Introduction

Comparing the Politics of Welfare

across Indian States

Louise Tillin, Rajeshwari Deshpande, and

K. K. Kailash

This book advances an agenda for the study of comparative

subnational politics and social policy in India. Indian states are at
the forefront of responding to popular expectations around welfare,
governance, and development at a time when political participation
has deepened among previously marginalized social groups and an
expanding middle class. Put simply, state governments are on the
frontline of resolving what has been posed across the global south as a
tension between the achievement of economic growth and the extent
to which it is socially inclusive (Bhagwati and Panagariya 2013;
Dréze and Sen 2013; Mkandawire 2005). Despite the importance of
subnational contexts in shaping India’s politics of welfare, the links
between regional political scenarios and the shape of emergent wel-
fare regimes are not well understood.
2 Politics of Welfare

This volume examines the politics of welfare across Indian states in

the Context of rapid economic growth, and against the backdrop of
new social legislation introduced by India’s central and state govern-
ments led by many different political parties since the mid-2000s.
Contributions to the volume demonstrate that there is substantial
diversity in how state governments implement, shape, and add to
central policy agendas as they seek to balance their approaches to
economic development and social policies while maintaining demo-
cratic legitimacy. State-level experiences show that the relationship
between growth and inclusiveness is not always a trade-off. Greater
revenues as a result of economic growth have increased the ability
of Indian state governments to invest in social provisioning, even
while tensions between pro-business and welfare agendas continue to
arise in areas such as land acquisition or unionization. The ways that
state governments have chosen to resolve such tensions vary. Much
of the most significant public policy experimentation has thus been
taking place at the state level in recent years. State-level approaches
reflect variation in historical patterns of socio-political dominance,
the organization of political parties and interest groups, bureaucratic
structures, economic context, and the agency of political leaders.
Through a series of specially commissioned comparative studies
conducted by individuals and collaborative teams of researchers, this
volume investigates the relationship between political environments
and social policies across Indian states within its multilevel federal
system. It thereby deepens knowledge of India’s multiple emergent
welfare regimes, challenging the national-level bias that is implicit in
much of the existing literature on social policy in India.

India’s New Politics of Welfare

After 2004, India saw a departure from several decades of retrench-

ment in the state’s welfare commitments. A national turning point
came with the election of the Congress-led United Progressive
Alliance (UPA) central government, supported in its first term of
office by parties of the Left Front. The Congress party chose to inter-
pret its electoral success as representing a revolt by the rural poor in
protest against the unequal distribution of the proceeds of economic
growth. This national electoral shift opened space for a transition
 Introduction 3

from the dynamics ofthe early years ofliberalization. The early 1990s
had seen a contraction ofstate expenditure on welfare, and the reform
of large programmes such as food subsidies to target provision only
to the populatfn deemed ‘below poverty line’. India was not unique:
declines in welfare spending were observed across the global south in
the 1980s and 1990s (Kaufman and Segura-Ubiergo 2001; Rickard
2012; Rudra 2002; Wibbels 2006). In many countries, the change
took the form of a shift away from more widespread or universal
schemes associated with an over-developed and rent-seeking state,
towards a goal of targeting scarce public resources to programmes
aimed at those in greatest need (Kabeer 2014).
A combination of civil society activism, judicial interventions, and
supportive political leadership—in addition to substantial increases
in state revenues in the context of faster economic growth from
2003 onwards—came together in the mid-2000s in India to forge
a new generation of ‘rights-based’ programmes. These programmes
sought to empower citizens to demand services as ‘rights’ which were
legally protected, rather than view them as patronage bestowed by
a selectively benevolent state. The reforms have been described as
‘audacious (Harriss 2013) because they opened up new mechanisms
through which ordinary people could demand rights and hold the
state accountable (Jenkins 2013; Ruparelia 2013). India’s central gov-
ernment thus oversaw a significant expansion of state-funded welfare
programmes during the second phase of economic liberalization.
The increased attention to social protection in India was mirrored
in other parts of the global south, reflecting a wider shift in attitudes
internationally and within donor agencies (Barrientos and Hulme
2009; Bender, Kaltenborn, and Pfleiderer 2013). These new activi-
ties in India also saw the projection of state power—and new modes
of governance—into marginalized spaces, such as parts of India's
Northeast, where state authority had earlier been circumscribed or
primarily coercive in nature (Dev, this volume).
Yet, as this volume shows, the realization of such visions of
transformed citizenship via new modes of welfare policies has been
partial, and—above all—contingent on local constellations of power
and historical trajectories of political contestation within Indian
states. Despite the fact that many of the new generation of social
programmes are sponsored by the central government, there has
4 Politics of Welfare

beer little systematic study of how national policymakers incen-

tivize or sanction states to adopt new policy agendas, or how local
political environments affect the shaping and implementation of
central programmes on the ground. This lacuna is curious since it
is this last mile of implementation at the subnational level which
critically determines the substantive outcomes $ of new policy agendas.
Furthermore, it is the states themselves that are frequently the origi-
nators of these new agendas.
Our analysis deliberately moves beyond single-state case studies to
engage in subnational comparisons which explain variation in public
policy formulation, implementation, and monitoring across space. By
adopting a comparative subnational perspective, the contributions to
the volume isolate what is distinctive or similar about the trajectories
of individual states. They thus avoid the twin pitfalls of false excep-
tionalism and universalism (Rose 1991) that bedevil single-state case
studies which continue to form a dominant approach within scholar-
ship on state politics in India more generally (Manor 2010; Tillin
2013). By comparing states within the common institutional frame-
work of the Indian federal system, and selecting carefully matched
pairs of states, authors are able to focus on the politics driving social
policy across states while controlling for non-political factors that
might otherwise explain differences between states.! In scaling down
to compare subnationally (Snyder 2001a), we are able to delineate
in this introductory chapter the emerging patterns among states that
can be considered more consistent ‘welfare regimes’, and those that
are less consistent performers.

' On the advantages of subnational comparison for studying the politics

of policy variation, see Snyder (2001b). On paired comparisons gener-
ally, see Tarrow (2010) and in the Indian context, see Jenkins (2004) and
Palshikar and Deshpande (2009). Kailash (201 1a) provides a detailed review
of traditions of subnational political analysis in India, delineating schools
of single state case studies (including the compendia of Frankel and Rao
1989; Narain 1967; Weiner 1968; Wood 1984); large-N quantitative com-
parisons (including Ayyangar and Jacob 2008; Chhibber 1995; Chhibber
and Nooruddin 2004; Khemani 2007; Sdez and Sinha 2010); and small-N
qualitative comparisons (for example, Chandra 2004; Desai 2007; Kohli
1987, 2012; Kumar 2011; Manor 2004a, 2004b, 2007, 2010; Mitra 2006;
Sinha 2005; Varshney 2002; Wilkinson 2006).
 Introduction 5

