Terrorist Safe Havens: Weak States vs. Failed States

During last week’s Somalia discussions, I argued:

“1- Weak states support terrorism better than failed states- As Dr. Ken Menkhaus has noted many times, failed states like Somalia are hard for everyone. It doesn’t matter if your AQ or Western peacekeepers. The cost of operating in chaos makes terrorism tough.”

In the comments, Petr posted some counterarguments noting:

You are basing this statement on more or less intuitive logic and two case studies (Somalia 92-94, Kenya 92-98). But if you see the broader picture (i.e. more case studies) it gets more complicated.
- it is problematic, even though not unsolvable to treat those countries as one entity and then classify the strength of the statehood.
- in my research it came out that not the strength of a given state but presence of a strategic ally (or radical islamist subculture) is the key variable when it comes to success of al-Qa´ida. I do not want to bother you with details, but simply to say al-Ittihaad could not provide AQ with the safe haven, unlike ash-Shabaab, which in my understanding is a good case of al-Qa´ida success.

The ‘weak state vs. failed state’ debate is one of my favorites (this makes me incredibly dull by the way).  I originally went into the HOA research following the “failed states equals terrorism” equation.  Having read the Harmony documents, spent some time in Kenya, and did some further research, I came to agree strongly with the claim that weak states support terrorism better than failed states thereby following the Menkhaus doctrine :

The case of Somalia suggests that external observers may have been mistaken
in our assumptions about the relationship between terrorism and collapsed
states. The reality is that, at least up to now, transnational criminals and terrorists have found zones of complete state collapse to be relatively inhospitable territory out of which to operate. There are certainly exceptions – the fiefdoms of
drug-lords and radicals in parts of Colombia, for instance. But in general, terrorist networks have instead found safety in weak, corrupted, quasi-states –
Pakistan, Yemen, Kenya, the Philippines, Guinea, Indonesia. Terrorist networks,
like mafias, appear to flourish where states are governed badly rather than not at
all.

Here are my reasons and I’ll address Petr’s notes above.

  1. Case Studies- I actually don’t see my case studies as two fold and defined to the dates mentioned.  I use Somalia (Failed) and Kenya (Weak) as two case studies extending from 1992 to the present.  Somalia’s safe haven support for AQ through the present day has been relatively weak and chaotic.  Sure, AQ has smuggled weapons, done some financing, etc.  But, major terrorist attacks on the West stemming from Somalia have not occurred.  Instead, Kenya has hosted a string of terrorist attacks and provided safe haven for AQ terrorists throughout the past 18 years.  During ’92-’94, AQ members transited through Nairobi airports, drove along the coast, and trafficked through ports in Mombasa, Malindi and Lamu.  The embassy bombings (’98) provide an obvious example of AQ action in Kenya.  AQ members did travel in and out of Somalia from ’94-’98, but they actually lived in Nairobi where they ran front charities and targeted the embassy.  After a brief departure from HOA, Harun Fazul, AQ’s East African commander, didn’t move back to Somalia.  In 2002, Fazul settled near Lamu, Kenya, built his own mosque, developed a fishing business and prepared for two more plots in Kenya: the 2002 Paradise Hotel bombing and the failed SAM missile attack on an El Al jet departing Mombasa.  Fazul has been seen in and out of Kenya for almost 20 years.  He might briefly stop in Somalia to move arms try to influence local groups, etc.  But the longer he stays there, the greater the chance CT folks or a rival clan will identify and interdict him.  The best example of AQ’s freedom of movement in Kenya is seen in the confessions of Omar Said Omar while in Kenyan custody.
  2. Three party problem in Kenya- Petr’s comments above approach only the terrorist side of the failed state story. Weak states provide greater safe haven than failed states because they impede counterterrorists.  In Kenya, there are three parties: AQ, Western CT forces, and the Kenyan government.  Kenya’s weak capacity permits AQ operations and limits Western CT efforts.  In Kenya, Western CT forces can’t interdict AQ and its affiliates militarily, use drones, or build intelligence without restrictions.  Weak state sovereignty requires the U.S. to use partners.  In Somalia, Western CT forces can act without restraint.  Individual AQ training camps or AQ leaders can be targeted.  Indigenous militias can be co-opted to counter AQ.
  3. Predictable Graft is better than Chaotic Graft- In Kenya, AQ operatives can navigate corruption fairly well. Graft is routine and predictable.  Legitimate businesses and charities can be established to generate revenue and augment illicit funding.  In Somalia, AQ’s costs are variable.  As seen in the Somalia Harmony records, operating costs in an austere environment, void of any legitimate transportation and exchange mechanisms, quickly soared to unsustainable levels.  Clan leaders extracted rents haphazardly and often.  Rarely did these clan payments result in AQ accomplishing its goals.

