Daveed Gartenstein-Ross is a senior fellow at the Foundation for Defense of Democracies, and a Ph.D. candidate in world politics at the Catholic University of America. He is the author or volume editor of eleven books and monographs, including Bin Laden’s Legacy: Why We’re Still Losing the War on Terror. Chris Albon earned a Ph.D. in political science from the University of California at Davis.
In late September, African Union (AU) forces surrounded the port city of Kismayo on Somalia’s southern coast. The AU troops stood on the threshold of capturing the city from the al Qaeda-affiliated militant group al Shabaab, which began the year as the dominant military power in southern Somalia. On October 2, when AU troops entered Kismayo and claimed control of the city after a long standoff, Shabaab had lost its final stronghold.
It is worth noting, though, that the AU troops were hit by a bomb blast as they entered, which was Shabaab’s way of saying that it was still a force to be reckoned with. As the group’s spokesman Sheikh Abdiasis Abu Musab said, “This is only an introduction to the forthcoming explosions.” As one of us noted at the time, in Foreign Policy, with the loss of Kismayo Shabaab was returning to somewhat familiar territory:
Al-Shabab will try to repeat a maneuver that already proved successful once before in Somalia. Back in 2006, an Islamist coalition called the Islamic Courts Union (ICU), of which al-Shabab is an offshoot, controlled most of the key strategic points in southern Somalia and had encircled the U.N.-recognized Transitional Federal Government in the south-central city of Baidoa. The ICU governed according to its strict interpretation of Islamic law. It executed people for watching soccer matches and imposed a number of other draconian restrictions (though wasn’t as harsh as al-Shabab would later become). As the year neared its end, many observers expected the ICU to undertake an offensive to wipe out the transitional government’s final sanctuary. Instead, however, Ethiopia launched an invasion of Somalia that not only received the approval of the United States, but critical military support as well. Ethiopian troops entered the capital, Mogadishu, on Dec. 28, 2006, and quickly reversed virtually all the ICU’s strategic gains throughout the country before the end of January 2007.
The ICU promised an insurgency, and one soon gripped Somalia. Al-Shabab split from the ICU during this period and eventually was able to become the dominant force in the country’s south. By early 2011, the situation looked much like it did prior to the Ethiopian invasion. Just as a few Ethiopian troops protected the transitional government in Baidoa in 2006, all that stood between the Somali government and certain death at al-Shabab’s hands in 2011 was the protection of an African Union force composed of Ugandan and Burundian troops.
Although Shabaab has now lost Kismayo, it is hoping for a repeat of what occurred from 2007 through 2011: that the country’s transitional government requires a foreign force to prop it up, which serves as an irritant and gives rise to a powerful insurgency. Whether they are able to do so, of course, remains to be seen.
Since African Union forces surrounded Kismayo, we have kept a database detailing every publicly-reported attack known or suspected of being carried out by the group and its sympathizers. This database runs from September 30, 2012, through December 5, 2012, covering a total of 68 attacks. In this article, we map the early part of Shabaab’s attempted post-Kismayo insurgency by providing a visualization of this data.
Overall, by our count 144 people have been killed in these attacks and 300 wounded. See Figure One for casualties caused by these attacks.

Further breaking down these figures, 125 were killed in Somalia and 19 in Kenya; 227 were wounded in Somalia, and 73 in Kenya. Our figures almost certainly underestimate the overall numbers of killed and wounded in Shabaab-related attacks (though perhaps not in Kenya, as we will explain subsequently), since we measure only attacks that were reported in the press. Not only do some attacks in Somalia go unreported, but also in many cases the press reporting did not include numbers of killed or wounded, and thus we could not add concrete numbers to our database.
The reason that the reporting on Kenya may not be an underestimation is because the responsibility of Shabaab or its sympathizers for attacks around Nairobi is often suspected but not known for a fact. Our database employs four different categories of attack perpetrators: known Shabaab, suspected Shabaab, known sympathizers, and suspected sympathizers. Of the thirteen attacks in Kenya, four were carried out by suspected al Shabaab, while nine were carried out by suspected sympathizers.
