Forum for Development Studies

ISSN: 0803-9410 (Print) 1891-1765 (Online) Journal homepage: http://www.tandfonline.com/loi/sfds20

MGNREGA, Power Politics, and Computerization in

Andhra Pradesh

Silvia Masiero & Diego Maiorano

To cite this article: Silvia Masiero & Diego Maiorano (2017): MGNREGA, Power
Politics, and Computerization in Andhra Pradesh, Forum for Development Studies, DOI:
10.1080/08039410.2017.1345785

To link to this article: http://dx.doi.org/10.1080/08039410.2017.1345785

Published online: 03 Jul 2017.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=sfds20

Download by: [University of Nottingham] Date: 04 July 2017, At: 03:40

Forum for Development Studies, 2017
https://doi.org/10.1080/08039410.2017.1345785

MGNREGA, Power Politics, and Computerization in Andhra

Pradesh
Silvia Masiero a and Diego Maiorano b
a
Lecturer in International Development, School of Business and Economics, Loughborough
University, Loughborough, UK; bLeverhulme Trust Early Career Fellow, Institute of Asia-
Pacific Studies, School of Politics and International Relations, University of Nottingham,
Nottingham, UK

Abstract The link between e-governance and accountability of state adminis-

trations for service provision has been problematized in the literature to date.
However, little is known about its application to anti-poverty programmes, of
which public workfare schemes are an increasingly important subset. In this
paper, we fill the gap with a study of the Mahatma Gandhi National Rural Employ-
ment Guarantee Act (MGNREGA), India’s largest workfare scheme, as it is being
computerized in the southern state of Andhra Pradesh. A state-level information
system was devised to ensure transparency of transactions, and hence combat the
illicit diversion of the programme’s funds to non-entitled recipients. But while
doing so, the system carries a policy of centralization, which concentrates
decision-making power in the hands of a limited set of actors rather than distribut-
ing it across the programme’s stakeholders. In particular the Field Assistants,
appointed officials responsible for the village-level management of the scheme,
have direct control on the information inputted in the system, which reinforces
their position of authority rather than challenging it in favour of greater empower-
ment of wageseekers. Furthermore, wage payments are traced by the information
system till they reach the disbursement agencies, but are prone to capture in the
‘last mile’ where workers collect their salaries, which results in greater vulnerability
for them. As a result, MGNREGA workers are constructed by the new information
system as sheer beneficiaries rather than active participants in the programme,
which concurs to crystallizing existing power structures rather than resulting in
wageseekers’ empowerment. Lessons are drawn for other states currently compu-
terizing their social safety nets.
Keywords: accountability; e-governance; anti-poverty policy; MGNREGA; India;
Andhra Pradesh

1. Introduction
Over the last 10–15 years, workfare programmes have acquired increasing relevance as
anti-poverty strategies. Narrative on these programmes has peaked in recent times,
potentially as a result of globally increasing reliance on them as means to tackle
# 2017 Norwegian Institute of International Affairs (NUPI)
2 Silvia Masiero and Diego Maiorano

poverty and unemployment (Barrientos & Hulme, 2016; Devereux and Vincent, 2010;
Subbarao et al., 2013). In particular, the last decade has seen a sharp increase in the
computerization of anti-poverty programmes, in all their phases from design to deliv-
ery. Computerization is designed to reduce leakage and ineffectiveness, and is hence
widely seen as a desirable technical fix to the most fundamental problems affecting
anti-poverty schemes (Bhatti, 2012; Muralidharan et al., 2014).
In India, the Mahatma Gandhi National Rural Employment Guarantee Act
(MGNREGA) is an integral part of the national anti-poverty agenda. By design, the
programme provides a legal guarantee of 100 days of employment in public works
per year to rural households who demand it. Though MGNREGA’s positive effects
on poverty reduction are widely recognized, wageseekers in the programme are
found to be disempowered by diversion of funds, clientelistic allocation of work,
and systematic delays in the scheme’s wage payments (Mookherjee, 2014; Narayanan
et al., 2016; Pai, 2013). These problems severely hinder MGNREGA’s downward
accountability, as they result in systematic denial of entitlements and hence directly
affect workers’ livelihoods (Chopra and Khera, 2012; Das, 2015).
Digital technologies have been increasingly taken up by Indian state governments to
improve service provision, strengthen rural development and increase effectiveness of
anti-poverty programmes. With respect to MGNREGA, a pioneer state is Andhra
Pradesh (AP), where a Management Information System (MIS) controls the
scheme’s functioning in all its phases (Murali, 2012). The system makes all
MGNREGA transactions, from procurement to assignation of works to villages,
visible on the Internet through the programme’s website. Beyond technical change,
this implies a shift in the distribution of information, which affects the politics of
MGNREGA and its accountability to wageseekers.
Having been in place since 2006, the information system for AP MGNREGA pro-
vides a solid basis for us to ask, what is the relation between the adoption of e-govern-
ance and the accountability of MGNREGA to its workers? To unpack this we
conducted a state-level case study, based on three field visits between November
2013 and September 2014. The study has involved actors responsible for AP
MGNREGA (senior officials, district-level bureaucrats, state- and village-level imple-
menters), software designers involved in its digitalization, wageseekers employed in
MGNREGA works, social auditors and community volunteers. Reliance on the analy-
sis of participants’ narratives has been paramount in answering our questions, and has
led to a clear picture of the systems’ functioning and its implications for programme
implementation. Interview data have been triangulated with notes from observation
of the information system at work, social audits, press releases and government docu-
ments on the programme.
Our findings can be classified in two clusters, referring respectively to design and
implementation of the information system for MGNREGA in the state. From the
point of view of design, the system’s functioning is explicitly oriented to transparency
of all transactions, as all MGNREGA operations are updated into a publicly accessible
 Forum for Development Studies 3

server in real time. All fund flows can be traced on the Internet, and the payment infra-
structure is currently being integrated with biometric recognition of users by disburse-
ment agencies. This results into a system in which transparency is used in a purposeful
and systematic way, tailored to combat the illicit capture of funds flowing through the
programme.
Yet, when it comes to implementation, workers’ narratives problematize the extent to
which the system makes the programme more accountable to them. Payments’ processes,
computerized only till funds reach the disbursement agencies, are then left in the hands of
banks and post offices: this makes it difficult to trace the most crucial stage of the process,
meaning the ‘last mile’ in which payments are disbursed. Ten years after the inception of
computerization, many workers still report late and incomplete wage payments, and
greater information on fund flows has not resulted in fostering their bargaining power.
Information displayed online is widely used by activists and community volunteers, but
can hardly be accessed by the often illiterate and non-connected MGNREGA workers.
The paper hence contributes to unpacking the link between e-governance and
accountability of the state for social assistance programmes, which some literature
depicts as a tool-and-effect relation (Bhatnagar, 2004). In the case of MGNREGA in
AP, technology reflects a political agenda that crystallizes the local hierarchies on
which the programme has been constructed. In particular the Field Assistants, respon-
sible for the village-level management of the scheme, can to an extent control the infor-
mation inputted in the system: this reinforces their power position, rather than
challenging it in favour of the systemic empowerment of wageseekers. Beyond the
single case study perspective, what the study suggests is a problematic nexus
between e-governance and accountability, directly informed by the political agendas
underlying digitalization of anti-poverty programmes.
The paper is structured as follows. Section 2 describes the functioning of
MGNREGA, and provides the rationale for studying its ongoing computerization.
Section 3 illustrates the information system designed and implemented for
MGNREGA in AP. Section 4, focused on system design, explores the historical and
political roots of the systems’ orientation to transparency of transactions. Yet in
section 5 implementation is explored, and the disconnections between the system’s
enactment and accountability to its users are identified. Section 6 concludes with
lessons for other states computerizing MGNREGA and other social safety nets
crucial to guaranteeing poorer people’s livelihoods.

