Two Cheers for Decentralisation: Unpacking Mechanisms, Politics and Accountability in the ICDS, Central India

Abstract

From long-term qualitative research this paper argues that Chhattisgarh’s decentralised mechanisms in implementation of the Integrated Child Development Services (ICDS) foster increased stake of local communities and local politics, and that these work to enhance accountability and programme quality. A larger number of actors with financial interests in the scheme at the level of the sub-block ICDS ‘sector’ and down to the village leads to wider distribution of financial gains from delivery of ICDS Services, thereby increasing local competition and political interest from lower tiers of governance. Decentralised mechanisms work to enhance checks and balances via formal and informal routes to governance and accountability. Chhattisgarh’s ICDS represents a ‘hybrid model’ between the short and long routes to accountability. While competing interests from local politics and institutions of governance work to improve ICDS accountability, they also work to appropriate the programme for political gain, or unfairly target ICDS workers. The paper unpacks mechanisms by which local politics relate with decentralised prescriptions in ICDS implementation. It gives the decentralised mechanisms a qualified two cheers.

Background and Overview

The idea of decentralisation occupies an important place in both the theory and practice of social development. Central to the agendas of accountability and good governance (Faguet, 2014; Agrawal and Ribot 1999; Véron et al. 2006), decentralisation is an important prescription of the World Development Report, which defines it as ‘the formal devolution of power to local decision makers’ (World Bank, 2003). The idea is rooted in the perceived economic advantages of federalism and a critique of economic centralism (Campos and Hellman 2005), and prescribes the devolution of political, administrative and fiscal powers to sub-national units of government (Bardhan 2002; Campos and Hellman 2005; Burki et al. 1999). The assumption is that a retreat of the centralised state and bringing of decision-making processes closer to the community would improve the relevance of policy decisions and hold public officials accountable (Khaleghian and Gupta 2005; Faguet 2014; Crook and Sverrisson 2001). Given the global prominence the idea occupies, several countries have undertaken reforms towards decentralisation and these have been in various capacities (Faguet 2014), and have displayed different trajectories, being also shaped by contextual state - civil society relations (Heller 2019). In the early 1990s India’s landmark constitutional reforms formally recognised a third tier of local governance at the sub-state level, that of the Panchayati Raj Institution (PRI) and Urban Local Body (ULB). For a country as large and diverse as India however, the actualisation of the ideals of decentralisation or institutionalisation of local governance units has been variable, and is influenced by the social and political context of the states (Rajashekhar 2021; Heller 2019; 2011; Manor 2011; Palanithurai 2017; Bhide 2017; Alok 2013, 2018; Ghosh and Kumar 2003; Crook et al. 1998; Johnson 2003; Johnson et al. 2005; Widmalm 2008; Harriss 2010; Lieten and Srivastava 1999).

While democratic decentralisation concerns federal politics or sub-state governance, decentralisation as an idea is broader in scope and extends to incorporation of its processes in implementation of social welfare programmes. Decentralisation is also argued to promote participatory development, a prescription also long commonplace, and the new development orthodoxy.1 Likewise, the formal incorporation of principals of decentralisation, or participation mechanisms, has been variable across sectors, across social welfare programmes, and across India’s states. Although decentralisation continues to occupy an important place in the theory and practice of development, empirical analysis on the operationalisation of decentralisation or on whether its assumed benefits materialise in practice is limited. What evidence exists presents a mixed picture for India with contested views about whether objectives of decentralisation are achieved. For rural India, evidence indicates that decentralisation contributes to elite capture, continuing power of political brokers or curtailment of the corrective hand of state agencies (Rajashekhar et al. 2018; Rajashekhar 2021; Sivanna and Veeresha 2021; Manor 2011; Khalingian and Gupta 2005; Sundar 2001), while other evidence points to its promises for democratic deepening, and what has been called, by Patrick Heller ‘a silent revolution’ via panchayats (Heller 2011; Abraham 2014; Manor 2010).

The related theme of public sector ‘accountability’ is central to the agenda of good governance and decentralisation. Here, the 2004 World Development Report has been influential in shaping the discourse on accountability frameworks, distinguishing two routes to accountability, the ‘long’ and the ‘short’ via the power relationships between citizens, the state and public service providers. The ‘long’ route is via authority delegated to political representatives who then govern bureaucracies by choosing policymakers who in turn manage frontline service providers. The ‘short’ route, in contrast, links citizens directly to service providers, through various ‘choice’ or ‘voice’ mechanisms (World Bank 2003). In India, we observe mechanisms such as ongoing elections or Public Interest Litigation, which are, via the long route expected to further accountability by shaping policy and staff performance, while mechanisms such as citizens’ complaints, social audits, the right to information act, citizens’ report cards or participatory budgeting are examples of short route mechanisms that may work in the more proximate term (Blair 2018). In their expansion of the World Bank accountability frameworks to include the role of civil society, Devarajan et al. (2011) point to the implications of the underlying political economy drivers of accountability, also for how civil society is constituted and functions. They argue that when higher level political leadership provides sufficient or appropriate powers for citizen participation in holding within-state agencies or frontline providers accountable, there is frequently a positive impact on outcomes. This paper contributes to scholarship from examining decentralised mechanisms of implementation in India’s flagship child development prescription—the Integrated Child Development Services (ICDS). It examines the entanglements these have with community participation, local politics, institutions of governance and accountability.

