This is a landmark study, beyond its immediate subject matter, which is India and the role of subnationalism in social policy in India. The book is likely to become a classic in social welfare research. India is infamous for utter destitution among large sections of its population, despite high rates of economic growth and areas of high-tech industries. Social policy, in ideas, law and practice, has been poorly developed, compared, for example, to other large countries like China or Brazil. While political rights are well established in this large democracy, social rights have mostly taken the backseat. How come? And why have some states in India developed much more successfully in social terms than other states?

This is also a problem for theory building. Since socio-economic, political and cultural conditions in most countries of the Global South differ from the conditions in North welfare states, the conventional theories of the welfare state do not easily travel to the Global South. How to analyse and explain “social policy in development contexts’? Prerna Singh, a scholar from Indian and US backgrounds, makes a major contribution to this field of research, which has been building up since the 2000s.

The idea of the book revolves around the notion of ‘subnationalism’, which means the identification by elites and possibly by the population at large with the state, envisaged as centre of political self-determination. Subnationalism implies a notion of ‘we-ness’, but not or not only in a cultural or even racial sense, but relating to the polity as embodiment of the common weal. The main thesis is that the rise of subnationalism in some Indian states has been conducive to the spread of social policies and improvements in social development.

In view of their size and their legislative competencies, Indian states are close to nation states. In fact, much of the argument of the book can be meaningfully applied to conditions of social policy in nation states. The link between nation building and the rise of social policy is familiar from the history of Northern welfare states. Examples include the introduction of social insurance by Chancellor Bismarck in the newly founded Imperial Germany in the 1880s, and periods of the renewal of the national spirit under the New Deal in the USA in the 1930s and in the foundation of the British post-war welfare state in 1948.

Singh's argument challenges three common theories found in the literature. First, the ‘diversity-development deficit’ thesis posits that high degrees of diversity (regarding religion, language or caste) impede development. Singh argues that this is not necessarily the case. Under certain conditions – linguistic homogeneity, a popular subnational movement, the absence of a separatist movement and the presence of a subnationalist party – subnationalism may emerge also in the face of diversity. Only entrenched cleavages and conflicts impede social development. Kerala is the key example: a state which historically had one of the worst varieties of the caste system and still is an extremely diverse state, that has also become a leader regarding subnational identification and social development. Kerala is the most pronounced example, but the quantitative comparative analysis shows similar, if weaker, effects for other states. Singh finds the diversity-development deficit thesis flawed because these studies on collective identities mostly refer to local social services rather than state services which matter for social development.

Second, Singh's argument challenges the role of class as emphasised in political economy approaches in welfare state research (Power Resource approach). The elites that promote subnational identification do not mobilize along lines of class, even if prima facie this seems to be the case. Against the dominant view of the role of the communist party in Kerala, Singh argues that the communist party was not so much a ‘left’ or working-class party, but acquired a broader support by pursuing a subnationalist strategy and promoting development for all sections of the population. By contrast, the communist party in West Bengal had a different character and did not succeed in the way the Kerala communist party did.

Third, by emphasising governmental policies in view of the common good, flowing from subnationalism, Singh also qualifies a narrow emphasis on self-interest in welfare politics. Interests matter, but in the case of subnationalism, fuelling subnational solidarity served the interests of the new challenger elites to attain power. In this way, interests were linked or even refined by ideas, namely by the idea of an overarching social responsibility of the state for all citizens.