2. Background and methods

2.1. MGNREGA: implementation and functioning
Approved by the Indian Parliament on 23 August 2005, MGNREGA was implemented
in 200 rural districts on 2 February 2006, and scaled up to the whole country in early
2008. The programme is based on the principle of non-excludability of the right to
4 Silvia Masiero and Diego Maiorano

work, and enacts this principle by providing a formal guarantee of employment to all
the residents of rural areas who demand it. By doing so it aims, at the same time, to
improve the infrastructural development of rural areas, which benefit from public
works (Khera, 2012).
The scheme operates on a very large scale. In 2016–17, 2.34 billion person days of
work were generated by the programme, covering around 51 million households, thus
providing income for about 250 million people.1 The scheme has contributed to poverty
reduction through positive impacts on consumption expenditure, intake of energy and
protein and asset accumulation (Liu and Deininger, 2010), as well as food security,
women’s empowerment, financial inclusion and mental health of the affected house-
holds (MORD, 2012; Ravi and Engler, 2015). MGNREGA is also found to yield posi-
tive effects on labour markets, through wage increases and reduction of distress
migration to urban areas (Imbert and Papp, 2015). The total cost of the scheme is
just above 0.3 per cent of India’s GDP (Mookherjee, 2014).2
It is important to note that MGNREGA is an Act of Parliament, not just a govern-
ment scheme, and this sets it apart from other Indian anti-poverty programmes (Jenkins
and Manor, 2017). This characteristic of the scheme makes it very difficult to withdraw,
as employment guarantees provided by it are an integral part of national law. Impor-
tantly, the right to 100 days of employment in public works per year (or to an unem-
ployment allowance, in case work cannot be procured) extends to all rural residents
of India, not simply to those formally recognized as below the poverty line. This
entails a mechanism that, while designed to facilitate self-selection of users, is wea-
kened where MGNREGA wages are low, and where beneficiaries are not fully
aware of the scheme’s nature as a legal guarantee (Shankar and Gaiha, 2013).
As in Jenkins and Manor (2017, p. 3), the programme is demand-driven, meaning
that ‘any person residing in a rural area can apply for work at their local government
office’. This confers a proactive role to wageseekers, making the scheme an occasion
for them to exert direct agency by requiring and obtaining employment in public
works. Furthermore the scheme is designed to involve village councils (panchayats)
in its management, enhancing the degree of participation of local bodies in anti-
poverty systems (Shankar and Gaiha, 2013). Yet, since states hold a degree of discre-
tionary power to adapt the programme to local needs and characteristics, diverse out-
comes emerge in terms of the effective degree of decentralization and participation
with which the scheme is implemented (Jenkins and Manor, 2017; Maiorano, 2014).
Notwithstanding the widely recognized positive impact on India’s rural poor, the
system is affected by a range of issues at the national and sub-national level. One

1 Full statistics are available on the national MGNREGA website: http://www.nrega.nic.in.

2 Critiques to the programme (e.g. Bhagwati & Panagariya, 2014) find their core argument in
the lower economic effectiveness of MGNREGA as compared to a cash transfers scheme.
This point, while central in the ongoing debate, does not dispute the contribution to poverty
reduction achieved by MGNREGA.
 Forum for Development Studies 5

relates to diversion of funds outside the programme’s supply chain, which reduces the
amount of resources available to beneficiaries.3 A second problem pertains to the dis-
cretionary basis of allocation of MGNREGA work: though the programme works as a
‘post-clientelistic’ policy, personal connections still play a role in assignation of entitle-
ments (Das, 2015; Maiorano et al., 2016). The issue is made more severe by shrinking
budgetary allocations for MGNREGA by the central government, which impose
additional financial constraints that limit the possibility to give work to anyone who
demands it (Maiorano and Das, 2015).
A third class of problems relates to wage payments, whose incomplete and delayed
delivery affects a rising proportion of workers (Mookherjee, 2014). Payments should be
generated for each work cycle in the scheme, and once a work cycle is complete, orders
are submitted to disbursement agencies, in which wageseekers’ accounts are credited
with the payments due. Implementing agencies are hence separated from disbursement
bodies, which increases the complexity of a supply chain already prone to capture. In
fact, depriving MGNREGA wageseekers of timely payments significantly affects
workers’ capability to sustain their livelihoods through the programme, and may ulti-
mately discourage them from seeking employment through the scheme (Bhatti,
2012; Muralidharan et al., 2014; Narayanan et al., 2016).4
To obtain a clear picture of MGNREGA, two more features should be considered.
First, the programme is inscribed in a so-called ‘new rights agenda’, a set of measures
that deeply changed the nature of rights-based policies by the Indian central govern-
ment (Ruparelia, 2013). This new agenda includes, measures such as the Right to Infor-
mation Act, the Food Security Act, the Forest Rights Act which provides new rights to
traditional forest dwellers, and the Right to Education Act which makes schooling com-
pulsory and at the same time entitles all children to attend school. By making the right
to work universal and non-excludable, MGNREGA takes active part in this agenda,
making employment guarantees a fundamental component of the pursuit of anti-
poverty action and socio-economic development.
A second feature of the programme is its variation between states, resulting in very
different outcomes across the nation. It is the case that state-level political dynamics,
including policy legacies and the nature of the social coalition supporting the ruling
party, shape implementation on the ground, influencing the making of social policy
(Deshpande et al., 2017). That MGNREGA’s implementation resents from political
dynamics is argued by a substantial stream of literature (e.g. Jenkins and Manor,
2017; Maiorano, 2014; Shankar and Gaiha, 2013), and this angle entails a focus on
state-level dynamics in the analysis of the programme. The state dimension is key to

3 Drèze (2014) shows, however, that corruption is significantly lower in MGNREGA as com-
pared to other workfare programmes, and has been declining over time.
4 Over fiscal year 2014/2015, only 27.38 per cent of payments were disbursed within the pre-
scribed 15 days, while more than 13 per cent were delayed by more than 90 days (Account-
ability Initiative, 2015).
6 Silvia Masiero and Diego Maiorano

observing the ground reality of MGNREGA, and this has concurred to the choice of
using a state-level study to answer our research question.

2.2. Rationale and methods

Technical hurdles, and roadblocks in the delivery of entitlements, should be seen in the
light of a broader debate on the politics of MGNREGA implementation. The recent
debate has been informed by the programme’s role in the space of social transform-
ation, moving beyond simple provision of short-term benefits to recipients. Studying
the structural effects of the programme, Carswell and De Neve (2013, 2014) highlight
potential to enhance wageseekers’ bargaining power towards high-class employers,
making MGNREGA instrumental for transformative social assistance (Devereux and
Vincent, 2010). In this perspective, if social protection is intended as combating the
root causes of poverty and vulnerability (Barrientos and Hulme, 2016), the good func-
tioning of MGNREGA concurs to tackling economic imbalances at their very basis.
The capacity of MGNREGA to reduce structural poverty is, however, a matter of
discussion. Social anthropological understandings of the programme, grounded on
in-depth study of its local politics, lead to observe that roadblocks to MGNREGA’s
effectiveness lie in structural power asymmetries, which prevent challenges to tra-
ditional class relations among the programme’s actors. In particular, Pattenden
(2011) suggests that the transformative social assistance agenda has not yet reached
a sufficient degree of transformativity, as it does not advocate a systematic analysis
of class relations. Similarly, Ghosh (2011) warns that development policies risk
failure if they do not tackle the structural and systemic processes that generate
poverty in the first place. Attention needs to be paid, as a result, to the specific social
contexts through which poverty is reproduced over time, and to the ability of extant pol-
icies to deal with it.
In this light, some considerations are required on MGNREGA specifically. First,
power asymmetry across actors in the programme is a known fact: the persistence of
clientelism is a tangible sign of it, and is found also in states where the programme
ranks well on most indicators (Das, 2015). Second, asymmetry is not neutral, but con-
notated along caste and gender lines (Maiorano et al., 2016; Pattenden, 2017), resulting
in systematic inequalities in access to key means of social protection. As revealed by
Pattenden’s (2017) class-relational approach to the programme, issues seem to be of
a structural, rather than incidental nature: power asymmetry reduces the bargaining
power of workers, making them helpless towards resource diversion or payments
that are missing or delayed for long times.
It is in this political perspective that the debate on e-governance in anti-poverty pro-
grammes should be placed. Computerized architectures, inscribed in the Indian anti-
poverty system, are constructed as part of the solution to problems of effectiveness
and accountability (Madon, 2009). As it enters anti-poverty schemes, technology
should not be playing a sheer technical function, repairing existing hurdles and ensuring
 Forum for Development Studies 7

