In an earlier post, commenter Brett makes an insightful point about corruption and raises the issue of feudalism:

BernardFinel.com » The Incoherence of COIN Advocates: David Kilcullen Edition

On the other hand, though, in an environment where any type of objective, impersonal “state” is non-existent, and the “government” basically consists of a whole host of informal and personal arrangements among local elites and power-brokers, corruption can be a useful tool for holding and consolidating a state. Think of how feudal societies’ leadership would ensure loyalty from local elites through land-grants and other forms of reciprocal deals.

I have been hesitant to use the “f-word” (feudalism) to describe Afghanistan for several reasons.  First, Afghan politics do not closely resemble the European period of high feudalism when rules and relationships had been highly formalized, and as a consequence I have been worried that people would misunderstand (and unfairly criticize) my arguments.  Second, the term feudalism carries various negative connotations, and I also wanted to avoid being drawn into a debate over whether Afghan politics were “better” or “worse” than those in modern industrialized democracies.

But that said, people would understand Afghanistan better — I suspect — if they had a better understanding of feudalism.  A few points:

The conventional view of feudalism as involving “land grants” in return for service is an oversimplification.  As a practical matter, feudal monarchs rarely “granted” land as much as they recognized effective control by others over parts of their nominal realm.  The service rendered by lesser lords was less as payment for the land as much as it was to solidify their ties to more powerful lords in order to dissuade and deter local challenges.  The biggest risk to a feudal lord was not that the monarch would nullify the land grant, it was that he would be displaced by a local rival, including those drawn from his own followers and relations.  But the essence of the point is that feudalism did not so much create local autonomy, but rather recognized it.  In practice, monarchs held in personal title and control as much land as they could effectively control and did not cede it just to ensure service.

These patterns were routinely upset as a result of insurrection and competing claims to various titles.  But again, as a general rule, the results followed patterns of effective control. A particular danger for local lords was choosing the wrong side in endemic squabbling among more powerful lords, but even then a pledge of fealty to the winner could undo even effective service to the losing side.

The result was the impression of massive instability in one sense — competitions over titles, frequent revolts, endemic conflict — and massive stability as well — individual families could hold title to particularly pieces of land for generation after generation, pledging fealty to an ever shifting set of liege lords.

Applying a model of feudal politics in Afghan would yield at least a few policy recommendations: (A) Identify and recognize key warlords as durable power brokers; (B) Pressure them into compliance by threatening them, not with the empty threat of central control, but with the threat to recognize and promote their ambitious subordinates in the their place; (C) Realize that shifting alliances and coalitions are a natural consequence of a polity that features both weak central control and endemic warfare; (D) Give up the illusion that key warlords and regional commanders can be bought off with a “job” — most will want some sort of local control and autonomy as the price for switching sides.

The other piece of the feudalism model that is worth consideration is whether there is any applicable model for transition from feudal political and economic structures to “modern” forms.

This is a complex topic, and I can only sketch out a few arguments here.  But here goes:

It is a mistake to conceive of “modernization” as some sort of linear process involving clear phases.  Yes, this is how some people conceived of the issue in the 1950s and 1960s, but the academic literature has since largely concluded that this is a flawed approach.  The point is, Afghanistan is not at an earlier stage on a path toward becoming an industrialized democracy along Western lines. It is, instead, what it is. It may indeed change into something different over time, but it is a mistake to assume a particular teleology in this case.

That said, a number of countries have indeed transformed from conditions roughly like Afghanistan (weak governance, subsistence economy) into stable, prosperous countries with effective governments. It usually takes a long time, and there is essentially no evidence that we have any way to hasten the process.  But more importantly, we need to at least apply the right model.  Some of the models include:

(1) A commerce-driven model, where trade creates centers of wealth usually in areas of strong central control. This effectively serves to impoverish (relatively) and marginalize rural lords, creating a new monarchical-bourgeois governing coalition that trends towards liberalism. Think Holland.

(2) An industrialization-driven model, where rural lords gradually build increased capital to invest in increasingly advanced economic activities, which creates a demand for a strong central government in order to safeguard investments and construct infrastructure to ease access to resources and markets. Think England, sort of.

(3) The classic extraction-coercion cycle where central control is linked to coercive extraction of resources (through taxation and effective conscription) that allows for further expansion of central control. Economic development is often driven by the creation of various bureaucratic mechanisms usually linked to command and control of military forces and industrialization to provide war materiel. Think France.

Which seems most likely in Afghanistan? If you were conceiving of a model of political centralization and economic development, which model seems most plausible? Probably the extraction-coercion dynamic, but two problems:

(1) We are uncomfortable with either the extraction or coercion parts of the equation. Instead we want to make the central government a source of largess and make human rights a key element of the government’s interaction with the populace.  It is a pleasant idea.  I can’t, however, think of a case where that has worked.  First time for everything, I guess.

(2) The problem with the extraction-coercion cycle is that it is not a simple process.  Actually, it isn’t a “process” as much  as a competition.  France was centralized at least in part in this manner, but there was no guarantee that France as we know it would emerge.  Instead of a France centered around Paris, it is easy to imagine scenarios in which the dominant force could have emerged out of Brittany or more likely Burgundy.  In short, even if Afghanistan does centralize through the impetus of warfare, it might be centralized around some power center other than Kabul.

Realistically, though, centralization is an unlikely outcome, so none of these models is particularly useful.  Instead, we should be thinking about Afghanistan as being — perhaps permanently — dominated by quasi-feudal political forms.

I’ll discuss some more modern models — particular Taiwan, South Korea, and Singapore — in a future post.  But those are not good models either for a variety of reasons.

Look, I get that this is complex stuff, and intelligent people can disagree.  But the problem with the COINdinista position is not that I disagree with their conclusions, it is that they demonstrate no understanding of any of the tricky issues revolving around the extension of political control and economic development. They don’t seem to have bothered to read any of the relevant literature.

UPDATE: The other issue I wanted to mention is that successful models of state-building include self-sustaining dynamics. Our current plans for Afghanistan do not.  Instead, every expansion of central authority increases the demands on the Afghan state. A state-building through largess approach becomes progressively more difficult as you reach out to areas where development and state control are weaker. This is precisely the opposite of successful processes which, once begun, often provide the resources needed for further consolidation.