February 1, 2010  

An alternative to COIN

It’s time to adapt our security strategy to leverage America’s conventional strengths

The U.S. military is a dominant fighting force, capable of rapid global power projection and able to defeat state adversaries quickly and at relatively low cost in American lives and treasure. Unfortunately, American leaders are increasingly trying to transform this force into one optimized for counterinsurgency missions and long-term military occupations. A fundamental problem with the adoption of population-centric counterinsurgency (COIN) doctrine as an organizing principle for American military operations is that it systematically fails to take advantage of the real strengths of the U.S. military.

It is true that not all political goals are achievable through the use of conventional military capabilities. However, “victory” in war is not dichotomous, and the cases of Iraq and Afghanistan — often seen as proving the necessity for COIN-capable forces as well as a commitment to nation-building — demonstrate in reality that the vast majority of goals can be accomplished through quick, decisive military operations. Not all political goals are achievable this way, but most are and those that cannot be achieved through conventional operations likely cannot be achieved by the application of even the most sophisticated counterinsurgency doctrine either.

As a consequence, I believe the U.S. should adopt a national military strategy that heavily leverages the core capability to break states and target and destroy fixed assets, iteratively if necessary. Such a strategy — which might loosely be termed “repetitive raiding” — could defeat and disrupt most potential threats the U.S. faces. While America’s adversaries may prefer to engage the U.S. using asymmetric strategies, there is no reason that the U.S. should agree to fight on these terms.

This essay argues the U.S. can largely defeat threats using conventional capabilities, and that what encourages a desire to engage in long-drawn-out asymmetric conflicts is not the elimination of threats, but rather the unattainable goal of trying to prevent threats from emerging in the future.

To demonstrate these points, I will first consider the nature of U.S. military power. Then I will identify the key war aims for the U.S. in Iraq, graph out when the aims were achieved, and show that the marginal expenditure of lives and resources to achieve maximalist goals is disproportionately large. I will further argue that the U.S. could have mitigated the negative consequences of potential developments in Iraq much more cheaply by embracing a willingness to reintervene rather than engaging in a drawn-out occupation. Finally, I will argue that in Afghanistan the Obama administration is likely to similarly pay too much to achieve at best marginal improvements in the current situation. In short, because of the nature of American power, the vast majority of the benefits from conflict come early and relatively cheaply, whereas the pursuit of additional benefits is increasingly costly and subject to diminishing returns.

A fundamental challenge in devising a strategy for the use of American military power is that the world has literally never seen anything like it. The U.S. today has military capabilities at least equal to the rest of the world combined. There is virtually no spot on the globe that could not be targeted by American forces, and at most a small handful of countries that could thwart a determined U.S. effort at regime change — and some of those only by virtue of their possession of nuclear weapons.

American military capabilities are not a potential form of power, subject to use only following a lengthy mobilizing and requiring a long campaign to achieve significant goals. Instead, the U.S. can destroy fixed locations in a matter of hours or at most days, and implement regime change in a matter of weeks or a few months.

Because this capability is so novel — dating only to the end of the Cold War — American strategists lack a clear framework to guide the utilization of this force. They have sought to match capabilities to conceptions of the use of force from a different era, one in which the Cold War made regime change unpalatable due to the risk of escalation and that tended to make localized setbacks appear as loses in a perceived zero-sum competition with the Soviets.

The reason, in other words, that the U.S. didn’t simply remove Fidel Castro from power was that after 1962, the international consequences seemed too high and the goal too risky. The reason American leaders felt compelled to engage in a lengthy counterinsurgency in Vietnam was the concern that a communist victory would have been a setback in the broader struggle. But imagine a world in which there were few or no international consequences to removing Castro from power, and imagine a world in which the commitment to Vietnam was strictly commensurate to the threat that the Vietnamese communists could pose to the U.S. That is the change in context that has occurred over the past 20 years, and the U.S. has not yet adapted.

Phrasing it another way, insurgents with small arms and homemade explosives cannot harm the U.S. and there is no reason to fight them directly. True, these types of forces pose a terror threat. However, the American experience in Iraq and Afghanistan — where insurgents have been able to build and deploy more than 80,000 IEDs while under occupation — calls into question the ability of occupying forces to root out terror networks. There is also compelling evidence that some terrorist networks, notably groups that carry out suicide attacks, are the result of military occupations rather than contained by such deployments.

However, significant threats to the U.S., ranging from the military capacity of regional powers to weapons of mass destruction development programs to significant terrorist infrastructures, can be targeted and destroyed by conventional military capabilities.