delivery of benefits. It should instead be entrenched in the political heart of the pro-
gramme, hence affecting the power imbalances on which ineffectiveness is predicated.
The theme of power relations needs further unpacking in the ongoing debate on
information and communication technologies (ICTs) for development. The link
between ICT and accountability may not work as expected, and discrepancies may
emerge between technology design and the reality for which systems are tailored
(Heeks, 2002). Frequent failure of projects in ICT for development, affecting the
public sector in particular, indicates the need to further problematize the link
between e-governance and accountability of anti-poverty programmes, including work-
fare schemes like MGNREGA. This is arguably the main intellectual problematique
relating to e-governance in the social protection space.
This leads us to ask, what is the relation between the adoption of e-governance and
accountability of MGNREGA to its workers? This question is at the root of the study
conducted here. MGNREGA’s problems of fund capture, illicit allocation of works,
and late/incomplete payments all contribute to generating an accountability gap
between the programme and its users, which ICTs may contribute to bridging.
However, social mechanisms linking e-governance to accountability of service provi-
ders are still unclear, and need to be elicited through dedicated observation of their
functioning at the state level.
Our main source of data consists in 68 interviews, conducted in a phase (November
2013–September 2014) in which the computerization of the programme had already
become established. A first cluster of interviews was conducted with senior officials
at AP’s Rural Development Department, and aimed at forming a clear picture of
MGNREGA in the state and the policy motivations lying behind computerization. A
second one was conducted at Tata Consultancy Services (TCS), Hyderabad, and
involved in-depth interviews with software designers and also practical demonstrations
of the system’s functioning. A third round was conducted with MGNREGA workers in
Visakhapatnam district, where the scheme has been implemented since 2008. Our inter-
views with wageseekers have revolved around specific questions about their perception
of technology in MGNREGA, their access to it, and the affordances related to its
utilization.
Visakhapatnam covers a largely rural and tribal area of north AP, whose internal
diversity makes it representative of the variety of conditions experienced by
MGNREGA workers within the state. The district is characterized by high rates of
rural and tribal unemployment, and is included in the Indian Government’s Backward
Regions Grant Fund Programme since 2010. Selection of the study district was
informed by the established presence of MGNREGA, as well as of volunteer associ-
ations that make the link between wageseekers and the institutions administering the
programme. The presence of volunteer groups has been instrumental for a study of
anti-poverty technologies, since these cannot easily be taken up by workers alone,
and often need mediation by dedicated pressure groups. Working closely with the vol-
unteer organizations, we have conducted research visits in six mandals of
8 Silvia Masiero and Diego Maiorano

Visakhapatnam district, observing the system at work from the perspective of the
workers operating its user interface.
The tribal context in which a substantial part of the research has been conducted
deserves attention. Scheduled Tribes, recognized as a special category by the Indian
Constitution, are among the most vulnerable socio-economic groups in the nation,
and derive their main subsistence from basic agricultural activities conducted in land
and forest. According to the 2001 Census of India, 61 per cent tribals in unified AP
lived below the poverty line (Maiorano and Buddha, 2015), and the economic fragility
of such communities entails the need for state governments to adopt special terms for
them. This calls for tailoring of anti-poverty programmes on the particular needs of
tribal communities, and has led us to conduct interviews and observation in a tribal
context whose approach to e-governance offered multiple cues for reflection.
Interviews have been informed by the concept of narrative interviewing as in Riess-
man (2008), in tune with the purpose of grasping the participants’ lived experience of e-
governance, as well as the policy rationale behind its implementation. Primary data
have also been collected through group interviews at work sites, in which technologies
have been collectively discussed by users as well as the mediating volunteers. To inte-
grate such data, observations of the system at work have been conducted at Tata Con-
sulting Services and in three Mandal Computer Centres, where Computer Operators
illustrated their use of the system in managing MGNREGA operations. This has
been complemented with secondary sources, in particular government documents on
programme implementation and social audit reports produced by the State’s social
audit organization.5

3. MGNREGA in AP and its computerization

AP is one of India’s top spenders of MGNREGA funds, and has consistently ranked
among the top providers of employment under the scheme.6 Below we detail the spe-
cificities of AP MGNREGA, and the properties of the digital system through which the
programme is implemented.

5 Beyond research conducted for this paper specifically, the second author has led extensive
work on MGNREGA in Andhra Pradesh, consisting of in-depth interviewing and participant
observation of works and social audits in the districts of Vishakapatnam, Srikakulam, Guntur,
Chittoor, Mahbubnagar, Karimnagar and Anantapur between 2012 and 2017. Part of this work
has preceded the bifurcation between Andhra Pradesh and Telangana in 2014. This work has
served to triangulate the data presented here, and to place observations in the broader context
of MGNREGA.
6 This was consistently true until 2012/2013. From 2013/2014 on, the ranking of Andhra
Pradesh dropped significantly. While what caused the fall in employment generation is not
clear, it is plausible that the agitation for the bifurcation of the state, occurred with the
Andhra Pradesh Reorganization Act in 2014, affected the implementation of the
MGNREGA too, as a result of the virtual paralysis of state administration post-bifurcation.
 Forum for Development Studies 9

3.1. AP MGNREGA: centralization and transparency

To explain the rationale behind the MGNREGA information system, it is crucial to
grasp the historical and political context in which the system has been developed.
Before the 2004 elections the incumbent chief minister, Chandrababu Naidu, was con-
sidered a model administrator by a variety of actors, including the press, business
circles and international funding bodies. Naidu actively contributed to sustained econ-
omic growth and a very business-friendly environment, to the point of being referred to
as the ‘CEO of AP’ (Suri, 2004). He also stressed the importance of ICTs for economic
and social development, actively contributing to AP’s rising reputation as a national
hub for technology development.
After India’s fiscal crisis in 1991, it was fundamental for state governments to
commit to generate steady economic growth. Naidu did so, strongly emphasizing the
importance of trade and openness to business for the state to thrive. However, his pol-
icies had systematic backlashes on the most vulnerable strata of the population, since
investment in the private sector was not matched by equal attention to anti-poverty pro-
grammes and their funding. According to commentators, this resulted in ‘the accentua-
tion of poverty, growing income inequalities, unemployment and unabated farmers’
suicides’ (Suri, 2004, p. 5493).
Y. S. Rajasekhara Reddy, popularly known as YSR, was a prominent politician
from the opposition, whose family was one of those that formed the backbone of the
Congress party’s organization during the anti-colonial times (Maiorano, 2014). Lever-
aging the weaknesses in Naidu’s popularity, he systematically accused Naidu of
neglecting the rural world and the state’s poor. He promised extensive welfare and
rural-oriented programmes, following the principle that implementing workfare
schemes successfully is an integral part of a populist politics (De Neve and Carswell,
2011; Price, 2011). YSR won the 2004 elections by a large margin, promoting policy
initiatives including a successful implementation of MGNREGA.
Yet, for the programme to have a chance of success, the scheme had to be
implemented very differently from employment programmes implemented previously
in the state, which were known for their high levels of leakage to the non-poor (Drèze,
2005). To maximize the new government’s political credibility, corruption in welfare
schemes had to be minimized and specific monitoring systems had to be designed
for social safety nets. In continuity with Naidu’s regime, YSR was able to centralize
power in his hands, with two objectives in mind: first, minimizing opposition
through the centralization of decision-making with respect to both the hitherto
faction-ridden Congress party and the administration of the state (Srinivasulu, 2009).
Second, acquiring a significant degree of control over the state machinery (Maiorano,
2014), as a way to monitor the implementation of welfare schemes to convey a strong
public image of corruption minimization.
The principles of transparency and centralization became visibly reflected in the
state’s management of MGNREGA. When it was launched, the state government
10 Silvia Masiero and Diego Maiorano