For example, on September 30, AFP reported on a grenade attack at a Nairobi church that killed a child and wounded nine other people. The article noted that the blast, which triggered reprisal attacks against Somalis, “came a day after Islamist Shebab fighters abandoned their last bastion in neighbouring Somalia in the face of an assault by Kenyan and other troops.” AFP reported the statement of Wilfred Mbithi, head of police operations in Nairobi, that witnesses “saw two men of Somali origin running towards the back of the church where the explosion occurred.” The article further notes:
No one has yet claimed responsibility for the church attack, the latest in a string of grenade attacks, shootings and bomb blasts that have rocked Kenya since it sent troops into southern Somalia in October 2011 to crush bases of Al-Qaeda-linked Shebab fighters.
On Saturday, the Shebab retreated from their last stronghold in Somalia, leaving the southern port city of Kismayo that has been a vital economic lifeline for the Islamists.
Shebab spokesman Ali Mohamud Rage had warned that the militia would remain a threat.
All of this means that the article is strongly implying that Shabaab sympathizers are the likely perpetrators of the attack; but without a claim of responsibility, that cannot be known for sure.
Figure One makes clear that there is no discernible trend in terms of casualties caused by Shabaab attacks: they cannot be shown to be either increasing or decreasing over time.
We also measured attacks by target in
Figure Two.
The overwhelming majority of attacks were against military targets. These include attacks against AU, Kenyan, and Somali forces. Civilians were the second most popular target, followed by government and police. There were also two attacks on churches, both of which occurred in Kenya (in Nairobi and Garissa), and both of which employed grenades.
We measured attacks by type in Figure Three.

The three most popular attack types by Shabaab and its sympathizers were IED attacks, ambushes, and shootings. Shabaab has also prominently employed grenade attacks, assassinations, and massed attacks. Only one suicide attack was found in this sample, a double suicide bombing at a Mogadishu restaurant in early November. However, there were also unconfirmed reports that pro-government forces captured a would-be female suicide bomber in late October as she was on her way to carry out an attack at “a busy location in Kismayo.”
We will continue to monitor the shape of Shabaab’s post-Kismayo fight against Somalia’s government, and may publish an update at some future point.
Marisa Porges has a forceful op-ed in today's NYT making the case for beefing up the capture component of U.S. counterterrorism efforts. Read the whole thing:
At the moment, the United States has nowhere to hold and interrogate newly captured terrorists. America just handed over control of its detention facility at Bagram Air Base in Afghanistan, a significant step toward transferring security operations to Afghans. And while Guantánamo Bay remains home to nearly 170 men that the United States believes are still a threat, no captured terrorist has been transferred there since August 2008. Yet in the past four years, drone strikes and airstrikes targeting Al Qaeda affiliates in Pakistan, Yemen and Somalia have increased dramatically.
Since 2010, there have been about 2,000 such strikes in Pakistan alone, with hundreds more in Yemen and North Africa. Meanwhile, only one alleged terrorist outside of Afghanistan — a Somali named Ahmed Abdulkadir Warsame — was captured, held and interrogated. He was later flown to New York to stand trial.
....
The fact that the United States now has nowhere to hold a terrorist — and no policy to deal with him once captured — means that a dangerous suspect might very well be let go. At present, there is no standard course of action approved by the president and relevant government agencies for what to do in the days and months following capture.
This situation creates disturbing incentives for troops on the battlefield. It encourages soldiers and policy makers in Washington to opt for the “five-cent solution” — a bullet. Rather than shooting people, we should be exercising due process, and bringing transnational terrorists to justice. That’s an approach that would help America maintain the moral high ground in the ongoing fight against Al Qaeda.
That there needs to be more human intelligence collection in U.S. CT is beyond dispute. So too are the issues currently wracking American detention policy. Warsame, for example, spent a good deal of time onboard the USS Boxer, which Spencer Ackerman fairly described as a "floating Gitmo" when put to this use. But that's not the worst of it. Warsame was lucky enough to make it to a U.S. courtroom, but as Jeremy Scahill has documented, Somalia's NSA and the CIA run some dreadful sounding facilities where not just fighters found in Somalia, but alleged terrorists from Kenya face interrogation and detention. If U.S. detention policy ultimately ends up relying on building Bagrams and Guantanamos across AFRICOM and CENTCOM, or else employing the U.S. Navy to this end, counterterrorism with a human face might not turn out all it's cracked up to be.