The gap, in short, between the threats the U.S. actually faces and the capabilities the American military possesses is much smaller than many suggest. The U.S. is not being forced to fight irregular conflicts. The U.S. is choosing to do so, deliberately fighting the kinds of wars America’s adversaries want to fight. While the American military is a tremendously potent force, even it cannot win by allowing its enemies to choose how, when and where conflicts will occur. Asymmetric “threats” are self-limited and should be largely ignored, and by doing so, the U.S. could free itself to apply military power more effectively in the types of conflict where American forces excel.


This argument can be demonstrated with a careful consideration of the war in Iraq. The U.S. had 12 distinct war aims during the Iraq War (Chart 1). They ranged from ending — or more accurately documenting the end of — Iraq’s efforts to acquire weapons of mass destruction all the way to the aspirational goal of “transforming” the Middle East by providing a model of democratic governance in opposition to authoritarianism and Islamism. Several of these goals emerged after the war began, and some after the occupation began to sour, but there is a great deal of debate about which goals were extant at any given time, and no useful process to distinguish between prewar and post-initial goals.

Many of these goals lack clear metrics to judge success, and as a consequence there may be some debate over whether and when they were actually accomplished. I have sought to be generous in my assessment of successes. A more pessimistic judgment about, say, our goal of containing Iranian influence would only strengthen my argument.

Furthermore, I have decided not to assign weights to the achievement of various goals, though clearly, in the run-up to the war, ensuring that Iraq stopped supporting terrorism and was blocked from pursuing weapons of mass destruction were considered much more significant than ensuring that Sunni, Shiites and Kurds could agree on a common constitutional framework. Again, my decision not to weigh the war aims provides the weakest case possible for my argument and the strongest for potential opponents.

What should be clear is that the majority of war aims were accomplished in Iraq within weeks of the beginning of the war (Chart 2). By the end of 2003, the U.S. had removed Saddam Hussein from power and assured that the Baathist regime — personalized as an incipient dynasty as in Syria — could not be restored. The U.S. had ended Iraq’s support for terrorism — largely Palestinian terrorism, but terrorism nonetheless — and had assured that Iraq could not threaten its neighbors. The intervention had also ended Saddam’s human rights abuses, and by getting the U.N. sanctions on Iraq lifted had removed another significant source of human suffering in the country.

All of this was accomplished at the cost of 486 American combat deaths and $53 billion (Charts 3 and 4). In short, at least 58 percent of the total goals were accomplished at the cost of roughly 11 percent of all fatalities suffered and 7.7 percent of total expenditures (through 2009.)

The marginal costs increased dramatically from that point forward. By the end of 2005, the U.S. had accomplished 69 percent of its war aims — a marginal improvement of 12 percent, but at the marginal cost of 1,695 American lives and $161.4 billion (Chart 5). At present, the U.S. has achieved — again using optimistic estimates — 88 percent of extant war aims, a marginal improvement of an additional 19 percent, but at the cost of an additional 2,149 lives and $469.6 billion.

The difference is between the types of operations. In 2003, American forces initially launched a conventional invasion of Iraq and sought to defeat regime elements and seize control of the country. From 2004 on, the occupation gradually fell into a quagmire of trying to achieve difficult and fuzzy goals using military tactics that fail to take advantage of areas of American comparative advantage in warfare.

The fundamental tragedy is that not only did American forces accomplish 58 percent of all goals in 2003, they accomplished those goals that were best supported by the American public. By the end of 2003, Saddam was out of power, Iraq was no longer supporting terrorism and the U.S. had ensured that Iraq would not use weapons of mass destruction against the U.S. or transfer them to state or nonstate enemies. By the end of 2003, the U.S. had won not just the majority of what American leaders ever hoped to win, but the vast majority of what the American public cared to accomplish. And it was done at less than one-tenth the total cost of the war.

Aside from trying to achieving increasingly more indirect war aims, a major factor driving the continued involvement in Iraq has been an attempt to shape the future of the country. The great fear was that an early withdrawal would lead to chaos, including, most significantly, the risk of state collapse, seizure of power by radicals, regional war or genocidal violence. Arguing that these outcomes were unlikely — though true — is insufficient. The more trenchant critique is that the U.S. paid much too much to insure against those outcomes because instead of relying on the ability to mitigate the consequences cheaply and at worst the capacity to re-intervene with decisive force in the event of severe security threats, the U.S. used costly and inefficient nation-building techniques to try to prevent them from occurring at all.