institutionalized a civic monitoring mechanism based on social audits, through the cre-
ation of an independent society (the Society for Social Audit, Accountability and Trans-
parency), that schedules and carries out audits in all villages on a rotation basis.
Although the efficacy of this system has been questioned (Afridi and Iversen, 2013),
no other state has sought to implement systematic social audits as it is prescribed by
the MGNREGA law. This is per se a deterrent against capture and misappropriation
of funds, to be added to the fact that audits are now considered integral part of the
MGNREGA model in AP (Aiyar and Samji, 2009).
As widely recognized, centralization surfaces more prominently in the Andhra
model than in other states. In fact, the national MGNREGA law envisages full involve-
ment of village-level elected institutions (Gram Panchayats) in the implementation of
the scheme. On paper, Gram Panchayats should be responsible for numerous tasks,
including selection of works and registration of wageseekers’ demands for employ-
ment. At the same time though, the Andhra government has excluded Gram Panchayats
from any meaningful role in implementation, delegating the management of
MGNREGA works to appointed Field Assistants who are in charge of decision-
making at the village level (Maiorano, 2014).
A further peculiarity of AP is that, differently from most other states,
MGNREGA workers do not obtain and conduct work alone, but are organized to
do so on a groups’ basis. In every village, workers are divided into groups of 15 –
20 members, headed by a leading actor referred to as a mate. As a result, all
members get employment as a group rather than as individual workers, which at
least on paper, should make it more difficult for Field Assistants to exclude wagesee-
kers from accessing the programme. With this model, interactions between state
administration and wageseekers are minimized, as the mate deals with all adminis-
trative procedures on the wageseekers’ behalf.

3.2. E-governance for AP MGNREGA

The implementation of MGNREGA in AP is based on an information system that, since
2006, structures all operations in the programme. The system at the core of the pro-
gramme has been developed and maintained by TCS. This choice constitutes an
anomaly, as in most Indian states, information systems for anti-poverty schemes are
developed by the local offices of a state agency, the National Informatics Centre. A
senior official explained the reason for this decision:

Experience and capabilities of TCS were a guarantee of reliability. The information

system is the main way to guarantee transparency of the programme. (Senior Official,
Rural Development Department)

Government narrative is adamant on the fact that, in AP, the information system is an
integral part of the functioning of MGNREGA. Its core purpose is that of enabling web-
 Forum for Development Studies 11

based accessibility (http://nrega.ap.gov.in) to all transactions in the programme, by

computerizing its five core phases:
Generation of works. Works are generated monthly for each mandal, on the basis of
local needs and infrastructural development priorities. The information system allows
the assignation of works, and calculates, on an automatic basis, the amount and nature
of labour and capital required for each.
Procurement of materials. Procurement is carried out through e-tendering, an online
procedure that allows every company to cast their offers and allows assignation to the
lowest bidder. The construction of e-tendering has been aimed at eliminating corruption
in the procurement phase, which was one of the most affected by illegal practices – as
most bids, in the non-computerized system, were assigned on the basis of political
connections.7
Measurement of works. Once works have been completed, outcomes are measured
by Engineering Consultants, and payments are provided on that basis. Engineering
Consultants and their staff use the information system to locate works, and to
measure the extent to which expected outcomes have been achieved.
Processing of payments. Muster rolls (registers of workers employed in each work)
are compiled at the work site, and uploaded weekly by the Mandal Computer Centre.
Once measurement of outcomes has occurred, payment orders are uploaded online and
sent to disbursement agencies. To collect their salary wageseekers need to visit the dis-
bursement agency, or wait for a banking correspondent to visit their area. The system
monitors the process till reception of payments by disbursement agencies.
Generation of reports. Since all operations described above are conducted through
the information system, all of them are visible through the AP MGNREGA website. On
the same platform, reports are automatically generated for each district, mandal and
constituency in which the programme operates. As summarized by one of the
system’s software developers,

The rationale of the system is that of generating internal accountability. The initial idea
was that of using the web to place MGNREGA in the hands of the citizens [. . .] the prin-
ciple is that of full transparency of the programme, to clear corruption and make citizens’
monitoring real. (Software Developer, Tata Consultancy Services)

Computerization of the payments’ disbursement phase, while not yet homogenized at a

state scale, is being led through two projects, Aadhaar-based recognition and a Smart-
card Programme:
Aadhaar, also known as Unique Identification project (UID), is a national pro-
gramme aimed at registering Indian residents through a 12-digit number and their bio-
metric details. The Aadhaar database currently constitutes the biggest project of
biometric identification ever conducted worldwide. In an increasing number of
mandals, Aadhaar-enabled payments are operated, and registered wageseekers can

7 Interview with a senior official, 30 August 2014.

12 Silvia Masiero and Diego Maiorano

therefore withdraw their wages on the basis of fingerprint identification. This is sup-
posed to simplify the recognition procedure, and ensure that payments are made to
genuine beneficiaries.
Smartcard Programme: beyond Aadhaar, AP hosted a state-level project of inde-
pendent biometric recognition, for payments of MGNREGA wages and social pen-
sions. In 158 randomly selected locations, biometric smartcards have been assigned
to entitled citizens to enable payments. Though the project has not been rolled out to
the whole state, there seems to be strong potential for biometric recognition to
enable more secure payments (Muralidharan et al., 2014).
Built on the principle of transparency for empowerment, implicit in the information
system, the payment infrastructure is complemented by biometric recognition at the
level of disbursement agencies. This configures a carefully designed e-governance
system, exceptional for a state whose general levels of ICT adoption in the public
sphere are still quite limited. Seeking to understand how technology intersects with
the politics of the programme, we present our findings in terms of system design and
implementation.

4. Transparency and the politics of system design

The system’s focus on transparency of transactions, fitting the anti-corruption agenda
set by YSR in the programme’s early days, is rooted in the politics of programme
implementation. The idea behind the system is that transparency should break the mon-
opoly of information retained by implementing officials, thereby reducing the tradition-
ally high discretionary power held by bureaucrats. As it emerged in interviews with
senior officials at the Rural Development Department, the stated objective was that
of reducing the influence of Members of the Legislative Assembly (MLAs) on the pro-
gramme, to create a public image of political accountability. Many Cabinet ministers
and MLAs reportedly resisted the introduction of transparency-enhancing technology,
while YSR’s rhetoric was very firm on the need for it, due to the instrumental value of
transparency for the political agenda that he advanced.8
MLAs displayed three key interests related to MGNREGA. The first one was that of
creating jobs in their constituencies, which was compatible with YSR’s strategy. The
other two interests, however, conflicted with it: one was that of diverting some of
the scheme’s funds to their patronized contractors, often owned by relatives or political
clients, and in some cases by themselves. These can be involved in two ways: first, con-
tractors can execute the works assigned to wageseekers, which is explicitly forbidden

8 It should be noted that YSR’s anti-corruption agenda was limited to welfare schemes and in
particular to the MGNREGA, which played a central role in building his political credibility.
In other spheres, such as the management of public revenues and the state’s food security
system, he allowed widespread corruption to occur, displaying sensibly lower levels of atten-
tion to the rigor of public procedures and their equality for all citizens.
 Forum for Development Studies 13

by the MGNREGA law. The information system makes this virtually impossible by
monitoring the flow of funds and sending payments to beneficiaries’ bank accounts,
so that even if works are executed by a contractor, wages would still be paid to the ben-
eficiaries listed in the official muster rolls.
Second, contractors can be involved in the procurement of works’ materials, and
MLAs put enormous pressures on the administration to assign procurement to their
contractors. This was a minor problem before YSR’s death, as the share of expenditure
spent of materials amounted to less than 10 per cent of the total. But respondents’ nar-
ratives agree on the view that after that, subsequent chief ministers were less able to
control their MLAs, who constantly pushed for higher material expenditure.9 The e-ten-
dering system, which assigns bids to contractors irrespective of their political connec-
tions, has been conceived to prevent this, hence creating a specific mechanism to reduce
diversion in the phase of the programme where it was observed more frequently.
Of course, e-tendering does not automatically ensure that the bid assignation
process will occur in a free and fair way. A senior official, interviewed in Hyderabad
shortly after the implementation of e-tendering, reported frequent attempts of local
counterparts (predominantly, bidders frustrated with the impartiality of the procedure)
to tamper with the system, seeking ways to bypass the automatic assignation of bids. It
is also true that technology, placed in a context characterized by a complex web of pol-
itical connections, cannot per se eliminate the incentive of bidders to pursue their own
interest of winning over other actors, and in principle local bureaucrats could find ways
to bypass the automatized decision made by the system. Technology is, however,
designed to minimize the discretionary power of bureaucrats, and it substantially
increases the practical difficulty of tampering with it to appease individual requests.
The third interest displayed by politicians is that of controlling the programme through
their network of officials in the state’s administration. For example, MLAs can ask Field
Assistants to provide more work to their voters than to supporters of the opposition. In this
case, YSR faced a difficult choice: he could have chosen to give Gram Panchayats the task
of implementing the scheme, as prescribed by the MGNREGA law. This is the common
praxis in other Indian states, and it would have meant empowering the councils’ heads, the
sarpanches, to provide jobs, significantly reducing the scope for discretion of the MLAs.10
YSR chose, however, a different route, that of implementing MGNREGA through
Field Assistants. This option was more prone to interference by MLAs, who put press-
ures at various levels to appoint their men as Field Assistants: as powerfully summar-
ized by a senior officer, ‘all field assistant posts are political posts’.11 Yet in 2012, the

