The Fragmentation of Afghanistan by Barnett Rubin
The Fragmentation of Afghanistan by Barnett Rubin

The Fragmentation of Afghanistan: State Formation and Collapse in the International System

Jun 25, 2002

Afghanistan is "hard to rule" for the same reason it's hard to conquer: it does not have many resources, the settlements are far apart, and there is not much water, contends Barnett Rubin.


JOANNE MYERS: On behalf of the Carnegie Council, I would like to welcome members and guests to our Author in the Afternoon. We're extremely pleased to have with us Barnett Rubin, author of The Fragmentation of Afghanistan.

When I first began to plan the programs for last fall, it would have been unimaginable to predict the nature of the journeys taken since that time. As this year comes to a close, it seems appropriate that our last program before we adjourn for the summer should focus on Afghanistan. After all, it is the country where years of misrule and abuse of power provided the fertile ground from which the terrorists were able to prepare and plan the horrifying attacks of last fall.

September 11th was a defining moment for all of us, and the war on terrorism which followed dramatically influenced our planning for this year's programs. Nine months have passed since that time, but nowhere in the world has a country been changed more than Afghanistan.

Although Afghanistan appears to be moving in a better direction, the task of rebuilding this country is far from complete. As we struggle to understand the events taking place there today, it is instructive to examine this nation's history in the hope that we will learn from our past mistakes and be able to participate more constructively in its future.

In the early weeks of the war on terrorism, policymakers, government officials, and journalists were hungry for information about this mountainous bastion of the Taliban. In their search to better understand the history of the country, they invariably turned to one of the one the world's leading experts on Afghanistan. It just so happens he is sitting next to me now.

But why Barney? Because Barney's understanding of the history of this region, his knowledge of the interrelationships of the various tribes, combined with an appreciation for the impact of the Soviet occupation and its withdrawal, and his insight into the failure of the state are unmatched.

In The Fragmentation of Afghanistan, he examines Afghan society and conflict from the 1978 Communist coup to the fall of the last Soviet-installed president. This revised edition has been updated to include material which reflects on the impact of the Taliban and of Osama bin Laden.

When his book first appeared in 1995, reviewers said that the contents were scholarly. Today I would say prescient as well, for in it he wrote that far from being an unchanged, traditional society living in a different time, Afghanistan has been thoroughly shaped by its interactions with the modern state system, torn as it was by its eastern neighbor of Pakistan and its western neighbor Iran, and of course the former Soviet Union, Saudi Arabia, and the United States.

He concluded by stating: "If the international community does not find a way to rebuild Afghanistan, a flood tide of weapons, cash, and contraband will escape that state's porous boundaries and make the world less secure for all." And that was said over seven years ago.

Barney's interest in Afghanistan began in 1984, when Geri Laybourne, the founder of Human Rights Watch, invited him to collaborate with her on a report about Afghanistan. This endeavor provided him with his first opportunity for meeting the refugees in Mujadeen and Peshawar. In 1985, with the founding of Asia Watch, he returned to the region, and has been back almost on a yearly basis since.

So what can we learn from our guest's many years of research and experience? Simply put, a great deal. I ask that you please join me in giving a very warm welcome to our guest, one of the leading thinkers on Afghanistan, Barnett Rubin.


BARNETT RUBIN: Thank you, Joanne, and thanks to you all for coming.

I thought that rather than give the usual update on the situation in Afghanistan, I would reflect a bit more on what my book has to say and how it is applicable to the situation that we face today.

When I go around the world - and I've worked on other conflicts besides Afghanistan - I find that there are two opposing views about conflict. The local people who are involved in a conflict tend to believe that it's the result of an international conspiracy, which is pitting them against each other; and the international actors believe that it is the result of the backwardness, ancient hatreds, poverty, or some other problem of the people involved and that the international community is there trying to solve it.

What I did in this book - and what I hope to continue to do in my work with Afghanistan and various other conflicts in this region and elsewhere - is to show how conflicts, but also peace processes and processes of peace-building, result from the interaction of these international forces and the local social structures. To some extent, the way that interaction takes place is not fully visible because there are certain things that they each take for granted. Let me be a little more specific.