But even leaving aside navigating the legal and logistical issue of where to put terrorists once we capture them - an issue that Porges readily acknowledges - there is an issue of how to bring back warm bodies from where we currently have drones buzzing overhead. This is a critical question, because the means the U.S. employs to capture terrorists and suspected terrorists will have a great impact on the costs, benefits, and relative merits and demerits of capturing HVTs as opposed to the current targeted killing campaign.
In Afghanistan, the massive conventional presence of U.S. forces was and still is a significant enabler for capture operations. Afghanistan's infamous night raids, now under the control of the Afghan military or specialized CIA-trained elements, are a prime example. Yet many familiar issues emerged. Civilians resented property damage, casualties, mistaken targets, lack of transparency or accountable due process, and increasing the role of the ANSF may not have significantly improved the situation.
In many respects though, Afghan night raids are easy. Special operations enjoy significant legal and operational freedom of movement. Large amounts of on-the-ground intelligence and conventional forces enable better targeting and mitigate the risks of raids. Try to pick up targets in, say, Somalia, and things get much harder. JSOC raids into that country required air and naval fire support, while the enabling conventional force in question was the Ethiopian military, which did not do much to win Somali hearts or minds. Penetrating Somalia has required a patchwork of often unsavory partner military forces, militia proxies, private contractors, and covert operations. While America has learned much from 1993's most infamous attempt to conduct HVT capture, its foes in Somalia continue to pose stiff security challenges - though fortunately Shabaab seems to be losing ground.
In Yemen, the U.S. has a number of options for conducting capture operations, none of them particularly appealing. It can rely on Yemen's government and U.S.-trained troops, whose political loyalty and human rights credentials are not great. Though drone strikes are destructive, so are smash-and-grab expeditions into ungoverned or hostile space, particularly with a partner state's less, well, delicate touch (this is the country that named its counterinsurgency against the Houthis Operation Scorched Earth, after all). While we should always remember that U.S. airstrikes - manned or unmanned - rely on significant theater basing and local covert ground presence, capture missions would likely increase the footprint of U.S. operations. In Yemen, geography is favorable enough to allow sea-based raiding, but maintaining raids at the tempo of drone strikes would likely mean a vastly expanded U.S. military presence in the area. Or else it might rely on the Yemeni government, the prisons of which helped radicalize an earlier generation of al Qaeda.
In Pakistan's Federally Administered Tribal Areas, too, we see similarly pressing problems. As C. Christine Fair rightfully points out, it's been Pakistani conventional offensives (which would provide the presumed enabling element for an increased tempo of capture raids) that have done the most damage and displacement to the region's population. A law enforcement approach's outlook is bleak because the entire region, whether the U.S. likes it or not, falls under the Frontier Crimes Regulation, a colonial piece of legislation which makes playing by host rules and occupying the moral high ground an ethical gymnastics act. While Pakistan is willing to tolerate, to some extent, drone strikes on its soil, will it be so willing to replace them with more cross-border activity from JSOC, the CIA-trained Counterterrorism Pursuit Team, or let the U.S. direct its own security forces to a degree amenable to U.S. interests?
In areas where government capacity is strong and politically pliant, using the FBI to capture terrorist suspects will likely remain viable. When the U.S. tried to capture the 1993 CIA headquarter shooter, Aimal Kasi, the FBI worked with the Pakistani government to render him to the U.S. But they could not capture him until he entered Punjab province, and even then the U.S. initially hid the extent of Pakistani government involvement due to the controversy of the extradition. This was one arrest in 1997 - conducting arrests and renditions at a high tempo today simultaneously demands a much larger host government role while straining the political space for it to participate.
All this said, on balance the U.S. still must reorient its HVT program towards collecting HUMINT. For pragmatic and ethical reasons, the U.S. also must do something to fix the current legal and logistical morass of its detention policy. Yet assessing the proper role of capturing terrorists, and the likely degree of practical and moral surplus derived from it, demands a frank assessment about the demands of substituting captures for kills, and the capacity and willpower of the U.S. to undertake such operations. Even with the legal problems sorted out, and a system of prisons without the lingering insidious reputation of Guantanamo, Bagram, or CIA black sites, we still have the matter of kicking down the doors of suspected terrorists in well-armed and unfriendly neighborhoods and spiriting them away to a host or foreign prison. This is a process that will still likely get civilians killed, families unjustly torn apart, and put armed men and military hardware in places where they are not wanted. Dealing in such generalities, it is extremely hard to say whether this would appear, to the broader population, more moral, more desirable, or less encouraging of radicalism than drone strikes, in part because it is already so difficult to accurately measure very much about drone strikes in these regions to begin with.