9 Interview, senior Rural Development Department official, Hyderabad, 9 December 2013.

10 Interview with the former Minister for Rural Development, Hyderabad, 5 August 2013. This
was confirmed by the second author’s field visits in Chittoor, Karimnagar, Mahbubnagar and
Vishakapatnam districts. In every single village visited, we found that Field Assistants had
important political connections.
11 Interview, senior Rural Development Department official, Hyderabad, 5 August 2013.
14 Silvia Masiero and Diego Maiorano

Rural Development Department introduced a new feature in the information system,

which made the appointment of Field Assistants automatic according to two criteria:
all must be former MGNREGA workers, and the system selects three names according
to their seniority, literacy and number of days worked under the programme. This does
not exclude the possibility that Field Assistants are co-opted into local power networks,
but it reduces MLAs’ discretion for influencing appointments.
It is important to observe the limits that automatic selection of Field Assistants
encounters when operated in practice. It is true, on the one hand, that the system gen-
erates a list of individuals eligible as Field Assistants, and mandal level officials can
only pick individuals that figure among the three top names of the list. But interviews
conducted with street-level officials, even beyond the scope of the information system,
have revealed the presence of cases in which, for instance, the three-name list is sent to
the local MLA for vetting. In other cases, political connections have resulted in the
three-name list being bypassed, and individuals being appointed as a result of their per-
sonal affiliation to local MLAs. As instances of clientelism persist, it is important to see
the power of technology in its right proportions: while the automatic selection of Field
Assistants adds a crucial layer of complexity to tampering, it still needs to be placed
within the web of politics that characterize the programme at the state level.
It should be noted that in other Indian states, where Gram Panchayats are in charge
of implementation, MLAs do not play such a significant role because the sarpanches,
who are democratically elected, are strong enough to prevent them from interfering
with the programme.12 However, in AP, the space for MLAs’ interference was much
more pronounced, and e-governance has hence been tailored in order to limit the per-
verse effects of this. As a result, MLAs are now being pushed to see the scheme as sup-
plementary spending in their constituencies, which can enhance their popularity.13
Furthermore, MLAs’ requests related to the MGNREGA are reported to be rather
rare, as they know that little money can be earned at a high risk, and that computeriza-
tion reduces the discretionary power of bureaucrats.
Resulting from a specific political context, system design physically enables greater
centralization of powers in the scheme, a policy that is sustained by the transparency of
the system’s internal transactions. Yet users’ narratives, explored below, provide a
different angle on the enactment of e-governance.

5. Computerization in practice: accountability of MGNREGA to wageseekers

How is system design related to accountability of providers to MGNREGA workers?
To answer we observe, through wageseekers’ narratives, the way in which the

12 This observation comes from the second author’s extensive fieldwork in Rajasthan’s rural
areas in January-February 2014 and from conversations with scholars working on other
states. See also Jenkins and Manor (forthcoming) for evidence regarding Rajasthan and
Madhya Pradesh.
13 Interview, senior official, 5 August 2013.
 Forum for Development Studies 15

system’s policy comes alive in practice. What emerges from these narratives, collected
from workers and the volunteers assisting them, is a twofold problem, which affects
both the accountability structure of MGNREGA and the mechanisms enabling wage-
seekers’ capabilities to obtain information and exert their voice.
As noted above, TCS’ monitoring focuses on the implementing agency, and finishes
just as payments are credited on wageseekers’ accounts. A first cause for concern is that
biometric authentication does not enforce downward accountability to wageseekers, but
operates the inverse process, as it assures the state that recipients are genuinely entitled
to payments (Khera, 2011). This also inspires the decision of paying wages on bank
accounts rather than in cash, whose untraceable nature does not allow to ascertain
the identity of the recipient and her entitlement to benefits.
The problem lies, though, in the other side of the accountability relation, that of the
reliability of the institutions in charge of payments. As it emerged from wageseekers’
narratives, it is exactly in the reliability of payments that the weakness of MGNREGA
seems to lie: the problem of missing, incomplete and late payments has persisted over
time, and computerization has not resulted into tangible changes in occurrence. In fact,
workers do not trace a connection between centralized computerization and more
reliable MGNREGA payments, and this results in perception of the system as unable
to solve the main problem encountered. The lack of monitoring of disbursement
agencies, the place in which wageseekers collect their payments, contributes to
making wage payments uncertain, and is frequently mentioned as one of the factors
that reduce effectiveness of the programme for beneficiaries.
Furthermore, biometric identification poses another problem, that of the recognition
errors in which the system sometimes incurs. Issues of recognition, while minimized by
programme staff, feature frequently in the narratives of wageseekers, and in those of
volunteer organizations protecting them. For example, tribal workers interviewed at
a work site openly displayed preoccupation about the system:

[Biometric recognition] very often does not work. If I go to the bank and it doesn’t recog-
nise me, I don’t get paid, and nothing can be done about it [. . .] for five months, I could not
collect any payments. This was not a problem when cash was paid. (Wageseeker,
Narsipatnam)

Since [the biometric system] has arrived, citizens have to queue for many hours. If they
are not paid, they have to go to the District Office, and hope to be luckier. Very often,
going to the District Office only results in a day of work lost. (Volunteer, labourers’ organ-
ization, Visakhapatnam)

It needs to be added that tribal users, living in areas that often lack basic infrastructures
and services, may need support when interfacing with technologies as new as biometric
authentication and recognition. This special need, currently addressed by several organ-
izations around the country, was not addressed with particular strength at the time of
our fieldwork, leaving several interviewees to reveal mistrust and suspicion around
16 Silvia Masiero and Diego Maiorano

such technologies, particularly towards their role in affecting access to entitlements

under MGNREGA and other anti-poverty schemes.
To date, incidence of non-recognition of users through biometric devices has not
been estimated in the state. Still, the extent to which this problem is reported in wage-
seekers’ recounts is a major cause for concern, because non-recognition prevents the
very reception of benefits through MGNREGA – and no back-up option, to use in
case of failure, has been found at the disbursement agencies. Failure of biometric
authentication, workers report, forces them to long trips to the district office, which
may easily end in frustration. From these narratives, with respect to the area in
which the research was conducted, technology seems to result in disempowerment of
wageseekers, as it concurs to blocking the reception of MGNREGA wage payments
instead of facilitating it.
Problems with downward accountability of disbursement agencies are matched by
two more gaps, found at the level of the implementing agency. First, during works,
wageseekers seem to have no means to ensure that their names have been included
in the muster roll, the register on which payment generation is based. As reported in
a group interview with wageseekers,

Whether one is, or is not, signed into the muster roll, depends on the FA [. . .] if one is not
signed in as present, no payment can be made [to them]. [In each work], there is no way to
know for sure that our names have been registered as present. (Worker, Visakhapatnam
district)