In order to demonstrate this point, we must stray into the realm of counterfactuals, but as with all speculation in this essay, I will argue from the worst-case assumptions for my argument.


The key questions we need to answer to understand the issue of potential state collapse in Iraq are: What would state collapse in Iraq have looked like? How much of a threat would it have posed to our interests? And how costly would it have been to eliminate the threat if it did materialize rather than seeking through prolonged intervention to avoid it outright?

It is difficult to conceive of outright state collapse in the case of Iraq because there were strong foundations of public order in place that did not require any intervention. It is wholly implausible that anything resembling collapse would occur in the Kurdish areas because they had been accorded tremendous autonomy even under Saddam due to protection. So those areas already had existing a civil society, a functioning security apparatus and established leadership. Similarly, though less clearly, the Shiite-dominated parts of the country were already under the influence of several powerful leaders, with strong tribal support, religious legitimacy and militias. We may have seen some fighting over turf, but because of limited power projection capabilities, it is likely that any such fighting would have been localized and small scale. State collapse, then, would most significantly affect the “Sunni Triangle” as well as the western province of Anbar and would create an ungoverned space surrounded by relatively effective buffer nations.

The scope of the problem can be assessed by considering the challenges posed by “failed states” such as the Palestinian Territories and Somalia. There are about 4.5 million Sunni Arabs in Iraq, which is only slightly more than the Palestinian population of Gaza and the West Bank. Somalia has a population twice as large. The challenge of an ungoverned space in the Sunni Arab areas of Iraq is bounded by challenges posed by Gaza and the West Bank on one hand and Somalia on the other.

Managing state collapse in those two regions is costly for the international community. Aid to the Palestinian territories totaled $836 million and to Somalia totaled $256 million in 2008. Both pose security challenges as well. The Israelis spend about $13 billion per year on defense, though clearly not all of that is in response to threats from Gaza and the West Bank, since Israel maintains a potent air force and a small but capable coastal navy. At most, half of that can be attributed to the direct costs of mitigating threats from the Palestinian territories. Somalia’s neighbors — Djibouti, Ethiopia, and Kenya — spend somewhere between $650 million and $1 billion per year on defense. An even smaller percentage can be attributed to threats from Somalia. Off the coast, piracy is estimated to have cost the world $60 million to $70 million in 2008.

Based on that, we can estimate the cost of mitigating state failure in the Sunni Triangle and Anbar to be on the order of $3 billion to $4 billion a year, mostly in transfer payment to Iraqi, Kurdish, Syrian, Jordanian and Saudi authorities, with perhaps an addition $1 billion in support for humanitarian assistance funneled through international organizations and nongovernmental organizations. For $5 billion a year — and no loss of American lives — the U.S. could have managed the collapse of governing authority in the Sunni Arab regions of Iraq. The actual cost, in short, was on the order of 50 times more to prevent a problem than the problem itself would have cost to mitigate had it occurred.

It is possible that terrorist groups might have been able to establish themselves in this sort of scenario. But terrorist groups are established in all sorts of places, from Colombia to Hamburg to London to Karachi. What they don’t have are large training camps, but large training camps were able to exist in Afghanistan not because they were impossible to target, but because the U.S. lacked the political will to do so on a consistent basis in the face of uncertain intelligence and the risk of collateral damage. Since 9/11, the policy consensus has gone completely overboard. Now, not only do American leaders have the will to eliminate any significant terrorist apparatus, especially one that would be protected with, at best, poorly armed militias, but American leaders have developed the compulsion to try to prevent even the possibility of such camps emerging by transforming countries such as Afghanistan into stable bastions of good governance, thus theoretically eliminating “upstream” factors like poverty. It is overkill. American forces can control the establishment of any large-scale terrorist infrastructure through a combination of raids and airstrikes. And networks that are hidden can’t be effectively targeted even if the U.S. maintains a military occupation — as the cases of IED networks in both Iraq and Afghanistan demonstrate conclusively.


The second category of challenges comes from the concern that an early withdrawal would open the door for a seizure of power by anti-American radicals. This case is even easier to manage than the state collapse scenario. The seizure of power by radicals is, in effect, the status quo antebellum. If the U.S. could remove Saddam, removing a new rogue regime from power would have been even easier, especially since any such new regime would have even less combat power than Saddam was able to deploy. In addition, American forces would have two major advantages in this scenario. First, the demonstrated capability to effect regime change in Iraq would serve as a powerful deterrent for a new regime — regardless of its ideology — to pursue an anti-American agenda. Second, having operated in Iraq proper, the military would be even better prepared for a new campaign there.

