Rebuilding Afghanistan

back to issue

THE political reconstruction of Afghanistan will involve three difficult tasks: reconciliation between groups that have from time to time fought each other; disarming and reintegrating combatants, and the welding of diverse ethnic groups into a single national community.

The last is challenging even when it does not follow two decades of war. As Afghans consider the lessons of their own history for their present choice of institutions and processes, their neighbours in South Asia also have a great deal to teach them about creating and consolidating a multi-ethnic polity. For fifty years, these states have been attempting to fashion themselves into national communities, and have experienced different degrees of success at different moments. From an examination of their histories, Afghanistan may be spared a repetition, not only of its own, but their mistakes.

Afghanistan has a unique geographical position and unique historical experiences, but it shares with South Asia, diverse ethnic populations whose kin live in adjacent regions across inter-state borders. These groups also occupy distinct parts of the country and have historically enjoyed varying degrees of autonomy. None of this need prevent Afghans from pulling together to rebuild their society and state. The experience of three South Asian states (India, Pakistan and Sri Lanka) suggests that such diversity is not an insuperable obstacle to cooperation, and offers the following insights to those involved in the reconstruction of Afghanistan.

The first thing to note about national integration is that it never really happens. Expecting that any contract forged at this moment will (a) be entirely and exactly realized, and (b) remain salient and valid for all times, is unrealistic. In reality, every contract, every vision, every arrangement is constantly renegotiated between all parties – whether between the ethnic groups that make up a state or the government and one or more of those groups. Anticipating such renegotiation goes a long way towards preventing conflict. Freed of the compulsion to enforce a single national self-image, the state is able to accommodate variations as they arise.

India, for instance, initially intended that Hindi should replace English as the official language within 15 years of the promulgation of the Constitution. In the face of protest from non-Hindi speakers, this provision was amended such that even after 50 years, English is still in use in government communications. This accommodation had the effect of defusing at least one ethno-linguistic group’s secessionist rhetoric.

Like Afghanistan, India, Pakistan and Sri Lanka all have ethnic groups that are concentrated in certain regions of the country. India has, over 50 years, accommodated a range of demands from these groups by creating linguistic provinces, by upgrading the powers of their unit within its complex federal structure, and by dividing or merging units as the need may arise. It has done this on a case-by-case basis, never establishing a single, rigid principle that must be upheld at all times, and allowing, as a result, these individual accommodations. Consciously or unconsciously, the resultant ‘ad hocism’ has helped India accommodate a variety of local and sub-regional territorial demands.

Constitutions that say less rather than more allow this renegotiation to take place in the political arena. They do not identify the state with one or another communal marker, their citizenship provisions do not explicitly include or exclude people and they guarantee political, civil and cultural rights to all their citizens equally. Not unlike the ever-shifting pieces of glass in a kaleidoscope, the more the central authority seeks to define the nature or identity of the national community, the further reality moves from that definition.

Pakistan was founded in 1947 as a homeland for the subcontinent’s Muslims. Attempts to make Urdu an additional marker of Pakistani identity alienated the majority Bengali speakers who ultimately seceded in 1971. In Sri Lanka, new constitutions adopted in 1972 and 1978 reinforced changes in citizenship, language and university admission policies that effectively restricted access to both political power and economic opportunity. Sri Lanka has seen three insurgencies in its independent history, at least two of which have drawn on ethnic grievances.

Keeping the detail to a minimum and expecting that you will have to make changes anyway ensures not just accommodation but early accommodation. It is easier to grant educational rights in minority languages, for instance, than it is to divide an existing province, and even that is easier than fighting a secessionist war. In retrospect, accommodations over language and even federalism seem less costly than the ethnic conflict that is bleeding Sri Lanka today.

The lesson for Afghanistan’s new rulers so far: set forth broad principles but leave out overly detailed definitions, expecting that they will emerge over time and be renegotiated repeatedly.

As for the distribution of authority within the state, history in every one of these countries (including Afghanistan) suggests a simple reality: tight, centrally controlled political relationships are unsustainable. Not only are people culturally different, but also local developmental needs vary greatly. Further, the terrain sometimes makes central control over every aspect of administration difficult, especially when the means of communication either do not exist or have been destroyed.

If some notion – variously coloured, variously detailed – of being Afghan or being Indian (as distinct from being Turkish or Portuguese or Chinese) has survived the centuries without the assistance of state projects, hermetically sealed territorial limits and essentialized cultural definitions, then creating and maintaining a modern polity can certainly survive decentralized regional authorities. Indeed, it may not survive in their absence.

The fact that federal solutions have worked in India and Pakistan to contain ethno-national secessionism suggests that redistributing authority away from the centre strengthens the polity as a whole. When Pakistan erased provincial boundaries in order to impose parity in the legislative representation of the two wings, thus disenfranchising the majority of its population, that violation of its original federal structure contributed to the breakup of the state. After 1971, when the provinces of the West were restored and to the extent that it has left the federal aspect of the polity undisturbed, Pakistan has averted a recurrence of those events.

If India’s success has lain in creating ad hoc federalizing solutions to sub-national movements within its borders, its great failures have arisen when the centre has repeatedly violated, through omission or commission, its end of the federal contract. In Kashmir and Punjab, this has been in the form of political interference in local politics, marginalizing local governments, bypassing them with central ordinances, directives and agents and through electoral malpractices. The centre also played one local faction against the other, making it impossible for them to function autonomously. In Kashmir and in the northeastern states, the central government has also failed to respond to the people’s socio-economic needs in such a way as to make being Indian worthwhile for them. Thus, the South Asian experience suggests that territorial pluralism along with devolution of as much authority as possible, works well if implemented in both letter and spirit.

The Afghan situation is complicated by two decades of civil and international conflict, and the proliferation of arms and armed groups. The new dispensation will have to take on the problems of disarming, demobilizing and reintegrating soldiers into peacetime life. This will be especially hard for those who grew up in the period of the conflict. ‘One size fits all’ solutions work even less in these circumstances.

Simple, ethnically neutral, decentralized arrangements have the advantage of allowing nuanced, contextual answers to individual cases. Accepting that perfect integration is an impossible dream makes it possible to view politics as a process of constant reconciliation and renegotiation, and there is no loss of face to making accommodations as and when necessary.



1. Seek agreement on founding principles but do not worry about details.

2. The less the constituting documents and agreements name and identify (with) specific groups, regions, languages and faiths, the more room there is to concede and accommodate at low political cost.

3. Anticipate that everything will be renegotiated over time. Therefore, creating good negotiating and feedback mechanisms is more useful than detailed arrangements that are meant to (but will not anyway) hold forever.

4. Establishing but consolidating a minimum common ground for integration, devolve everything that can realistically be handled away from the centre.

5. A framework of universally applicable rights works better than integrative mechanisms that seek to include specific groups and thus inevitably exclude others.

6. Stick to what is agreed – in letter and spirit. The first violations should not come from the centre and the first violations outside the centre should be interpreted as simply the beginning of new negotiations.


Swarna Rajagopalan