Just look at the Phoenix Program, the massive effort to capture suspected foes in Vietnam to dismantle VC infrastructure. As William Rosenau and Austin Long explain in their invaluable report on its relevance for modern operations, the Phoenix Program unduly gained a lasting reputation as an "assassination" campaign of marauding "death squads" - a reputation so widespread that even President Nixon thought this was what the CIA-handled Provincial Reconnaissance Units were really aiming for. Whether using local governments, proxy forces, special operations, or some other element, snatching somebody from their home at night at gunpoint is a risky proposition for seeking political kudos. Particularly when placed alongside host governments that engage in disappearing opponents, brutal methods of counterinsurgency, and generally repressive practices, the perceptual and counter-radicalization benefits of a similar-tempo capture campaign might rapidly wane.
Doubtlessly, expecting all of this from an op-ed is a curmudgeon's (and a blogger's) game, but shifting the frame somewhat is necessary from a policy perspective. We must at least broach the question of what kind of force commitments and operational guidelines we need to effectively conduct a capture campaign is essential, as well as when and where we ought to employ such means. While it's undeniable the HUMINT value of capture operations are higher, the costs of undertaking them may well reduce or even eliminate the presumed ancillary benefits.
As is wont to happen, the current forms of warfare the United States in engaging in and preparing for lend themselves easily to misrepresentation and simplification. As the U.S. appears to wind down the era of large scale U.S.-led land operations, particularly ones in which the U.S. is bearing the brunt of combat against insurgencies, the new form of U.S. operations against non-state actors has unsurprisingly been described in terms such as drone wars or components of an offshore or counterterrorism strategy, while conventional platforms and capabilities are viewed in reference to the apparent "pivot to Asia" and AirSea Battle. However, recent events in Yemen demonstrate that such these sorts of small war operations, while they have a significant covert component and often involve the use of remotely piloted aircraft, also involve boots on the ground and the use of what are often conceived of as conventional military platforms such as naval and aerial ISR and strike assets.
On Sunday, a U.S. special forces trainer embedded with Yemeni troops suffered a serious combat wound. This came on the heels of a LA Times piece last week which noted that several dozen U.S. military personnel were on the ground in Yemen embedded with Yemeni forces and assisting with targeting for U.S. strike capabilities. Also notable is a recent story by David Axe highlighting the work of bloggers who have publicized the presence of a unit of F-15Es based out of Djibouti. All of these developments reveal some uncomfortable truths about the nature of U.S. power projection.
Firstly, while drones are undoubtedly a significant portion of U.S. operations in its counterterrorism campaigns across the Indian Ocean rim, they are but one platform in one prong of the effort. The phrase "drone war" might be accurate for Pakistan, where by tacit agreement with the Pakistani government and military the U.S. has restricted its strike operations to drone activities. But even in Pakistan the presence of proxy forces such as the Counterterrorism Pursuit Teams and significant amounts of on-the-ground CIA and JSOC personnel supporting targeting are a major portion of the campaign. In Yemen and Somalia, where the U.S. has more flexibility diplomatically and geographically, expansions in strike campaigns have meant more U.S. forces operating on the ground, as well as the use of manned aircraft and naval vessels.
Despite the hype, even in politically sensitive counterterrorism options drones remain but one instrument in the U.S. arsenal. The primary perceived advantage of drones, that they keep service members out of harm's way, is really not a significant concern in these theaters. Pakistan is not in the habit of seriously defending its western airspace, nor does Yemen or AQAP have the will or capability of imposing significant military obstacles to U.S. standoff strikes. All these sentiments hold even more true for Somalia. Indeed, as the crash of the U-28 in Djibouti recently demonstrated, pilots of fixed-wing aircraft are at a greater risk for accidents than they are of being shot down in the Indian Ocean "shadow wars." And indeed, manned aircraft are frequently employed - AC-130 gunships were frequent features of U.S. operations in Somalia, along with helicopter gunshipsand JSOC assets. So too has the U.S. used cruise missiles to strike targets in Yemen and naval gunfire to support U.S. special operations on the ground in Somalia.
