The successful drone strike in Yemen on Monday against Nasser al-Wuhayshi, leader of al-Qaeda in the Arabian Peninsula, has sparked a renewed conversation (here, here) about whether so-called decapitation strikes are effective. From a now decade-long campaign against Taliban commanders in Afghanistan and Pakistan to high-profile efforts against Osama Bin Laden and Abu Bakr al-Baghdadi, leadership assassination has emerged as a key — and sometimes the only — tactic against terrorist and insurgent organizations.
But does it work?
Not surprisingly, political scientists have increasingly weighed in. And, also not surprisingly, there’s little consensus. Ranked from most to least optimistic, we’ve found:
- Analyzing 119 attempted assassinations, Patrick Johnston has found that killing terrorist leaders is associated with shorter campaigns, higher odds of defeating insurgencies, and lower overall violence in 90 campaigns (1975-2003).
- Daniel Byman has argued that Hamas’s attacks against Israeli citizens became less lethal after successful leadership strikes, with only a short-term hardening of Palestinian attitudes against Israel.
- Brian Phillips has demonstrated that whacking leaders of Mexican drug trafficking organizations is associated with both short-term decreases and long-term increases in violence; reductions only occur when leaders are arrested (not killed) and midlevel (not top commanders).
- Examining 298 leadership targets (1945-2004), Jenna Jordan has found that decapitation did not increase the rate of organization collapse; in fact, it may have actually prolonged these organizations
These findings are, of course, contradictory. Why don’t we have more definitive results and, ideally, policy recommendations? It’s worth stepping back a moment to see how three key problems stand in the way of figuring out what works.
First, the data problem looms large. Ideally, we would have data on all decapitation strikes, whether successful or unsuccessful, and whether the intended leader was actually killed. This seems a pretty low bar, yet for many of these attempts, especially from drones, the identity of the individuals remain in doubt (here, here). We also need to peek inside the machinery of these organizations to understand their leadership structures (hierarchical? decentralized?) and factions. And, on top of that, we’d want to know how the locals view these organizations (here, here). Taken together, these requirements are formidable; without these pieces of evidence, however, assessing effects will be tough.
Second, we need to abandon simplistic “up or down” judgments on whether this tactic “works.” Instead, we need to embrace nuance and, above all, the idea that these efforts have cross-cutting effects. Decapitation strikes may, for example, destabilize a group, complicate its logistics, deter would-be recruits (and leaders) from signing up, and reduce its violence temporarily. Yet these same strikes may also create incentives for new, more radical, leaders to try their hand, may loosen command and control structures, and may ultimately make it harder to reach a political settlement. Similarly, decapitation strikes may be both celebrated and mourned by locals depending on their relationship to the organization — and whether they’re in the crosshairs for reprisals for “diming out” leaders. When we decide to measure effects — a day after? a month? a year? — will also shape our judgments. What’s “effective” today may be disastrous a year from now.
Third, we should ask ourselves: effective compared to what? Decapitation is a tactic typically embedded in a broader strategy. As such, its effects should be assessed relative to other instruments of power, including night raids, arrest and reintegration programs, among others. Even this task, however, is tricky. High-value targets have their own sets of rules, especially about inflicting civilian casualties, that may not apply to other tactics. In Afghanistan, for example, drone strikes targeting leaders have different rules of engagement and are typically used in different situations than manned airstrikes, raising issues of comparability. In some ways, decapitation strikes are tools of last resort, opportunistic efforts against hard to reach targets where the odds of success are comparatively low.
In the end, decapitation efforts are unlikely to be war-winning silver bullets. Instead, they are likely to have conditional, local-level, unpredictable, effects that rarely affect whether wars are won or lost. Rather than asking “do they work?” we should be asking “under what conditions can decapitation achieve the desired political objectives?” In many cases, perhaps most, it is likely that the best shot is the one not taken.