I looked at different periods of the development of Afghan conflicts, starting with the background, and in particular the formation of the state in the territory that we now call Afghanistan. I explain how various changes took place by examining the flows of resources - money and commodities to political actors, including foreign aid. I looked at how the state and other political projects were financed, both on the real side - actual economic activity - and also on the monetary side, because printing money is one way that states are financed too.

Concretely, what I saw in Afghanistan was the following:

  1. In terms of resources, expenditures in a weak state government like Afghanistan never seem to be more than about 10 percent of the gross domestic product, whereas in most countries it's 20 or 30 percent. And the domestic resources were an even smaller proportion of that domestic product because the government became increasingly dependent on foreign aid, and then, when foreign aid decreased, it became dependent on sales of natural gas to the Soviet Union.
  2. The government or rulers used those resources to create institutions of violence and coercion - that is, the military and police, which are the core institutions of any state. We aren't so conscious of them in well-developed states because the actual use of violence is relegated to the margins of society and to external affairs. But for the consolidation of a state, the organization and monopoly of violence is a very key element. Thus the debate about warlordism and creating a national army, because the state in Afghanistan today doesn't have a monopoly of the means of violence.
  3. I examined the nature of political elites - that is, people who had key positions, whether in the government, in the Communist party, in the Mujahideen parties, or later in the Taliban, and now in the interim administration. I collected data on this, such where they were recruited, what type of education they had, where they were educated, and what are their ethnic, regional, and tribal backgrounds.

I reached certain general conclusions about why this turmoil overtook Afghanistan. First, not only was it a weak state, which everyone recognized, but it also had always been a territory inhospitable to the formation of states or strong governments because it is very expensive to govern. It's hard to rule for the same reason it's hard to conquer: because it doesn't have many resources, the settlements are far apart, and there is relatively little water. At times, it was on the periphery of empires. When it became a state, rulers were only able to control it because of the foreign assistance that they received.

This is very much the case today because even though the Taliban were, relatively speaking, a weak state, it was only through the intervention of the United States that it finally became possible to defeat them, and it's only the international presence and assistance that is keeping the Islamic Transitional Government from falling apart, which it would likely do if it were left on its own without the resources to sustain itself.

Second, the political elites, through their education and dependance on foreign resources, developed an increasing social distance from the majority of the country's population. If you look at the education of the Cabinet members under the old regime, the vast majority of them were educated in the West. This is a country where the literacy rate was probably under 10 percent. The educated people were concentrated in Kabul and parts of a few other cities and most of the foreign assistance went into these cities. The educated, once they graduated, gravitated toward those cities or a few provincial centers because that's where the government jobs were. There was very little in the way of government penetration or government services provided to most of the population.

The government had very little capacity. It wasn't a predatory or extremely oppressive government either. What repression it exercised was mainly against rival political elites in the capital; it didn't touch most of the population.

In this context, you don't develop much of a political system in the sense of articulated interests on a national level or in political parties because the national government is irrelevant to most people. That doesn't mean that they don't have a diffuse loyalty to it or that they were trying to rebel against it, but they weren't actively engaged in trying to change it or to participate in it either. That meant that when, for a number of reasons having to do with the interior politics of the elite in Kabul and foreign intervention, the government was overthrown, that old elite did not have the organizational or political resources to resist and reorganize.

Part of the reason that the resistance to the Soviets was overwhelmingly dominated by the Islamist or so-called Fundamentalist Party was because those were the parties favored by Pakistan and, therefore, funded by the United States and Saudi Arabia, for reasons having to do with Pakistan's conception of its security interests (in particular that the religious parties didn't make territorial demands against Pakistan in the same way that the nationalist parties did.

But, at the same time, the elites from the previous regime dispersed very quickly and without much of a struggle and went to the West, because that's where they had been educated. This was manifest in the Bonn talks in December. When the Rome Delegation, supposedly representing the interests of the former King, sent their list to the United Nations, it didn't contain the name of a single person who was living in Afghanistan. The UN asked them if they could please add some people, and they came up with two other names, Hamid Karzai and Pacha Khan Zadran, who is now the warlord in southeast Afghanistan who is wreaking havoc and resisting the attempts of the government to assign a governor to the region.