While social inequalities and manifold determinants underlie poor child development outcomes, direct interventions can play an important role in enhancing child wellbeing. India’s ICDS is the largest scheme of its kind globally, and while the centrally funded scheme was launched in 1975 on an experimental basis, it has since expanded considerably in scope and geographical coverage. Following collective advocacy from civil society groups, since 2001, India’s Supreme Court has issued multiple interim court orders on ICDS universalisation, and India’s National Food Security Act of 2013 enshrines in law nutrition support for women and young children through the ICDS.2 The ICDS seeks to lay the foundation for long-term growth and development through provisioning a set of interrelated services with a focus on the 0-6 years of life. These include supplementary nutrition, growth promotion and monitoring, nutrition and health education, immunisation, health services, referral and pre-school education. Given its community-oriented activities operating from a village-based centre or aanganwadi which in Hindi means a ‘courtyard centre’, the name emphasising its habitation or neighbourhood location, the ICDS has since inception held central the idea of participation. The idea is, however, usually limited to user participation in service use rather than wider community participation in programme implementation, governance or accountability. Critics argue moreover, that the ICDS has had poor participation even by standards of service use as the programme is often viewed as providing merely a hand out or dole, while its wider impact on lives of mothers and children is overlooked (Gragnolati et al. 2006; Saxena and Srivastava 2009). Although ICDS ‘coverage’ by criteria of sanctioned centres and staff has expanded significantly, several pressing questions remain about programme effectiveness or implementation quality. Studies reveal significant shortcomings in delivery of ICDS promises and considerable variation in the range, quality and reach of services, utilisation of funds, or embezzlements therein (Jain 2015; Planning Commission 2011; Adhikari and Bredeenkamp 2009; Gragnolati et al. 2006; Chanchani 2017b). Moreover, states demonstrate dissimilar interest or additional investment in the programme and this has resulted in profoundly different implementation outcomes, as is demonstrated, for instance, from comparative research in districts across 6 states (FOCUS 2006), which was repeated after 10 years (POCUS 2016).

Although Chhattisgarh is commonly viewed to be poor and is clubbed with states of India considered backward by generally accepted indicators of social development, the state has developed a reputation for good governance over the past two decades.3 This reputation remains even as intra-state disparities exist in administrative reach or quality, as do differences in normative understandings of development or welfare ideals - differences most sharply visible in forested regions of the naxalite–maoist conflict in southern regions of the state.4 Chhattisgarh is an interesting state to examine ICDS governance and accountability as although it yet has high levels of poverty, it is distancing itself from the reputation of being developmentally backward. Moreover, the state has taken noteworthy initiatives in improving design and implementation of social welfare programmes. Chhattisgarh has, for instance, been lauded for its Public Distribution System (PDS) reforms that have improved last-mile delivery of foodgrain, and the community health worker, Mitanin programme was a frontrunner for the similar, national, ASHA programme.5 With respect to the ICDS, Chhattisgarh has adopted a distinct approach by decentralising implementation of important programme components relating to procurement and supply of supplementary nutrition to the common women’s Self Help Groups (SHG), or locally samuhs at the level of the village or the sub-block ICDS sector. This is salient because the supplementary nutrition component of the ICDS is also the most substantial by budgetary allocations, making it a particularly lucrative avenue for misappropriation of funds by vested interests, or in other words, with potential for ‘leakage’—a ubiquitous term used to describe corruption or embezzlement of public funds in India. The bureaucracy–politician–contractor nexus in the supply of ICDS supplementary foods is well recognised, and as a norm, most states have awarded food supply contracts to a single firm or a small set of firms favoured by the establishment for centralised supply across the state.6 Following advocacy from activist groups across the country, in 2004 India’s Supreme Court passed a series of orders for the removal of contractors and the decentralisation of supply of ICDS supplementary nutrition by women’s groups and village communities. While most states have openly violated the Supreme Court orders on decentralisation more than a decade after they were passed, Chhattisgarh is among the very few that followed orders on the removal of contractors and decentralised supply of ICDS foods to women’s groups relatively soon after the court directives.7 Furthermore, in Chhattisgarh, the local units of governance - the Panchayati Raj Institutions - play other important roles in, for instance, provisioning of ICDS building infrastructure. This paper examines the implications of these decentralised processes and their interplay with local politics for programme accountability.

Over the past few decades, a body of scholarship has examined the anthropology of the state and local development politics. Anthropological literature argues the importance of seeing the state as culturally embedded and discursively constructed (Sharma and Gupta 2009), while emphasising that the state is not static or unitary as a conceptual or empirical entity, but is instead translocal, multi-layered, multi-centred and fluid. Literature moreover points to the blurring of boundaries between the state and society, or a collapsing of distinctions between representatives of the state performing their roles as ‘public representatives’ and ‘private individuals’ (Gupta 1995; Harriss-White 2003; Fuller and Benei 2009). In particular, anthropological literature argues for moving beyond the macro-level institutional analysis of the state, to directing attention towards ethnography of everyday bureaucratic practices of the state and to the nitty-gritty of subjective engagements at sites of interaction with government processes (Mathur 2012; Sharma and Gupta 2009, 27; Gupta 1995, 2012; Fuller and Benei 2009; Corbridge et al. 2005). This paper seeks to contribute to scholarship in this area from observing everyday processes of ICDS implementation from long-term engagement in the region of Chhattisgarh’s rice-growing plains. It contributes to literature on how processes of decentralisation may influence programme delivery and accountability and to literature on the everyday state as it relates to society. It further contributes to implementation research in public nutrition.

In the “Research Setting and Methods” section that follows, I describe the research setting and methods including here the politics of access in researching the everyday state. I then illustrate, in “Decentralisation, Grievance Redressal and the Improvement of Quality”, how decentralisation influences community participation and relationships of accountability. I unpack the ways in which decentralisation sets in play grievance redressal mechanisms and investigate implications these have for accountability relationships, service delivery and equity. The concluding “Discussion” section unpacks the findings for the socio-cultural context of Chhattisgarh and highlights how this research contributes to literature.