The cross-state comparisons focus on states across the breadth of

the Indian subcontinent, including states from southern, central,
eastern, northern, north-eastern, and western India. We have also
included a c@mparison of two cities, Delhi and Chandigarh, to
consider the effects of multilevel governance structures on policy
processes in urban settings—an area of growing importance as
India urbanizes. We have also selected a cross-section of areas of
social policy for close engagement. These include social assistance
policies (health insurance and social security for informal sector
workers), rights-based programmes (Mahatma Gandhi National
Rural Employment Guarantee Scheme [MGNREGS] and the Right
to Education [RTE]), as well as attempts to reform a classic subsidy
programme (Public Distribution System [PDS]). This breadth allows
the volume as a whole to produce a uniquely synoptic account—
drawn together in this introductory chapter—of state-level political
regimes as contexts for welfare provision across India.
The introductory chapter is organized in two main parts. The
first section looks at how the volume innovates by bringing together
political and policy analyses, seeking to move beyond a simple pes-
simism about the contribution of electoral politics to public policy.
It goes on to assess the changed nature of welfare policies in India in
the context of economic liberalization, and asks—with reference to
the wide-ranging literature on ‘welfare states’, and the role of the state
in development—what kind of welfare regime(s) can be identified in
the country. We use our analysis to present a preliminary classifica-
tion of subnational ‘welfare regime’ types across Indian states. We
then provide a justification for treating the state or federal sub-unit
as our unit of analysis. The second part of the introduction turns to
the major themes treated within the chapters of the volume, drawing
out the principal findings with regard to the linkages between politics
and policy processes. This is followed by a description of some of the
major lines of differentiation between states in terms of their overall
social policy mix, and an overview of the argument of each chapter.
It should be noted that the primary analytical focus of the volume is
on how regional political environments shape the formulation and
implementation of policy. While the chapters also use secondary data
to consider welfare outcomes, we do not present a systematic evalua-
tion of policy effects since this would take us beyond the scope of our
6 Politics of Welfare

empirical research. Furthermore, while we remain alert to the fact

that structural economic transformations involving employment and
labour markets may be as important to welfare outcomes as social
protection and social programmes (see also United Nations Research
Institute for Social Development [UNRISD] 2010), our focus here
i.
is primarily on social policy interventions.

Subnational Comparative Politics and Welfare

Regimes in India
Analysing the Interface between Politics and Policy
There is, as numerous scholars have observed, something of a curi-
ous divide between the study of public policy and the study of
political parties and electoral politics (Kitschelt 1999; Sinha 2004,
p. 72). At one level such a divide reflects scepticism about whether
electoral politics is a primary, or even prominent, determinant of
policy. On a more prosaic level, this division also reflects disciplin-
ary and departmental divisions of labour. As Mooij and de Voos
(2003, p. vii) note: “There are very few political scientists, sociolo-
gists or anthropologists [in India] focusing on public policies.... The
issues and questions, for instance of why policies are formulated
and designed in particular ways in the first place, and the political
shaping of policies “on the ground”, do not receive much attention.’
Melo, Ng’ethe, and Manor (2012) remark similarly that ‘[d]espite
their huge importance, politicians in less-developed countries have
received remarkably little attention in analyses of development and
poverty’.
It is worth stressing that a divide between the study of electoral poli-
tics and public policy is not a uniquely Indian phenomenon. Scholars
of European public policy have pointed to a gulf between studies of
political economy and of political parties, or to the pre-eminence of
‘anti-politics’ understandings of public policy formation. Increasing
ideological convergence between political parties in advanced capital-
ist economies has meant that elections have no longer appeared so
pivotal in producing major changes to policy environments (for a
critique of this view, see Boix 1997; Castles and McKinlay 1977),
although elections are frequently seen as influencing the timing of
 introduction 7

expenditure. Recent research comparing states across India has also

suggested that factors connected to the election cycle, or the effective
number of political parties contesting for power in a given state, are
more significagt than partisan or ideological factors in explaining lev-
els of expenditure on public goods (Chhibber and Nooruddin 2004;
Saez and Sinha 2010).
The other reason for a divide between the study of domestic
politics and public policy is the increasing integration of the global
economy. Closer economic integration has raised questions about
how far the policy options of governments have been constrained,
particularly to pursue policies favoured by organized labour or class-
based political movements within individual countries. Explanations
for overall declines in welfare expenditures in developing countries
in the 1980s and 1990s have highlighted the weak organization of
labour in these countries (Rudra 2002; Wibbels and Arce 2003);
pressure for fiscal retrenchment from international financial institu-
tions in the context of structural adjustment; as well as the fact that
governments in the global south are more sensitive to income shocks
that lead them to cut social spending at precisely the times it is most
needed (Wibbels 2006). As Wibbels (2006, p. 438) argues, ‘[g]iven
the need to fiscally retrench in the face of recessions, governments
face tradable [export-oriented] sectors whose long-term interest
in social spending conflicts with their short-term interest in fiscal
balance’.
Such fluctuations in welfare commitments have given rise to a
more general concern that as developing countries, including India,
have become more integrated with the global economy, the quality of
their democracy has been compromised because more disadvantaged
groups have been less able to organize and exert influence to protect
their interests (Yadav 1999, p. 2398). Beyond this, however, also lies
a deeper cynicism about the world of politicians and politicking as a
morally ambivalent, frequently corrupt, sometimes criminal, zone of
activity driven by the pursuit of self-interest. As Vasudha Chhotray
(2011, p. xv) writes, ‘such cynicism... [affects] a generally negative
perception of what democratic politics and politicians can deliver’
(emphasis added).
In India, Partha Chatterjee’s resurrection of the Gramscian binary
of civil and political society influentially suggested that in the context
8 Politics of Welfare

of economic liberalization, crucial areas of policymaking on economic

issues remain isolated from the rough and tumble of electoral poli-
tics (Chatterjee 2004, 2008; see also Nayyar 1998). The poor make
transactional demands of the state from within ‘political society’,
but in Chatterjee’s thematic schema they do not—or cannot—make
sustained demands to pursue policy change’ through their engage-
ment with politicians and political parties. Relatedly it is argued that
elections do not act as an opportunity for voters to choose between
competing policy platforms offered by different parties (Chandra
2004; Keefer and Khemani 2004; Kitschelt and Wilkinson 2007a).
Instead, political parties offer a mixture of (a) targeted benefits and
services to smaller groups of voters, often defined by caste identity,
through clientelistic ties; and (b) what are commonly described as
‘populist’ schemes designed to win broader-based support, often
through pre-election ‘give-aways, such as promises of cheap food,
electricity, TVs, or laptops. In a slightly different formulation, Pranab
Bardhan (1984) in his classic analysis saw India’s democracy as cap-
tured by dominant proprietary classes with a vested interest in the
proliferation of unproductive subsidies to the longer term detriment
of development and poverty reduction.
The concept of clientelism, so prevalent in the analysis of dis-
tributive politics in lower income democracies, is premised upon
the notion of the contingent exchange of material goods and/or
access to social programmes between political parties or brokers,
and voters (see, for instance, Hicken 2011; Kitschelt and Wilkinson
2007b; Stokes et al. 2013). Clientelism stands in contrast to more
programmatic approaches to public policy in which transparent cri-
teria determine the distribution of and access to social goods. It is
typically understood as a mode of political socialization that militates
against the provision of public goods and the improvement of overall
welfare outcomes. As Stokes et al. (2013, p. 6) state, ‘[t]he story of
the demise of clientelism and machine politics is... the prehistory of
the welfare state.’ These facets of developmental interventions initi-
ated by the state, and the uncertain articulation of the welfare claims
of the poor within political life, have given rise to what Deshpande
(this volume) describes as a key puzzle for Indian democracy: the
apparent (im)possibility of the politics of the poor, despite their
increased political participation (see also Jayal 2013).
 Introduction 9

In this volume we urge a deeper look at the political and eco-

nomic dynamics that shape engagement with social policies in
different ways across states, and the ways in which social policies are
constructed agd implemented differently from the past. In doing
so, we complicate the narrow lens through which the influence of
electoral politics on public policy is often read. We find substantial
variation across states in the design of social programmes, modes of
distributive politics including the linkages between the state, political
actors, and voters, as well as in their outcomes. In presenting our
analysis, we insist that the link between politics and policy processes
across India cannot be reduced either simply to clientelism or to
expansionary fiscal distortions linked to the election cycle (although
these things are not to be dismissed either). Rather we insist that
different regional political settlements and leadership play a central
role in determining the shape that policies take in different places. As
we will outline, it is possible to identify a distinct cluster of emergent
welfare regimes as well as a cluster of states that are less consistent
performers often because clientelistic modes of political organization
persist more overwhelmingly, or because more exclusive pro-business
priorities overshadow the political realm.