Ultimately, to a non-state threat in a conventionally permissive environment, a submarine, frigate, Strike Eagle or an AC-130 gunship is just as invulnerable as a drone, and offer a variety of other strike options drones cannot provide. Additionally, the use of manned aircraft such as the F-15E has likely allowed the U.S. to conduct mysterious airstrikes in support of Kenya's "Linda Nchi" incursion last fall, or conduct airstrikes which seemed implausible for Yemen's organizationally beleaguered air force. Nor are they necessarily significantly more costly. As Winslow Wheeler pointed out in an excellent series on the MQ-9 Reaper, when the cost of the additional infrastructure on the ground necessary to operate drones is factored in, similar overall maintenance costs make them fiscally competitive with many manned platforms.
Secondly, offshore strike campaigns are simply one prong of these so-called shadow wars. Supporting offshore strike, of course, are personnel on the ground, and often ones acting in support of foreign partner or proxy forces. As Daveed Gartenstein-Ross and I noted earlier, in Somalia, for example, strike campaigns are simply one component of a larger effort that involved supporting Somali intelligence, a proxy war using Somali armed groups, and supporting partner nation counterinsurgency efforts. All of the thorny dilemmas of dealing with COIN remain, even if most of the blood price of grappling with them is passed on to foreign soldiers. While it may take longer for the U.S. to embroil large formations of conventional forces directly into another insurgency or civil war, the dilemmas of how to balance U.S. strike campaigns without endangering counterinsurgency efforts still remain.
Additionally, any change in one prong of the strategy necessarily effects the importance or execution of the others. For example, in controversy about Yemen, the issue of signature strikes has prompted understandable concern. However, improving intelligence products contributing to offshore strikes requires greater resources being put towards clandestine intelligence or special operations units acting on the ground, or towards greater work with partner nation intelligence services. However, partner nations often have their own agendas, and particularly in an era when the U.S. is at least publically reticent to work with strongmen and unsavory regimes for the sake of "stability," trading strikes for greater support of regime security and intelligence services may end up having even more debilitating - and long lasting - effects.
While offshore strikes in and of themselves are far less costly and resource intensive than large formations operating in land campaigns, they are not all that cheap or small, and they exist only as part of a larger constellation of programs to feed intelligence and address partner nation concerns. Counterterrorism is no more a strategy than is counterinsurgency. It is a capability, a set of tactical building blocks, aimed at political objectives which, more often than not, require the employment of other capabilities in a broader war strategy.
The new U.S. force posture does not mean the death of U.S. support of COIN campaigns, but it does change their character. The U.S. discomfort with the Arab Spring's outbreak in Yemen is emblematic of this reality. Because the conception of counterterrorism required working with partner nations, it required accepting regime stability, and accepting regime stability required involvement in the regime's efforts to maintain power against peaceful and violent attempts to overthrow it.
Thirdly, the prosecution of multi-pronged military operations from an offshore and covert posture is a reminder that rhetoric about AirSea Battle and the Pivot to Asia notwithstanding, conventional military forces will remain major assets in combat against irregular assets, and that these conflicts will continue to rage outside PACOM, political rhetoric notwithstanding. This should not, of course, be all that surprising. Hopes that reductions to land forces would somehow starve the beast and reduce the U.S. appetite for waging wars against irregular threats were obviously misplaced, and indeed the Asian "Pivot" really serves to increase U.S. freedom of action for prosecuting such campaigns by providing a host of platforms capable of projecting power into navally-accessible regions and reducing the level of political attention and controversy surrounding the Middle East and American activities there.
The supposedly offshore shadow wars in fact involve major ground operations by partners, an active ground presence by the U.S., and large amounts of conventional military assets. Rhetoric and planning aside, for the near future the U.S. will likely remain militarily engaged in and around the Africa and the West Eurasian rimlands against irregular foes. These operations will likely also likely involve greater degrees of ground combat troops in the future. As Brett Friedman explains in an excellent post at the Marine Corps Gazette, the USMC will likely take a larger role in these small wars, as it did during the early 20th century. A survey of British history also demonstrates that despite the portrayal of Britain as dispassionately concerned with offshore balancing, it frequently engaged in onshore warfare across its colonial area of interest.