Research Setting and Methods

The data I present in this paper is generated from rural and peri-urban regions of Chhattisgarh’s rice-growing plains and from engagements with the state-level bureaucracy in the capital city Raipur. The rice-growing plains of central Chhattisgarh are densely populated and of enormous political and economic significance to the state as they are home to major industrial cities of Raipur, Durg and Bhilai, and because rice farming remains the bedrock of the rural economy. The population of the plains is predominantly made up of caste groups, and the region has a relatively small population of adivasi communities or tribes, who have a larger presence in the northern hills of the state and the southern Bastar plateau. The caste groups of the central plains are primarily classified into Other Backward Classes (OBC) or Scheduled Castes (SC), and the region, like Chhattisgarh as a whole, has a very small population of privileged castes, or caste groups commonly categorised as upper or forward castes.8 My observations on bureaucratic processes and ICDS programme implementation draw on fieldwork in rural regions of Raipur districts, and two blocks I call Hasaud and Giroha therein. I also observed the programme in Dhamtari district, 100 kms from Raipur, the capital city, with a view to build a larger experience base and rule out any major disparities that may result from closer proximity of the rural research blocks of Raipur district to the capital city. I, along with a research associate, made multiple visits to the region between April 2015 and June 2016, mostly to Giroha and Hasaud. The 14-month window allowed us to observe both one-off and ongoing ICDS activity, and while certain villages were visited repeatedly for longer-term observation, a wider set of villages within the blocks Hasaud and Giroha were visited to examine a range of aspects. The analysis I present draws on this engagement and also on my longer-term research engagement in the region over a period of 12 years.9 Primary data were generated from interviews with informants such as from aanganwadi workers, ICDS users, representatives of the elected local government, staff of the ICDS operating at all levels of programme organisation—the state, district and sub-district units, representatives of non-government organisations and representatives from the bureaucracy. Likewise, data are generated from observation of ICDS centres and the range of related activity in villages and government offices. As with study blocks, I have anonymised the identity of villages and respondents in the text.

In qualitative research on the Indian state, especially research that intends to examine implementation of public programmes, negotiating the politics of access to information is a central challenge. Gaining admission to observe programme implementation settings and bureaucratic process requires delicate negotiation with gatekeepers, and at the level of the district, the key sub-state administrative division, initial introductions shape how research objectives are perceived, and the nature of information that local actors share. Although taking the pathway of seeking official sanction from senior bureaucrats would ease initial or ongoing ‘contact’ with district level officers, such instruction would, it was envisaged, have expressly shaped and circumscribed the nature of the field interaction. This is because state employees routinely, and increasingly, exercise control over what ‘field’ visitors such as ourselves encounter and what narratives we are exposed to. If we had carried introductions from higher-level bureaucratic superiors, as is said ‘from above’ or uparse, our block-level research is likely to have been perceived to be more ‘official’ or of greater importance and with potential for feedback to administrative superiors and to authorities beyond. This, it is readily envisaged, would entail significant gatekeeping on settings or informants the research team might access. While officers across the system nonetheless exercise control on what information they expose to outsiders-to-the-system such as ourselves, if we were perceived as individuals of importance or influence, efforts towards controlling the narrative are expected to have been greater. Given this, although the research purpose was upfront as was our identity as researchers - affiliated with an international university who were conducting a study on the ICDS and nutrition - we were often walking the tightrope between gaining official sanction for our inquiry while maintaining an informal purpose and position. Although some authorisation from senior officials was important for ICDS staff members to take us seriously and allow repeated contact, it was also crucial to build independent relationships with more junior officers or service providers and make some field visits not mediated by line staff. ICDS staff in Chhattisgarh were overall open to our inquiry and presence.10

Decentralisation, Grievance Redressal and the Improvement of Quality

The Case of the Rotten sabji

On a day in January, 2016, Mr. Sathe the unusually forthcoming Community Development Project Officer (CDPO) of Giroha block, the nodal ICDS officer at the block level, had a matter to attend to in the community. He had received a recent complaint that the village-level women’s group or samuh responsible for food provisions to two aanganwadis in Tiloda village of the Giroha Nagar Panchayat or Municipal Council was unmindful of quality, and had in the past week supplied vegetables or sabji that was rotten and unfit for consumption. This meant that children attending the centre had to forgo, over two days, the valued and nutrient-rich sabji element of the aanganwadi meal. There was also a related matter to resolve, that of renting new premises for the aanganwadi, as its current location was inappropriate and in unhygienic surrounds.

Mr Sathe visited the aanganwadi in question, in a temporary location. He interacted with the gathered children, put them at ease and engaged in friendly conversation. He noted that the vegetable the samuh had supplied, capsicum or green pepper, was indeed unusable. He pointed out programme implementation shortcomings to the aanganwadi worker or karyakarta and offered some achievable suggestions to enhance quality. His manner was non-patronising and respectful.11 Next he proceeded to examine the suitability of the premises identified as a potential new location for the aanganwadi. The property appeared appropriate. It was located centrally and had essential infrastructure in place—two rooms, a toilet and a courtyard. The building did not have electricity and the floor was not tiled. However, Mr. Sathe deemed these elements could be overlooked, as other essential criteria were met, and there were no immediately available alternatives that were better.

At the meeting held in the office of the village school, Mr. Chaubey the Block Education Officer (BEO), was also present. Apart from two aanganwadis, the samuh was also responsible for provisioning supplies for the school mid-day meal. Given this, the education department too had received a complaint about poor quality sabji. Other participants at the meeting were the headmaster, a teacher, the parsad or elected member of the nagar (municipal) panchayat who had put forward the complaint, and the ICDS supervisor for the region Ms. Hemlata. Five women from the samuh stood in a row at the entrance to the room. First the relatively straightforward business of securing the new aanganwadi premises and negotiating rent was resolved, and following this the meeting turned to the matter of the rotten sabji. Mr. Chaubey, the BEO had already addressed the samuh before Mr. Sathe’s arrival. At his time, Mr. Sathe took a firm stance about food quality albeit adopting a polite approach. He simultaneously both reprimanded the samuh while also seeking to inspire or encourage better quality.