What Kind of Welfare Regime(s) Does India Have?

Under the UPA government elected in 2004, a new National Advisory
Council (NAC) was appointed to play a leading role in designing
welfare initiatives as part of the government’s common minimum
programme. The NAC crucially provided an institutional channel
for civil society involvement in the formulation of new rights-based
policies (Arora and Kailash 2014; Chopra 2014). There was a pro-
nounced change in the scale and composition of social protection
expenditure in this period, particularly following the introduction
of the Mahatma Gandhi National Rural Employment Guarantee
Act (MGNREGA) which promised paid employment for 100 days
annually to every rural household that demanded it. Legislation was
passed enshrining a Right to Education, Right to Employment (via
the MGNREGA), Forest Rights Act, and although it is not directly
considered within the ambit of welfare policies, Right to Information,
which nevertheless has implications for how citizens interact with the
10 Politics of Welfare

State to secure their rights. In addition, a National Food Security Bill

passed in late 2013 provided legislative backing for a Right to Food,
albeit not.a universal one. These are all policies for which civil soci-
ety organizations (CSOs) have campaigned, in part through Public
the Supreme Court.
Interest Litigation (PIL) pursued through
In addition to rights-based legislation, there have been new
approaches in other areas of social policy, many of which began at
the state level and which the central government has attempted to
scale up. These include experiments with health insurance (Rashtriya
Swasthya Bima Yojana [RSBY], which followed the state-level scheme
Aarogyasri in Andhra Pradesh); the National Rural Health Mission
which picked up elements from state schemes; reform of social assis-
tance programmes such as the public distribution of subsidized food
through the National Food Security Act passed in late 2013 (build-
ing on the reforms adopted in states such as Chhattisgarh and Tamil
Nadu); and experiments with cash transfers or ‘direct benefit transfers’.
The categorization employed by Indian public finance statistics
does not permit straightforward analysis of expenditure on social
policy, yet indicative analysis suggests a substantial expansion in
social sector expenditure across this period. As Figure 1.1 suggests,
social sector spending as a proportion of the total government expen-
diture rose from 23 per cent in 2000-1 to 29 per cent in 2013-14.
This figure provides an estimate of social spending based on the
major budget categories of ‘social and community services’ including
education, health, family welfare, and social security in addition to
the separate category of rural development. The calculation does not,
however, include expenditure on food subsidies which are considered
an example of social protection in this volume, but are classified as
‘non-developmental expenditure’ within the budget.
It is worth remembering, however, that India’s overall social expen-
diture remains low in comparative perspective. For instance, India
spent only 4 per cent of GDP on health in 2012 compared to 4.7 per
cent in Kenya, 8.8 per cent in South Africa, and 9.3 per cent in Brazil.
It spent 3.4 per cent of GDP on education, compared to 6.6 per
cent in South Africa and 5.8 per cent in Brazil (World Development
Indicators 2014).? Social protection expenditure by the central

* These are accessed online via the World Bank website: hetp://data.
worldbank.org/products/wdi (accessed 23 August 2014).
 Introduction 11

1990-1
2000-1
2006-7
2007-8
2008-9
2009-10
2010-11
2011-12
2012-13
2013-14

0 5 10 15 20 25 30 35
® Central government expenditure mf State government expenditure

Figure 1.1 Social Sector Spending as Proportion of Total Government

Expenditure
Source: Data derived from Tables 1.1, 2.1, and 3.1, Indian Public Finance Statistics
2014-15, Ministry of Finance at http://finmin.nic.in/reports/IPFStat201415.pdf
(accessed 15 February 2016).
Notes: Social sector spending includes total expenditure on budget categories of
social and community services plus rural development. Central government expen-
diture includes the totals reported on these budget categories in Table 2.1 plus the
estimated share of non-apportionable central Plan grants spent by the states on the
social sector. The states’ share of total social sector expenditure therefore provides
an estimate of the amount that states spend on the social sector from the untied
revenues at their disposal (a combination of their own revenues and share of central
taxation). Our thanks to Indira Rajaraman for assistance with the data. Any errors
are Our OWN.

government—including MGNREGA and the PDS—accounted for

only 1.7 per cent of the GDP in 2009 compared to 3.6 per cent in
Thailand or 5.4 per cent in China (Asian Development Bank 2013,
pp. 13-14). The breakdown between central and state government
expenditures is also telling. While India’s total social sector expen-
diture has been growing, the central government's commitment as a
proportion of total expenditure has fallen in the most recent years.
The overall increase in recent years is driven by the growing propor-
tion of state-level expenditure committed to the social sector.? This

> Central government expenditure has increased in absolute terms but

fallen as a proportion of total government expenditure.
12 Politics of Welfare

further underlines the importance of the states as the crucial location

for'understanding trends in social policy.
The overall shift in welfare policies and expenditure outlined so
far raises the question of whether a recognizable ‘welfare state’ in the
‘mainstream’ sense of the term has now emerged in India, and if so,
what kind of a welfare state it is.* Research on Indian social policy
provides no clear roadmap to answer this question: much of the
existing research is not informed by specific theoretical concerns or
analytical frameworks, and many studies are primarily descriptive in
nature. While there is substantial literature that examines particular
policies from different perspectives including evaluating their aims,
goals, and effectiveness, such studies are often more focused on pre-
scription than on understanding the overarching nature of welfare
regimes, and the politics that produce them. Studies of social policy
in the Indian context ask different questions, use varied methods,
and frequently do not position themselves with regard to broader
literatures (Kailash 2013). This stands in contrast to Europe, where
there is not only an established tradition of examining the politics
of welfare but the comparison of welfare states across the region
is itself a veritable industry. It is therefore not surprising that the
dominant approaches, indicators, and measures to both understand-
ing and comparing welfare states in the political science discipline
have been drawn from the experience of advanced European welfare
states (Amenta 2003). There is, as Thandika Mkandawire (2005, p.
4) states, ‘little in social policy studies in developing countries that is
as heuristically potent as Esping-Anderson’s work on welfare regimes
[in Europe]’. This theoretical thrust has had the effect of foreground-
ing the experience of the global north and frequently privileging a
focus on state-provided welfare over other forms of social assistance
(Midgley 2004, p. 217).

