Rao Reference Heller and Rao2015). The design of this study accounted for regional differences in socioeconomic inequality within northern India. Inequality was considered both in terms of the agrarian economy, by incorporating land-ownership patterns (Bardhan, Ghatak, and Karaivanov Reference Bardhan, Ghatak and Karaivanov2007), and with respect to gender and caste relations (Agarwal Reference Agarwal1997). My comparative field research gave rise to four theoretically distinct local-level patterns, or what I call “regimes” of coproduction (Figure 9.1). The matched-pair comparison of HP (Chapter 5) and Uttarakhand (Chapter 6) identified differences in coproduction within the western Himalayas, a region with relatively less inequality. Deliberative bureaucracy in HP supported inclusive coproduction, broad-based societal participation and complementary investments from communities, helping to bind households to government schools. In Uttarakhand, by contrast, legalistic bureaucracy provided a platform for societal input, but gradually induced communities to seek non-state services, a pattern of substitutive coproduction. Community efforts to hold government schools accountable were frustrated, compelling households to seek private tutoring and exit for non-state services. Village collective action in Uttarakhand did not diminish, but instead shifted away from government services to support private and “hybrid” forms of education provision (Cammett and MacLean Reference Cammett and MacLean2014; Post, Bronsoler, and Salman Reference Post, Bronsoler and Salman2017). Legalistic bureaucracy is thus shown to induce privatization.Footnote ⁷ The divergent patterns of coproduction observed in the two Himalayan states are more striking given that the case-study villages had similar levels of social capital, evidenced by membership in village-wide women’s associations, shared religious practices and other forms of social cooperation (Krishna Reference Krishna2003). Village collective action was found to enhance education services in both states, but differences in bureaucratic norms had a significant impact when it came to sustaining community participation and investment in government schools.

Figure 9.1 Regimes of coproduction

In the Gangetic plains, where inequality is higher, marginalized citizens encountered greater hurdles to participation in service delivery. Compared to the Himalayan region, social relations in UP villages were more fragmented and unequal on economic, caste and gender lines. Households still came together at various junctures to monitor primary schooling, but they were discouraged by their experiences with VECs and district administration. Collective action around schooling did not cease entirely, but proceeded in spurts, following a pattern of episodic coproduction. These observations from UP echo Ostrom’s (Reference Ostrom1996) findings about the fragile nature of coproduction in rural Nigeria. As she observes, “[v]illages that had demonstrated their capabilities to engage in collective action were discouraged by government officials from active engagement in the education of their children” (1996, 1076). However, comparative findings from within UP suggest the possibilities for an alternative path. The study of Mahila Samakhya’s programming on girls’ education in Chapter 7 demonstrated how deliberative bureaucracy helps to sustain the robust engagement of marginalized groups. Frontline workers mobilized Dalit women and countered village hierarchies to promote public deliberation, a process that induced social conflict and backlash from upper castes.Footnote ⁸ Through deliberation, Dalit parents learned that their engagement in school meetings meant something. The pattern of coproduction nevertheless remained conflictual. Mahila Samkhya frontline workers mediated conflicts, serving as “pressure groups,” to counter societal inequalities and challenge conventional norms in the bureaucracy.

The four coproduction regimes that I have outlined capture theoretical differences in local state–society relations impacting policy implementation. They also point to the bureaucratic mechanisms that help invigorate coproduction in hierarchical settings. Where coproduction is conflictual, for example, frontline officials must take an activist stance to support marginalized groups when they make claims on the state. The significance of bureaucratic activism in support of disadvantaged groups is further demonstrated in comparative research from Latin America. Abers (Reference Abers2019) shows how activism took hold in Brazil’s environmental bureaucracy, motivating officials to challenge rules and support contentious causes. Similarly, Rich (Reference Rich2019) finds that the health bureaucracy in Brazil fostered societal activism, forging a state–civil society coalition to combat AIDS. These and other Latin American cases reinforce the point us that coproduction is a political process requiring active support from the state (Baiocchi Reference Baiocchi2003).