So we can see that there are some real flaws and weaknesses in the state structure of Afghanistan. Furthermore, this structure is not just a traditional state structure. State/society relations have been shaped by this interaction with the international system.

In the last twenty-five years, much has changed there. This state was destroyed, but some elements of the structure remain still. The basic administrative structure of provinces and districts is still present and people adhere to it in some way. During the recent Loya Jirga elections, people found it quite easy to integrate themselves into a process that was organized around those districts and provinces. But they don't have functioning capacities in most cases. Furthermore, to the extent that they do have these capacities, they are under the control of these regional power holders, or warlords.

Who are these warlords? We can answer this question if we again examine resources, military/political structure, and the political elites.

First, the warlords are power holders who accumulate resources by controlling revenues that by law ought to go to the central government, in particular the customs revenue. The warlords who control the key customs points at the borders keep this revenue for themselves to pay their own military police and administration, all of who are recruited from their own region. So the old Afghan military was a national military, in the sense that it was ethnically mixed. The officer corps was more homogeneous - the top level was composed of certain tribal groups that were very close to the Royal Family. This is the pattern throughout states that are not very rich and have highly institutionalized forms of government. This is not the case in India or Pakistan, but it is so in many African countries or in Syria, for instance. The top military will be from the same group as the ruler to ensure him more confidence in his security.

But, aside from that, the troops and units themselves were ethnically mixed, and they were assigned all over the country. So now the military units belong to the region, and the ethnic group of their commander. And the political elites, similarly, who are running things are also homogeneous.

While none of these warlords challenges the political legitimacy of the Afghan state - none of them has raised any demand for secession - the way that they exercise power and accumulate wealth guarantees that the state is today fragmented into these somewhat larger bits.

The key problem in the reconstruction of Afghanistan, then, is to integrate these regionally based powers into a central power that will have the capacity to deliver all those goods that people normally think of as reconstruction: building schools, providing health care, providing security, and maintaining the financial and transport infrastructure that is necessary for basic economic activity.

The way these warlords got this power was primarily through the U.S. and Coalition Offensive, which destroyed their rivals, the Taliban and al-Qaeda, who have become increasingly integrated in the past couple of years, and also eventually led to their being armed and encouraged to take over the cities. These are more or less the same people who were exercising power in these areas between 1992 and 1996, but it isn't true, as many people say, that Afghanistan has always been ruled by these regional warlords. It's true that at the very local level - below the district level - the state didn't have much penetration, although it turned out that the Taliban did to some extent, because they managed the remarkable achievement of wiping out opium growing for the 2001 harvest, which did not, however, outweigh some of their other characteristics. Nonetheless, there was generally no state apparatus in the village level, whereas the Taliban had the mullah network, which is how they penetrated the villages.

But that didn't mean that there were large-scale warlords. And people generally do not consider them legitimate tribal or ethnic leaders. I've been in Afghanistan twice in the past three or four months. I was mostly in Kabul, although on my last visit I was also in Jalalabad. People prefer to be oppressed by warlords and commanders from their own group rather than by those from another group because they tend to be more ruthless. But the universally expressed preference was for being ruled or governed by a central authority that obeys some laws and where the district administrator can be changed by the central government if he or she doesn't do the right thing, because they describe their present situation as being the "rule of the gun" - or "rifle culture" or "rifle rule," as they call it. It's what they universally want to get rid of.

Some of the rulers could become legitimate representatives. There hasn't been any way to be a legitimate leader in Afghanistan for the past twenty years, so you can't exactly criticize people for not standing in democratic elections when you had a system that was based on violence and anarchy. But no one really tries to legitimate the system as it exists now.

If we come back to the international community again, how and whether international assistance is disbursed in Afghanistan will be one of the key issues in how or whether a new Afghan state is formed. Here are some things that people don't always realize.

For instance, the predominate school of thought in modern and contemporary development practice is that the goal of development is to strengthen civil society - that development aid should go to NGOs and not to oppressive state structures.

There was a debate after Kabila took over the Democratic Republic of the Congo in 1996 over whether to aid the government, even though he was still not permitting an investigation into atrocities that had taken place, or just aid civil society. Governments, of course, are on the first side; human rights organizations are on the second. Some groups were in the middle, though, like Oxfam, which was in favor of some limited support to the government, on the grounds that you had to build government capacity.

