What's Wrong With Weak States?

By Janine Davidson

Tom Johnson and M. Chris Mason have an excellent short piece, "All
Counterinsurgency Is Local
," in the latest Atlantic magazine.

They critique the NATO counterinsurgency strategy in Afghanistan for its ill-conceived emphasis on strengthening national-level governance and its disregard for the smaller districts, where the real center of gravity is for Afghan society.

Politically and strategically, the most important level of governance in Afghanistan is neither national nor regional nor provincial. Afghan identity is rooted in the woleswali: the districts within each province that are typically home to a single clan or tribe. Historically, unrest has always bubbled up from this stratum-whether against Alexander, the Victorian British, or the Soviet Union. Yet the woleswali are last, not first, in U.S. military and political strategy.

This is a simple, yet not-so-obvious observation. Despite headlines that emphasize military operations and chasing bad guys in Afghanistan and Iraq, at the end of the day counterinsurgency is about armed competition for governance. Thus, good counterinsurgency strategy should focus at the level of society where governance takes place.

In contrast, classic counterinsurgency theory, combined with the U.S. emphasis on democratization and mirror imaging, has led to a conflation of counterinsurgency with nation-building. The thinking goes that we need to help countries govern, by which we mean develop the capacity to carry out their obligations -- both internally with respect to their citizens and internationally with respect to other nations. The intervening force (us) can't leave until local systems are able to take on these key tasks of governance and resist further subversion and rebellions. Ideally, we'd like to leave behind a nation-state with which we can sustain diplomatic relations. Thus, we identify governance from the perspective of the Westphalian international system and a Weberian bureaucratic structure.

But Johnson and Mason make an interesting point. How good is this Westphalian/Weberian approach in a society that looks to local-level authorities for its basic needs?

Beyond Afghanistan, consider Somalia, which the international community has been trying to help transform into a proper nation-state for decades. As Ken Menkhaus, of Davidson College, points out in an excellent article for International Security, power elites at the local level may reject a stronger national-level system because they don't perceive it as in their best interest. Locals who look to these elites care more about their basic needs than whether or not they have a passport, or a national army, or are otherwise taxed for such things from a far away government. And so nationalization goes nowhere.

Likewise, the 2007 re-focus at the tribal level in Iraq has done more to quell the violence, and therefore promote human security, than have the efforts at the national level to build the Iraqi state and all its national-level ministries.

But the question we need to examine is about tradeoffs. What are we sacrificing from a national or international security perspective when we focus on human security at the local level, as Johnson and Mason suggest? Westphalia provides the foundation for the international legal structure we have today. The keystone to this system is the sovereign nation state. What might an international system with weaker nation-states look like?

Do we have to choose between strengthening the local over the national level systems? Is it possible to have both? And can we help build both simultaneously, or should we focus on the local level and then eventually aggregate efforts up to a national level?

In the United States, the founding fathers spent a great deal of time debating the need to unite under a strong national-level system. There were many who saw little to commend a national government that might undermine local (state) authority. Still, consensus formed around the idea that a weak confederation of states would be vulnerable as an international actor -- militarily and economically. And so the Constitution replaced the Articles of Confederation.

Johnson and Mason give us much to consider. They definitely make a strong case for shifting our focus to the local level in certain cases. But of course, each society is different; each conflict has its own logic. And there is an international system to consider in an increasingly globalized world. Is the best we can say, "It just depends"?

By Marisa Katz |  September 24, 2008; 12:00 AM ET
Previous: A Crisis of Competence | Next: LPC Marksmanship*


Please email us to report offensive comments.

Maybe much more local than you imagine:


And not just counterinsurgency. It appears much of our mission/problems over there now are counternarcotics. One does not always equal the other.

Well, they're all "evil-doers", I guess you can make that work. Even the ones that will be starving soon.

Quite a conundrum - how do you create "states" that can behave well in the global political and economic sandbox if there are two many local chiefs? So you foist some nationalistic effort, but not too much, lest they get any non-compliant bents. That's usually related to some outlier charismatic individual, and we have a good record for finding some local to take them out.

That said, foisting nationalism upon lines some Brit or Belgian or what-not drew 90 years ago in the sand is pan-generational. Why we're not serious about this "War". When I see republicans grandkids in Pashtun language elementary school classes, I'll start taking them seriously.

Politics and politics-by-other-means just won't work here. Only economics will.

Posted by: srv | September 24, 2008 11:33 AM

Loyalty has to come from the bottom up, it can't be imposed.

We have to start w/ improving local governance. In terms of the national institutions, they will come when the locals are ready for it. In the mean time, we can depend on the native Army and higher education to start the process of national identity. Many countries had no national identity beyond their armies and a titular gov't for a long time, and I expect Afgh can work with that as well.

Posted by: Jimmy American Mohist | September 24, 2008 4:09 PM

Intel-Dump, your link seems to be broken. You can reach the article at:


Posted by: Carl Prine | September 24, 2008 6:49 PM

A) The US system was based on a confederation of STRONG states that had well developed internal governments, constitutions and economies. And it still is.

B) For COIN to work we need to grow LOCAL support, governments, policing and economies. Only strong regional powers who are unchanged within their boundaries and who have functioning local economies have the excess capacity, time and energy to contribute to a central federation.

C) The model that the British used in India and else where was good. Letting local boundaries and tribes continue, with self rule and local governance so long as national laws were followed and individual rights were protected, (same as US court system), but with oversight and forces to be dispatched if things got out of hand. (See US civil rights, or election law enforcement, or British enforcement of laws against immolation.)

Vetnam is a good example of why you work on the local level and not through an unstable and infective national government that has little influence out side the capital, other than what it can purchase with money.

Posted by: jamesm2 | September 25, 2008 10:18 PM

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company