* This is not to imply the absence of state commitment to welfare goals

prior to 2004. Immediately after Independence, the state sought to provide
a host of social services for the welfare of the people via centralized planning.
The community development programme, for instance, aimed to use locally
available resources to create socially useful assets. The programme linked
a commitment to social welfare, even if partially realized, with a focus on
economic development common to other nation-building efforts elsewhere
(see also Midgley 2004, p. 226).
 Introduction 13

Within the literature on welfare systems in Organisation for

Economic Co-operation and Development (OECD) countries,
much of the focus is on aspects of policy design that regulate citi-
zens access to gocial benefits and provisions (for example, Clasen
and Clegg 2003; Clegg 2002; Cox 1998; Esping-Andersen 1990;
Nelson 2012; Robinson 2000; Scruggs and Allan 2006; Walters
1997; Whiteside and Salais 1998), or on public expenditure (Castles
2002; Castles and Obinger 2007; Wilensky 1975). Yet by themselves,
neither legislation nor budget provision will tell us much about the
real impact or outcome of a particular policy, particularly in contexts
where the size of the informal economy makes income maintenance
and labour market interventions less feasible (Gough 2004b). In its
assessment of policy outcomes, the capability approach (Sen 1999)
further questions all resource-based approaches to welfare, positing
that measuring resources is fundamentally different from measuring
the capabilities of differently situated human beings.
The typology of welfare state regimes outlined by Esping-Andersen
(1990), which forms the starting point of much of the comparative
literature on this question, assumes an advanced capitalist context
in which—as Karl Polanyi (1944) set out—the ‘commodification
of labour has necessitated state intervention to ameliorate socio-
economic dislocations caused by the market. Esping-Andersen
identifies three clusters of welfare state regimes that are classified
in terms of variations in the interrelationships of state, market, and
family. Anglo-Saxon ‘liberal welfare states’ adopt a mixed model of
modest social transfers, including socially stigmatizing rules for ben-
efits, while encouraging the market to provide for growing middle
class demands for welfare. In conservative, ‘corporatist’ regimes such
as Austria, France, or Germany, the state marginalizes the market as a
provider of welfare; there is a strong commitment to the preservation
of family and the influence of the Church. Lastly, ‘social democratic’
regimes, such as Scandinavia, are characterized by a fully universal
welfare state, commitment to full employment and income protec-
tion, with high levels of taxpayer contributions, and a high premium
placed on equality. As Gough (2004a, p. 33) argues, the notion of a
welfare state regime in Esping-Andersen’s discussion assumes a mini-
mum ‘set of conditions where people can reasonably expect to meet
(to a varying extent) their security needs via participation in labour
14 Politics of Welfare

markets, financial markets and the finance and provisioning role of

a “welfare state”’. These conditions circumscribe the applicability of
the welfare state concept, and raise questions about the extent to
which it should be used in the Indian context.
The relevance of conventional understandings of the welfare state is
thus limited in India for reasons that include the fact that the labour
market is characterized by the existence of a large informal economy.
In many parts of the country, ‘public’ policy is not very well articulated,
and across India, the state is frequently not sufficiently autonomous
from social or economic power structures to be able to effectively imple-
ment redistributive public policy (or to oversee economic development
more generally [Herring 1999]). Furthermore, there are comparatively
low levels of taxation and low public expenditure on welfare overall,
as well as weak outcomes on a range of measures of welfare and well-
being. These structural features of the Indian economy, as well as the
social embeddedness of the state, lie at the heart of the challenges of
realizing on the ground what appears as a well-articulated shift at the
national level to a rights-based conception of an Indian welfare regime,
with a corresponding increase in expenditure on social policy.
Despite these limitations, some attempts have been made in recent
years to explore the relevance of welfare state conceptualizations in
countries with low income, even where there is higher insecurity than
would be associated with classic welfare state models (see, for exam-
ple, Bender, Kaltenborn, and Pfleiderer 2013; Gough et al. 2004;
Koehler and Chopra 2014; Rudra 2007; UNRISD 2010). Rudra
(2007) identifies three clusters of welfare state types across less devel-
oped countries: ‘productive welfare states (where efforts focus on
encouraging wage labour and engagement in the market); ‘protective
welfare states’ (which focus on protecting individuals from the mar-
ket); and ‘dual welfare states’ (with elements of both protective and
productive regimes). ‘Protective’ regimes are more likely to emerge
in countries which have historically been shielded from the effects
of competition in international markets, and where protection has
been given to both firms and workers in domestic industrial sectors.
Despite appearances, such regimes may not promote redistribution
because welfare benefits are directed towards a small, protected work-
force in the formal sector. ‘Productive’ welfare states are more likely to
emerge in places such as parts of East Asia where active participation
 Introduction 15

in export markets has been encouraged, and therefore governments

have been forced to surrender some control over the economy. In
such places there is a more limited range of social policies because
governments ang constrained by the interests of employers (Rudra
2007, p. 384). In Rudra’s analysis, India (at an aggregate level) is
classified as a ‘protective’ welfare regime, with countries such as
Korea, Costa Rica, Sri Lanka, or Thailand as examples of ‘produc-
tive’ regimes. Yet there is considerably greater variation within India
than a simple national-level attempt at categorization allows for. This
variation is not driven simply by levels of economic development
across regions in India, but suggests the existence and consolidation
of distinctive clusters of relationships between state, markets, society,
and the global economy in different states—expressed in a range of
political settlements—with implications for the nature of welfare
policies. In the next section we thus begin to sketch a clustering of
patterns of welfare politics within Indian states.

Regional Political Settlements and Subnational

Welfare Regimes in India
There is a body of work on comparative social policy internation-
ally that attests to the importance of local political settlements in
determining the shape of welfare regimes, and their perpetuation
over time (Collier and Collier 1991; Esping-Andersen 1990; Gough
et al. 2004; Haggard and Kaufman 2008; Swank 2002). Both the
literature on OECD welfare states and on the global south have
asserted the continuing relevance of domestic politics in shaping
distinctive clusters of national regimes, despite globalization (Rudra
2007; Swank 2002). Similarly, despite their location within a com-
mon institutional framework, India’s welfare scenarios are shaped
by regional politics. A major focus of the comparative literature on
this question has been the extent and means by which urban labour
and the poor in rural areas are incorporated or excluded by politi-
cal elites. Kohli (1987, 2012) explores similar themes in the Indian
national context to explain the differential success or commitment
of state governments to tackling poverty. His early work identified
the importance of well-organized, left-of-centre regimes which were
able to maintain a degree of autonomy from propertied classes to
16 Politics of Welfare

steer state intervention in the interests of lower classes. These regime

features shaped what he describes as India’s more social democrati-
cally inclined states, such as Kerala and West Bengal. But these states
are fairly exceptional within the wider Indian context, and the recent
history of West Bengal has also diverted. from the earlier ‘ideal type
Kohli sets out. Harriss (2003), building on Kohli’s insights, plus the
comparative insights of the major study coordinated by Frankel and
Rao (1989) into patterns of social dominance across states, offers a
more variegated picture of Indian states and their record on pov-
erty reduction. He highlights the crucial variables of how early and
fulsomely the Congress party dominance (and with it upper caste
and class dominance) was challenged across states, and the extent to
which lower castes and classes are represented in political regimes. As
well as the states identified by Kohli (1987), Harriss (2003, p. 205)
highlights states such as Tamil Nadu and Andhra Pradesh that have
institutionalized populist, welfare-oriented politics and have done
better at reducing poverty than might otherwise have been expected.
In Kohli’s more recent work, he identifies two other ‘ideal type’
regimes—neo-patrimonial and developmental—in addition to social
democratic regimes. He identifies ‘neo-patrimonial’ regimes (using
Uttar Pradesh and Bihar as examples) which lack public purpose and
where ruling elites use their power for personal gain or to provide
benefits to particular segments of the population. In such states,
patron-clientelism predominates and politics lacks the cohesion that
might be provided by well-organized political parties, regional nation-
alism, or organized class politics (Kohli 2012, pp. 145 and 152). In
these places, ‘the zero-sum quality of politics makes those excluded
from power feel totally excluded’ (Kohli 2012, p. 153). On the
other hand, ‘developmental’ states—such as Gujarat, Maharashtra,
Andhra Pradesh, Karnataka, Punjab, or Haryana—are characterized
by governments that work closely with business in order to promote
private-sector-led economic development. These states, he suggests,
tend to have a clear public purpose defined more narrowly in terms
of the pursuit of growth, and can see more volatile or exclusionary
politics (Kohli 2012, pp. 146 and 155). They also have mixed records
on poverty reduction.
In this book we go further than the existing typological accounts
to capture the mechanisms and dynamics within political and
 Introduction 17