Indeed, efforts to build a strong Iraqi state are arguably the biggest source of risk for U.S. interests. Just as Iran was able, after 1979, to confront the region with a potent American-trained and -equipped military, a radicalized Iraq is a larger potential risk thanks to efforts to train and equip their forces. A significant percentage of Iran’s air force — to this day — comes from prerevolution sales to the shah’s regime. American efforts to insure against a resurgent, anti-American Iraq may have decreased the likelihood of the contingency, but increased the challenge it would pose if it came to pass.

Had the U.S. left sooner, the worst case we would face in this regard would have been to repeat the campaign of 2003, though at a reduced cost. Instead, American leaders again overpaid dramatically to try to prevent the scenario from occurring.

A significant criticism of this argument is that had the U.S. left it would be difficult, if not impossible, to intervene a second time. There is no historical basis for this claim. The U.S. repeatedly intervened in the affairs of Latin American countries for more than 60 years from the Spanish-American War until the Dominican Republic intervention in 1965. The U.S. fought Germany in 1917 and again in 1941. The U.S. intervened in Lebanon and 1958 and 1983. The U.S. fought wars against Iraq in 1991 and 2003. And those are just the repeat cases of open deployments of force.


It is a famous cliché of international politics to say that nature abhors a vacuum and then use it to justify some sort of interventionist policy. As a conceptual matter it is not clear that the statement has any meaning, and as an empirical matter it is not clear that whatever meanings might be ascribed to it are accurate. There are many, many cases of weak states coexisting with strong ones. Canada is a military vacuum compared to the U.S., but has not faced an invasion from the south in nearly 200 years. Costa Rica does not have a military, and yet it does not find itself under constant siege from Nicaragua and Panama, both of which have had radical leaders in the past 20 years. Happily we no longer live in a world where military weakness inherently tempts aggression. Instead, in order to assess the risk of regional conflict, we must consider the specific issues or dynamics that could have provoked conflict.

There are essentially three plausible scenarios for a regional war involving a weakened post-conflict Iraq. These are:

å An attempt by Kuwait or Saudi Arabia to seize oil resources.

å An attempt by Iran to establish control over Shiite parts of Iraq.

å Involvement of Turkey to quash Kurdish independence or Saudi Arabia to come to the aid of Sunnis as a result of sectarian or tribal affinity.

Kuwait, Saudi Arabia and Turkey are American allies. It is an extraordinary claim to suggest that the U.S. needed to occupy Iraq for the better part of a decade in order to prevent it from being invaded by American allies. Surely, vigorous diplomacy should be able to head off this possibility. Indeed, the biggest threat comes from Turkey, which would likely find itself under tremendous pressure for restraint by both the U.S. and the European Union. The larger challenge is Iran, but again, there is no reason to assume that an explicit guarantee of Iraq’s sovereignty would be insufficient to deter Iranian aggression.


Perhaps the most compelling reason for remaining in Iraq was the possibility of genocidal violence breaking out in the immediate aftermath of an American withdrawal. There is a lot of research on the causes of genocides and mass killing, and the research suggests that Iraq was not particularly at risk. There are two main sets of arguments about the causes of genocides and mass killings. The first comes from an analysis of contextual factors in past genocides, and the other comes from a focus on strategic motivations.

There are several key risk factors in the outbreaks of genocidal violence. These include:

å Existence of a genocidal ideology with explicit efforts to dehumanize “the other.”

å Scapegoating dynamics, where a minority group is blamed for societal hardships. This is especially potent when the minority is associated with an external enemy.

å A history of previous mass killing, which generates an impetus either to complete genocide or to avenge previous killings.

å My research also suggests the importance of a dominance of force.

There are also two common strategic motivations for genocide:

å A desire to dispossess a group by seizing its assets or territory.

å An effort to coerce a population, particularly in a brutal counterinsurgency campaign, as in the recent case of Sri Lanka.

An interesting — and underexplored — question in the case of Iraq was who were we expecting to kill whom? Saddam’s violence against the Kurds was an oft-cited precedent for the possibility of mass killings in Iraq, but that was a case of an authoritarian minority regime with a large superiority of military power engaging in a brutal counterinsurgency campaign. Ultimately in Iraq, the most likely cases of mass killings were the possibility of Sunni Arabs engaging in mass killings of Kurds or Shiites, or of Kurds seeking revenge on Sunnis. The Kurds might have sought revenge on Sunnis for Saddam’s mass killings, but that seems unlikely given the existing autonomy of the Kurdish region. Why place it at risk in order to exact indirect revenge?