‘If we receive a complaint about gunvatta (quality) then usually we remove the samuh’, he warned. ‘I must not get another complain about the quality of food for children’. This language of admonishment was accompanied by simultaneous encouragement of samuh members to view the provisioning of good food as their kartavya or duty, explaining that village children accessing the aanganwadi were indeed their own, and it was ‘their own children’ who suffered poor quality food. Hemlata, the ICDS supervisor responsible for the ICDS sector of this region, echoed this stance. The samuh women, keen to retain ICDS contracts, assured Mr. Sathe that the slip up was an isolated incident and that they would not be found wanting again. Mr Sathe additionally encouraged the samuh to ‘keep an eye’ on the aanganwadi, as mothers or residents of the village who were also service users with a view to ensure that children were receiving their entitled services. On this latter point, about their role as a ‘vigilance’ or nigrani committee, members from the samuh responded in one voice that the elected representative or ‘parsad’ had admonished them for seeking to keep a watch on the ICDS. The parsad had maintained that any ICDS ‘monitoring’ should be left to him.

The Case for Decentralisation and Community Involvement

Chhattisgarh has fostered processes of decentralised implementation and community involvement across many of its departments,12 and as noted, has implemented orders from India’s Supreme Court on the removal of contractors responsible for the centralised supply of ICDS supplementary nutrition foods. The supplementary nutrition component of the ICDS has a considerable budgetary allocation for two types of food-based interventions—hot meals to be provided at the aanganwadi, and Take Home Rations, henceforth THR, which are nutrient supplements, usually in powdered form to be distributed for consumption by pregnant women and younger children between 6 months and 3 years of age.13 Given the large budgets and the related scope for corruption and profiteering, the distribution of contracts for THR supply is commonly fraught with political interference.

Chhattisgarh was formed as a new state in the year 2000, and in 2007 replaced the three contractors then responsible for the supply of THR foods for the entire state, with a decentralised food supply model that involved over 3000 women’s Self Help Groups (SHG) or samuhs, each with contracts to provide THR for a sub-block region, or ICDS sector. Likewise, for ICDS cooked meals at the village-level aanganwadi, local samuhs were made responsible for supplying ingredients for meal preparation, the latter being the responsibility of the aanganwadi staff. Both these decentralisation measures have critically enhanced the stake of community members in the ICDS, as beyond their role as service users, a wider set of individuals now have economic interest in the ICDS. This then has also increased local political attention to the programme and changed the dynamic of relationships between local elected representatives and ICDS line managers. The samuhs are expected to be community-based providers of food supplies and have impermanent contracts from the ICDS.

Complaints about Tiloda’s aanganwadi samuh emerged during a 9-day ‘quality campaign’, or Gunvatta Abhiyan carried out by the ICDS department. The campaign sought to foster community involvement with the ICDS and examine implementation shortcomings, while mandating involvement of panchayat representatives, government employees, and officers from both outside and within the village. In the case from Tiloda, while the vegetables provided by the samuh were indeed of inferior quality, the elected parsad who had put forward the complaint, perceived in this shortcoming a larger political opportunity, beyond the concern about food quality for the children. Given that samuh contracts with the ICDS are not permanent, he sought retraction of contracts from the current samuh, with a view to make available the economic opportunity to another samuh, which was affiliated to him. A common narrative in rural Chhattisgarh, central to the complaints process in the ICDS as indeed to other programmes, relates to village-level jealousy or jalan about the economic position of others. In the words of a CDPO from Rimapur of Dhamtari district for instance, “Nobody is happy to see another person succed. If they see someone succeeding they want to bring them down. If they see a samuh making money, they think—why not a samuh affiliated to me?”.

The village-based women’s samuh is now ubiquitous, but in the early years when contracts for supplying provisions for village-based ICDS meals were first handed out across Chhattisgarh, this was not the case. Given this, few groups met the required conditions for participation as suppliers in the ICDS programme. In Tiloda for instance, as we note above, that one samuh held supply contracts for two of its three aanganwadis, as well as for supplying provisions for the meal at the village school. With progression of time as samuhs multiplied and it was observed that village women benefited financially, the competition for ICDS contracts intensified. Such contracts for food provisioning or other low-rank village-level work such as of a health assistant or school cook are impermanent, but sought after in an environment that offers few other opportunities for village-based income not dependent on manual labour. Given this, control over such jobs or contracts become central to political negotiation, clientelism or score settling.

In the case of Tiloda, Mr. Sathe, the senior most authority of the ICDS department in the block was required to adjudicate on what was, in essence, a political complaints process. The case displays performative elements by Mr. Sathe, who seeks to complete procedural requirements of examination of the aanganwadi premises; admonishing the samuh in front of other officials, issuing a warning while also encouraging better compliance with requirements. He also speaks to the larger participatory aims of community involvement by encouraging the samuh to play a vigilante role. This public course is in keeping with requirements of the inquiry and his role. In subsequent conversation back in his office, however, Mr. Sathe shared the backroom processes. He shared that he himself did not wish to revoke the supplementary nutrition contract from the samuh unduly as he knew the case of the sub-optimal sabji was a one-off incident. However, he admitted that the parsad being ‘against’ the group, the word against used in English, could force its removal. It was expected by all that Tilodas samuh, after the reprimand, would provide better quality ingredients in the self-interest of protecting ICDS contracts.