Understanding why bureaucratic norms and cultures diverge across settings requires us to engage with the comparative history and politics of state-building (Slater Reference Slater2010; Soifer Reference Soifer2015; Centeno et al. Reference Centeno, Kohli, Yashar and Mistree2017).Footnote ⁹ Historical evidence suggests that the political conditions for deliberative bureaucratic norms to thrive are relatively rare. At its normative core, deliberation involves a submission to the constraints of public reason (Rawls Reference Rawls1996, 212–254). The political commitment to deliberation means, at minimum, that bureaucracy must be willing to listen and respond to reasonable demands from the public. However, state elites are reluctant to submit to such demands, more so when they are made by the least advantaged (Acemoglu and Robinson Reference Acemoglu and Robinson2006). In HP, deliberative bureaucracy emerged from the historically specific process of state-building, where elite cooperation was spurred by political marginalization and the acute need for central government resources. Elsewhere in India, programmatic social movements have been critical to the forging of deliberative institutions, as in the case of Kerala (Heller Reference Heller1999). Explored briefly in Chapter 8, Kerala’s deliberative bureaucracy has supported mass education and the deepening of local democracy. The success of these projects was also contingent on historical social movements led by subaltern groups as well as civil society organizations (Isaac and Heller Reference Isaac, Heller, Fung and Wright2003). Where social movement politics has not materialized, deliberative bureaucracy’s reach is more likely to be restricted to geographic and administrative pockets of the state, as witnessed in the case of Mahila Samakhya in UP. Additional research is needed to identify the political factors that enable deliberative bureaucracy to take root and flourish more fully.

This book’s arguments pertain mainly to bureaucracy and public services in developing countries. However, increased global attention to economic inequality and social discrimination suggests that inclusive development has also become a matter of concern for advanced economies (Hacker Reference Hacker2006; Lieberman Reference Lieberman2011; Case and Deaton Reference Case and Deaton2020). My theoretical construct of deliberative bureaucracy builds on scholarship examining how deliberative and pragmatic governance arrangements can address complex policy problems in North America and Europe (Cohen and Sabel Reference Cohen and Sabel1997; Fung Reference Fung2006; Ansell Reference Ansell2011; Sabel and Zeitlin Reference Sabel, Zeitlin and Levi-Faur2012; Moffitt Reference Moffitt2014). Public administration scholars have similarly called for a rethinking of how bureaucracy can support vulnerable groups in these regions (Nabatchi Reference Nabatchi2010; Maynard‐Moody and Musheno Reference Maynard‐Moody and Musheno2012; Brodkin Reference Brodkin2016). Extending my argument to OECD countries, Chapter 8 offered a comparative sketch of school education in Finland and France. The comparison proposed, albeit provisionally, that differences in bureaucratic norms help account for variation in the performance of school systems across the two countries, leading to differences in student achievement. Finland’s superior and more equitable achievement results were tied to deliberation by Finnish Student Welfare Groups as well as education and health agencies, encouraging “individualized service delivery” (Sabel Reference Sabel, de Mello and Dutz2012). In France, meanwhile, legalistic bureaucracy has pressed for centralized control and uniformity of education delivery, lessening the school system’s capacity for local adaptation. More research is needed to establish these propositions. The comparative study of bureaucratic norms opens up new avenues for future research on inclusive development within and across countries.

9.4 Policy Implementation and Societal Feedback

My argument about bureaucracy’s conditioning effects on citizens contributes to scholarship on policy feedback. This rich body of work has demonstrated the downstream political effects of social policies, which gradually can transform the landscape of mass politics (Pierson Reference Pierson1993; Mettler Reference Mettler2002; Campbell Reference Campbell2005; Kumlin and Rothstein Reference Kumlin and Rothstein2005). I extend the insights of this literature to the terrain of public services in developing countries, where policy implementation generates its own politics (Grindle Reference Grindle1980). Further, my emphasis here is on mundane forms of participation beyond elections, driven by the collective needs of underprivileged groups (Davies and Falleti Reference Davies and Falleti2017). Though relatively rare, studies of policy feedback in developing countries have the potential to transform our understanding of local governance and state–society relations.Footnote ¹⁰ Holding India’s national primary education policy framework constant, I have shown that subnational differences in bureaucratic norms lead education officials to interpret the same policies in radically divergent ways, which in turn leads to very different citizen experiences of local bureaucracy. Examining this causal chain helps us understand why similarly placed social groups in India come to “see” the state so differently (Corbridge et al. Reference Corbridge, Williams, Srivastava and Véron2005).

