The Afghan Model

We are preparing to go to war in Iraq under the assumption that making the peace will be harder than winning the conflict. Many commentators have offered more or less sloppy analogies with the situation in Afghanistan, most of which have little value. But if there is anything to apply from one reconstruction effort to another, it is the importance of learning how the society in question is organized and how power has traditionally functioned within it. This is just what we have failed to ascertain in Afghanistan.


By now, Americans are familiar with the standard themes of articles on Afghan politics. We have read arguments for and against strong central government, the ethnic composition of the current government, the role of the “warlords,” and various approaches to reconstruction. We have come to realize that Hamid Karzai has a great deal more power in Kabul than outside. We have, lately, seen a lot of articles lamenting the slow pace of rebuilding, the continued stagnation of the economy, the difficulties with forming a national army and disarming “warlords” and the fact that in material terms little has changed for the better in many Afghans’ lives.

But no matter how much money we pour into reconstruction and education — and I believe we should pour quite a lot — it won’t be spent wisely if we continue to think about power in Afghanistan along the standard lines. The terms of the debate on Afghan politics are wrong.

The dirty secret of Afghan society is not just that Karzai has surprisingly little power outside Kabul, but that the so-called warlords governing the rest of the country have surprisingly little power in civil administration. The central government versus regional autonomy discussion overlooks an essential third set of players which American military and reconstruction efforts have by and large neither addressed nor included.


This is the group of influential Afghans one might call the khans, following the American anthropologist and writer Whitney Azoy in his seminal study Buzkashi: Game and Power in Afghanistan originally published in 1982 and reissued in 2002 by Waveland Press. Khans are men who do not have to work for a living — precisely how the English once defined gentlemen. In both cases, the source of unearned income is usually land.

They are difficult to define because power in Afghanistan is unusually diffuse and decentralized, but everyone knows who they are. They represent a ruling class, but one not nearly as distinct and conscious of itself as, say, the British ruling class. And it is important that American policy acknowledge their existence and incorporate their influence, if we are to achieve our goals in reconstruction and development.

The central organizing principle of Afghan society is the group of bacha-e-kaka, or paternal first cousins. It is the couple of hundred khans and their few thousand bacha-e-kakas who constitute the Afghan power elite. Bacha-e-kaka are allies in both political and economic terms, as well as close friends and also in-laws: The optimum marriage in Afghanistan is between paternal first cousins. (It is also true, as Azoy notes, that first cousins can become deadly rivals for inheritances and family leadership, but I am going to talk here about the positive ways this group functions.)

With large families and, even today, a fair number of affluent provincial men with two wives, an average Afghan man may have seven or eight or nine aunts and uncles on each side, each with a similar number of children. These cousins constitute an Afghan’s social network. Afghan society, while superficially egalitarian, is in many ways closed.

There is little reason for an Afghan to “network,” since everyone is born with the basic contacts an American goes about making. Your doctor and your contractor, your auto mechanic and grocer are all cousins, or the employees or bosses of cousins, depending on your social position. So is your go-to man in local government and your contact in the armed forces.

What separates the middle class from the upper class is not so much the trappings of wealth, which are pretty thin on the ground anyway, a matter of a Toyota Land cruiser, a satellite phone and a satellite TV, but the quality of cousinship. A middle-class man’s hundred first cousins are teachers and businessmen and engineers and doctors and army officers. The upper-class man would claim government officials and major landowners and bigger businessman, provincial governors and mayors, generals, and, perhaps, a stray member of the royal family.

Reliance on these cousinship accounts for the lack of development of formal political structures — which in turn makes for decentralized government. What constitutes legitimacy in Afghan political life has never been well defined. Before the Soviet invasion, Afghanistan had muddled along with a king whose authority diminished as distance from Kabul increased, provincial governors from the khan class, and the clans of a hundred cousins below them. As long as a reasonable degree of security prevailed, the system worked — if a country at the bottom of the list in terms of public health, literacy, and many measures of wealth “worked.”