administrative processes that afford insights into the actually exist-

ing life of social policies as implemented across a representative
range of states. Mechanisms and dynamics that we scrutinize in
depth include t@e nature of political leadership; the capacity and
autonomy of the bureaucracy; the strength of CSO, organized
interest groups and their relationship with the state; as well as the
design of different programmes. These mechanisms are typically
more significant than the partisan or ideological orientation of
political parties. They need to be understood within the context of
the evolving political economy of different states. Such dimensions
of the policy process are rarely systematically studied in macro-
accounts of regime-types and poverty reduction, but they are
critically important for understanding the nature of policy environ-
ments and outcomes at the subnational level. We identify here six
broad types of subnational regimes within two main clusters that
help to frame the analysis presented in the body of this volume. The
two clusters capture variation in political conditions, policy design,
and their outcomes.
The first cluster of states are those that are somewhat more consis-
tent performers as welfare regimes, even if they do not approximate
coherent ‘welfare states’. These are political regimes in which either
broad-based political support through well-organized political parties
or consistent interventions by political leadership provide backing
for social policies that tend towards more universal and rule-bound
implementation. States within this cluster fall into three broad
categories. The first group of states includes those that have been clas-
sically described as ‘social democratic’ in thrust, with well-organized
left-of-centre regimes (such as Kerala and, historically, West Bengal).
The second group has been described as having ‘competitive popu-
list’ regimes (such as Andhra Pradesh and Tamil Nadu [see Harriss
2003; Subramanian 1999; Swamy 1998]). These are states in which
competing political parties project welfare policies to appeal to the
public without a significant degree of differentiation based on iden-
tity. They exhibit relatively strong policy continuity across changes in
government suggesting that there is deeper underlying social support
for redistributive social policy, accompanied by some trends towards
universal provision. The third set of states we label ‘incorporation-
ist’ regimes (such as Chhattisgarh or Odisha). These are states that
18 Politics of Welfare

coatinue to be dominated by a narrow socio-economic elite and in

which lower castes/classes are weakly organized in politics but that
nevertheless implement substantial social protection programmes for
the rural poor.
States in this cluster as a whole are more likely to favour broadening
eligibility for government programmes and moving beyond narrowly
targeted approaches. Preferences for more universal coverage reflect
the greater extent to which coalitions of political support exist—or
have been created—for social sector programmes which underpin
their longer term sustainability. The initiatives may sit alongside
moves by senior political leaders to limit the discretion that lower-
level politicians or officials can use in influencing access to certain
social programmes, or what Manor (2010) has described as ‘post-
clientelist’ approaches to service delivery. Some of these states have
also protected elements of the public provisioning of social services,
such as public hospitals under health insurance schemes in Kerala and
Tamil Nadu; or the ‘de-privatization’ of fair price shops and transpor-
tation networks in Chhattisgarh’s PDS. Such moves are also consistent
with more universal conceptions of—or goals for—welfare.
The second cluster of states are less consistent performers as welfare
regimes. These states tend to demonstrate shallower commitment to
redistributive social policy with patterns of political competition or
leadership that militate against longer term investment in social pro-
visioning. Public policy related to welfare plays a relatively marginal
role in electoral competition in these states. The states within this
cluster vary a great deal. Some are rich states with pro-business eco-
nomic policies that place them closer to what Rudra (2007) defined
as the ‘productive’ end of welfare strategies focusing on labour mar-
ket participation rather than social safety nets or redistributive public
policy. They may be more responsive to organized business interests,
and have more labour-repressive policies. Other states are among
the poorest in India, and demonstrate very different pathologies of
public administration and state capacity.
States within this second cluster can also be separated into three
broad groups. First of all, ‘pro-business’ states (such as Gujarat) privi-
lege economic growth and have a weak articulation of welfare policies.
This is accompanied by lower-than-expected social sector expenditure
and middling records in reducing poverty. A second group can be
 Introduction 19

labelled ‘competitive clientelist’ regimes (such as Uttar Pradésh and

Bihar). Parties in these states—since the politicization of lower castes
and breakdown of Congress party dominance—have competed to
build electoral coalitions among different sections of the popula-
tion, principally defined by caste identity. There is weak cross-regime
policy stability, and there tends to be a stronger commitment to the
maintenance of targeted rather than universal approaches to social
protection. Government programmes are more likely to be predomi-
nantly predicated upon clientelistic modes of service delivery in which
local bureaucrats and politicians are able to exercise strong discretion
over who can access goods/services. The final group comprises the
more extreme case of predatory states (such as Jharkhand) in which
political power is personalized and extremely localized, political parties
lack organization, and regimes lack a sense of public purpose. Public
office is closely associated with private gain, and there is weak enforce-
ment of the rule of law. These features can be particularly pronounced
in natural-resource-rich states. This last category of states also exhibit
strong versions of clientelism and the discretionary authority of local
political and bureaucratic agents, which influences the implementa-
tion of public policies on the ground. All the states in this cluster
are places where government programmes and centrally sponsored
schemes are implemented, but in which local political and political—
economic settlements frequently circumscribe their outcomes.
These clusters present a snapshot of attributes of different regimes.
In reality, many states do not fit squarely into one type or another.
States may also share characteristics of both the clusters. All states
see instances of the corruption of public office for private gain. Even
where states have adopted ‘post-clientelistic’ strategies in some policy
areas, they will see the perpetuation of patron—client relationships in
others. The clustering here is, however, intended to capture the extent
to which such features of political life overwhelm the capacity of the
state to deliver more purposive forms of social policy. States may also
shift in character over time. This has happened in the case of West
Bengal, where the nature of the political regime stultified in later years
of Left rule moving away from a ‘social democratic’ ideal type.
In the final part of this introduction, we set out some of the critical
mechanisms which produce interstate variation and chart important
differences in the design and implementation of palicies across states.
20 Politics of Welfare

But before that we set out an agenda for the comparative analysis of
state politics that contributions to this volume build on. We begin
by looking at the rationale for treating Indian states as discrete units
of analysis as a result of a pronounced shift in the democratic and
economic landscape from the 1980s.