Ultimately, fixation on specific platforms or their elemental nature (land, sea, air), or specific prongs and their shorthand typology (counterterrorism, counterinsurgency, etc) obscures more about the nature of these conflicts than they reveal. Ignoring this weakens quality and utility of the debate the debate for both proponents and detractors of America's present military undertakings.
I apologize for not writing on the blog this week. I have a lot of posts in my head but have been busy with other activities -- and writing for other websites.
Judah Grunstein of the World Politics Review commissioned a debate over the future of counterinsurgency and got responses from Steven Metz, Bing West, Michael Mazarr, and Starbuck. I contributed a piece arguing that the debate really misses the larger issue of what the hell our ground forces -- and especially our army -- are supposed to do.
Judah wryly noted that my article was itself textbook counterinsurgency: "Redefine the center of gravity (not COIN, but US Army); secure it from unnecessary collateral damage of kinetic ops; and construct narrative to encourage buy-in from on-the-fencers."
Anyway, I am always proud to participate in such debates, especially with other thinkers I very much admire. My article bears strong resemblance to a talk I gave earlier in the week at the Council on Foreign Relations in New York in which I did my best to occupy the middle ground on a panel discussion with Gian Gentile and Max Boot.
***
On another note, SEAL Team 6 is doing its very best to make the president immune to Republican attacks that he's Jimmy Carter. My analysis of the hostage rescue operation in Somalia can be read on the website of the BBC.
MOGADISHU, Somalia (Reuters) — Dutch commandos freed 20 Yemeni hostages on Saturday and briefly detained seven pirates who had forced the Yemenis to join them in attacking vessels in the Gulf of Aden, NATO officials said.
The Dutch forces, operating under a NATO antipiracy mission, then released the pirates, a NATO commander said, because NATO has no “detainment policy.”
Meanwhile, gunmen from Somalia seized a Belgian-registered ship and its 10 crew members farther south in the Indian Ocean. A pirate spokesman said the vessel, the Pompei, would be taken to the coast.
Somali sea gangs have captured dozens of ships, taken hundreds of sailors prisoner and made off with millions of dollars in ransoms despite the presence of foreign warships in waters off the Horn of Africa.
Lt. Cmdr. Alexandre Fernandes of NATO said the 20 Yemeni fishermen were rescued after a Dutch Navy frigate on a NATO patrol responded to an assault on a Greek-owned tanker. The tanker had been attacked by pirates firing assault rifles and grenades.
Commandos from the Dutch ship chased the pirates, who were on a small skiff, back to their mother ship, a hijacked Yemeni fishing dhow.
“We have freed the hostages, we have freed the dhow and we have seized the weapons,” Commander Fernandes said, speaking on board the Portuguese warship Corte-Real. “The pirates did not fight, and no gunfire was exchanged.” The Corte-Real is also on a NATO antipiracy mission.
We need to deal with this problem from the beach side, in concert with the ocean side, but we don’t have an embassy in Somalia and limited, ineffective intelligence operations. We need to work in Somalia and in Lebanon, where a lot of the ransom money has changed hands. But our operations in Lebanon are a joke, and we have no presence at all in Somalia.
The operation to rescue Capt. Richard Phillips involved dozens of Navy SEALs, who parachuted from an aircraft into the scene near dark Saturday, landing in the ocean. The SEALs were part of a larger group of Special Operations Forces involved in the effort, according to military officials.I do not have the time to explain the training, missions, and capabilities of our nation's special operations forces. To even those without a security clearance or any relevant military or policy background, the value of these forces should be gobsmackingly obvious. And anyone who has closely read what I have written knows that I -- far from being "obsessed" with special operations forces -- have been quite critical about their employment in operations in Iraq and Afghanistan. This criticism is based on both personal experience and a careful study of policies and operations.U.S. military observers believed that Phillips was about to be shot. SEAL snipers, who were positioned on a deck at the stern of the Bainbridge, an area known as the fantail, had the three pirates in their sights. The on-scene commander gave the SEAL snipers authority to fire.
"As soon as the snipers had a clear shot at the guy who had the rifle, they shot him and the other two in the hatches," said the senior military official.
A member of the Special Operations team slid down the tow line into the water and climbed aboard the lifeboat. Phillips was then put in a small craft and taken to the Bainbridge.