Other examples further illustrate how awareness of the rules because, as an officer in Dhamtari puts it, “Everybody is jagrut or ‘awake’ these days leads to heightened observation of the ICDS and enhances accountabilty. The government has long been making people aware of their entitlements. Noteworthy in Chhattisgarh for example is the high compliance aanganwadi workers display to regular opening hours. Over multiple unannounced visits to a village it is unusual to find ICDS centres not open during work hours or for the aanganwadi worker to be absent without prior permission. This is in contrast to what is commonly observed in comparably poor states in north India. Although children in Chhattisgarh return to their homes soon after the ICDS hot meal is served around noon, and the aanganwadi offered little by way of organised early years education in encouraging them to stay, aanganwadi workers routinely remained, alone in an empty aanganwadi until the official hour when they lock the aanganwadi and return home around 3:00 pm. This maintenance of official hours, even after the key activity for the day was complete and children had left, is noteworthy, and aanganwadi workers consistently report the fear of a shikayat or a complaint against them as a reason for their compliance to official hours. In the words of the aanganwadi worker from village Dolchidani “There is no dearth of people to complain about you”. While aanganwadi workers who may have affiliations with political opponents of current local elected leaders such as the village sarpanch or parsad of the urban municipal corporation have more reason to fear a shikayat than those affiliated to the current elected leaders, the overall nature of grassroots competitive local politics has contributed to an environment where compliance to aanganwadi hours is maintained. In the words of an aanganwadi worker from Pannenpara, “no matter whether the sarpanch supports you or not, there will certainly be someone who is ‘against’ you”.

The above cases illustrate the mechanisms by which decentralisation and grievance redressal processes work in an environment of insecure employment or service provision contracts that are expected to encourage participation. Local government employees routinely assert that the high level of jagrutta or ‘awakeness’ in the populace has resulted in a system that is routinely responding to complaints, and that this pressure on the system has resulted in greater accountability and improvements in gunvatta or programme quality. The view is summed up in the words of the ICDS supervisor of Giroha block, which are expressed as a self-evident truism “Where there are complaints (shikayat), there quality (gunvatta) improves”. While such complaints stem from aspects relatively easy to monitor — such as aanganwadi opening hours or food received by a child, rather than the more intangible elements such as quality of health or early childhood education, or factors more dependent on participation of the larger community, it is envisaged that the complaints and grievance redressal processes contribute to ongoing and overall quality improvements. While there are positive influences on ICDS quality, local elected representatives may unfairly target workers or have inordinate power, which also work against equitable distribution of services, as the following discussion illustrates.

Building of aanganwadi bhavans: Too Much Power with the Panchayat?

Building a bhawan (centre) in a village remains a constant problem for the (ICDS) department and it is usually the sarpanch who has the last say. However, what they (the village people) do not realise is that the centres would not benefit us (the ICDS line department) but instead children from their villages. From our end, we feel it is the duty of the supervisor (the ICDS employee who the aanganwadi workers report to) to assess the location allotted by the sarpanch and to judge how appropriate it is to community needs. There are cases when the sarpanch refuses to budge and we may have to consult a tehseeldar (revenue officer), however, like always, the sarpanch creates some ‘commotion’ and gets his way around things.

[Mr. Sathe, CDPO, Giroha Block, Aarambh]

The majority of aanganwadi centres in Chhattisgarh operate from their own building or bhavan. This means that the premises of the aanganwadi centre belong to the ICDS department and are not rented from the village panchayat or another entity. Provisioning of ICDS bhavans is high on the government’s priorities, and the department of Women and Child Development has been innovative in securing funds for centre construction from various schemes, including from the Sarva Shiksha Abhiyan (SSA) of the department for education. Although there is inter-centre variation, aanganwadis in Chhattisgarh are commonly secured by a boundary wall, have dedicated indoor and outdoor space for children to gather, space to store equipment and provisions, a kitchen and a toilet. A ‘model’ aanganwadi centre also has more substantial play equipment such as a slide or a swing. Although proactive in making available construction funds, the ICDS department has limited say in defining the physical location of bhavans in the village, as this is the remit of the sarpanch - the elected head of the village panchayat, or the equivalent in the nagar municipal panchayat. This is also because the local body is responsible for allocation of land for the bhavan and responsible for its construction.

We observe a clear pattern of political considerations mediating decisions for location of the centre or bhavan, and these often work against equitable geographical distribution of ICDS services. Although the availability of vacant village common land poses a problem, the primary considerations shaping bhavan location are political. Factors such as the geographical location of a ruling sarpanch’s support base; the relationship between the sarpanch and aanganwadi worker; or the political affiliation of the aanganwadi worker, all influence land allocation. While in aggregate terms ICDS coverage in Chhattisgarh by criteria of centres by population norm is considered good, centre distribution within villages is uneven. Furthermore, while the location of centres operating from rented buildings could be temporary, a centre having its own building makes the location more permanent, with lasting implications for unequal service distribution. To illustrate, in Pattipurwa, a large village with a population of 4500, money was allocated to build a bhawan with the intent to make services available to a neglected basti or hamlet called manibasti, in the words of the informant, “the bhavan was sanctioned in the name of the children of manibasti”. However, the sarpanch was reluctant to assign space for the centre close to the said habitation, and instead allotted land at the village periphery, overlooking the concern of the ICDS Supervisor that the distant location would negatively influence aanganwadi attendance. The sarpanch’s decision was shaped by his wish to have the aanganwadi worker, who he did not favour, be forced have to move away from the current main habitation or central location of the rented centre, to a less desirable location of a new bhavan at the edge of the village, thereby also reducing her overall influence in the village. Likewise, in village Nariar, although a final decision on centre location was yet to be taken, the sarpanch viewed the building of a new centre as an opportunity to shift the work jurisdiction of the longstanding village aanganwadi worker he was opposed to, away from the main village habitation. This was from a consideration to further personal political scores with a previous sarpanch while also establishing that a newly appointed aanganwadi worker politically aligned to his family operated in the main basti. This was expected to shape future power and control that the two aanganwadi workers might have over village residents, or budgets that may be allocated by population served. In another example, from village Adhipara, two aanganwadis were located adjacent to each other on the main street of the village, as the sarpanch wished to save money by building a common wall between the centres, irrespective of the reality that this left another village location underserved. As funds for a third centre became available, the sarpanch was keen to construct this third centre adjacent to the first two. Given the obvious geographical concentration, the ICDS Supervisor had suggested another location, however, the sarpanch was inflexible to moving the centre from the habitation of his choice.