I’ll stop with these three large reasons.  To clarify, I think AQ operates in both Somalia and Kenya.  But, similar weak state issues can be seen in Yemen, Pakistan, and the Sahel today.  Meanwhile, the U.S. has dismantled terrorist operations in Afghanistan and Iraq where it has freedom of movement and no weak state limitations.  Although Iraq may be entering a weak state era soon.

Lastly, Vahid Brown wrote a great biography of Harun Fazul which provides an excellent account of terrorists taking advantage of weak states.  Also, I encourage all interested in the “Failed vs. Weak” state debate to read Dr. Ken Menkhaus article in The Journal of Conflict Studies entitled “Quasi-State, Nation-Building, and Terrorist Safe Havens.”  He explains this much better than I.

I’m taking sometime off for Festivus but I’ll chime in later in the week reference new Somali clan alliances.

6 comments

  1. Fair points, but I am still not convinced. I do not claim to be right, but I present my argument as follows:
    - ad. 1) I agree with you that Somalia has not been used as a staging ground for an AQ terrorist attack, but this does not mean that AQ is not successful there. Ash-Shabaab (AS for short) is the most powerful (non-state) actor in the theatre and it has staged jihadist attacks beyond Somalia, AS from a certain (sic) point is AQ – all of the leaders of are AQ-trained, it has proclaimed the same goals and even to follow Ayman az-Zawahiri and Usama bin Ladin, the former even praised them for that act. Creating such an ally in Somalia a great success for AQ. And it is not a mere coincidence that AQ operatives are reported to run AS operations, for example Fazul Muhammad you´ve metioned (also abu Taha as-Sudani). It is single source reporting mostly, but there is some evidence.
    - ad. 2) we agree on this point, the formal sovereignty hinders some of CT operations. But you´ve metioned Pakistan as a weak state, yet in Waziristan drone attacks are conducted on weekly basis (Here a sub-state approach would probably help, it is usually some locations (North Waziristan, Mombasa, not the whole territory)).
    - ad. 3) certainly.
    I would put forward a slightly different approach it is not just the state strength (failed, weak, strong) which affects the equation but we need to insert two other factors – the willingness (I borrow this from Daniel Byman) to combat terrorism and that was precisely Kenyan´s problem till quite recently. Or UK is another strong case (al-Muhajiruna) till 2005, where the gov has reassessed its relations to radical islamists subcultures. Finally it is the presence of an ally who can protect AQ assets or to quite simply pursue AQs goals perhaps instead of it or with some sort of cooperation.

    I would thus aruge that AQ is most successful where the state strength is low, as its determination to fight it and where it can rely on a strong ally. It is not as neat as your argument and it is much more problematic to operationalize it, but I believe it captures the reality better.

    I hope I did not go for too long, if you want I can send you the argument to email, so you will have it in a more coherent way.

    • Here are a few more points.

      Ash-Shabaab (AS for short) is the most powerful (non-state) actor in the theatre and it has staged jihadist attacks beyond Somalia, AS from a certain (sic) point is AQ – all of the leaders of are AQ-trained, it has proclaimed the same goals and even to follow Ayman az-Zawahiri and Usama bin Ladin, the former even praised them for that act. Creating such an ally in Somalia a great success for AQ. And it is not a mere coincidence that AQ operatives are reported to run AS operations, for example Fazul Muhammad you´ve metioned (also abu Taha as-Sudani). It is single source reporting mostly, but there is some evidence.