If the international agencies go into Afghanistan now with the idea that they're going to work with civil society and build up powers outside the government, they will destroy the effort, because civil society can only function when it has a legal structure that provides security and a framework for functioning. There are civil society groups struggling to gain a foothold in Afghanistan, and they deserve some support, but the fundamental problem is security, administration, and strengthening the government. So development and reconstruction projects should be financed and implemented by the Afghan authorities. That means that although there will be corruption and waste, there will be less corruption and waste than if foreigners do it, because it's much less expensive.

I have an anecdote in mind which I'll tell you, even though I don't know if it's true. The Governor of Nangarhar Province, Haji Qadyr, told me this story. We could safely say that he is a warlord.

He was in Germany, where he has another residence. He saw some people in the German Government. They said, "We have a project to clean the Jalalabad Canal," (the main irrigation canal in his province) "and it will cost $3 million." They expected him to be very happy. They said, "Aren't you happy with this?" He said, "No, I'm not happy."

He went back to Nangarhar. The Germans were going to do this project by giving a contract to a German firm. But Qadyr claimed that he then went to his director of irrigation for Nangarhar Province, and said, "I want you to clean the Jalalabad Canal. Give me a proposal." The guy came back in a few days with a proposal to clean the Jalalabad Canal for $25,000. So he gave him $25,000 and he claimed forty-five days later that the guy had cleaned the canal - and he returned $7,000. And not only did they have a clean canal, but they had some new equipment and some upgraded old equipment.

Maybe the figures aren't true. Was it up to the same standard as if the German company had cleaned the canal? I don't know. But you could clean it one or two times. But in any case, you build up the capacity of the local administration.

The U.S. and other major powers have refused to extend the International Security Assistance Force, which has provided the umbrella that has made it possible to establish a government, at least in Kabul. This has to some extent reduced the power of the main warlord group there, the Panjshiris, Masood's people. Sometimes they're called the Northern Alliance, but this is not the whole Northern Alliance; this is just the core group.

What the United States has done instead is to send small special forces to monitor the major warlords. But they also have their civil affairs teams working in those areas providing benefits. There was an article about this in the New York Times recently; USAID is funding some of those activities.

The humanitarian organizations are concerned about these because of non-uniformed military providing assistance, which they feel will compromise their status. But, besides that, the appearance is that the United States is providing military assistance to the local warlord and then providing assistance under that aegis rather than going through the central government. It will be tempting for many organizations who are looking for opportunities to make deals with local power holders and provide assistance, which again could undermine the efforts of the government to establish itself by becoming the point that people have to go to for resources, and which would enable it to set up a structure that provides security.

It sounds as if I'm advocating a centralized government for Afghanistan as the best solution. When you talk to people in Afghanistan, a very large number of them say that that is what they want, because they associate local control with control by gunmen and warlords. In fact, the international experience with devolution/decentralization is that if you try it in a rather chaotic situation without a strong legal system, it is a formula for corruption rather than effective self-government.

Nonetheless, there are ways of setting up local control or local participation in development projects. There's a project that UNDP and the World Bank are doing on community development that is being implemented through a part of the Afghan Government.

There are many reasons to think that a more decentralized, if not federal, system would be better for Afghanistan, but the first step will have to be establishing the basic institutions of authority, demobilizing, and integrating the major extra-constitutional power holders.

I am not optimistic. I'm certain that all of this will not go easily according to the scheduled in the Bonn Agreement, under which within two years we're supposed to have free and fair elections. I'm quite confident that that will not happen.

It will be difficult to do if there isn't a continual international presence. At the same time, that international presence cannot try to take over in too-heavy-handed a way and substitute itself for the Afghan government. The UN Mission in Afghanistan has a very delicate balancing act. Sometimes it is undermined both by the institutional imperatives of the United Nations itself, at other times by the actions of donors, and at other times by the institutional incapacities of the Afghan Government, which is not always very susceptible to using aid in a way that international donors can find acceptable.

My research certainly shows, and the argument of the book as applied to contemporary politics shows, that the key to stability is building up those principal state institutions, without which other desirable things won't follow.

JOANNE MYERS: Thank you very much. We'll open the floor to questions.