Based on the narratives collected on the field, the picture that emerges is that high dis-
cretionality of Field Assistants and mates is a consequence of the centralized MGNREGA
policy in the state. In such a system, the perverse dynamics of purposeful exclusion is not
formally detected by the information system. The philosophy behind the technology,
which matches every worker with a job card number, is designed to prevent the inclusion
of ‘ghost beneficiaries’ in muster rolls, but does not prevent exclusion of genuine ones. As
a result, while inclusion errors are tackled, exclusion errors are likely to remain unsolved,
and this further weakens the accountability structure into place.
Secondly, payments are based on the measurements conducted by Engineering
Consultants on works. If measurements reveal that, in a given work, expected results
have not been reached, payments will be lower than those initially determined. The
problem here is that the system, while making measurements technically possible,
does not guarantee their fairness, and hence there is still room for unjust practices in
measurement, as explained by a member of a volunteer organization:

The terrain is rocky, which makes it very difficult to dig in the first place. Still, workers
have been digging it for months, but they keep being assigned to the same pond – where
nothing is left to dig, they have hit rock bottom. The ECs know it, but no action is taken
[. . .] workers have been paid 40 – 50 [Rs.] a day instead of 160, because the measurement
says they are less productive than they should be. (Volunteer, Narsipatnam)
 Forum for Development Studies 17

Hence, disbursement agencies are not the only part responsible for missing payments of
MGNREGA wages. Even the implementing agency presents gaps in accountability,
which are found both in the construction of muster rolls, and in the measurements
leading to determination of payments. The structure of MGNREGA in the state,
flowing from the policy agenda that the system embodies, is not designed to achieve
downward accountability to users, and seems therefore suboptimal in guaranteeing
prompt payments.
The problem may in fact be running even deeper than that. For wageseekers to be
able to claim their payments, two more mechanisms are needed: first, citizens need full
information on their entitlements under the programme (e.g. wage determination,
timings and conditions of disbursement). Second, the capability of exerting voice –
lodging complaints, and having them attended in reasonable times – is needed in
case of missing, incomplete or late payments, so that grievances can be redressed.
With respect to information on entitlements, the picture is mixed. On the one hand,
AP is one of the states in which awareness of MGNREGA workers’ rights is highest in
India (Aiyar and Samji, 2009), and evidence of awareness is consistently found in
workers’ narratives. In some cases, wageseekers even manage to obtain work in
spite of lobbying action by big farmers, whose interests converge towards the minimiz-
ation of MGNREGA work in their areas (Jakimow, 2014). A different matter is the
capability of accessing work, which depends largely on Field Assistants: workers
may still be subjected to exclusion, but their ability to demand paid work in
MGNREGA is based on awareness of the programme.
On the other hand, an exception to the above is constituted again by tribal areas,
where poverty levels are very high and literacy levels are low. In these areas, not
only lacking awareness of entitlements is still a reality, but wageseekers are sometimes
induced to see MGNREGA as offered by contractors, rather than as a right guaranteed
by law. While the diffusion of such misleading narrative is difficult to estimate, the
problem has been reported by community volunteers:

Many workers do not know it is their right to demand work [. . .] so, they do not know their
entitlements. Without the intervention of volunteers, they would not at all know that they
are due certain payments, at some specific regular times. (Volunteer, labourers’ organiz-
ation, Visakhapatnam)

Furthermore, even where awareness of the programme is high, issues persist with
respect to information on exact amounts of MGNREGA wages. In most cases,
workers are not aware of how payments are determined, and are therefore unable to
claim the amounts due. This also makes the incidence of incomplete payments hard
to measure, as wageseekers are not given precise estimates, which makes them
unaware of the extent to which the process may be fair.
This brings us to the heart of the problem: has the system not been created exactly to
make MGNREGA information fully accessible? On paper, it is so: yet, wageseekers are
18 Silvia Masiero and Diego Maiorano

in large majority illiterate, and unable to access computers and other information tech-
nologies. Information provided by the system is not constructed for the wageseekers to
access it: it is, instead, made available to organizations that are able to use data in a
meaningful way. This includes the social audit organization discussed above, and the
volunteer organizations which use the information system in their work.14 But infor-
mation does not, at present, cater directly to MGNREGA’s recipients, and this is a
strong limit to its effectiveness.
Similarly, the system does not empower workers to lodge complaints, or to guaran-
tee that their grievances can be redressed. Many workers reported the intention to lodge
complaints, but were unable to identify an institutional means or mechanism to do so.
Complaints do not only address missing payments directly, but revolve around the
process through which they are determined, as in the above-mentioned case of unfair
measurement:

It has been months since we reported the problem [of assignation of workers to the same
pond, where nothing is left to excavate] to the RDD. Still, after months, we had no
responses. We asked to classify the terrain [of this mandal] as rocky, which influences
EC measurements, but even that request has never been listened to. (Volunteers,
labourers’ organization, Visakhapatnam)

To be sure, MGNREGA staff reported the presence of several redressal mechanisms,

such as local call centres and programme offices to turn to, in case of any issues
with benefits under MGNREGA. However, little evidence of awareness and/or usage
of any form of grievance redressal has been found in wageseekers’ narratives. As
reported in a group interview at a work site:

They send us from one office to the other, till the APO [Additional Programme Officer]
tells us to go [to complain] to the computer centre. Here, they [the Computer Operator] tell
us that they cannot help, because it is the bank that decides on payments [. . .] it is very
unclear who we should turn to. (Worker, Visakhapatnam district)

As it seems, the Computer Operator acts on a strategic nodal point of the programme,
and enjoys full powers on a crucial stage of the payment process, that of uploading the
muster rolls. His ineffectiveness in facing payment grievances is met with frustration:
once again, in situations of this type, technology is viewed as disempowering, as the
agent in charge of it is unable to address complaints effectively.
Crucially, the accountability structure examined here seems to derive directly from
the centralized logic with which the programme is implemented. It is due to centraliza-
tion if accountability structures focus chiefly on internal transactions, rather than on

14 Observation conducted at the Society for Social Audit, Accountability and Transparency, as
well as at three different volunteer organizations engaging in protection of MGNREGA
workers, has confirmed the frequent use that these organizations make of reports produced
by the information system.
 Forum for Development Studies 19

relations with wageseekers. Weakness of downward accountability mechanisms seems,

therefore, a side effect of centralization, and a direct consequence of the top-down logic
inscribed in the information system. Ultimately, the power politics implicit in the pro-
gramme and in its implementation seem to be reinforced by the information system,
which perpetuates the existing centralizing logic rather than challenging it in favour
of greater empowerment.
It should be noted that workers’ alienation, reviewed above through the voices of
wageseekers, is not found solely in the problem of incomplete and late payments. It
finds tangible manifestations in at least two more issues, which recur systematically
in wageseekers’ narratives. First, workers do not really have a voice in influencing
the choice of works allocated to them: these should be decided by the Gram Sabha
(an assembly consisting of all adult members of the Gram Panchayat), but commonly
turn out to be chosen by the Rural Development Department. As noted by a worker
group’s leader,

[The workers’ group] would rather engage in jungle clearance. This is the type of work
that is needed in this area, because it is from jungle clearance that their households,
their economy, would really benefit. But they are always assigned to irrigation works
[. . .] because these are the works that they get from the government. [RDD] never asks
which works would be better for them. (Worker, Visakhapatnam district)

This was again a tribal community, where jungle clearance was viewed as more impor-
tant than irrigation, resulting in a scale of priorities that the Rural Development Depart-
ment reportedly did not recognize. In the best cases reported, workers are consulted on
the works that they perceive as most important, but it is still the Department that makes
the final choice. It must be noted, however, that leaving the decision to the Gram Sabha
would not by itself imply ensuring wageseekers’ participation, as in many instances
across India, the Gram Sabha depends strategically on the most powerful people
within villages.
Another problem is found in the dynamics internal to groups, in which the discre-
tionary power enjoyed by the mate is very high. Mates are in charge of voicing the
group’s stance when complaints are to be made, hence their role is characterized by
responsibility rather than sheer authority. But it is undeniable that mates exert substan-
tial power, as consistently reported in wageseekers’ interviews:

The mate is very independent, and collaborates with the FA [. . .] if he wishes, the mate
can withhold payments too, by not inserting names in the muster rolls [. . .] payments
often depend on the mate’s decision. (Worker, Visakhapatnam district)

Hence, MGNREGA workers in the state find themselves objectified not only by the
problem of payments, but even by the groups’ structure, in which they are de facto sub-
jected to the mate. For how the programme is constructed, workers seem to be excluded
from participation in all its main aspects, and an alienating process of objectification
20 Silvia Masiero and Diego Maiorano

surfaces in payments, works’ assignations process, and even in the internal dynamics of
workers’ groups.
To sum up, the picture emerging from the data seems to suggest that the centralized
logic of MGNREGA in AP has been thought out to enhance the effectiveness of the
scheme. However this logic constructs wageseekers as passive beneficiaries of the pro-
gramme, rather than as active participants to it. Rather than subverting existing power
relations, the information system seems to embody a policy that crystallizes them,
which is reflected in the programme’s design and implementation. As it emerges
from wageseekers’ narratives, the system results in higher degrees of transparency,
but does not result in the greater downward accountability that an effective check on
the power of Field Assistants would entail.

