A U.S. airstrike killed ISIL spokesman and strategist Abu Mohammad al-Adnani on August 30th, 2016. In May, the United States killed Mullah Mohamad Akhtar Mansour, emir of the Taliban. These are just two of the most recent and prominent kills the United States has notched. In addition to high-profile leaders, numerous other jihadists have been the quarry of U.S. drone strikes or special forces raids.
Targeted killing, a term used loosely in press and undefined in international law, can be understood broadly as the premeditated use of lethal force against an individual or group of individuals. The definition is expansive, covering the targeting of senior leadership or ground-level combatants, the use of conventional military force or drone strikes, the goal of force protection overseas or prevention of attacks on the homeland. Despite innovations in the tools used to kill, the practice of targeted killing is nothing new, and it’s not just Americans who do it. To flag just a handful of examples: Israel has killed Palestinian leaders since the 1970s, the Russians (formerly Soviets) targeted Chechen leaders for decades, and Iran hunted Kurdish leaders in the 1990s. For its part, the United States has killed leaders of the Viet Cong, al Qaeda, and the Taliban, among others. Fifteen years after September 11 and the launch of the Global War on Terror, President Obama, mostly through drone strikes against suspected militants in Pakistan, Somalia, Yemen, Libya, Iraq, and now Syria, has approved more targeted killings than any president in modern history.
In short, targeted killing is a popular strategy for states fighting enemies outside their borders. And it’s not hard to see why.
Targeted killing is easy for governments to “sell” to casualty-averse domestic populations because the strategy minimizes risk to their own military personnel. Targeted killing from the air is also pitched by its practitioners as morally preferable to conventional alternatives like ground invasions, which, they argue, would cause much more collateral damage. Finally, the strategy seems to make sense. Targeted killing is believed to operate through three mechanisms: it disrupts terrorist organizations by forcing them to run for cover and replace leaders instead of plot against their enemies, it degrades them by destroying their top talent, and it deters them from plotting additional attacks by raising the perceived costs of doing so.
The logic seems straightforward. And yet, as Max Fisher pointed out a few weeks ago, scholars have struggled to find evidence that targeted killing makes any difference at all. A lack of evidence, though, is not the same as evidence against, and targeted killing is a particularly tough study for three reasons. First, it is implemented at the same time as other counterterrorism strategies, so it is tough to isolate its specific effects. Second, the counterfactual is difficult to estimate — even if terrorist attacks rise, who’s to say that they would not have risen higher in the absence of the targeted killings? And third, what does “effective” even mean? Does targeted killing have to reduce the number or lethality of terrorist attacks to be deemed effective? Attacks against whom? In what geographic radius? Can a reduction in terrorist membership count as effective? What about turning the tide of local public opinion? The metrics for gauging effectiveness of any counterterrorism strategy, targeted killing included, are anything but clear.
Targeted killing is an open question because it triggers processes that work in conflicting directions. It triggers mechanisms that work in the intended direction — deterring, degrading, and disrupting terrorist activity — yet at the same time triggers mechanisms that work in the exact opposite direction — radicalizing populations and inciting terrorist escalation. But which direction wins?
In the absence of compelling empirical evidence regarding the relative effect sizes of the mechanisms, it is important to carefully trace the causal logic behind each. The exercise provides a framework for understanding the conditions under which targeted killing would be more likely to fuel terrorist escalation than to reduce it.
Targeted Killing Deters, Degrades, and Disrupts, but How Much?
Deterrence hinges on raising the perceived costs of a certain action so that they outweigh the perceived benefits. Targeted killing certainly imposes costs on terrorist organizations for continued operations, but the benefits remain high: militant groups’ continued survival, hinges on recruiting new members, and groups such as Hamas and ISIL have at times used specific types of violence, such as suicide bombings and broadcasted beheadings, respectively, as recruitment tools. That means that the perceived costs for continued attacks — targeted killings — may very well never outweigh the perceived benefits of continued attacks — bolstering membership (or, to take ISIL at its rhetorical word, destroying the heretics). In short, militant groups have a variety of strong incentives for conducting attacks, all of which should be considered as part of the cost-benefit scale that targeted killing is intended to tip through deterrence.
Targeted killing of militant leadership is degrading to the extent that eliminating experienced, skilled, or “charismatic” leaders durably handicaps operations. The degradation mechanism hinges on the presumed irreplaceability of militant leaders. But is it really so difficult to replace militant leaders with equally effective successors? When companies replace managers, successors could be less effective at the helm, but they can also be more effective. There is no reason the same basic principle should not apply to terrorist leadership succession. Perhaps it takes longer to replace a highly skilled bomb-maker than an experienced or charismatic leader. Maybe replacing a bomb-maker takes a week, maybe a month, maybe a year, or maybe the group can never find a replacement and has to change tactics. Or maybe a smart group has a bomb-maker trainee ready to go in case their top bomb man is targeted.
The extent to which targeting low-level militants (as opposed to leadership) degrades organizations also varies entirely depending on the context. For instance, it is much easier for al-Qaeda in the Arabian Peninsula to replace killed members than it is for al-Qaeda, an Arab organization operating in Pakistan.
In sum, targeted killing degrades a terrorist organization only to the extent that the leaders or low-level members targeted are difficult to replace. Because the difficulty of replacing members varies entirely depending on the specifics of the target and the operational context of the organization, there is reason to question optimistic assumptions about the power of the degradation mechanism. In fact, Ioannis Koskinas characterized Mansour’s replacement with Akhunzada as “seamless.”