Questions and Answers

QUESTION: Any government, like the current transition, is totally dependent on foreign aid. To have any legitimacy among the locals, a government must have resources. And what potential resources are there for a central government in Afghanistan? We know from experience that foreign aid is very fickle. How long will this last? From that perspective, where is this national entity going in the next few years?

BARNETT RUBIN: The government issued its first budget in March when the Afghan fiscal year starts. The budget for this year will be 82 percent funded by foreign assistance and 18 percent by domestic resources.

The 18 percent of domestic resources is, first of all, the customs revenue. Currently the government has not been receiving customs revenue, except from Kabul. However, at the recent Loya Jirga, Hamid Karzai carried out negotiations with the principal warlords. Several of them agreed, in return for getting certain types of positions, that they would turn over their customs revenue. Haji Abdul Qadyr from Nangarhar agreed that he would turn it over from the customs post on the road from Peshawar. Dostum agreed that the would turn it over from the north - although I'm not 100 percent sure that he controls that, because there are three rival warlords in that area. Khalili from central Afghanistan agreed, but there's no money there, so it doesn't really matter - it's in his interest to make such a pro forma agreement because he needs aid. I'm not sure about Qandahar. The trouble is that the customs post is in a district that is extremely unstable, where they couldn't even have elections to the Loya Jirga.

But the main issue is Herat, because, for various reasons, Afghanistan's foreign trade has shifted from primarily Pakistan to Iran, and the head of the National Customs Service estimated that 70 percent of Afghanistan's imports are now coming through Iran. That means that the Governor of Herat, Ismael Khan, controls a large share of Afghanistan's customs. He was offered vice presidency at the Loya Jirga in return for turning over the customs duties, and he would only accept if he could still be the Governor of Herat. They said no, so he has gone back to Herat. He wanted to be defense or interior minister and they didn't give him that.

There is no oil extraction in Afghanistan, but there is natural gas. However, the gas wells were capped in 1989 when the Soviets left, and since then all the pipelines have been looted. The pipelines would go to Uzbekistan, as they used to. It's not clear whether it's worth the investment to put those in at that time.

If a pipeline were built from Turkmenistan to Pakistan to carry Turkmen gas to Pakistan, which all of those countries want to do, then Afghanistan stands to make about $150 million a year in transit fees. This has given rise to many conspiracy theories about how the whole U.S. campaign is the result of the manipulation of the Bush Administration by oil companies. But I'm not clear that there is much interest in this project by oil companies. So far, there is no financing available, and it's considered pretty risky and not necessarily profitable compared to other possible ventures.

There's another item that some people have talked about, which is customs duties on the import of luxury goods by members of international organizations. This touches on the status agreements, at least of UN organizations. It might not apply in the same way to international NGOs.

There have been some disputes about this with the transitional government in East Timor. There was one recently involving a floating hotel in the harbor, where the government - still under UN control - attempted to collect excise tax from that hotel, which tried to sail away instead. There's no danger of anyone sailing away from Afghanistan. It's worth doing a study.

What has already been set up now is telecoms, wherethere is some potential. The problem is getting past the tremendous corruption that that industry tends to generate in such circumstances to the point where contracts will be awarded competitively and the government will tax it, because it's much cheaper to bribe people than to pay taxes. I don't have the impression that the government is exploiting that.

QUESTION: I'm interested in your reaction to an editorial in the New York Times, in which they characterized the Loya Jirga as "a lost opportunity" because the United States and the United Nations both played too heavy-handed a role in persuading the king not to seek a more influential post other than father of his country; and also that Karzai did not have the courage to select Hazaras and Uzbeks in the government, so they felt that that was to the detriment of the government, and that this was a lost opportunity. I'd be interested in your reaction.

BARNETT RUBIN: Everyone who was involved with the Loya Jirga in one way or another - including me, because I was in the Bonn Accords on the UN team when we drafted the Accord - believed that although the government at Bonn was flawed in certain ways since it was a non-representative government formed by a non-representative meeting, it nevertheless represented what was happening and what was possible at that time. In six months you'd have the Loya Jirga and you would again have, not a government that was democratic or fully representative, but a better one.

What you have is a different one, and maybe it's better in some ways, and maybe in some ways it's not. Those criticisms, however, miss the point.