My argument for how bureaucratic norms induce varied societal responses during implementation brings policy feedback into conversation with public administration scholarship (Moynihan and Soss Reference Moynihan and Soss2014). I find that legalistic bureaucracy triggers compliance-oriented school monitoring and administrative grievance procedures, creating opportunities for parents to register complaints. However, these channels impose administrative burdens that are disproportionately stacked against the least advantaged, making it harder for them to hold public service providers to account (Goetz and Jenkins Reference Goetz and Jenkins2005; Herd and Moynihan Reference Herd and Moynihan2018). I also find that legalistic bureaucracy reproduces caste inequalities and stigmas in the process of monitoring services, which can dissuade marginalized groups from following up on complaints. As Chapter 4 showed in the case of UP’s Midday Meal Programme, local bureaucracy’s rule-based interpretation of food “hygiene” comingled with dominant village narratives of caste-based pollution, which disempowered Dalit women and rendered futile their collective efforts to monitor school meals. By contrast, ordinary experiences of the local state enhanced the confidence of rural women in HP as well as among Dalit women attached to UP’s Mahila Samakhya program.

Attention to local bureaucracy’s conditioning effects on societal participation invites a corrective to the conventional view that residents of northern India are apathetic about primary education and other collective goods. Studies of public service provision in UP, Bihar and other Hindi belt states point to a deficit of political “awareness” among rural residents, whose “apathy” and “inertia” are cited as a cause for implementation failures (Drèze and Gazdar Reference Drèze, Gazdar, Drèze and Sen1996; Lanjouw and Stern Reference Lanjouw and Stern1998; World Bank 2003, 44–45; Singh Reference Singh2016, 168–172). My argument reverses the direction of causality. Field research in villages across UP suggests that poor households were aware of deficiencies in government primary schools, that local monitoring was taking place and that parents did communicate grievances to the local state. However, their voices were muffled by a thicket of bureaucratic rules and procedures that weakened school accountability.Footnote ¹¹ Societal “apathy” was not the cause of implementation failures but the consequence of concerned parents engaging in “active citizenship” through which they acquired practical knowledge of the local state.Footnote ¹²

To be sure, the obstacles to societal monitoring of public services in the Hindi belt remain significant. The findings of this study concur with prior research identifying the difficulties in achieving substantive participation in collective institutions, especially among marginalized groups (Mansuri and Rao Reference Mansuri and Rao2013). As Appadurai (Reference Appadurai, Heller and Rao2015) notes, behind a chain of participation, there often lies a string of performative failures. The question is whether such failures bring incremental gains that enlarge the aspirations of the poor. In that sense, this study suggests that cautious optimism regarding the possibilities of state-induced societal feedback in settings of inequality is warranted. Chapter 7 demonstrated that Mahila Samakhya’s frontline workers encouraged village-level deliberation and supported Dalit women during public conflicts. Dalit households experienced greater confidence in their abilities to participate in school governance.Footnote ¹³ These positive dynamics filtered down into classrooms as well, encouraging Dalit girls who had previously dropped out of school to aspire for a better future.

This book’s comparative examination of societal feedback across northern India suggests the benefits of combining field research with an historically situated analysis of subnational bureaucracies. I build on the methodological approach advanced by MacLean (Reference MacLean2010) in her study of the state and citizenship in Africa. Drawing on village-level fieldwork and oral histories, Maclean identifies how the state in Ghana and Cote d’Ivoire spawned vastly different informal institutions and practices. In a similar way, my comparative ethnography of rural India identifies divergent societal responses to state bureaucracy. In the absence of historical data and written records, interviews and focus group discussions with village elders were a window into the histories of collective action in villages.Footnote ¹⁴ Without intensive fieldwork and attention to the temporal dimension of societal feedback, the political activities of marginalized groups can be missed.