The ICDS line department, from the state to the district and below, recognises that the aanganwadi location is commonly inappropriate. The additional director responsible for ICDS in the state states, for instance, ‘‘Even though the money has been released to build more centres, the task of allocating space for the centre has been given to the sarpanch, who does what he pleases… We agree that the location of the centre needs to be re-evaluated, and it should be made more need-based”. He went on to describe that while the line department might recognise what would serve child development interests and could make recommendations, the local line department had little power to ‘insist’ (word expressed in English) on a location of its choice. In virtually every situation of contest, the choice of the sarpanch prevails, also because the ICDS department or supervisors do not wish to work, over the long term, in an environment of conflict or antagonism with local centres of power such as the sarpanch. Furthermore, the sarpanch or his supporters may find ways to frustrate efforts to build a centre in a location they do not favour. They may, for instance, cause delays from refusing to complete necessary paperwork or create local opposition. Moreover, the interests of the line department are to utilise construction funds within a normative timeframe before unutilised funds are stipulated to be returned, which would make the department fall short of its aanganwadi construction targets. These factors together result in the ICDS department habitually conceding to demands of local elected leaders on centre location. While advocates have long argued for ICDS to have its own infrastructure, in some villages, a centre operating from a rented building is perhaps better located and used than a centre operating from its own building, in a less appropriate location.

I turn briefly now to the other aspect of decentralised supplementary nutrition provisioning, to samuhs responsible for the supply of THR, who operate at the level of the ICDS ‘sector’ at the sub-block level. As discussed, Chhattisgarh is among the handful of states that has moved the production of THR from 3 contractors previously holding contracts for the entire state to over 3000 women’s groups, each typically serving an aanganwadi sector, which has between 15 and 25 aanganwadis. The THR production units run by women obtain supplies from the market, clean grain by hand, measure quantities, roast ingredients over an open fire and grind ingredients together, package the product and supply the THR to aanganwadis. Given the large budgetary allocations and scope of profit, THR contracts remain a subject of political tussle through the hierarchy down from the state to the district and below. Interests seeking a reversion to more centralised contracts for THR provision base their objections to the decentralised model through technocratic discourses of food quality. They argue that multiple, widely distributed, small-scale production units managed by samuhs produce foods of irregular quality - that the units are unable to adhere to exacting norms of nutritional composition; unable to fortify their products with essential micronutrients; and unable to maintain hygiene standards in processes that depend on foods being physically handled by humans through the production process.

They contend that children would benefit instead from a mechanised production process where foods are produced in processes ‘untouched by hand’, with stricter adherence to exacting normative standards of nutrition composition. Acknowledging the ongoing tussles, many within the bureaucracy and civil society groups argue that block-based samuhs rather than central contractors much removed from the community are likely to be more responsive to feedback on quality improvements, given samuh interests to retain contracts. Beyond questions of production politics however, there remain important questions on the nature, relevance and cultural acceptability of THR as a nutritional supplement for maternity or for young children, as the THR is rarely consumed in the recommended form (Chanchani 2017b). While the current research does not address the relevance of THR to child development outcomes, the decentralised models of samuh-led THR production and impermanency of contracts therein display similar tussles of political control.

Official norms dictate that in addition to signatures of the ICDS Supervisor (responsible for aanganwadis in an ICDS sector), the monthly accounts of the samuh must be signed by the sarpanch of the village where the group operates. While this step may work to exert quality checks on samuh production of THR, it is also opens it up to administrative discretion. In an illustration, ‘Sewa Adhipara Dham Samuh’ of village Adhipara, Hasaud, responsible of producing THR foods for 26 aanganwadis faced minimal interference from the community or panchayat in the 4 years of its operations, until the previous year, when a new sarpanch was elected. In the words of the secretary of the samuh “Since then we have faced unwarranted harassment… there are occasions when he (the sarpanch) would refuse to sign our supply vouchers and ask us to leave his place. We still don’t know what he wants. From our end, we informed the CDPO (block level ICDS head) who advised us to take signatures from another member of the panchayat for the time being, but that is only a temporary solution to the problem”. Exploring this further it was understood that here, as in other illustrations we note before, the reason behind the hostility of the new sarpanch was that a key member of the samuh holding the THR contract was a family member of the previous sarpanch.

Discussion

The case of Chhattisgarh is interesting because the state has, from within, fostered processes of decentralisation and community participation across many of its departments and welfare programmes. These are evident across the wide geography of the state, apart from pockets affected by the longstanding Naxalite–Maoist insurgency, where the state is accused of repression.14 Relative to many states in the country, in Chhattisgarh we witness devolution of powers to local elected representatives, models encouraging citizen participation in service delivery across a range of programmes, and forums for citizen engagement in governance, and for feedback to political representatives or line departments. We witness, for instance, initiatives such as the Gram Swaraj Abhiyan, a campaign by the state for information provision on the range of government services, and also department-centric campaigns for improvement of programme quality in specific schemes. Although the nature of state engagement varies across villages, as does the involvement of vulnerable communities, it is observed that state-led quality or engagement campaigns are widespread and remain important forums for citizen engagement. Likewise, Chhattisgarh has instituted mechanisms for decentralisation and community participation across a range of programmes, beyond ICDS. This is through, for instance, the community-based health worker or Mitanin programme and through enhancing panchayat control of the Public Distribution System for subsidised foodgrain, both setting examples for national programmes or reforms therein.