The Significance of States as a Locus for Analysis

Until the 1980s, political scientists tended to view the politics of the
states as a residual category. The smooth functioning of what Rajni
Kothari (1964) defined as the ‘Congress system’ saw the Congress
party remain the dominant party at both the Centre and in the states
for much of the period from the first elections until the 1980s. The
euphoria of the nationalist project served to grant a degree of priority
to national-level elections and to the analysis of ‘all India phenomena
and categories. It was only in the decade of the 1980s and in the
context of the crisis of state following the short-lived period of emer-
gency rule from 1975-7 that the terrain of state politics became an
autonomous and ‘interesting-in-itself? point of enquiry for students
of Indian politics. Leading studies of this period (Frankel and Rao
1989; Kohli 1987; Wood 1984) deepened these enquiries as they
researched the nuances of regional politics to understand and deci-
pher national-level political puzzles.
It was however the decade of the 1990s that truly belonged to the
states in many senses. As Yadav and Palshikar (2008) write in their
seminal essay, during the past two decades, state politics has ‘broken
free of the logic of national politics and has acquired a rhythm and
logic of its own’. State politics developed an autonomous logic as the
political choices of people from different states varied, as also the
fortunes of the all-India political parties across space. The vote share
of national political parties consistently declined in parliamentary
elections held during 1985-2009. This trend continues despite the
success of the BJP in 2014. The combined vote share of the regional
parties (all parties excluding the four main national parties—
Congress, BJP, Communist Party of India, and Communist Party of
India [Marxist])—went up from 22 per cent in 1989 to more than 45
per cent in the elections of 2014. The necessity of coalitions ensured
that state-based parties could dictate terms at the national level.
 Introduction 21

Most importantly, the post-1990 period saw the crystallization of

regional identities and—in much of India—the consolidation of
distinct political communities within the geographical boundaries of
the states. All ghese processes led to the arrival of the state arena as
the primary, effective arena of Indian politics, even if the national
domain was not entirely eclipsed (Chhibber 2009; Kailash 201 1b).
The changing dynamics of state politics unfolded at three inter-
related levels. The first was that of party politics and electoral
competition and constituted the mainstream arena of democratic
politics. The second consisted of the social bases of power and the
resulting patterns of mobilizations. The third was the changing nature
of regional political economies, policy discourses, and reconfigura-
tion of interest groups in the wake of these changes. At all these levels
the twin possibilities of the opening and shrinking of democratic
options were evidently at play.
The steady rise of state-based parties since the late 1980s coincided
with the decline of the Congress and the arrival of a post-Congress
polity and therefore was appreciated by many as leading to fed-
eralization of the polity (Arora 2000, 2003; Verney 2002) and to
the deepening of democratic processes (Nigam 1996; Rangarajan
1999). The new state-based parties consolidated regional identities
(Jharkhand Mukti Morcha, Telugu Desam Party, Shiv Sena); upset
dominant equations of social power in the electoral realm (Bahujan
Samaj Party, Rashtriya Janata Dal, Samajwadi Party); and/or posed a
challenge to the Congress as some parties were formed by factional
leaders of the Congress in particular states (Trinamool Congress,
Nationalist Congress Party). In all these diverse avatars smaller par-
ties contributed to the regionalization of politics and complicated the
nature of the new party system.
The rise of state parties shared a close link with the ‘post-Mandal’
moment of Indian politics in which caste arrived as a dominant
idiom. This followed the implementation in 1990 of the recom-
mendations of the Mandal Commission for affirmative action for the
broad swathe of lower castes known as the Other Backward Classes
(OBCs). The sudden transformation of the OBC category from a
legal or administrative category to a robust mobilizational political
category symbolized a new era of caste politics and signified ‘India’s
silent revolution’ (Jaffrelot 2003). The OBC politics or the politics
22 Politics of Welfare

of the bahujan (‘majority’) was a prominent factor in the shaping

of what Yadav famously christened the ‘second democratic upsurge
(Yadav 1999, 2000). Post-Mandial caste politics ambitiously aimed
at establishing an ideological and political alliance of the OBCs and
Dalits to alter the frameworks of politics ar the national level. And
yet, since caste always operated in politics through a regional prism
(Palshikar 2003), OBC politics consolidated itself mainly at the
regional level. This gave way to the rise of several state-based parties
that overtly pushed an (anti)caste agenda, particularly in north-
ern India, such as Rashtriya Janata Dal in Bihar, Bahujan Samaj
Party and Samajwadi Party in UP, and the Bahujan Mahasangha in
Maharashtra. The rise of caste-based parties influenced the dynamics
of state politics in several important ways. It not only made politics
more competitive but also helped in broadening its participatory
base as new democratic claims on behalf of marginalized sections
of the population found new articulation. It also challenged the
established norms of representation as OBCs demanded a share in
political offices and as new elites were able to enter politics through
state-level caste-based parties. The entry of the new elites also forced
the all-India parties to undergo ‘obcization’ of their party organiza-
tion (Vora 2004).
The findings of the National Election Studies (NES) indicated that
the patterns of political participation changed significantly during
1990-9 to include voters from more diverse communities. During
this period, voters from marginalized groups participated in politics
to a greater degree compared to the past (Palshikar and Kumar 2004).
However, the same studies noted that the participatory norm was
arrested and moderately reversed during 1999-2009 phase (Kumar
2009, 2013). The noisy politics of caste and communal mobiliza-
tions had helped to partially conceal the rapid ascendance of the
middle class as a locus of political agency. If the earlier success of the
Hindu nationalist BJP in forming the central government in the late
1990s was partly explained as the result of ‘elite revolts’ against the
intrusion of plebeian interests in public life (Corbridge and Harriss
2000), the impressive victory of the BJP in the 2014 elections may
be seen as a major democratic articulation of middle-class agency in
Indian politics. These changing political scenarios at the state level
have critically influenced the environment for the implementation

NM
 Introduction 24

of, and formulation of, public policy—especially the social policies

that are the focus of this volume.
7
Major Themes in this Volume
Comparing the Politics of Policy Processes
The chapters that follow comprise a series of paired comparisons
between purposively selected states, in which an element of control
is exercised by holding certain features constant in order to explain
variation in a given policy field. ‘The states selected for comparison
encompass a representative sample across regions, levels of income, and
political traditions: Kerala and Tamil Nadu, Chhattisgarh and Madhya
Pradesh, Maharashtra and West Bengal, Rajasthan and Madhya
Pradesh, Jharkhand and Meghalaya, and Delhi and Chandigarh.
The chapters broadly address the tension identified at the out-
set of the introduction: how do states strike a balance between
the democratic compulsions of responding to popular demands
for improved welfare with pressures that arise from expectations
that state governments should also facilitate business development
through land acquisition, taxation policies, and other forms of sub-
sidy. But while identifying this as a central tension for Indian states
in an age of high growth, it should be clear by now that we do not
privilege reductionist political economy explanations for changes in
the policy environment. Rather, we maintain that political agency
can be critical in explaining when and why new paths are adopted
or not, and how different regimes respond to similar problems. The
first theme that our studies thus pay particular attention to is the
shape of longer term paths of historical development and why, in
certain circumstances, political leaders can institute breaks with ear-
lier trajectories. In this way a state such as Chhattisgarh, where upper
caste and class interests retain dominance within its political regime,
has overseen far-reaching social protection policies which have been
driven by its political leadership (Tillin, Saxena, and Sisodia, this
volume). Alternatively, longer term histories—even where progres-
sive in an earlier era—can help to reproduce exclusions that render
regimes relatively impermeable to newer constellations of marginal-
ization. Thus West Bengal with a history of trade union politics and
24 Politics of Welfare