9.5 Street-Level Bureaucracy and the Institutions of Collective Welfare

Scholarship on street-level bureaucracy has shed important light on the quotidian practices of the state (Lipsky Reference Lipsky1984; Brodkin Reference Brodkin1997; Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2003; Piore Reference Piore2011; Zacka Reference Zacka2017). My argument advances this literature by situating the micro-politics of frontline service work within a multilevel comparative analysis of bureaucratic institutions. Tracing implementation processes over multiple tiers of the Indian state reveals the two-way interactions between citizens, frontline agencies and higher-level authorities.Footnote ¹⁵ This approach complicates the well-worn dichotomy between “top-down” administrative control and “bottom-up” discretion posited in the implementation literature (Sabatier Reference Sabatier1986; Matland Reference Matland1995).Footnote ¹⁶ The behavior of local officials, I find, is contingent on bureaucratic norms established upstream in the administrative hierarchy. Supervisory officials stationed in state capitals (or district headquarters) reinforce norms through their routine communications with subordinates, influencing how the latter understand their duties.Footnote ¹⁷ As shown in Chapter 4, monitoring of the Midday Meal Programme by local officials in UP reflected a narrow, “accounting-based” interpretation of the policy, as propagated by state planners in Lucknow. By contrast, Mahila Samakhya’s frontline workers in UP (Chapter 7) saw health and nutrition programming as integral to girls’ education, a more capacious interpretation, encouraged by senior officials through gender-oriented training. I also find that street-level bureaucrats help sustain norms collectively from below. Local education officials in HP (Chapter 5) supported bureaucratic deliberation by seeking out village women’s groups during adult literacy campaigns. These same officials later coordinated with state planners to experiment with models of school governance involving mother teacher associations.

The empirical findings also suggest the ways that frontline worker commitment and conflict supports processes of institutional persistence and change (Mahoney and Thelen

2010). Institutionalist scholars of different theoretical persuasions have recognized that the rules of the game are not merely applied, but also interpreted by officials, who can have conflicting ideas and interests (March and Olsen Reference March and Olsen1989; Mantzavinos, North, and Shariq Reference Mantzavinos, North and Shariq2004; Thelen Reference Thelen2004; Schmidt Reference Schmidt2008). In that sense, not only do street-level bureaucrats achieve policy outputs, they also enact norms and practices that can reinforce, enhance or subvert institutions over time.Footnote ¹⁸ In developing countries, where informal rules often have greater purchase, the interpretive work of frontline agents helps to build and consolidate state institutions (Helmke and Levitsky Reference Helmke and Levitsky2004). For example, the institutional “layering” of the Mahila Samakhya program within the education bureaucracy sparked conflicts inside the state and between women’s associations and local administration. While mediating such conflicts, frontline workers advocated for local-level changes in norms, which over time contributed to a shift in formal rules that expanded residential schooling for disadvantaged girls. In a contrasting case, local administration’s commitment to legalism in Bihar was found to have frustrated the efforts of policymakers to institute innovative pedagogical reforms in classrooms. These findings accord with prior work on institutions of collective welfare, such as Falleti’s (2010b) study of health reforms in Brazil, which point to gradual processes of change. Future research may go even further to connect street-level bureaucracies with comparative and historical institutional processes and probe the linkages between frontline work and the ideas and strategies of political actors located in the higher echelons of the state.

9.6 Policy Implications

This study of bureaucratic norms carries policy implications, particularly for the reform of education and other public services. Developing countries throughout the world are in the midst of a learning crises (World Bank 2018). As the UN Millennium Development Goals program reached a close in 2015, 193 countries extended the pledge for “inclusive and equitable quality education” under the Sustainable Development Goals (SDG) program. To realize SDG 4 (Education), governments and international donor agencies have committed to ensuring that all children, including first-generation learners, acquire foundational skills in the early years of schooling (Beeharry Reference Beeharry2021). These pronouncements are noteworthy and have the potential to transform the lives of millions of children. They also present daunting challenges for policymakers and practitioners. Meanwhile, the COVID-19 pandemic has taken an immeasurable toll on human life and economic livelihoods, while wreaking havoc on school systems worldwide (UNESCO 2021). The pandemic has also brought attention to the critical need for strengthening frontline state capacity.

Amid these concerns, the policy advice given to developing countries has also undergone important changes. The earlier impulse to recommend a universal policy blueprint (or the “right” institutions) was criticized for imposing unrealistic demands on developing country governments, while ignoring their sociopolitical contexts (Evans Reference Evans2004; Grindle Reference Grindle2004). More attention is now accorded to the historical conditions of institutional development (Acemoglu and Robinson Reference Acemoglu and Robinson2012). Micro-level research has aimed to guide reforms incrementally through local experimentation (Banerjee and Duflo Reference Banerjee and Duflo2011). International donor agencies have also put more resources into strengthening community participation in development (Mansuri and Rao Reference Mansuri and Rao2013). These important shifts notwithstanding, the policy advice for public sector reform in developing countries has remained focused on technical, efficiency-oriented fixes, performance metrics and the tightening of formal rules. This advice stems from an intellectual paradigm favoring top-down control and private sector incentives, with an aim to limit shirking and the misuse of discretion (Brinkerhoff and Brinkerhoff Reference Brinkerhoff and Brinkerhoff2015).