Proponents of decentralisation and participation argue that such processes promote pluralist pressures that along with local elections will provide the mechanisms by which direct and accountable relations can arise (Heller 2011; Abraham 2014; Manor, 1999; 2010; Crook et al. 1998; Fischer 2016, 2021; Nagarajan et al. 2014). However, there remains limited understanding of how governance processes or power relationships that transform them are affected by decentralisation (Fischer and Ali 2019; Faguet 2014). This paper seeks to contribute empirical material to illustrate how the assumed benefits of decentralisation may operate and how accountability relations, or power relations may play out in response to decentralisation measures. Moreover, it highlights how accountability relationships of grievance redressal triggered by decentralised measures often operate within the setting of contractual, informalised work, which is often the form that community, or more often, women’s participation takes.

We see in the multi-actor accountability relationships at the village and sub-district level, a tension in the balance of power and authority between the state agency, here the ICDS department, and elected leaders of local government. These tensions unfold in the context of direct financial stakes for the local community, which transform power relations and have striking implications for the interest that the local community displays in the ICDS. This is in contrast to what is witnessed across many states in India. To illustrate, district-based civil servants such as the collector, who is the key administrative officer, commonly remark that they receive few grievances from the community about functioning of the ICDS programme when compared to grievances they receive about functioning of other services such as the village school or state distribution of food rations. This lack of complaints, the civil servants claim, demonstrates the common indifference of the community to the ICDS or its services when compared to greater community interest in other services.15 The evidence from Chhattisgarh, where active local politics and decentralised implementation measures have recast community financial stake, indicates very much the opposite. There are here, as is expressed, ‘many people to complain’. The complaints process triggers what Mathur (2012) refers to as the ‘paper state’, the marked reliance and centrality of documentation, paperwork, files and the written word, added to other mundane bureaucratic practices of the Indian state. The grievance redressal machinery in Chhattisgarh is compelled to respond to a written complaint. That said, it is noteworthy that grievances are motivated commonly by monetary interests and political rivalries rather than quality of service or community displeasure in them. We find, for instance, that citizen participatory committees, such as nighrahi or ‘vigilance’ committees, conceived as user or more commonly ‘mothers’ groups responsible for monitoring the aanganwadi or enhancing community participation, play relatively unimportant roles as it is local political actors rather than service users who become the vigilante. In the long term, however, with improvements in service delivery and the continued experience of some consistently delivered ICDS services,16 service use and expectations of the community from the ICDS are also observed to increase. It remains important nonetheless to continue to scrutinise the implications of the political interest in the ICDS for the power-balance between frontline service providers, the ICDS department and local political representatives, and for fairness to employees or service providers such as samuhs. Raising a cautionary voice, a body of literature has argued that decentralisation needs to be carefully designed if it is to avoid the many pitfalls as a result of factors such as elite capture, lack of local accountability mechanisms, or policy fragmentation from weakening of central government agencies to the extent that their corrective hands are tied (Khaleghian and Gupta 2005).

In seeking to address the question of the balance of power between elected representatives and line ministry employees, Manor (2011) points to the need for both ‘horizontal’ and ‘upward’ accountability with a view to guard against possible abuse of overweening power by either camp—misuse of power by local leaders against line ministry employees, or, on the other hand, low accountability to local communities by line ministry staff, resulting in problems such as absenteeism or neglect of responsibilities. He suggests that such ‘balance’ might be achieved by giving significant powers over the future prospects of line ministry employees to both elected local bodies and to their bureaucratic superiors (2011, 11). Observations of relationships in Chhattisgarh reveal this balance and a degree of interplay. The line department has more direct control over the job security and prospects of both key workers of the village-based aanganwadi, the main worker—the aanganwadi karyakarta, and the assistant called the aanganwadi helper or cook. While in both these positions, the workers are conceived not as full employees, but as individuals who receive an honorarium for community services, the positions are secure, in that it is virtually unheard that an appointed aanganwadi worker or helper loses her employment within her expected service tenure. While the employment may be secure, local elected leaders can yet influence future prospects of village-based workers in terms of the prestige or honour they enjoy, additional opportunities that may become available to them, and the scope of, or support to their work. The influence of local elected leaders on permanent government employees is more variable and depends on the particular configuration of the power dynamic and support of politicians that local elected representatives may enjoy. While ICDS line department employees—the ICDS supervisors or CDPO—commonly negotiate to protect both their staff and ICDS programme interests, they are oftentimes unwilling to antagonise elected local leaders in a geography where they expect to work in the long term. It is where contracts are more informalised—such as those of village-based samuhs, as food provisions providers to aanganwadis, or samuhs managing THR production at a larger ICDS sector of 15–20 aanganwadis—that the balance of powers between the elected representatives and line employees becomes critical in shaping prospects of contract holders.