Targeted killing disrupts to the extent that it forces terrorist organizations to spend time trying to prevent or recover from targeted killings that they would otherwise have devoted more efficiently to plotting and executing their attacks. The logic, particularly in Jake Shapiro’s book showing how targeted killings force terrorist organizations to use more clandestine, less efficient modes of communication, is powerful. The logic is also lent empirical credence by communication collected in the Abbottabad raid, in which bin Laden and his lieutenants lament the disruptive impact on Al-Qaeda operations of drone strikes. However, Jenna Jordan’s study shows that as groups age, grow, and decentralize, they become more resilient to the disruptive effects of targeted killing. Age and experience yield improved routines for operating under fire. As they grow larger, day-to-day operations increasingly function without close management by group leadership, enabling continuation of operations despite leadership death. Similarly, decentralized groups (as opposed to hierarchical) comprised of networks of horizontal cells are deliberately designed to operate independent of group leadership. Groups capable of operating in the absence of slain leaders would be less likely to pause daily operations until successors are selected. In short, militant groups, to varying degrees, develop the experience and the structure to mitigate (if not by a long shot eliminate) the disruptive effects of targeted killings.
It is not that targeted killing fails entirely to deter, degrade, or disrupt. On the contrary, targeted killing almost certainly does all three. Terrorist organizations, however, take a variety of measures to weaken these mechanisms with varying success. And the distinction is critical because targeted killing also triggers mechanisms that work in the countervailing direction.
Targeted Killing Incites Escalation
A number of scholars and servicemen have argued, for instance, that targeted killings, and repressive counterinsurgency measures more broadly, fuel grievances among local populations against the counterinsurgents and in favor of the terrorist organizations that fight them. If the aim of the targeted killings is to disrupt imminent plots against the United States, local blowback may be considered a small price to pay. However, not all (and arguably only a small minority of) targeted killings are meant to disrupt imminent plots against the United States. Even in those cases, the key question remains: Does killing militants involved in plotting against the United States trigger more plots than it disrupts? Are we playing a game of whack-a-mole, or do two heads grow in the place of every one we cut off? The fact that targeted killing might make enemies is not a problem in itself, but it becomes a problem if making enemies through targeted killing leads to attacks against the United States and its allies. So-called “grievance” or “radicalization” theory operates through at least three discrete mechanisms.
- Local populations radicalized by targeted killings (which can be frightening, destructive, and cause collateral damage) are more likely to join the organizations that fight against the government that perpetrated the targeted killings. Increased membership strengthens the organization that the targeted killings were intended to weaken, facilitating increased terrorist activity that the targeted killings were meant to reduce.
- Radicalized local populations, even if they do not take the extreme step of joining the militant group, are less likely to cooperate with counterinsurgent personnel, undermining the intelligence collection upon which effective counterinsurgency (and peacekeeping) hinges. The binary choice between joining or not joining a terrorist group is a fallacy. Individuals make choices to act anywhere along a fluid spectrum, ranging from joining a terrorist group, to providing resources to the group, to refusing to cooperate with counterterrorism efforts, to remaining neutral, to actively assisting counterterrorism efforts. Targeted killings affect these choices, and these choices impact terrorist group strength and activity.
- Radicalized local populations, even if they do not join the terrorist organizations, may encourage escalation. Outcry by the local population, upon which terrorist organizations depends for membership, resources, and legitimacy, may pressure leadership to launch attacks even when the leaders would otherwise have thought it wise to temporarily cease operations. It is true that some militant leaders may operate without much concern for local opinion. Many leaders of militant organizations, however, are to varying degrees beholden and responsive to the demands of their local populations, just like the representatives of American political parties are to their constituencies and the managers of publicly traded companies to their shareholders. In keeping with cost-benefit analysis theory, in the wake of targeted killings, surviving terrorist leadership may deem the costs of ignoring local demands for retaliation higher than the costs imposed by external actors for the same behavior.
The question is not whether targeted killing disrupts, degrades, deters, or incites. It does all four. The question is how much of each? Is the net sign positive or negative? A close look at the causal mechanisms helps to explain why targeted killing may not be as effective as we think, and why, in certain circumstances, it might fuel escalation.
Too often, counterterrorism strategies are sustained because they are more politically viable, less expensive (in blood and treasure), or more tactically effective than alternatives. Too rarely do policymakers step back and ask the larger questions: what is the end game? Is what we’re doing working? Policymakers concerned with the strategic efficacy of targeted killing should start by defining their long-term counterterrorism objectives, and then, with close attention to organizational context, consider the following questions:
To what extent does targeted killing deter? Policymakers must bear in mind that organizations conduct cost-benefit analyses and that targeted killing might raise the costs but do nothing to reduce — or perhaps even raise — the perceived benefits of sustaining, or even escalating, operations.
To what extent does targeted killing degrade? Policymakers should try to gauge the degree of difficulty specific organizations would have replacing the individuals targeted. The estimate should be founded on assessment of the uniqueness of the individual’s skill, and the pool available to the organization for replacement.
To what extent does targeted killing disrupt? Policymakers should consider the organization’s adaptability to operating under conditions of extreme clandestinity and in the absence of close leadership oversight.
To what extent does targeted killing incite? After reaching an assessment of the strength of the mechanisms working in the intended direction, policymakers must likewise ask — how many locals will join or provide resources to this organization because of targeted killings? How many locals will refuse to provide intelligence cooperation to Americans and American allies because of targeted killings? Will the locals demand retribution from the targeted organization against the American aggressor? How beholden to local demands for retribution is the targeted organization?
These questions are difficult to answer, but they are where strategic assessment should start.
Rachel Tecott is a PhD student in political science at MIT. She studies strategy and decision-making in inter- and intrastate war. Find her on Twitter @racheltecott.
Image: U.S. Air Force photo by Staff Sgt. N.B.