6. Discussion and conclusion

In this paper, the link between uptake of e-governance and accountability of anti-
poverty programmes has been observed through the case of MGNREGA’s computer-
ization in AP. Studying the information system related to it, we found that the systems’
design carries an accountability structure that acts on the transparency of internal trans-
actions, but not on the direct involvement of wageseekers. Our findings reveal the
inscription of a specific policy agenda in e-governance, for which the historical
phase and political context of information system implementation have been crucial.
This problematizes the ‘tool-and-effect’ logic in the relation between e-governance
and accountability, as the technology-led promotion of a centralized system results
in suboptimal levels of accountability to wageseekers.
While based on a single workfare scheme, central to India’s anti-poverty agenda,
implications may be envisaged at a more general level, concerning the globally
rising use of ICTs in social safety nets for poverty reduction. Adding to the rapidly
expanding literature on this topic, we find that e-governance does not per se speed
up existing accountability mechanisms, but seems, at least from the experience of com-
puterization recounted here, to embody the assumptions and priorities inscribed in the
state government’s political vision. In the case of AP MGNREGA, transparency of
transactions and centralization of responsibilities arise as core principles advanced by
the government, and are reflected in an information system that is rooted specifically
on their promotion. But the system also embodies the drawbacks of this agenda, rele-
gating workers to a non-participatory role and making it impossible for illiterate wage-
seekers to access information without intermediaries.
If this is so, adverse class politics in the scheme are not challenged, but reinforced
by ICT implementation. This is because the information system is designed in such a
way to consolidate existing power roles, not to modify them in favour of greater
empowerment of wageseekers. The vision of technology as a carrier of policy, found
in Cordella and Iannacci (2010), is hence reflected in our findings, as the information
system embodies policy decisions made prior to computerization. The view of ICT as
 Forum for Development Studies 21

capable to subvert existing power relations is thus problematized, leaving room for a
vision of technology as instrumental to enacting the will of the political actors
behind it.
In turn, this problematizes the view of computerization as a desirable technical fix to
the issues affecting delivery of anti-poverty programmes. As noted by Corbridge and
Srivastava (2013), the function of anti-poverty programmes is largely developmental,
and goes beyond the short-term task of transferring social benefits to their recipients.
Social safety nets are a direct expression of development policy, and are functional
to promoting the principles of social protection behind them (Sabates-Wheeler and
Devereux, 2007). Rather than being viewed as standalone technological units, digital
artefacts hence need to be observed in continuity with the policy agendas that they
put forward.
Beyond conceptual contributions, three orders of practical implications emerge
from the study. First, we have learned from wageseekers’ experience that technical
glitches, such as non-recognition of workers by biometric devices, may result in
denial of benefits to entitled recipients. This is particularly crucial at the present day,
as biometric recognition is becoming hegemonic in anti-poverty schemes worldwide
(Gelb and Decker, 2011). If this is the case, the perverse effects of failure in biometric
recognition, and more at large in technology for authentication of beneficiaries, need to
be tackled as a matter of urgent priority, because problems lying at the technical surface
level may result in the actual disempowerment of beneficiaries.
Second, technical changes in anti-poverty infrastructure may be inspired by deep-
seated aims of policy reform. In India, Aadhaar is becoming increasingly connected
to the delivery of social benefits, a move supported by the Central Government’s
decision to enact deep reform of the anti-poverty infrastructure (Government of
India, 2015). Rather than simply combating leakage, Aadhaar-based identification is
now being used to enact a policy shift to cash transfers, to be implemented in the
place of the subsidies that dominate the current national anti-poverty agenda (Bhatia
and Bhabha, 2017; Khera, 2016). If enforced, this move will result in the radical recon-
struction of India’s anti-poverty system, with deep consequences for the entitlements of
current beneficiaries and for the ways in which these will be delivered.
As a final remark, ethical considerations emerge when intersecting technology
adoption with anti-poverty programme delivery. Exclusion errors, predicated on non-
recognition of entitled beneficiaries, remind of the importance of ensuring alignment
between technical infrastructures and the reality lived by users. As Aadhaar enrolment
is becoming increasingly crucial for social benefits, including MGNREGA wage pay-
ments, the ethical importance of ensuring smooth delivery of anti-poverty benefits to all
entitled users cannot be overstated. This calls for more research on technologies that,
beyond combating leakage, ensure the minimization of exclusion errors that deprive
genuine beneficiaries of their entitlements.
MGNREGA is deeply entrenched in Indian anti-poverty strategies, both at the
national and at the state level. The construction of better accountability structures,
22 Silvia Masiero and Diego Maiorano

based on appraisal of existing drawbacks, can increase responsiveness of the pro-

gramme to wageseekers, which is vital for them to access their entitlements as pre-
scribed by the scheme. We hope that our suggestions, based on the case presented
here, may be a useful basis to integrate e-governance in the pursuit of this objective.

Acknowledgements
We would like to thank three anonymous reviewers, whose constructive comments have enabled
us to substantially improve the quality of the paper.

Notes on contributors
Silvia Masiero is a Lecturer in International Development at the School of Business and Econ-
omics, Loughborough University. Her research focuses on the use of information and communi-
cation technologies (ICTs) in the field of socio-economic development. Her current work
revolves around the embeddedness of ICT artefacts in development policy and governance,
with a specific interest in their participation in the politics of anti-poverty programmes.

Diego Maiorano is a Leverhulme Trust Early Career Fellow at the Institute of Asia-Pacific
Studies and the School of Politics and International Relations, University of Nottingham. His
research focuses on India’s politics and development and on political and economic change
in developing countries, with special reference to the themes of poverty and inequality. His
current research focuses on India’s largest anti-poverty programme, the Mahatma Gandhi
National Rural Employment Guarantee Act. He is the author of Autumn of the Matriarch -
Indira Gandhi’s Final Term in Office (Hurst/Oxford University Press/Harper Collins, 2015).

References
Accountability Initiative, 2015, MGNREGS Budget Brief 2015/16, New Delhi: Accountability
Initative.
Afridi, F. and V. Iversen, 2013, ‘Social audits and MGNREGA delivery: Lessons from Andhra
Pradesh’, India Policy Forum, pp. 1 –47.
Aiyar, Y. and S. Samji, 2009, Transparency and Accountability in NREGA: A Case Study of
Andhra Pradesh, New Delhi: Accountability Initiative, CPR.
Barrientos, A. and D. Hulme, (Eds.), 2016, Social Protection for the Poor and Poorest:
Concepts, Policies and Politics, London: Springer.
Bhagwati, J. and A. Panagariya, 2014, ‘Rural Inefficiency Act: Despite protests about diluting
NREGA, the PM is right to confine it to 200 poorest districts’, Times of India, 23 October.
Bhatia, A. and J. Bhabha, 2017, ‘India’s Aadhaar scheme and the promise of inclusive social
protection’, Oxford Development Studies, Vol. 45, No. 1, pp. 64– 79.
Bhatnagar, S., 2004, E-government: From Vision to Implementation – A Practical Guide with
Case Studies, New Delhi: Sage.
Bhatti, B., 2012, ‘Aadhaar-enabled payments for NREGA workers’, Economic & Political
Weekly, Vol. 47, No. 49, pp. 17 – 20.
Carswell, G. and G. De Neve, 2013, ‘Women at the crossroads: Implementation of employment
guarantee scheme in rural Tamil Nadu’, Economic & Political Weekly, Vol. 48, No. 52, pp.
83– 89.
 Forum for Development Studies 23