The recognition that bureaucracies have their own informal norms and cultures prompts a different approach to institutional reform. To achieve high quality public services, bureaucracies need to embrace norms that encourage vigorous deliberation and problem-solving. Rather than simply tightening rules, a more productive approach would be for policymakers to first take stock of the range of administrative tasks associated with implementation for a given policy context, anticipate points of ambiguity, interdepartmental coordination and societal engagement, and the power differences and conflicts these may occasion.Footnote ¹⁹ To overcome implementation hurdles, especially around complex tasks, bureaucrats need room to “puzzle,” which means having opportunities to discuss and learn through iterative experiments and partial changes (Heclo Reference Heclo1974). The institutional space for deliberation must extend beyond elite policy circles to include street-level bureaucrats, as well as citizens and non-state agencies. Local administration in particular needs to be empowered to take risks, communicate ideas, and interrogate status quo arrangements. By supporting problem-oriented discussion on the front lines of the state, policymakers can integrate local knowledge, what Scott (Reference Scott1998) has referred to as metis, into the implementation process.

Deliberation, as described here, may be an alien practice that conflicts with existing bureaucratic norms. In such cases, initiating collective discussions in well-defined institutional spaces would allow officials to gain familiarity with deliberation, before extending it widely. A related approach would be to identify positive deviants within an agency and enable them to build deliberative spaces (Andrews Reference Andrews2015). It should be noted, however, that positive deviation can also spawn conflict, and reformers may therefore need greater support and flexibility from external agencies. Where bureaucratic deliberation has arisen organically as a norm, policymakers should exercise care not to disrupt it. Reforms in these cases may focus on reinforcing and scaling up deliberation, for example, by institutionalizing discursive spaces between geographic and administrative units of the state. The opportunity to expand deliberation is likely to be greater in moments of political stability, when the climate for politicians and bureaucrats to cooperate for state-building is more favorable. It may be tempting to seize these moments to execute reforms rapidly, in a wholesale manner. However, the lessons of this book caution against introducing deliberative reforms with haste.Footnote ²⁰ State-building is a political process that unfolds over the longue durée. Building deliberative bureaucratic norms is not a one-off intervention but requires incremental and continuous engagement. Moreover, to sustain a shift in bureaucratic norms, participation from across the administrative hierarchy is needed, including from frontline agents.

While these policy implications focus on informal norms, formal authority and incentives clearly matter as well. Well-designed monitoring and incentive schemes can, in some cases, improve the quality of services (Muralidharan and Sundararaman Reference Muralidharan and Sundararaman2011; Duflo, Hanna and Ryan Reference Duflo, Hanna and Ryan2012). However, an exclusive emphasis on formal rules and pecuniary rewards may dampen the intrinsic motivation of bureaucrats (Banuri and Keefer Reference Banuri and Keefer2013; Rasul and Rogger Reference Rasul and Rogger2018).Footnote ²¹ Top-down managerial pressures may induce perverse behaviors as well, as Honig (Reference Honig2018) observes in his study of foreign aid organizations, leading frontline agencies to suspend their own judgment and instead comply with rules and narrow performance metrics. Such risks loom large in the education field, where pressures to adopt global “best practices” amplify the seduction of “looking like a state,” while camouflaging the underperformance of school systems (Pritchett, Woolcock and Andrews Reference Pritchett, Woolcock and Andrews2013). Greater attention to bureaucratic norms may help identify the institutional dynamics that support (or discourage) the uptake of incentives and performance measurement schemes, enabling such initiatives to achieve their potential.Footnote ²²

