How to Improve the Drones Debate

We are delighted to welcome the following guest post by Princeton PhD student Omar Bashir on how political scientists can help improve the drones debate. We have linked to his excellent article in Foreign Affairs on accountability over drone strikes before.


The Washington Post special report in late October about the practice of targeted killing set off a new wave of commentary about the ethics of American drone campaigns.  It was good to see so many outlets, in print and especially online, focus attention on the topic.  Yet it seemed as though the most serious criticisms were difficult to pick out amidst invective and otherwise distracting lines of argument.  This is a problem if you believe that debates which converge on and identify the central failings of government behavior are more likely to bring about changes in policy.

With this in mind, the current post describes six ways that participants in the drones debate can make it better.  It is meant for a wide audience even though it is addressed to political scientists who choose to participate in public discourse.  It suggests that our skills and subject knowledge can be clarifying forces, even if we do not work on the topic of targeted killing specifically (see one example of this work here and a literature review here).

Clarify and organize ethical objections regarding the conduct of strikes. The number of concerns raised in critiques of targeted killing must be dizzying for any citizen trying to navigate the moral terrain.  Political theorists have a role to play here by highlighting the tools philosophers and practitioners have developed to assess the morality of organized violence.  Three “just war” principles seem most controversial in the conduct of drone attacks: necessity, discrimination, and proportionality.  Almost every in bello criticism is associated with one of these three criteria.  For instance, those worried about “signature strikes” are expressing concern about discrimination, since these particular strikes risk targeting civilians (worse than killing them unintentionally). Those worried that alternatives to killing may not get enough attention are making a criticism based mostly on necessity. Similarly, commentators are talking about proportionality when they argue that too many civilians are killed even when dangerous militants are targeted successfully.

Several other concerns can be categorized in this way; doing so can help suggest solutions.  If the most important critiques seem to be about necessity, this suggests that oversight proposals should be focused on the process by which kill/capture decisions are made (formation of the “kill lists”). Alternatively, if the discrimination critiques seem most troubling, perhaps oversight should focus downstream, on the process by which judgments are made about “patterns of life.” Organizing ethical issues in this way also reminds people that we may not need to reinvent the wheel, ethically speaking.  Though drones are new, principles already exist to guide the conduct of violence; the debate may be better if it focuses on ways to get governments to demonstrate that they are taking these principles seriously.

Question characterizations of affected peoples and territories. To caricature the prevailing view in drone commentary, Pakistanis and Yemenis are so primed for vengeance that exposure to civilian casualties leads them like lemmings incapable of moral deliberation into the arms of groups that aim to kill yet more non-combatants.  The question of whether or not drone strikes are counterproductive is surely important because it bears on proportionality. But, given the difficulty discussed below in designing a study to answer this question with confidence, we should be skeptical of anyone who asserts a conclusion.  Some efforts to collect relevant data, though imperfect, at least raise doubts about the conventional wisdom.  One example is Christopher Swift’s article entitled “The Drone Blowback Fallacy,” which is based on structured interviews with 40 tribal leaders, politicians, and clerics across 14 of 21 provinces in Yemen.  Another is the “Understanding FATAsurvey conducted in Pakistan; despite the author’s recurrent criticisms of drone use, the data in the report suggest  that (1) citizens in the tribal areas prioritize several other complaints about life in FATA (“Talibanization,” for instance) over objections to drone strikes (p. 59); and (2) opposition to drone strikes in areas like North Waziristan, where the strikes occur most frequently, is lower relative to other areas (p. 89).

Still, there are reasons to worry from a public diplomacy perspective.  Swift raises the point that drones may increase anti-Americanism not necessarily because of civilian death, but because drone campaigns constitute an affront to national pride when citizens do not see their own government solving its own problems.  Elite-driven criticism of drone attacks may have an even larger role in Pakistan; outrage over drones is an unambiguous political winner, especially in urban areas far from strikes.  This means that drone attacks, regardless of how carefully they are conducted, may provide a long-term source of anti-American sentiment for political opportunists who wish to tap it.  (Reuters journalist Myra MacDonald, author of the linked piece just above, has in my view the most thoughtful things to say about the drones debate and how it plays out in Pakistan; more of her work can be found here and here.)

Note that there is room for more empirical work to help us understand whether proportionality is satisfied.  For instance, research might explain which of the pathways above, if any, are most consequential for militant recruitment.  Of course, studying recruitment as a dependent variable poses special challenges because of the difficulty of data collection and the presence of confounders (e.g. strikes may not cause higher recruitment if there is a third variable that causes both higher recruitment and more frequent strikes in a given area).