Carswell, G. and G. Neve, 2014, ‘MGNREGA in Tamil Nadu: A story of success and transform-
ation?’, Journal of Agrarian Change, Vol. 14, No. 4, pp. 564 – 585.
Chopra, S. and R. Khera, 2012, ‘Cutting delays in MGNREGA wages’, Ideas for India, http://
www.ideasforindia.in/article.aspx?article_id=62.
Corbridge, S. and M. Srivastava, 2013, ‘Mapping the social order by fund flows: The political
geography of employment assurance schemes in India’, Economy and Society, Vol. 42, No.
3, pp. 455– 479.
Cordella, A. and F. Iannacci, 2010, ‘Information systems in the public sector: The e-Government
enactment framework’, The Journal of Strategic Information Systems, Vol. 19, No. 1, pp.
52– 66.
Das, U., 2015, ‘Does political activism and affiliation affect allocation of benefits in the rural
employment guarantee program: Evidence from West Bengal, India’, World
Development, Vol. 67, pp. 202– 217.
De Neve, G. and G. Carswell, 2011, ‘NREGA and the return of identity politics in western Tamil
Nadu, India’, Forum for Development Studies, Vol. 38, No. 2, pp. 205 – 210.
Deshpande, R., K. K. Kailash and L. Tillin, 2017, ‘States as laboratories: The politics of social
welfare policies in India’, India Review, Vol. 16, No. 1, pp. 85– 105.
Devereux, S. and K. Vincent, 2010, ‘Using technology to deliver social protection: Exploring
opportunities and risks’, Development in Practice, Vol. 20, No. 3, pp. 367– 379.
Drèze, J., 2005, ‘Loot for Work Programme’, Times of India, 1 July 2005.
Drèze, J., 2014, ‘Learning from NREGA’, The Hindu, 23 August 2014.
Gelb, A. and C. Decker, 2011, ‘Cash at Your Fingertips: Biometric technology for transfers in
developing and resource-rich countries’, Centre for Global Development Working Paper no.
253.
Ghosh, J., 2011, ‘Dealing with “The Poor”’, Development and Change, Vol. 42, No. 3, pp. 849–
858.
Government of India, 2015, ‘Wiping every tear from every eye: The JAM Trinity number sol-
ution’, Economic Survey 205– 2016, http://indiabudget.nic.in/es2014-15/echapvol1-03.pdf.
Heeks, R., 2002, ‘Information systems and developing countries: Failure, success, and local
improvisations’, The Information Society, Vol. 18, No. 2, pp. 101 –112.
Imbert, C., and J. Papp, 2015, ‘Labor market effects of social programs: Evidence from India’s
Employment Guarantee’, American Economic Journal: Applied Economics, Vol. 7, No. 2,
pp. 233– 263.
Jakimow, T., 2014, ‘“Breaking the backbone of farmers”: Contestations in a rural employment
guarantee scheme’, The Journal of Peasant Studies, Vol. 41, No. 2, pp. 263 –281.
Jenkins, R. and J. Manor, 2017. Politics and the Right to Work – India’s National Rural
Employment Guarantee Act, London: Hurst & Co.
Khera, R., 2011, ‘The UID project and welfare schemes’, Economic & Political Weekly, Vol. 46,
No. 9, pp. 38– 44.
Khera, R. (ed.), 2012, The Battle for Employment Guarantee, New Delhi: Oxford University
Press.
Khera, R., 2016, ‘MNREGA: Technology vs. technocracy’, Ideas for India, http://www.
ideasforindia.in/article.aspx?article_id=1599.
Liu, Y. and K. Deininger, 2010, ‘Poverty impacts of India’s national rural employment guaran-
tee scheme: Evidence from Andhra Pradesh’, Selected Paper prepared for presentation at the
Agricultural & Applied Economics Association 2010 AAEA, CAES, & WAEA Joint
Annual Meeting, Denver, Colorado, July 25 – 27, 2010.
24 Silvia Masiero and Diego Maiorano

Madon, S., 2009, E-Governance for Development: A Focus on Rural India, London: Palgrave
Macmillan.
Maiorano, D., 2014, ‘The politics of the Mahatma Gandhi National Rural Employment
Guarantee Act in Andhra Pradesh’, World Development, Vol. 58, pp. 95– 105.
Maiorano, D. and C. Buddha, 2015, ‘MGNREGA in Andhra Pradesh’s tribal areas’, Economic
& Political Weekly, Vol. 49, No. 51, pp. 15 –17.
Maiorano, D. and U. Das, 2015, ‘Is the MGNREGA fund crunch making the programme clien-
telistic?’, Ideas for India, http://www.ideasforindia.in/article.aspx?article_id=1497.
Maiorano, D., S. Thapar-Björkert and H. Blomkvist, 2016, ‘The paradoxes of empowerment:
Gendering NREGA in the rural landscape of India’, Development in Practice, Vol. 26,
No. 1, pp. 127– 134.
Ministry of Rural Development (MORD), 2012, NREGA Sameeksha, New Delhi: Oriental
Blackswan.
Mookherjee, D., 2014, ‘MNREGA: Populist leaky bucket or successful anti-poverty pro-
gramme’, Ideas for India, http://ideasforindia.in/article.aspx?article_id=292.
Murali, A., 2012, ‘India: Technological innovation for effective Management Information
Systems’, World Bank Group, http://elibrary.worldbank.org/doi/pdf/10.1596/
9780821389683_CH10.
Muralidharan, K., P. Niehaus and S. Sukhtankar, 2014, ‘Building state capacity: Evidence from
biometric smartcards in India’, National Bureau of Economic Research (NBER) Working
Paper No. 19999, October 2014.
Narayanan, S., U. Das, Y. Liu and C. B. Barrett, 2016, ‘The ‘discouraged worker effect’ in
public works programs: Evidence from the MGNREGA in India’, SSRN Working Paper,
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2882149.
Pai, S., 2013, ‘Delayed NREGA payments drive workers to suicide’, Hindustan Times, 29
December.
Pattenden, J., 2011, ‘Gatekeeping as accumulation and domination: Decentralization and class
relations in rural south India’, Journal of Agrarian Change, Vol. 11, No. 2, pp. 164 –194.
Pattenden, J., 2017, ‘Class and social policy: The national rural employment guarantee scheme
in Karnataka, India’, Journal of Agrarian Change, Vol. 17, No. 1, pp. 43 – 66.
Price, P., 2011, ‘Re-visiting policy and political behavior in India with a view of the National
Rural Employment Guarantee Act’, Forum for Development Studies, Vol. 38, No. 2, pp.
197– 203.
Ravi, S. and M. Engler, 2015, ‘Workfare as an effective way to fight poverty: The case of India’s
NREGS’, World Development, Vol. 67, pp. 57– 71.
Riessman, C. K., 2008, Narrative Methods for the Human Sciences, Thousand Oaks: Sage.
Ruparelia, S., 2013, ‘India’s New Rights Agenda: Genesis, promises, risks’, Pacific Affairs, Vol.
86, No. 3, pp. 569– 590.
Sabates-Wheeler, R. and S. Devereux, 2007, ‘Social protection for transformation’, IDS bulletin,
Vol. 38, No. 3, pp. 23 –28.
Shankar, S. and R. Gaiha, 2013, Battling Corruption: Has NREGA Reached India’s Rural
Poor?, New Delhi: Oxford University Press.
Srinivasulu, K., 2009, ‘Y. S. Rajasekhara Reddy: A political appraisal’, Economic & Political
Weekly, Vol. 44, No. 38, pp. 8 – 10.
Subbarao, K., C. Del Ninno, C. Andrews and C. Rodriguez-Alas, 2013, Public Works as a
Safety Net, Washington, DC: The World Bank.
Suri, K., 2004, ‘‘Andhra Pradesh: Fall of the CEO’ in arena of democracy’, Economic and
Political Weekly, Vol. 39, No. 51, pp. 5493 – 5497.