From interactions with state- and district-level ICDS officers, we notice that there is considerable awareness of the political power tussles at sub-state levels, and by and large, the department seeks to play a mediating role with a view to further ICDS performance, and protect employees and samuhs against unfair targeting. However, we witness also that the department often submits to demands of local political representatives, especially on matters of infrastructure such as of building bhavans. Notwithstanding a degree of inter-district variation, the Chhattisgarh administration and state agencies are not overly captured by feudal or partisan interests, and display both the capacity and intent to mediate towards furthering ICDS objectives and protecting frontline workers from unfair targeting. We see this in the attitude of Mr Sathe as he adjudicates on the matter of Tilodah’s samuh, in ‘workaround’ measures adopted to protect samuh THR contracts, in negotiations and awareness displayed by ICDS state-level line authorities towards bhavan construction or in upholding ICDS child development objectives. We observe, then, that state agencies do seek to exert that ‘corrective hand’. It may, however, for some aspects, be tied.

Given the blurred-boundaries between citizens and elected representatives of local self-governance, what we see in Chhattisgarh’s ICDS is perhaps a hybrid form of the world bank ‘long’ and ‘short’ route mechanisms to accountability, enmeshed with local panchayat politics, decentralised implementation contracts and, importantly, informalisation and impermanency of samuh or other ICDS contracts. While locally elected leaders are not direct makers of ICDS policy, they organise to influence the system, determine infrastructure location, choose local-level implementers and act, furthermore, as the voice of citizens in what is often a ‘political’ complaints process. Mechanisms such as the state-initiated Gram Swaraj Campaign, where the complaint about Tiloda’s samuh emerged serve here a purpose similar to a social audit or short-route ‘voice’ mechanism, in eliciting complaints and seeking to improve accountability.

Challenging the national bias in research on comparative welfare regimes, Tillin (2022) highlights the operation of significant regional autonomy and sub-national welfare regimes in India’s federal system between the period from 2009 to 2014. In this regard, the relatively widened coverage of social welfare programmes in Chhattisgarh, when examined comparative to several other poor states, is apparent. I would often remark about how impressed I was by the initiative and commitment displayed by the Chhattisgarhi state, across party formations, to a range of social development programmes, and to the relative openness the state displayed to civil society input. This was to me, especially notable in contrast to what is commonly witnessed in rural areas of India’s comparably poor northern states, most notably Uttar Pradesh, where the predatory nature of the state, or its dysfunction in serving ordinary citizens is apparent from even cursory observation. In one such casual conversation, a friend with long experience in the field of social development in the state remarked, ‘perhaps this (Chhattisgarh) was always a less-rotten part of the country’. This comment has struck me as particularly relevant in thinking about implications of Chhattisgarh’s underlying socio-political fabric for development programmes or governance reforms it has been able to undertake.

Devarajan et al.’s (2011) expansion of the World Bank accountability framework highlights implications of the underlying political economy drivers of accountability, also for how civil society is constituted and functions. They underline the positive impact of higher-level political leadership in holding within-state agencies or frontline providers accountable, and also the importance of historic institutions of poverty and inequality, or of ethnic identity in shaping public-interest action. They speak further of how such action is heavily constrained by existing systems and institutions of the state. In a similar vein, theoretical literature has pointed to the role of long-term institutions, power relations and social inequalities in shaping development outcomes (see for instance Rajan 2009; Acemoglu et al. 2005). Chhattisgarh displays curious social and gender characteristics and relatively egalitarian gender relations that defy easy generalisation or fit with northern or southern stereotypes (Chanchani 2019). These relations are also reflected in survival outcomes by sex, where the state displays near neutral adult sex ratios and more females to males in the child sex ratio (ibid). Furthermore, it is observed that the social structure, even in the caste group dominated central plains, is relatively flat, with a large population of Other Backward Classes and Scheduled Castes (SC), and a very small population of upper castes - under 3%, who are commonly considered immigrants into the region from the north. There are few, if any, readily observable economic differences between the Satnami classified as SC and other numerically dominant OBC castes such as the Yadav or Sahu. Although benefits to the Satnami from quota or reservation in jobs and education have contributed to this, observation of inter-caste interactions also indicate that social relations are, relative to other northern states, less feudal, and inter-caste inequalities not stark or acute.17 The relatively egalitarian social composition, reflected also in state employees of the ICDS and other departments, lays the ground for the nature of governance and accountability relationships or indeed civil society engagement that Chhattisgarh has been able to cultivate. Returning to Devarajan Khemani and Walton’s view of civil society, or citizen engagement, we see that ongoing political pressures notwithstanding, Chhattisgarh has often created the space for civil society groups to inform and engage with the state government, including allowing substantial consultation or roles in programme design, and oversight. This is readily observed in the departments of health; education; and women and child development—responsible for the ICDS. Recognising an apparent gap, the state has, for instance, partnered with civil socity expertiese to enhance the quality of the early childhood education offered as a part of the ICDS.

By way of a generalisable insight then, the evidence suggests that in such an environment, where both the state administrative systems and panchayats are strong, decentralised mechanisms—which enhance significantly the number of individuals with financial or political stake in the ICDS—force the system at various levels and especially at sub-district units, to a degree of attentiveness which works to enhance accountability and programme quality, while encouraging wider community participation. This is promising for a programme that has often lacked desirable levels of community engagement for multiple reasons. These positive indications go alongside the cautionary findings with respect to misuse of power by local elected representatives that have readily envisaged implications on programme quality, and other, perhaps inadequately understood, implications of panchayat politics on the social fabric.

For their enhancing of participation by a wider range of stakeholders, albeit in the form of informalised contracts which remain attractive to samuhs, and for their enhancing of competitive pluralist pressures which improve accountability, I argue that the decentralised measures in Chhattisgarh deserve a qualified, ‘two cheers’. That said, while everyday accountability on certain easy to measure indicators may be visible in the short term, child wellbeing or development outcomes also depend on a range of other factors—including programme design, state capacity, investment and community participation.

Funding

This study was supported by Department for International Development (Grant No. LANSA).

Declarations

Conflict of interest

The author declares no potential conflict of interest.