The book starts with an introductory chapter which summarises the argument. The second chapter explains the subnationalism-social development nexus in theoretical terms. Chapter 3 goes back in history, sketching the rise of challenger elites in four Indian states. Chapters 4 and 5 present the qualitative case studies of the (two) successful and (three) unsuccessful Indian states. Chapter 6 operationalises the key variables and presents a quantitative multivariate analysis across India with more cases. The final chapter discusses the findings and puts them in a wider context, including lessons for policy making. This study is extremely ambitious and encompassing, covering a vast ground: the book spans 200 years; it is comparative, covering 5 states (qualitative analysis) and 15 states (quantitative analysis); it combines qualitative and quantitative research in a ‘nested’ way; it covers two areas of social welfare, health and education; and it combines theories and empirical analyses, through a most comprehensive and in-depth discussion of various strands of the literature that matter for the argument. Regarding the comparison, India provides a good testing ground, since the political institutions are the same in every state, and this turns the explanation of differences into a puzzle. The data includes quantitative data on socioeconomic and institutional development, a broad variety of historical documents and a comprehensive reception of secondary literature on the states analysed. The book is fascinating reading also for non-specialists, and it is written in a crystal-clear way.

All in all, Singh makes a convincing case. A study which covers 200 years of history and a vast territory with a theoretical focus on one key driver (subnationalism and its repercussion on state social policy) is liable to criticism by area specialists. But the author goes to some length to make her research design very clear and to discuss in detail alternatives regarding measurement, theories or statistical modelling. I see three points that call for a discussion.

First, the definition of social policy and social development: Singh defines these concepts in a deductive way, by choosing health and education as key areas, rather than reconstructing the actors’ conceptions of social policy. A more constructivist approach (which is used in other respects in the book) would have allowed reconstruction of the changing concepts of social policy and social development held by relevant actors. An analysis of changing concepts would have deepened the analysis, to trace the rise of social policy not only in terms of more or less social activity, but also in terms of varieties of social policy. Considering that five states are analysed over a period of 200 years, the effort of such a discursive analysis would have been tremendous, though. The choice of health is well-taken, although the other key area of social policy, income transfers, is left out. The other area chosen by the author, education, obviously matters, but serving both social and economic ends, the rise of education is less unlikely than the rise of genuine social policies.

Second, the measurement of social policy: Singh uses social spending and outcome indicators (literacy, child mortality) as quantitative measures, and exemplary references to statutes, political initiatives and symbols as qualitative measures of social development. Both social spending and outcome indicators do not measure the institutional design and the quality of welfare programmes. While spending only measures the magnitude of welfare efforts, outcomes (what Singh calls ‘social development’) are only loosely linked to welfare programmes. To measure the quality of the social concerns of a state and the legal commitments to promoting welfare for all citizens, social rights would be a more suitable measure. But Esping-Andersen's (1990) measure of social rights, decommodification, is not applicable to development contexts since there is little commodification in developing countries (moreover, Esping-Andersen's measure only refers to income transfers). Data for other measures are difficult to come by. Coverage of welfare programmes is an indicator which captures an important aspect of social rights (extension to all citizens), and data is available for man countries. But Singh argues that enrolment figures in education in India are grossly misleading and useless. Regarding health, meaningful measures of real coverage are difficult to construct.

Third, regarding the core of Singh's argument, the subnationalism-social policy nexus: this nexus – the argument that subnationalism tends to entail universalist (rather than particularist) social policies (‘progressive social policy’) – remains surprisingly underspecified. One could think of countries in which the elites and the population at large strongly identify with their nation state, but do not push for more far-reaching ‘social’ engagement by the state. India as a whole, for example, has a strong identity as a democracy, but social issues were not prioritised until the mid-1980s, as Singh herself asserts. The ‘national question’ has been on the agenda ever since India's independence in 1947, but it has eclipsed rather than promoted social concerns. Similarly, the USA have a strong national identity, and this includes the idea of individual and social progress, but not primarily in terms of welfare programmes. Such counter examples to the (sub)nationalism-social policy nexus seemingly do not figure among the Indian states analysed by Singh. To model the nexus in a more differentiated way, one may have to examine and compare more closely national concepts and traditions of the state and of the state-society relationship. Singh's key term ‘solidarity’ is located at the interface of subnationalism and social commitments, but the term remains vague throughout (and is not included in the index) and rather obscures the relationship between subnationalism and social commitments. Singh calls upon Adam Smith's Theory of Moral Sentiments, which posits sympathy for others and mutuality as bases of social relationships, and which could indeed be a starting point for a closer investigation of the meaning and the potential of solidarity.