Focusing on conflict and hatred, and reading Afghanistan in terms of the Balkans or Africa, we have also attributed too much importance to ethnic divisions. Intermarriage is far more frequent than the stereotype would have you believe, and to the extent that ethnic groups socialize mainly with each other, it is because Afghans in general socialize mainly with their families. (Otherwise, it is difficult to receive visitors in mixed company, owing to taboos on women meeting unrelated men.) In my admittedly limited experience of six weeks in the country, Afghans make far fewer comments about ethnicity, and fewer derogatory comments, than people just about anywhere I have visited.

So, you may ask, what is their problem? Why the decades of civil war, and why the fragility of the peace? Why the glacial pace of reconstruction, outside of the efforts of the U.N. and foreign governments and NGOs?


A popular Western media answer is: the warlords. (Afghans, as writer Robert Young Pelton points out, call them “commanders,” which conveys the flavor better: less absolutist, and less bellicose. Though Karzai, perhaps translating back from English, calls them jang-saalaar, which means “warlord.”) And if you ask local people in why the government has made few or none of the basic improvements in infrastructure and education which are supposed to be underway, they will tell you that the commanders does not have the money.

That is to say, they do not have the money as long as they are maintaining thousands of men under arms. That is where the money goes. Now this is not entirely a bad thing, for at least these men are getting paid, and getting fed, which stimulates the economy. Sometimes more the latter than the former; you hear of areas — including Kabul — where the soldiers are basically working to eat.

On my first visit to Afghanistan, I stayed at the guesthouse of General Dostum in Shiberghan and heard his people’s take on local politics. And one thing Dostum’s detractors and admirers agreed on was power. I took this for granted; I was more interested in how the man differed from the reputation, the “brutal warlord” of the American and British press. Instead, I found an open, blunt-spoken man, egalitarian by nature and with the endearing naïveté common even to the toughest Afghans.

On my second trip I found myself staying with a family of khans. And as I spoke with a greater diversity of people, I noticed that Afghans generally do not assume that government — whether the central government or the commanders — is the way things get done. As an Afghan friend jokes, “Afghans are natural Republicans!”

The dirty secret of Afghan politics is that the “warlords” are very far from the absolute rulers we take them to be. In fact, they have far less power in civil administration than elected politicians in the United States. And herein lies the tragedy of commanders who, like Dostum, seem well intentioned and democratically inclined. Since the end of the war, Dostum has been saying that he wants to become a civilian leader, concerned with building the economy and democratic institutions. This is what he must do in order to improve the lives of the people living under his rule.

The problem is that five or ten or fifteen thousand men — depending on whose numbers you trust — have to be paid and fed, and kept in working equipment, gasoline, and so on. Little is left for civilian purposes, and that little is often spent in ostentatious public buildings that proclaim the power of the builder, rather than meaningful improvements in public services. Dostum has renovated the stadium in his base of Shiberghan, but the roads and electric supply and schools are still in terrible shape. Why is his “reconstruction” so largely symbolic?

Because in many ways, his power is. Why doesn’t Dostum simply disband these forces, perhaps channeling them into civil service, building projects and so on? Because without them, he is not important. He has the wrong hundred cousins, so he needs an army.

Dostum, born into poverty in a small village, almost by definition had no useful family alliances. His remarkable rise was through his own talents. Like many other gifted men in developing nations, and in our own, he joined the army and benefited from its fairly meritocratic structure. As Americans, we find the self-made man admirable and natural. But we do not realize how vulnerable his background makes him in this very different society.


Power in Afghanistan comes from the family network, or from having men under arms. It does not come from being an efficient technocrat or administrator or media celebrity. Controlling 10,000 men engaged in public works does not translate into political power the way controlling ten thousand soldiers does. Dostum’s revenues come from the Koda Barq area and electrical plants and from customs duties at the Uzbek border. But without military might, these would not be under his control. Atta or another commander would seize control of Koda Barq, and the central government in Kabul would take the customs revenues (as it has long been trying to do).

