Elite Agency and Governance in Changing Societies: India in Comparative Perspective

Abstract

The article uses a rational choice, neo-institutional model to analyse cross-region and over-time variation in governance in India. Governance is measured through qualitative indicators (perception of law and order, measured by a national opinion survey) and quantitative indicators (murders and riots per million of inhabitants). The article raises both specific and general questions. Why did India succeed in achieving a high level of governance compared to the majority of post-colonial states? Why has governance declined in some parts of India during the past decades? Which policies and administrative and legal structures promote governance? How do India's new social elites—many of whom have come up through elections and policies of positive discrimination—transform rebels into stakeholders and agents of law and order? The article concludes that governance remains high when decision-making elites respond to challenges to order through policies that combine sanctions with strategic reform and the accommodation of identity.

Keywords:

• Governance
• India
• Elite Agency
• Neo-institutionalism
• Rational Choice
• Identity

Acknowledgements

An earlier version of this article was presented at the panel on Political Sociology, IPSA, Fukuoka, 2006. The author would like to thank Karsten Frey, Malte Pehl, Clemens Spiess and anonymous referees for their comments on the preliminary draft, Ivo Bielitz for his technical assistance, and the CSDS (Delhi) for providing access to the national public opinion survey (NES, 1996).

Notes

1. The detailed discussion of the methodology of the survey can be found in Mitra and Singh (1999). For the logical basis of the arguments and supplementary data from analytic narratives of governance in six Indian regions, see Mitra (2005).

2. The article is empirically located in post-Independence India where the actual research has been conducted through detailed investigations in six Indian States, in-depth interviews with 150 regional elites, a survey of the Indian electorate and examination of aggregate data and key public documents.

3. See Mitra (1999) for further illustrations of this model.

4. The literature on rules and neo-institutionalism is vast but the two key references that are crucial to this point are North (1990) and Baldwin (1995).

5. These are not the only reasons for breaking rules. When the existing rule and the government upholding it are seen as ‘evil’ or, symbolically, as a moral affront to the identity of the actor (e.g. Gandhi's satyagraha against British rule, Sikh opposition to Indira Gandhi after the army action against the Golden Temple in Amritsar, Punjab in 1984), rule-breaking might be perceived by actors as a ‘good in itself’, in the manner of Gandhi's satyagraha, the courting of ‘mass arrests’ by Sikh militants in Punjab at the height of the campaign for Khalistan—a separate territorial state for the Sikhs in Punjab—and the familiar rhetoric of Islamic terrorists who assert Jihad as a ‘goal in itself’.

6. I am grateful to Karsten Frey (personal communication) for pointing this out to me. Thanks are also due to the anonymous referee (communication via editor, AJPS) for additional references—Bender (1993), Bender and Swistak (1997), Scholz (1984a, 1984b), Ayres and Braithwaite (1992, Chapter 3) and Miller (2004) where the implications of the PD game for governance are worked out in greater detail.

7. See Fearon and Laitin (1996) for self-policing, and Varshney (2001, 2002) for inter-ethnic relations as a method of peer pressure leading to governance.

8. Rousseau's concept of community can be seen as a solution to the anarchy that potentially prevails in all interpersonal situations. Axelrod has theorised it in terms of some specific empirical conditions. See Axelrod (1984).

9. In the voluminous literature that has grown around governance, one notices more a pendulum-like mood swing between optimism and pessimism with regard to the state of governance, reflecting the political reality of the day, than the cumulating of analytical rigour and methodological precision. The tendency in some recent studies is to lean towards a pessimistic prognosis. In contrast to his earlier prognosis, Kohli's recent The Success of India's Democracy (2001) endorses the achievements of India's institutional arrangements.

10. In consequence, Indian developments are not adequately explained by models that consider democratic governance as an outcome of class conflict where a victorious bourgeoisie, drawing its political momentum from the industrial revolution, establishes a shared interest in terms of rules of transaction that respect the rights to property and participation (Moore, 1966; North, 1990).

11. Around the ‘core’ explanatory variables are also others such as harmony or ‘vote has effect’ which emerge as not significant. This is not to say that they do not play a role, for in fact they do, when their effects are measure in multiple correlations. However, in a regression model, when several explanatory variables are correlated amongst themselves, the multicollinearity wipes out the contribution of all but one in the results with regard to regression coefficients.

12. See Mitra (2005) for detailed breakdown of perceptions of the state of governance in India's regions.

13. The continuation of the administration of rights of various groups of interests that form part of the Jagannath Temple and religious communities within the secular State of Orissa, with the Gajapati King of Puri playing a significant role both in administration and ritual provides an interesting regional model of the accommodation of the secular state and the sacred beliefs of important groups of people (see Mitra, 1994).

14. Consider, for example, the difficulty that Indologists face in describing Hinduism both as a religion with a cohesive core, and as a lived-in category which makes it appear as a cluster of cults (see Heinrich von Stietencron, 1989).

15. When asked ‘Suppose there were no parties or assemblies and elections were not held—do you think that the government in this country can be run better?’, 69% of Indians argue in the opposite. But the number of Muslims, at 72%, making the same argument in favour of retaining the democratic structure, is even higher than the average (Mitra and Singh, 1999).

16. See Mitra and Schoettli (2007: 21) for an illustration of how to combine domestic and international constraints in a single model.

Additional information

Notes on contributors

Subrata K. Mitra

Subrata K. Mitra, PhD, is Professor of Political Science, South Asia Institute, University of Heidelberg, Germany; and Visiting Fellow, Centre for the Study of Developing Societies, India