Irrespective of criticism that might be raised, the study makes a significant contribution to the theory of social policy, in three respects.

First, a contribution to research on social policy in development contexts. The study confirms that conventional explanatory models of welfare state development have to be modified for the Global South, and Singh points out new directions. Class analysis is limited in the face of the huge informal sector of the economy and large rural sections of the population. The capital-labour divide is not as dominant as in the North. Singh mentions, for example, that poverty, which is often rural and signals absent or jobless de-agrarianization, is positively correlated with the rise of social policy, contradicting the conventional logic of industrialisation which operated in the history of Northern societies. Moreover, the great 19^th century European ideologies emphasised by Esping-Andersen – social democracy, conservatism, liberalism – are absent in the Global South or they are blended with local thinking and take on a different meaning, as demonstrated by Singh in her analysis of the communist party in Kerala. At the same time, Singh draws attention to a factor which operates both in North and South, (sub)nationalist consciousness and collective identification. Singh's study can be taken as a contribution to a broader, generalised theory of the rise of state social policy pertaining worldwide (a one world theory).

Second, a contribution to the role of the state: the study can be read as a vindication of the state. The very term ‘state’, rather than ‘government’, indicates the emphasis on the state as an enduring political-administrative structure and sociocultural entity. Singh convincingly argues that welfare production in a mass society requires the state as key actor, even considering the scope of informal welfare, the weakness of states and the extensive role of local movements and associations (like SEWA in India) emphasised by students of development. Even welfare provided by voluntary associations and private providers relies on support by the state, through funding, legal regulation, promotion and creation in the first place. Some might see this perspective as euro-centric. Strangely, even in Northern welfare state theory, the state itself has not received much attention (Kaufmann, 2013). Esping-Andersen (1990) models the output of welfare states (social rights, decommodification, stratification) and the conditions of social politics (political parties, ideologies, power resources), but leaves open what the welfare state ‘is’. Few authors pursue the latter question. These authors emphasise the sociocultural foundations of the state, conceived as political self-determination of a citizenry. This leads to the third contribution of the book.

Third, a contribution to a sociocultural theory of social policy. Descriptive and political economy approaches to social policy refer to the sociocultural or ideational foundations of the state only in very broad terms, mostly by way of the three great ideologies, social democracy, conservatism and liberalism. There is a burgeoning literature on the role of ideas in social policy, but it mostly applies to policy analysis rather than to research on the welfare state as a whole and welfare state development. Singh's use of ideas has an affinity to historical institutionalism, but even more to the (few) socio-cultural approaches to welfare state analysis (e.g. by Robert Pinker or Franz-Xaver Kaufmann), which emphasise awareness, ideas and attitudes by elites and the population at large towards the state as provider of the common weal. Subnationalist identification is not only one aspect of social policy among many, but signifies the crucial role of socio-cultural conditions that have to be met for modern state social policy to emerge.

What can we learn from Singh's great study? First of all, Singh shows that welfare state building is a demanding historical process, that it relies on sociocultural conditions, namely the emergence of a (sub)national feeling of belonging and related solidarity. These sociocultural conditions of state welfare are not met in many countries of the Global South, and they are also absent or weak in some countries of the Global North, especially in countries like the USA that rely on markets and civil society and in clientelist societies like Greece. Second, Singh herself draws the important conclusion that the locus of public responsibility for welfare provision ought to coincide with the unit with which people most strongly identify, and that, to promote welfare and social development, identification with this unit ought to be strengthened. These conclusions can be transferred from subnational to national contexts. The question, for example, what role supra-national units like the EU or even global norms and institutions ought to play in social policy, has to be thought through in this light. Singh's book leaves us not only with lessons for theory building, but also with a ‘ray of hope’, because she shows that even deep-seated social cleavages and particularistic social structures and legacies can be overcome, even in India. Social development can be advanced, when the elites and the citizenry succeed in creating a sense of we-ness and a responsibility for the welfare of all citizens.