a well-organized left-of-centre regime has evolved a policy regime

that today accords near invisibility to the demands of informal sector
labour (Deshpande, this volume).
The second major theme that runs throughout the chapters is
the pattern of electoral competition at the state level. There is, how-
ever, no unidirectional flow in this pattern. The pressure for welfare
commitments may be generated within the context of party competi-
tion or arise in a top-down fashion. In Tamil Nadu and Kerala, for
instance, competitive two-party or two-front politics has ratcheted up
welfare commitments over time, in regional contexts with a grow-
ing middle class. Broader political and policy cultures—that are
reasonably resilient across changes in government—help to explain
why state governments have adapted health insurance schemes to
ensure a degree of protection for public-sector hospitals (Kailash and
Rasaratnam, this volume). Whereas in Chhattisgarh, a chief minister
saw the opportunity to reach out more effectively to the rural poor
through a subsidized food programme (Tillin, Saxena, and Sisodia,
this volume). The pressure in Chhattisgarh did not emanate from
within organized party politics, although a by-election defeat raised
questions about the performance of the PDS. In each state, new social
policies may have an impact on electoral competition in the future
if subsequent challengers prefer to augment rather than dismantle
new policy regimes. Other chapters focus on the implementation of
centrally sponsored schemes within particular states; there may have
been little or no impetus from within those states for such policy ini-
tiatives. Such was the case in Jharkhand and Meghalaya, where Rajesh
Dev studies the implementation of MGNREGA. But the potentially
disruptive effects of MGNREGA on the local customary power struc-
tures in areas governed by the Fifth and Sixth Schedules may affect
electoral competition in the future. As Jenkins and Manor remind
us in their chapter, while politics shapes the design and operation
of social programmes, such programmes also ‘impinge on politics’,
including through their influence on the political capacities and con-
nections of the poor.
The third major theme that threads through the volume is the
nature of bureaucratic structures, agency, and internal politics.
The two studies on MGNREGA in the volume (Dev; Jenkins and
Manor) outline the power of block- and district-level bureaucracy
 Introduction 25

to micromanage the programme. This was evident in the quite

different settings of Jharkhand, Meghalaya, Madhya Pradesh,
and Rajasthan. Both chapters document the tensions that emerge
between bloclglevel bureaucrats and local councils, which are, in
theory, the primary agency for managing the local operation of the
scheme. But the capacity of local bureaucrats may also play a critical
role in determining the overall ‘success’ of the scheme in providing
employment, and providing some protection against local structures
of exclusion at the village level (Dev, this volume). Pressure from
higher levels of the political and/or administrative system helps in
some places to reinforce the extent to which ‘street level bureaucrats’
(Lipsky 2010) feel compelled to conform to the law and/or rules
governing administration of the programme. The absence of those
sanctions in other places can effectively facilitate the undermining of
a programme by local-level bureaucrats (see chapters by Dev; Jenkins
and Manor; Tillin, Saxena, and Sisodia). Yet in some states the
disciplining—of both block-level bureaucrats, as well as local poli-
ticians—is possible because of the marked centralization of power
within the chief minister's office. Such situations give rise to other
concerns about the checks and balances on the executive, as well as
the downward accountability of local bureaucrats and politicians to
their constituents. The design of programmes is also significant: some
have stronger in-built mechanisms for accountability and transpar-
ency than others.
Another recurrent line of analysis with regards to the bureaucracy are
the ambiguous lines of accountability and responsibility that emerge
from the involvement of multiple ministries in single policy areas,
or across different parts of the policy cycle. This comes out particu-
larly starkly in the chapter by Vaid on the procedures for monitoring
the Right to Education (RTE) Act. While the Ministry of Human
Resource Development is responsible for the RTE, the major bureau-
cracy selected for its implementation consists of that established for
the earlier Sarva Shiksha Abhiyan which functions as a ‘society’ with
autonomy from the line ministry. Further complicating the scenario,
the responsibility for monitoring the RTE lies with another ministry
altogether: the Ministry of Women and Child Development. The
multiple and overlapping lines of responsibility create weak chains of
accountability and ownership. Leadership (political or administrative)
26 Politics of Welfare

can be a critical variable in explaining why some states do better in

facilitating inter-agency cooperation than others.
The fourth theme is the relationship between state and non-state
actors in the formulation, implementation, and evaluation of poli-
cies. In states such as Rajasthan (Jenkins and Manor, this volume) and
Chhattisgarh (Tillin, Saxena, and Sisodia, this volume), productive
working relationships between CSOs and government actors have
played an important role in shaping MGNREGA and food security
policies, respectively. But Madhya Pradesh—the state for comparison
in both these studies—has been more selective in its openness to civil
society actors. This mirrors the situation in many states where state
governments have sought to keep CSOs at an arm’s length, in some
instances to actively suppress them, or elsewhere sought to selectively
appropriate their agendas while limiting effective involvement in the
policy arena (Deshpande, this volume).
Fifthly, the backdrop to all of the studies is the fact that state gov-
ernments are embedded in a series of external relationships within
a multilevel governance structure which play an important role, if
not a deterministic one, in shaping their policy—and fiscal—space.
This includes the central government, other state governments,
international organizations, as well as other foreign governments.
Thus the states are not simply independent units of analysis but
have a degree of interdependence on other levels of the federal—and
international—system (see also Snyder 2001a, p. 104).

Comparing the Policy Mix across States

The overall social policy mix also varies across states, with dif-
ferent types of emergent ‘welfare regime’ as states grapple with
similar problems in different ways; make different decisions about
the implementation of centrally designed programmes; and have
varying needs arising from the prevailing patterns of socio-economic
development. As Box 1.1 shows, our analyses present five major
lines of comparison in the approaches taken by state governments
to shape social policies or oversee the implementation of central
government programmes.
In the first place, the studies identify a difference between states
that have opted to move closer towards universalization, reversing the
 Introduction 27

Box 1.1 Comparing the Approach of States to Welfare Policy

* Universal versus targeted approaches

* State as funder versus provider of social services
* Ability to regulate the informal sector labour market
* Clientelism versus ‘post-clientelism’
* Measurement of outcomes and determinants of ‘success’

target towards ‘below poverty line’ population imposed on states for

certain schemes in the early years of economic liberalization. Many
developing countries moved towards targeted social policies in the
1990s in response to fiscal pressure and market-based administrative
reforms. But especially in contexts of high societal inequality, rather
than leading to more effective public expenditure, targeting has
frequently been associated with the effective exclusion of the most
marginalized groups from social provisioning altogether (see, for
instance, Sen and Himanshu 2011). Recognizing the shortcomings
of targeted approaches, some state governments have expanded access
to social programmes. For some this arises from an explicit concern
to improve the welfare of the poorest, while for other political lead-
ers it functions primarily as a means of building broader political
coalitions. A move towards universalism can help to mitigate against
exclusion errors that arise from the operation of local bureaucratic
or political discretion. It can also help to build stronger coalitions of
political support for social policies. As Echeverri-Gent (1993, p. 13)
observes, ‘[b]y reducing the discretion of local officials to determine
the beneficiaries of a program, it [universalism] limits their opportuni-
ties for corruption. At the same time, universal eligibility for benefits
makes programs more politically viable by creating broad coalitions
of support’. While some new programmes such as the MGNREGA
were universal by design, in other areas—such as the PDS or health
insurance—some state governments have expanded access above and
beyond the ‘below poverty line’ population subsidized by the central
government.
Secondly, we see states grappling with the role of the state as funder
versus provider of social services. In the health and education sectors,
28 Politics of Welfare