First, about the king and the United Nations and the United States: no one can criticize Lakhdar Brahimi for putting himself forward too much. In fact, he remained completely invisible during this whole affair. That doesn't mean that he didn't do anything, but you didn't see his picture in the paper and you didn't see him giving press conferences announcing what the result was before the king announced it, or he wasn't there sitting up with the king during the press conference. It was Zalmay Khalilzad, the special envoy of the United States, who did that. He wanted to show support. But announcing it before the king did was a mistake.

The decision itself I don't consider to be a mistake, because, in fact, the king did not want to have an office where he would have political responsibilities. But the king is a rather malleable person who doesn't like to say no to people, and there are people around him who want him to have those positions. The king could not possibly exercise those functions, so they would, even though they weren't chosen for them, and they wouldn't be accountable to anyone.

It is true that many people in Afghanistan did want the king. Omar Zakhilwal, who was a delegate, who wrote this Op-Ed in the New York Times, said that people thought that only the king could create a more representative government and curb the power of the Panjshiris. But there's no way that he could do that when he couldn't even control his own entourage, let alone the people organized by Ahmed Shah Masood. By helping the king to articulate what he did want, they did the right thing.

Is the government more or less representative? We can do the arithmetic. The biggest change, in terms of ethnic balance, from the former government to the current one is that there are more Pashtun technocrats and fewer people, technocrats and mullahs, because now there are two Pashtun mullahs too - I'm counting the Chief Justice and the Minister of Haj and Mosques - and fewer people from the King's group.

The presence of the Northern Alliance as a whole is about the same; that is, about half of all of the positions. It still includes most of the powerful positions. Qanooni's leaving the Interior Ministry is somewhat vitiated now that he will have a position in charge of security, which might still leave him in charge of the Interior Ministry, given that the Interior Minister is eighty years old, lived in Los Angeles for the past twenty years, and everybody who works in the Interior Minister who's a Panjshiri is loyal to Qanooni. So in that sense, the power of the Panjshiris has not been significantly lessened.

It's interesting that everybody in the government who's in charge of money is Pashtun, including the Minister of Finance, the new Director of the Central Bank, and also some other ministers.

If you have a lot of guns, you can get the money from the people who have the money.

But the real issue is that having a meeting, even a good meeting, even a better meeting than this one, can't change the relations of power in Afghanistan. Maybe the expectations were too high for the Loya Jirga, because Afghanistan is not at a stage of institutional development where you can determine who has power by holding an election, because the people who really have power will not leave just because of an election, even an indirect election like the Loya Jirga.

There might have been more that Karzai could have done. People say he's a very poor negotiator and that he didn't really press his people to give up as much as he might have. Maybe he could have got more out of them.

The main issue will be what the government does. When I spoke to the delegates who came to the Loya Jirga elections in Jalalabad and asked them, "What do you want from the Loya Jirga?," the first thing they said was not "we want the king back" or "we don't want the king back" or "we want more Pashtuns in the government." When I asked them specifically about those things, they said that they wanted the government to provide them with certain services, with security, education, that was not run by people holding guns.

If the government in partnership with the international community manages to deliver something, people will be less concerned about the ethnicity of the ministers. And if it doesn't, it wouldn't matter much if the ministers had the right ethnicity. It would just be another source of conflict.

You may also like

MAY 23, 2024 Podcast

U.S. Election 2024 in a Post-Policy World, with Tom Nichols

"Atlantic" staff writer Tom Nichols returns to "The Doorstep" in its penultimate episode to discuss the lead-up to the 2024 U.S. presidential election.

ChatGPT homepage on a computer screen

MAY 15, 2024 Article

Forecasting Scenarios from the Use of AI in Diplomacy

Read through six scenarios and expert commentaries that explore potential impacts of AI on diplomacy.

MAY 15, 2024 Podcast

Beneficial AI: Moving Beyond Risks, with Raja Chatila

In this episode of the "AIEI" podcast, Senior Fellow Anja Kaspersen engages with Sorbonne University's Raja Chatila, exploring the integration of robotics, AI, and ethics.

Not translated

This content has not yet been translated into your language. You can request a translation by clicking the button below.

Request Translation