Development in Multiple Dimensions: Social Power and Regional Policy in India

One of the most serious obstacles to economic growth and human development in poor countries is that their governments often refuse to provide basic public services such as education, health care, roads and electricity. Many observers have suggested that such failures are a consequence of the social inequalities prevalent in many poor countries, which create powerful elites who manipulate state institutions for their own benefit. However, such a conclusion ignores a crucial aspect of the societies of most poor countries: that social inequalities are not single, but multiple, and that a single region may have several competing elites, with sharply different policy preferences and sources of political power.

This book will explore the complicated, chaotic, distributional politics that result from multiple patterns of reinforcing, and non-reinforcing, social inequality. It will show why some elite groups are excluded from political power, while others wield disproportionate influence. On the outcome side, it rejects simplistic notions of “high” or “low” performing states, and the related idea that development policy has no programatic component. On the contrary, there are important compositional differences in the types of public services governments supply, with some favoring infrastructure over social services and others taking the opposite approach.

This theory, and the empirics that test it, are drawn from India and the Indian political experience. It examines one of the great problems of South Asian studies: Why the states of India, despite virtually identical formal institutions, differ so dramatically both in the levels and in the types of services provided by their governments? They show that different states have pursued very different paths to the ideal of development, and that these paths are closely associated with the local distribution of social power.

The book makes two contributions to the study of the political economy of development in India in general, and South Asia in particular. The first of these is a reconceptualization of variation in government performance. This book argues that public goods distribution in India varies in two dimensions, corresponding to social and infrastructure goods, and, ultimately, economic and human development. Some states, like Gujarat, produced excellent roads and power plants but poor schools and hospitals, while other states, like West Bengal, have social services that are much better than their infrastructure. Such a distinction allows a richer conversation about what politics in poor countries looks like, adding a greater awareness of policy disagreements to a literature focused on unidimensional “performance.”

The second contribution is a theory of the cause of these differences, based on the importance of multiple sources of social inequality. Indian states are dominated by powerful social groups, which are overrepresented in the political system and able to exercise disproportionate influence on policy. The power of these social groups comes a variety of sources, most importantly land (and the clientelistic ties that come with it), urban education, and simple group numbers (in democratic settings). The distribution of these sources of power relative to one another determines the political influence of each group, and in turn the types of resource transfers that politicians will find attractive. Areas dominated by large landed groups, such as Gujarat, Maharashtra, and Punjab, will favor the provision of infrastructure goods, while areas dominated by large educated groups, such as West Bengal and Kerala, will favor the provision of infrastructure goods. Areas dominated by small, educated landed groups, such as Bihar and Uttar Pradesh, will provide few of either kinds of public goods, while areas dominated by groups combining all three sources of power, such as Himachal Pradesh, will provide both types of goods.

Sample Chapter

From Hierarchy to Ethnicity: Constructing Caste in Colonial India

In much of the colonial world, the decades before independence saw a rapid increase in the political importance of ascriptive identities among nascent political elites, with groups large and small forming organizations, petitioning government bodies and distributing propaganda. Even more interesting than the general rise in ethnic or caste consciousness was its uneven distribution across groups, with many individuals disdaining narrow ethnic appeals in favor of the broader rhetoric of imperial loyalty or incipient nationalism. This rise was especially notable in India, where the last years of colonial rule saw an upsurge in the salience of caste activism and caste-organizational activity that shaped political patterns both at the time and after independence.

At the same time caste identities were becoming more salient, they were changing in nature. While most scholars today think of ethnic groups as “conceptually autonomous” categories, there are (both in South Asia and elsewhere) many cases of groups that relied on external legitimation and emphasized their similarities to high status groups over their own distinctive characteristics—where upwardly mobile members of poor groups sought to assimilate into rich ones rather than challenge them. Such “ranked” identities were common in many parts of the world before the industrial revolution. In India, where ranking was very noticable during the colonial era, the gradual evolution of a very different “ethnified” view of identity was one of the key events of the 20th century.

This book will examine the causes of the upsurge in colonial caste activism in India, and the strategic choices made by caste activists, in particular the leaders of upwardly mobile poor groups who had to chose what role caste, and the ranking norms with which it was associated, should play in their political careers. This resolves itself naturally into two questions. Firstly, why do some identities become the focus for elite activism?. Secondly, why do some activists participate in maintaining existing ranked identity systems by rejecting opportunities to create a conceptually independent identity of their own?

The empirical evidence shows that, overall, group size (the variable most often mentioned in the existing literature) is less important than the socio-economic status of castes as a predictor of ethnic rhetoric. While educated groups tend to mobilize more than the poor, the most educated groups hardly mention their identity at all. Among groups that do mobilize, the embrace of social ranking is conditioned on the political context. Ranked rhetoric is useful in building patron-client networks, and is thus common in areas with patrimonial institutions, while unranked rhetoric is useful in building large blocs of supporters, and is thus more useful in areas with participatory political institutions. The theory implies that as education levels rise and elections became more common, we should see both an increase in ethnic mobilization and a decline in the degree to which it used ranked rhetoric. Overall, it suggests that the “modern” concept of an ethnic group as an unranked identity is a product of specific, historically determined, institutional circumstances.

The empirical base of this project is a large panel dataset of the petitions filed by caste groups with the Indian census authorities, a type of engagement widely discussed in the secondary literature on India. The specific goal of these petitions was to change the way in which the census referred to the group—a goal in which they were almost invariably disappointed. However, petitions provide a window into the complex processes of identity formation that are usually hidden from the historical record. In particular, they represent an index of the presence of an activist group and the rhetoric of that group, one of the first attempts to measure this key process at the elite level. The measures of ranked rhetoric, indeed, are the first attempt to measure this concept at all. The panel structure of the data enables comparisons of petitioning behavior within groups or categories of groups, a crucial factor given the many plausible cultural and historical differences between groups.

Sample Chapter