In the literature on institutional engineering in multi-ethnic democracies, ethnic diversity is usually characterized as a “threat” which must be “defused”, “contained” or “managed.” Institutions are usually seen as one of the main tools of such threat management. But the characterization of ethnic diversity as a problem, and the literature on institutional design as the solution, suffers from two important biases: a primordialist bias and a statist bias. It assumes that ethnic groups are fixed and exogenous to democratic politics, and that the entity whose interests are to be served through institutional design is the state.

 

In previous work (Chandra 2008), I have tried to develop a constructivist principle for institutional design in multi-ethnic democracies. I term this the principle of ethnic “invention.” It departs from previous work by starting from that assumption ethnic identities can be fluid, multiple and endogenous to institutional design, although they need not always be, and by starting from a perspective that gives primacy to the interests of individuals over states. It suggests that inclusion can be best achieved in an ethnically divided democracy by first designing institutions of interpretation such as the census which allow individuals who find themselves in an ethnic minority the option of reinventing themselves as members of at least one majority ethnic category – and then designing institutions of competition such as party and electoral systems which encourage ethnic self-reinvention in the political sphere.  This principle produces institutions which do not so much “manage” ethnic differences as facilitate and encourage their political expression.

 

My talk today combines the theoretical insights from this previous article with an empirical paper showing how this principle is realized (or not) in South Asia.  In this working paper, I evaluate the extent to which the institutional framework governing the “management” of ethnic differences in three South Asian countries – India, Pakistan and Bangladesh – enable ethnic invention. I evaluate, in other words, the extent to which they encourage the mobilization of multiple ethnic categories in political life. I focus on three specific institutions: the census, the constitution and electoral laws. In India, these three institutions combine to actively encourage the mobilization of multiple ethnic majorities. In Pakistan, these institutions have a more limited effect on ethnic invention: while they do not enable the invention of new ethnic categories to the same degree, they do enable political mobilization along multiple dimensions of ethnic identity. In Bangladesh, these institutions actively prevent ethnic mobilization. The institutional framework in Bangladesh is most closely described by the principle of ethnic ““banning” – the attempt to prevent ethnic mobilization altogether.

 

The result, I suggest, is that the mobilization of ethnic differences does not pose a threat to democracy in India – if anything, it sustains it. In Pakistan, the limited institutional encouragement of ethnic mobilization also dilutes the threat supposedly posed by the electoral mobilization of ethnic differences. While there are certainly threats to democracy in Pakistan, the mobilization of ethnic differences by political parties is not among the most significant. Paradoxically, it is Bangladesh – widely seen as the most ethnically homogeneous country in South Asia – where the threat posed to democratic stability by ethnic differences may be the strongest.

 

This paper is self-contained, but for those interested more in the theory and less in the empirics, I attach also a copy of the earlier published article. I very much look forward to your comments and especially to the insights of colleagues working in other geographic areas!