the de facto opting out of publicly provided services has been sup-
ported—for those who can afford it—by flourishing private markets.
Furthermore, New Public Management approaches from the 1990s
helped to spread responsibility for social assistance among mul-
tiple agencies: public, private, and third sector (see also Jayal 2013,
pp. 176-7). But, as Kailash and Rasaratharh show in this book, the
introduction of state-funded health insurance has presented states
with a choice about how to manage a market for healthcare con-
taining both public and private providers. Some states have used the
advent of health insurance to protect a role for the state as provider,
as well as funder, of tertiary healthcare (see also Kannan and Breman
2013a). We also see states making a variety of decisions related to the
involvement of private sector or non-state actors in the delivery of
other social programmes.
Thirdly, we identify the differential degrees to which states have
sought or been able to intervene to provide social security within
a labour market dominated by the informal economy in which
jobs are insecure and not underpinned by wage rates or conditions
of employment available to formal sector workers. This is a crucial
element of any welfare regime analysis since, as discussed, classic
theoretical models of the welfare state are premised upon the ability
of the state to intervene to ameliorate insecurities caused by the mar-
ket. But in an economy dominated by informalization, and where
much employment is small-scale and home-based, the challenges of
regulating employment conditions and ensuring social security are
substantial (Deshpande, this volume; Kannan and Breman 2013b;
UNRISD 2010, p. 44). Four of the chapters focus on schemes intro-
duced specifically to improve social security in the informal sector:
through a rural public works programme (MGNREGA), health
insurance schemes, and social security for domestic workers.
Fourthly, states reflect different degrees of reliance on clientelistic
modes of political organization in which the delivery of social services
is contingent upon locally mediated political relationships. Some
schemes are, as Jenkins and Manor (this volume) suggest, explicitly
designed to be ‘post-clientelist’ in nature such that by design they
limit the possibility for local patrons to derive spoils from them.
Others, such as a classic subsidy scheme like PDS, remain vehicles for
patronage in many states. However, with clear political management,
 Introduction 29

they can be reformed in ways that make it more difficult for local
political and bureaucratic actors to exercise discretion over access to
and quality of services (Tillin, Saxena, and Sisodia, this volume).
Lastly, the stgtes present differences in terms of how the monitor-
ing and evaluation of programmes works or does not (see especially
Vaid, this volume). The latter is critical because it links to the over-
arching question of how we should understand the relative success
of government programmes designed to enhance well-being and the
types of emergent welfare regimes we identify in this volume.

Overview of the Chapters

Each of the chapters presents a comparative study of the politics that
produce divergent policy choices across states. The first chapter, by
Kailash and Rasaratnam, addresses the implementation of publically
funded health insurance in Tamil Nadu and Kerala. Health insurance
has been promoted by international and national policymakers as
a means of expanding access to healthcare by utilizing the growing
capacity of the private sector. However, the authors show that in
both Tamil Nadu and Kerala, policymakers deliberately introduced
modifications to the health insurance schemes that facilitated the
flow of insurance funds to the public sector at the expense of the
private sector. The chapter links these unusual decisions to the ‘policy
frames’ shared by politicians, bureaucrats, and academics alike, that
have emerged from long-term patterns of political mobilization and
institutional development despite the different health economies of
the two states.
Investigations into the historical legacies of state politics lead
to another kind of enquiry in the second chapter by Rajeshwari
Deshpande that focuses on social security for domestic workers in
Maharashtra and West Bengal. Domestic workers gained a sudden
visibility when the central government enacted the Unorganized
Workers’ Social Security Act in 2008. Deshpande contrasts the
enthusiasm of policymakers in Maharashtra for reaching out to
domestic workers to the absence of any such initiatives in the state
of West Bengal, despite the state’s official ideological commitments
to labour welfare. She argues that diverse patterns of the formation
and consolidation of the middle class in the two states, as well as
30 Politics of Welfare

the configuration of political competition and political culture, have

shaped the politics of policies for informal sector workers. While a
competitive politics of patronage emerged in the context of party
competition in Maharashtra that offered greater visibility if not sub-
stantive inclusion to domestic workers, in West Bengal there was very
little attempt to mobilize domestic workers at all.
If the first two chapters mainly focus on issues of social security,
the third chapter by Tillin, Saxena, and Sisodia addresses the reform
of a classic subsidy programme, the PDS, in the adjoining states of
Chhattisgarh and Madhya Pradesh which were bifurcated in 2000.
The chapter focuses on the different approaches taken by the states
to reform the PDS in the decade following the Supreme Court inter-
ventions in the ‘right to food’ case which placed certain obligations
upon states. Madhya Pradesh has sought to improve the accuracy of
targeting via the PDS, and outsourced the development of biometric
ration cards to a private company. Yet the programme remains beset
with ‘leakages’ and exclusions that prevent its effective contribution
to food security. Chhattisgarh by contrast ‘de-privatized’ ration shops
and placed them in the hands of local elected councils, self-help
groups, and cooperatives; expanded access to the PDS to make it a
‘quasi-universal’ programme; made the reliable provision of heavily
subsidized foodgrains an executive political priority; and took steps
to improve the delivery network. The chapter highlights the role
of political leadership and its relationship with the bureaucracy, to
explain the different outcomes in the two states.
The next chapter by Rajesh Dev compares the local dynamics of
MGNREGAs implementation in the two ostensibly ‘tribal’ states
of Meghalaya and Jharkhand. Dev compares the implementation
of ‘Narega’ (using the colloquial expression) in the Fifth Schedule
areas of Jharkhand where elected panchayati raj institutions exist to
the situation in Sixth Schedule areas of Meghalaya where structures
to implement the scheme have been grafted on to customary gov-
ernance institutions. He refers to the multiple entanglements that
policy initiatives like MGNREGA introduce between traditional
authorities, NGOs, insurgent groups, claims-making citizens, and
the State. Increasing contestation between customary leaders and
newly enfranchised groups has left scope for greater bureaucratic
intervention in the functioning of the programme.
 Introduction 31

The chapter by Jenkins and Manor offers a different reading

of the implementation of MGNREGA in the regional settings
of Rajasthan and Madhya Pradesh. They argue that the agency of
political leadergin both these states has been important in insulat-
ing the programme from clientelistic interference by local political
actors in most—if not all—districts of the states. They also argue
that CSOs were able to influence the shaping and implementation
of the programme in significant ways in both states, particularly in
Rajasthan from where civil society activists also played a leading
role in the crafting of the national-level legislation. The chapter
underlines the political implications of MGNREGA in these states,
and others, as a programme which has expanded the political
capacities of the rural poor.
The aspect of monitoring and continuous assessment of policies is
often neglected in policy studies. The last chapter in this book focuses
on this question through a comparative study of educational policy.
Vaid scrutinizes the operation of monitoring mechanisms under the
RTE Act in the two urban centres of Delhi and Chandigarh. She
highlights the obstacles to effective monitoring created by the over-
lapping involvement of multiple agencies, as well as the inconsistent
support of bureaucratic leadership—a problem that has been more
acute in Delhi than Chandigarh.
The comparative study of public policy at the state level is an area
where more research is needed. With states at the centre of the politi-
cal and economic life in India, it is likely that states will continue to
demand and exert greater policy autonomy. This may unleash a larger
variety of policy models and experimentation in the field of social
policy. Changing approaches by the national government will also
graft on to longer histories of policy development at the state level.
The chapters in this volume offer an exploration of how local political
environments shape the context for policy implementation; a critical
endeavour in order to appreciate the opportunities and frustrations
inherent within India’s emergent welfare regime.

References

Asian Development Bank. 2013. The Social Protection Index: Assessing Results
for Asia and the Pacific. Manila: Asian Development Bank.