      I don’t believe that AQ holds complete direction over Shabab. I think AS actions are the natural outcome of clan conflict. While the Uganda and Kenya AS attacks show growth, I’m not convinced this is directed by AQ. Zawahiri and Bin Laden have been calling for jihad in Somalia for years, but there still is not a massive influx of support beyond Somali Diaspora members. AQ’s other theaters have seen massive support of foreign fighters and funds, while Somalia has been a trickle. If AS were to really expand its operations, Western CT operations would accelerate quickly and unimpeded.

      But you´ve metioned Pakistan as a weak state, yet in Waziristan drone attacks are conducted on weekly basis (Here a sub-state approach would probably help, it is usually some locations (North Waziristan, Mombasa, not the whole territory)).

      This supports my argument. The U.S. has wanted to conduct direct military strikes into Waziristan for more than 8 years. The U.S. cannot because of Pakistani sovereignty. If the U.S. could attack in Waziristan directly, this might have ended long ago. See this NY Times article from this week, “U.S. Military Seeks to Expand Raids Inside Pakistan.”

      I would put forward a slightly different approach it is not just the state strength (failed, weak, strong) which affects theequation but we need to insert two other factors – the willingness (I borrow this from Daniel Byman) to combat terrorism and that was precisely Kenyan´s problem till quite recently. Or UK is another strong case (al-Muhajiruna) till 2005, where the gov has reassessed its relations to radical islamists subcultures. Finally it is the presence of an ally who can protect AQ assets or to quite simply pursue AQs goals perhaps instead of it or with some sort of cooperation.

      I understand you’re trying to get at the cultural support argument. But, I don’t think it fits as well in Somalia as other places. I also think you should examine why Kenya, Pakistan and other weak states might want to keep a low level of terrorism. If terrorism was solved in these weak states, funding from the U.S. would dry up quickly. Dr. Jacob Shapiro has studied a lot of this and I believe it helps explain why CT efforts can’t get to the finish line in places like Kenya and Pakistan. Counterterrorism is very lucrative for weak states. The day Bin Laden is caught, the Pakistani governments aid will get cut. Terrorists know this too! So there often is cooperation to keep the cultural support for terrorism going to keep the base of counterterrorism funding coming in.

  2. “I don’t believe that AQ holds complete direction over Shabab.”

    We probably slightly diverge on what we understand as AQ, I assume that for you it mostly stands for AQ core and those groups who signed formal cooperation, I prefer AQ as a social movement (social movement theory is very useful when analysing AQ, anyone who acts like AQ, have same goals and understand itself as part of the global jihad movement). I agree with you, that AS is probably not steered by AQ and the few operatives rather function as advisors an to radicalize the leaders, they are probably completely out of touch with the AQ core, this still does not mean that it is not success for AQ. But again, that is because we probably understand AQ differently.

    “I think AS actions are the natural outcome of clan conflict” I do not think so, this could be partially valid for al-Ittihaad, who was a completly different ideological based actor then its rivals at its time. But AS leaders has been in contact or at least influence of AQ from the very beggining, all trained in Afghanistan, there is nothing natural in takfiri jihadism in Somalia. Clan politics clearly play a very important point in the conflict, but Shabaab is not really just an Abdgal(Hawiye) thing.

    ” …This supports my argument. The U.S. has wanted to conduct direct military strikes into Waziristan for more than 8 years…” That is a good point, eventhough I understand drone strikes (which GoP denied till 2008) and the opertions of US special forces (see wikileaks – Islamabad) as a rather direct insertion of US power in Waziristan. But that does not invalidate your point, thats true.

    “The money effect” – that is in fact the factor of willingness to fight terrorism restated, but I would argue that it does not concern only weak states. The fact that UK was a jihadist hub and most of the jihadist attacks in Europe had some connection to the UK jihadist scene is quite telling. It is not always the lure of money, which make a state to be indecisive in CT. The problem of the money argument is that you can not proove what a state has for motivation, unless you have a very good access to the key decision makers. I prefer not to rely on intuitive arguments. In case of Pakistan I would argue it is not that much money, as power politics aiming at Afghanistan, but that would be a for a different discussion.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>