The Sunnis had the strongest motivations, but the least capability. Indeed, the risk of potential genocide in Iraq was constrained by this factor. The Shiites and Kurds largely had what they wanted in terms of control over the government and local autonomy, respectively. But while the Sunnis may have settled upon a genocidal policy had they possessed a unified leadership structure and a significant military capacity, their lack of either is what led to a disorganized insurgent campaign based largely on using the threat of indiscriminant violence as a way of coercing political concessions from Baghdad.

This analysis suggests that genocide was always unlikely, and that the most likely scenario for mass killings would be in the context of either Shiite government forces or Kurdish militias engaging in mass killings in order to suppress a Sunni insurgency. The problem with using this fear as a justification for the American presence is that this is precisely what happened even with the American presence. It happened on the American watch, and forces were unable to prevent it because it occurred slowly, over time, as Sunni terrorism and Shiite and Kurd militia death squad retaliation effectively “ethnically cleansed” the country. In short, the U.S. remained and was unable to prevent what was, most likely, the worse case anyway.


America’s extended occupation of Iraq was based on two fundamental fallacies. The first is that if U.S. forces remained, they could ensure a measure of predictability over developments in the country. Unfortunately, the reality is that political reversals occur quickly. In January 1978, no one expected that the shah of Iran would be overthrown in less than a year. The collapse of the Somoza regime in Nicaragua was also unforeseen even a year in advance. In 1986, anyone who suggested that within five years the Soviet Union would have suffered a military coup and then been dissolved would have been considered a lunatic.

In international politics, it is difficult to foresee even a year into the future, much less control developments 10 years out. But the only possible justification for our extended presence in Iraq is that, somehow, by doing so it can guarantee a positive future.

The second fallacy is the notion that leaving and returning is more costly than remaining in place. This is not true. Unless the occupying force is brutal in extracting resources from a country under its control, the cost of an extended occupation is always going to be higher than of a second (or third) intervention. This is particularly true for the U.S., whose military is built for power projection, not for occupation. It is a unique capability and a source of extraordinary strength. The current fascination with COIN systematically seeks to erode this capability in favor of a more costly, less capable approach to safeguarding American interests.

The U.S. has the capacity to mitigate the consequences of most negative developments through the judicious use of decisive military force. If anything, an insistence that any use of force must, as its corollary, imply a costly and pointless decade-long occupation is likely to make American leaders and the public overly reticent to use force in defense of American interests. Insisting that military actions require a follow-on nation-building adventure is likely to serve as a major self-deterrent to the rational employment of our military capacity.


Afghanistan is a more difficult case than Iraq for this argument. Because Afghanistan is so remote, it is true that the relative cost of reintervening is higher than in the case of Iraq. Nonetheless, its remoteness cuts both ways. Afghanistan may be harder to reach, but threats emanating from that country are also less severe. As devastating as was 9/11, the reality is that the consequences of Taliban control and al-Qaida safe havens were perceived as so low at the time that it was impossible to generate a policy consensus to risk any American lives or collateral Afghan damage to eliminate the situation — an assessment shared by the Clinton and the Bush administrations until the attack. It is certainly true that American leaders underestimated the risk prior to 9/11. But it seems equally likely that they overestimate the risk today.

As in Iraq, one of the worst cases in Afghanistan would be the status quo antebellum — a restoration of Taliban control giving safe haven to al-Qaida. But having chased the Taliban from power and scattered al-Qaida to the four winds once, what would prevent the U.S. from doing it again? And as in the case of Iraq, American forces accomplished most key war aims in the first few months of the war.

Military operations began roughly on Oct. 7, 2001. Kabul fell on Nov. 13, 2001. Tora Bora was captured on Dec. 17, 2001, and al-Qaida leaders were chased from the country. The remaining significant al-Qaida and Taliban forces in the country were defeated in hard fighting in March 2002. Within six months the Taliban had been devastated and al-Qaida was shattered.

American forces have, ever since, been fighting, dying and paying to achieve smaller and smaller goals. In five years, when this analysis takes on the same clarity as the Iraq case, we will again be left scratching our heads, wondering how the nation could have spent so much to achieve so little over what was obtained a mere six months into the conflict.

BERNARD I. FINEL is a senior fellow at the American Security Project, where he directs research on counterterrorism and defense policy. He was a professor of military strategy and operations at the National War College from 2004 to 2006.