My study also has implications for institutional reforms aimed at fostering societal participation in development. These reforms, which include decentralized governance and community-based development, promise to enhance citizen voice and local accountability, though evidence of their impact is mixed (Bardhan and Mookherjee Reference Bardhan and Mookherjee2006; Mansuri and Rao Reference Mansuri and Rao2013). An experimental study of VECs in eastern UP illustrates the challenges of community mobilization in unequal settings (Banerjee et al. Reference Banerjee, Banerji, Duflo, Glennerster and Khemani2010). After it provided information on learning outcomes to parents and employed “best practice guidelines” to stimulate their engagement, the intervention yielded no improvement in monitoring or the quality of education services. These results may be read as a failure of societal participation. Taking bureaucratic norms into consideration, however, the lackluster results are not surprising (Fox Reference Fox2015, 349). Under a legalistic bureaucracy, voicing complaints through VECs will tend to generate administrative burdens on marginalized citizens, while doing little to mitigate the social stigmas and reprisals that may come with complaining. To make participation more meaningful, attention to what Fox (Reference Fox2015) calls the “vertical” dimension of accountability is required. Citizens need institutional backing to question frontline service providers, which means that local agencies must work to mitigate practical obstacles that come in the way of such questioning. Otherwise, VECs are likely to achieve little more than rule compliance and “thin accountability” within the school system (Dubnick and O’Kelly Reference Dubnick, O’Kelly, Frederickson and Ghere2005).

To make local accountability “thick,” higher-level bureaucracies have a critical part to play in strengthening societal input. It is not by coincidence that some of the more effective social programs in developing countries assign a prominent role to central bureaucracies, including Mexico’s Progresa and Brazil’s Bolsa Família, both of which are administered by federal agencies (Sugiyama and Hunter Reference Sugiyama and Hunter2013; Weitz-Shapiro Reference Weitz-Shapiro2014, 153–164). Central agencies may help curb local clientelistic interferences, but also guide and support bureaucratic problem-solving, lessons that Tendler (Reference Tendler1997) identified in her research on public health reforms in Brazil. Through deliberation with lower-level agencies, central agencies reinforce the purpose of local participation, helping to align local practices with programmatic goals.

Turning specifically to India’s primary education system, this study offers policy guidance for Indian states as they gear up to implement the new National Education Policy (NEP). Enacted in 2020, the NEP is a bold policy document. It acknowledges the shortcomings of government schools, evidenced by the rapid growth of low-fee private schools targeting poor households (Kingdon Reference Kingdon2020). Replacing the previous policy of 1986, the new NEP calls for an overhaul of the education system, with changes to school curriculum, pedagogy, governance and regulation. Foundational literacy and numeracy in the early years of schooling are emphasized, along with systems of assessment to measure and track learning outcomes. The NEP recognizes the need to shift classroom pedagogy and assessment away from the traditional emphasis on rote learning and compliance with overambitious school curricula (Pritchett and Beatty Reference Pritchett and Beatty2012). Instead, it encourages the mastery of concepts and analytical skills. These classroom-level changes will require updates to teacher training as well as continuous administrative support. Noting the importance of early child development, the NEP calls for universal preschool services. Centers specializing in childcare and nutrition support under the Integrated Child Development Services (ICDS) program will be repurposed to provide pre-primary education. Anganwadi workers, frontline agents managing ICDS centers, will be assigned new responsibilities for delivering preschool education.

Like so many social policies crafted in New Delhi, realizing the NEP’s promise will turn on the actions of state governments and local administration. The policy will also require investment in augmenting the state’s capacity to deliberate. For example, pre-primary education will require interagency coordination between the Departments of Education and Women and Child Development. Moving past conventional school curricula to support foundational learning means that teachers will need to adapt pedagogical approaches to the diverse needs and learning levels of different students. From the experience of past reforms, it appears that legalistic bureaucracy in India can institute changes involving compliance with rules and well-specified targets, but it faces difficulties in instituting pedagogical innovations, which belie easy codification through rules (Aiyar et al. Reference Aiyar, Davis, Govindan and Kapoor2021). Detailed policy blueprints are unlikely to achieve the necessary shift in bureaucratic understandings and behaviors, and may even have the opposite effect, as seen in prior attempts to adopt “Teaching at the Right Level” (TaRL) in Bihar. The adoption of TaRL faltered, not due to poor policy design or even a lack of political will within the Bihar government, but because frontline administration and schoolteachers were ill-prepared to embrace pedagogical practices that conflicted with existing administrative rules and curricular requirements.Footnote ²³ Looking ahead, a more promising approach would be to encourage state governments to adapt NEP rules according to their varying contexts and also grant local districts and schools the flexibility to experiment with curricular changes based on the practical needs of students and teachers. Central agencies may continue to serve important functions, providing guidelines for local adaptation and institutionalizing processes for deliberation and learning across states and local districts.

