In this fine book, Anoop Sadanandan argues that “democracy deepens when central leaders share power with elected local politicians to use their intimate knowledge of voters to mobilize political support” (p. 7). He adds rigor and substance to a familiar formulation: that decentralization not only deepens democracy but also improves service delivery and accountability, and bolsters popular confidence in local governments and the democratic process. He also shows that where decentralization is weak or absent, development suffers, breeding mistrust in democratic institutions.
In an inspired opening passage, Sadanandan vividly illustrates this contrast by comparing local politics in two contiguous villages on opposite sides of the line dividing two Indian states: West Bengal (with strong decentralization) and Bihar (where it is weak). He then draws upon an admirable range of sources and his solid field research to develop a fresh, challenging explanation of why some high-level leaders decentralize substantially while others do little or nothing. He argues that in Indian regions that have seen positive socioeconomic changes, people tend to decide for themselves how to vote, free from influence by traditional intermediaries. By contrast, in less developed regions where people are more influenced by intermediaries who often play upon ethnic (or, more properly, caste, religious, or regional) identities, democracy is shallower and citizens are more influenced by those intermediaries, so that in his view they tend to vote en bloc.
Upward information flows loom large in this analysis. Since senior leaders in the first set of regions cannot rely on traditional intermediaries for information, they tend to compensate by devolving power onto elected local councillors who can provide it. Leaders in regions where intermediaries are more influential and supply information on local trends tend to avoid it. Sadanandan also extends his analysis beyond India by discussing decentralization elsewhere.
There is substance in these findings, and analysts of decentralization—in and beyond India—must now consider them carefully when they assess the benefits of decentralization and especially the conditions that trigger senior leaders’ decisions for or against it. Sadanandan carefully notes (p. 148) that his argument is not the only reason for democratic deepening in every case. But amid his understandable delight with his new insights, he fails to give enough attention to their limitations. Consider four such points, the first two of which have to do with his view that Indians in less developed areas tend to vote as a bloc.
First, John Huber and Pavithra Suryanarayan (“Ethnic Inequality and the Ethnification of Political Parties: Evidence from India,” World Politics, 68(1), 2016) have found that voters tend to coalesce behind parties that stress ethnic identities when there are high levels of economic inequality. Their argument at times supports Sadanandan’s, but inequality can—and often does—increase amid high levels of development.
Second, several studies more directly challenge the notion that Indians tend to vote as a bloc. Oliver Heath, Gilles Verniers, and Sanjay Kumar (“Do Muslim Voters Prefer Muslim Candidates? Co-Religiosity and Political Support in India,” Electoral Studies, 38(1), 2015) show that Muslims are more likely to vote for Muslim candidates only when the latter have realistic chances of winning. They often fail to vote as a bloc. Similar findings emerge from studies by Yogendra Yadav (“Five Myths about the Muslim Vote,” The Hindu, 20 April 2009), Mohd Sanjeer Alam (“Whither Muslim Politics?” Economic and Political Weekly, 26 September 2009), M. Shakir (“Congress and the Minority Vote,” in Ramashray Roy and Richard Sisson, eds., Diversity and Dominance in Indian Politics, vol.2, 1990), and Irfan Engineer (“Politics of Muslim Vote Bank,” Economic and Political Weekly, 28 January 1995). The belief in bloc voting—by Muslims who have been extensively analyzed, but also by other social groups–is highly dubious. Pradeep Chhibber, Harsh Shah, and Rahul Verma (“Who Swings the Vote in India?” The Hindu, 5 December 2017)—and, in a separate analysis, Tariq Thachil and Adam Auerbach (“How Clients Select Brokers: Competition and Choice in India’s Slums,” American Political Science Review, forthcoming)—reinforce the point by showing that intermediaries have very little ability to trigger bloc voting.
A third limitation is evident from material that Sadanandan presents. For example, a map of India (p. 69) accurately ranks states by levels of socioeconomic security or vulnerability. But it raises doubts about the influence of these things on decisions by state governments about decentralization. Kerala and Tamil Nadu both have high levels of economic security, but leaders in the former have generously decentralized, while both main parties in the latter resist that process. Two states (Kerala and West Bengal) are marked by high socioeconomic security, while a third (Madhya Pradesh) ranks low. But all three have strong decentralized systems—an outcome that undermines the author’s basic case.
Finally, while as Sadanandan convincingly argues, socioeconomic conditions influence decisions about decentralization, they are often trumped by something else: politics. He acknowledges its importance, but it has greater autonomy from socioeconomic forces than he suggests. There are several strands to this argument. The interplay of political forces within governments always affects decisions about decentralization. In nearly every case, certain powerful politicians and bureaucrats—and, almost universally, legislators—resist the devolution of powers and resources that they see as rightfully theirs onto elected bodies at lower levels. This may thwart or limit decentralization—and this has little to do with socioeconomic conditions. International actors sometimes apply leverage in different ways. For example, the World Bank first colluded with the Zambian government to enable it to avoid devolution, but later pressed Cambodian leaders to undertake it.
More important have been senior leaders’ ideas, calculations, and imaginings that have little connection to socioeconomic circumstances and information flows. The Malawian government (among others) was inordinately fearful that decentralization would create chaos and threaten national unity, and so it abandoned the idea. The African National Congress in South Africa was so distrustful of locals that it imposed impossible technocratic checks and tasks on elected local councils, crippling their new system at birth. Dictators (like Hossain Mohammed Ershad in Bangladesh) who badly need grassroots support—which is more important to them than information from below—have cultivated elected local councillors by giving them substantial powers. They use local democracy as a substitute for the same thing at the national level.
Enlightened leaders’ ideas also matter. Successive Brazilian governments since 1995, as well as governments in Karnataka after 1985 and in Madhya Pradesh after 1993, all devolved generously in order to catalyze participation from below that would deepen democracy, produce better development outcomes, and enhance their own popularity. These decisions occurred in places with diverse socioeconomic circumstances. So while Sadanandan provides us with a valuable argument that must be carefully weighed by analysts of decentralization, its limitations must be borne in mind.
