Federalism and Ethnic Conflict in South Asia
Katharine Adeney


Editors Introduction How successful were the federal political experiments that created the states of India and Pakistan? In this lecture Katherine Adeney-- ex-tutorial fellow in the department of government at the London School of Economics and Political Science and now at Balliol College, Oxford University--considers the relationship between constitutional design and the regulation of ethnic and national conflict in India and Pakistan. Although such investigations might seem narrowly focused, Adeney argues that they are of great importance to the exercise of state and nation-building in a volatile region.

I am going to focus on the relationship between federalism, secession and autonomy, specifically between constitutional design (particularly federal design), and the effectiveness of that design in regulating ethno-national conflict in divided societies.

Federalism can be viewed in two ways. On the one hand you can look at it as being an independent variable, a pre-existing set of institutional structures which therefore condition to some extent the behaviour of elites and masses. On the other, you can look at it as being a dependent variable, whereby elites who are setting up a new state, or are just seeking to reorganise their existing state, use federal structures with the specific intention of creating certain types of behaviour amongst their masses or other elite groups. Now, in terms of its relationship to fragmentation and coherence in Asia, I think one would certainly say that apparent fragmentation promoted by federalism actually leads to coherence, but only in certain forms, and that is what I want to focus on.

Federalism has a pretty bad track record generally. Forty-six percent of federations in the twentieth century have failed. That is a pretty devastating critique. I think that its failure, especially in regulating conflict in ethnically divided societies, has been primarily due to the nature of the federal design rather than something inherent in federalism itself. Second, when ethnic groups are territorially concentrated, lumped together as one, then I think that homogeneous provinces encompassing the whole of this ethnic group are going to be better than, say, a federal design that seeks to split up these territorially concentrated ethnic groups. Third, I don't think that the creation of homogeneous states in a federation is in itself enough to regulate and to prevent ethnic conflict.

Ethnicity and language
India and Pakistan provide excellent examples for a comparative analysis because they have so many similarities. But they also have one important difference in relation to their federal structures. So what are their similarities? The two states are very ethnically heterogeneous, both at the time of partition and independence, and remain so today. They are products of the same colonial regime, the British Raj. Even though they didn't have identical institutional frameworks during their time under the Raj, both states also had exactly the same percentage of religious minorities--about 14-15 percent.

Despite the idea that partition was going to be creating a Muslim homeland for Pakistan, 13 percent of the population of Partition Pakistan was actually Hindu. The important difference from India was that this Hindu minority was almost exclusively concentrated in the eastern wing of Pakistan, the area which is now Bangladesh. Twenty-two per cent of the population of East Pakistan was Hindu.

If you look at the two states divided by language you will see that India is incredibly heterogeneous. There is no one dominant group: the Hindi speaking group is not even a majority. It possess between 30 and 40 per cent of the population of India, depending on which census results you take and depending on how exactly you classify a Hindi speaker. If you consider Pakistan, the biggest single group before the secession of East Pakistan are the Bengalis. But in contrast to India, not only is there a very large linguistic group, there is also another very large linguistic group opposing it, and these are the Punjabi speakers, who are concentrated in the western wing of Pakistan.

After the secession of East Pakistan this changed radically, and Punjabi speakers quickly transformed themselves into the majority. This is important, because in terms of dominant groups and relationships between ethnic groups, you get a very different relationship depending on whether there is any group that can oppose another one. So, the salient point is that where in India there is not a very clear dominant group, in Pakistan there is, both before and after the secession.

Independence and reorganisation
At independence in 1948, both states had very similar constraints. They needed to pursue economic development. They also had the task of state building, as well as the weighty task of nation building, which was incredibly competitive during the partition struggle, but also after the partition struggle. There was no single definition of a Pakistani or an Indian nation. There were always different identities within these entities. But both of these states provide examples of competitive features of federal design, specifically the reorganisation of their provincial units within their federal systems. This might seem a very narrow thing to concentrate on, but it is actually at the heart of their nation building, their state building, and you can look at it in terms of their multicultural policies and their assimilations policies. I think they are key to understanding both India and Pakistan.

The organisation of both countries was designed to manage their ethno-national diversity, but this was their only similarity. India's State Reorganisation Commission in 1955 recommended the division and the reorganisation of provincial units based on language. Unity and integrity of India were paramount, but ultimately that is what it ended up recommending. It denied those based on religion, despite the fact that due to the lack of territorial concentration very few religious states could have been created. Even when language coincided with religion, it shied away as the religious angle was shunned as a consequence of partition.

Pakistan, in contrast, rejected the recognition of language as a basis of identity, and embraced the one-unit plan in 1965, and fused her western wing into one provincial unit. So she rejected language as a basis for recognition and identity in provincial reorganisation in her western wing, but of course thereby completely institutionalised it in her eastern wing, where 95 percent or more of the population spoke Bengali. So it was something of a paradoxical merger.

Both of these reorganisations are contested within India and Pakistan today. Unsurprisingly, the one-unit plan is widely derided for causing the break up of the federation in 1971. What is more surprising, possibly even staggering, is the following quote from India Today in 1998: "Four decades ago, the country upturned every tenet of good governance by carving out new states on the basis of language rather than administrative convenience." Clearly there is the sense that India should not have done this. So this is very much a live issue and something which should be explored and understood.

This feature is taken from a lecture delivered at the London School of Economics and Political Science on November 21, 2000. Copyright The London School of Economics and Political Science.