A final set of implications pertain to education reforms in response to the COVID-19 pandemic. Owing to the public health crisis, school systems have faced prolonged closures, compounding inequalities in access and learning. Disproportionate learning losses have been experienced by children from low-income backgrounds, particularly those lacking access to technology, online resources and educational support at home (UNESCO 2021). India, which has experienced one of the longest school closures in the world (17 months or longer in some states), has seen a decline in student participation and learning, among girls in particular (ASER 2021). Unless mitigated, this crisis may cause irreparable damage to the education and life chances of millions of children. Given the pandemic’s unequal impacts, education researchers have argued for more personalized instruction for children of different social backgrounds (Reimers Reference Reimers2021). Relatedly, there are calls to institute remedial education as a regular feature of schooling, to ensure that all children gain foundational skills (Beeharry Reference Beeharry2021). For personalized and remedial education to succeed, policy efforts must also go toward augmenting bureaucratic capabilities to deliberate and adapt local programming. Equally, the pandemic has made clear that public health, education and other social services are interlinked. Building societal resilience raises the need to strengthen deliberation between different frontline services and agencies.

At the same time, the pandemic has opened spaces for reform, encouraging new practices and the uptake of educational technology (EdTech). India is becoming a leader in digital learning tools, with sizable investments going into EdTech start-ups.Footnote ²⁴ We have yet to see where the growth of EdTech ultimately leads. I would caution against either celebratory or cynical readings of technology’s potential to transform public services. In India, the reality is that almost a third of rural households with children of school-going age do not own smartphones (ASER 2021). There is a digital divide even within households owning smartphones, as girls and younger children face greater barriers to access. Meanwhile, household access to smartphones has been improving and some Indian states have adopted digital platforms to monitor government schools and foster communication between teachers and parents. Others are using digital technologies to provide teachers pedagogical training and support for customized learning. While some technological interventions appear promising, we are only starting to learn about their impact on teachers and students (Muralidharan, Singh and Ganimian Reference Muralidharan, Singh and Ganimian2019; Cilliers et al. Reference Cilliers, Fleisch, Kotze, Mohohlwane, Taylor and Thulare2022). From a system-wide governance perspective, digital communication technologies may support deliberation within schooling systems, though the quality of interactions they facilitate remains an open question. Technologies may also weaken deliberation, for instance, by reinforcing a culture of compliance with technical standards and metrics for thin accountability (Honig and Pritchett Reference Honig and Pritchett2019). Along with improving access, much will depend on the social practices surrounding the use of technologies as they become embedded within education systems.

9.7 In Closing

This book has put forward a new argument for what makes bureaucracy work, especially for the least advantaged. In contrast to explanations emphasizing formal rules and structures of the state, I have highlighted differences in bureaucratic norms, an underappreciated but critical feature of public institutions that shapes how officials behave and relate to citizens. I developed and tested my argument connecting the divergence in bureaucratic norms to variation in the implementation of primary schooling across northern India. Based on a multilevel comparative analysis, I have shown that bureaucratic norms have meaningful effects on the delivery of primary education, with further consequences for how citizens experience the local state. This concluding chapter has shown how my argument may advance scholarly debates in comparative politics, public administration and development. I have explored policy implications as well and suggested how reforms might encourage bureaucratic deliberation in a sustainable manner to help address the need for quality education. There has been a profusion of research on distributive politics and public goods provision in developing countries. Much of this work paints a bleak portrait of bureaucracy, and understandably so. In India, ordinary experiences of the state are often disappointing, more so in the case of services like primary education. Yet, as I have tried to show, bureaucracy has multiple faces. The same agencies that succeed on some tasks can flounder on others. Agencies can also go beyond the call of duty and achieve results that exceed our expectations. From a distance, the vagaries of bureaucracy can appear arbitrary, but when investigated up close, significant patterns emerge.

We still have much to learn about when, why and how bureaucracy works in developing countries, and why some agencies do so more effectively than others. I have suggested that a comparative, multilevel approach to the study of policy implementation can illuminate important aspects of these puzzles. The politics of implementation covers a vast terrain and multiple policy arenas. This book has explored the understudied arena of primary education, a public service that touches many people’s lives. As I have shown, primary schools are political sites, where state bureaucracies project authority and impart lessons of citizenship (Bruch and Soss Reference Bruch and Soss2018). The delivery of primary education is also intimately tied to the political development of modern states (Ansell and Lindvall Reference Ansell and Lindvall2020). To grasp these political dynamics requires us to appreciate bureaucracy as an institution and take cognizance of the norms that bind it.