The approach adopted by the American government and United Nations is crucially flawed. The U.N. role is especially important, since it is spearheading most of the educational and food aid efforts. If you speak with United Nations officials in the provinces, they will tell you that they work with three sets of powerbrokers: the so-called warlords, the provincial governors, and the mullahs. To get something done, they gain the cooperation of the first two men, and ask the mullahs to preach in favor of it in the mosques. They will even tell you how surprisingly cooperative they are.

A word on this. Islam pervades the fabric of life to an extent perhaps analogous to Christianity in medieval Europe, and is officially taught in the schools (the only Afghan textbooks the U.S. government is not paying for are the religion textbooks). But unlike other parts of the Muslim world, Afghan Islam is not centered around the mosques, but around the home. Even at the major holiday of Eid, the end of Ramazan, total attendance at prayers at three prominent Kabul mosques was less than a thousand people. Sermons preached by mullahs will fall upon few ears — most likely the old, poor and idle, for whom the mosque is a social outlet.

Of course the mullahs are cooperative with the U.N.: It’s the first time anyone has ever considered them important enough to enlist in an effort of any kind. As Whitney Azoy points out in his book, in the villages the mullahs are often just as poor as their congregation, and are dependent on the local khans. The analogy might be with village curates in 18th-century England, who were dependent on the livings bestowed upon them by the lord of the manor.

The United Nations is largely ignoring the khans, except in cases where they happen to coincide with positions like that of provincial governor, as they often do. But the khans are doing more to improve the conditions of their fellow citizens than the “warlords” are.

The khans do seem motivated by something like noblesse oblige, which, of course, enhances their prestige. They will donate a classroom to a school ($500) or land for a public park, a prized local amenity.

The neighborhood of Baba Yodgar in Mazar is an interesting case of a settlement of families who fled insecurity in Faryab Province in the war years to live in the relative peace of Dostum’s sphere of influence. The links between Maimana and Mazar are strong; wealthy Maimana families also have houses in Mazar. And so the natural-gas supply, power lines, and wells for this neighborhood were all provided by these families as a service to the people of their hometown. Land has been donated for a better school, and a deep well has been dug that will — if $20,000 can be found for a pump and plumbing-supply water to the houses.

When the khans undertake an initiative like this, it happens, because they have the connections all the way down and across — a cousin is the school principal, another is the town engineer. In business terms we might term some of this corruption, but the fact is that everyone in Afghanistan is someone’s cousin, views himself as such, and operates in a more or less potent version of the same network.


What are the implications of these facts for American policy? To actually achieve progress in reconstruction — better termed, construction, since some of the improvements we are trying to make have never existed — we must work with the khans as well as the warlords.

When a new project is planned, we must meet with these men and win their cooperation. Building alliances between the United Nations, NGOs, local commanders, and the traditional khan power structure is necessary for getting things done. It is also the only chance of getting men like Dostum and Atta — whatever one may think of either or both of them — to disband their militias and thus our best hope of an end to fighting in the provinces. It is true that the khans do not represent an elected government, but neither do the “warlords.” The power of the cousinship is no more or less legitimate than that of the militias.

It is fashionable to wring one’s hands over the difficulties of the task we have set ourselves in Afghanistan, but instead we should rejoice at a situation where with little relatively little money and effort we can make a huge cultural impact and win ourselves friends for generations. The key, however, lies in appreciating the diffuse and complex nature of power in Afghanistan, and the role of the khans, little studied when Afghanistan was not in the headlines, and lately overshadowed by the colorful “warlords.”

— Ann Marlowe is a New York writer who has written about Afghan politics for NRO and the New York Post.

Tags: , , , ,

Leave a Reply