Cut through distracting rhetorical tactics to find better questions to argue about.  After learning about the “disposition matrix” described in the Washington Post, I wanted to understand whether or not the matrix—a database “meant to map out contingencies, creating an operational menu that spells out each agency’s role in case a suspect surfaces in an unexpected spot”—was a positive development.  Instead, most commentators rushed to highlight the Orwellian/creepy/alarming/troubling name of the database and evaded discussion about whether the existence of a contingency table is preferable to ad hoc decision-making when it comes to kill-or-capture decisions.  For instance, though there are fears that institutionalization may lead to “endless war” (discussed below), might it be a positive development in terms of accountability that the administration has cobbled together a road map of its policy that can be criticized, re-evaluated, or leaked?  We continue to see a similar pattern in “kill list” commentary.  Name notwithstanding, is it better for the administration to delegate the setting of target selection criteria to military or intelligence organizations as in wartime, or to raise this type of decision-making to the executive level in tandem with congressional oversight?

Set aside the answers to these questions for a moment.  The point is that optics-inspired hyperbole distracts us from thinking clearly about the alternatives in these cases and in others (e.g. “drone pilots are like video gamers” obscures thoughtful comparison to the ethical constraints on pilots of manned aircraft).  If you can link a given policy to dystopian imagery, you can more easily convince people that the policy is misguided without explaining why it is worse than other options.  The same goes for drawing comparisons to Bush administration practices when the relevant moral comparison is to current policy alternatives.  Since reasoning about alternatives and counterfactuals is second nature to most political scientists, we can help push back against these critical tendencies and raise the profile of more substantive analysis where we see it.

Unpack “extrajudicial” and “illegal.” These terms are often employed as tools for building a narrative about the government’s disregard for constraints.  But it is hard to organize a push for compliance if it is not clear which legal standards are thought to be at issue.  There are three main bodies of law to consider: (1) domestic law, (2) the international law of armed conflict (rules about the conduct of violence), and (3) international human rights law (rules that protect rights but which can be suspended in war zones).  Assertions about illegality on these fronts may be more problematic than critics assume.

But I want to argue a different point: we need to bring perspective to discussion about legal violations.  For instance, if you believe that drone killings are extrajudicial with respect to domestic law because the Obama administration claims it can kill American citizens, then you need other reasons to label as extrajudicial the vast majority of drone killings for which non-citizens are the targets.  Perhaps your claim is based instead on the observation that targeted killing seems increasingly to be taking place away from traditional battlefields. Thus, the strikes may be extrajudicial because international human rights law should apply instead of the more permissive law of armed conflict.

But you should at least acknowledge that the “it’s not really war” criticism begs a question about international law.  Might we want to allow for the targeting (or detention) of members of groups who plan and intermittently act to rob non-combatants of their right to life, even if the groups’ bases of operation are far from war zones? (Before reacting to the difficulty in establishing criteria regarding which groups would be targetable, remember that “just cause” and other ad bellum principles that that we use to evaluate the resort to traditional war are useful while also being highly ambiguous and subject to interpretation.) To make it more concrete: if schools are targeted by militants in northern Pakistan years after NATO ends its war in Afghanistan, will we be glad that international law calls for inaction and thus buck-passing to a Pakistani military that is known to be far less capable of limiting civilian casualties?

Scholars of international law can be more vocal about how standards have changed throughout history to adapt to certain technological changes, moral shifts, and new forms of political organization; the advent of modern transnational terrorist groups is an example of the third.  These scholars can also point out that human rights advocates have at times found themselves arguing against proponents of international law in its contemporary form, as when British abolitionists cheered their navy’s unilateral search and seizure of slave ships in the early 1800s.


Make the right claims about precedent.  There are two main fears about international precedent that come up in the context of drones, but one withstands scrutiny better than the other.  First is the idea that targeted killing will erode state sovereignty.  To begin with, it is not clear whether legal sovereignty is violated in the first place when host governments give consent, as they likely have done for many strikes in Pakistan and especially Yemen.  Regardless, Stephen Krasner’s classic international relations work entitled Sovereignty: Organized Hypocrisy (1999) chronicles how sovereignty has endured in world politics even as it has been repeatedly violated.  You may still be tempted to ask a leading question about how Americans would feel if another country launched drone strikes in U.S. territory, but make sure you think about other ways to frame the same question: “I wonder how I would feel if my own village in Scotland were overrun by the Taliban – would I want rulers in London to ‘give peace a chance,’ for the British Army to shell us with artillery, or American drones to kill the Taliban leadership?”

The second and more serious concern about precedent seems to be that other governments, even if legitimately involved in war, will conduct drone strikes in a way that is reckless (or opaque) and that criticism of this behavior will ring hollow because of the way the U.S. is currently running its own campaigns.  Foreign leaders will be able to point to the Obama administration’s precedent of “reckless” drone use.  This characterization may not be accurate, but, as it stands now, nobody has any way to show convincingly that today’s strikes are executed with appropriate care.

Instead of raising yet more questions, help propose answers.  Most news articles about drones cover some new development, claim to raise new ethical questions, and mention superficially the need for greater transparency and/or accountability.  Specific recommendations for change are rare or rarely helpful (this recent editorial calls for strikes to be subject to congressional review, but they already are).  There may be an opportunity for political scientists to contribute by formulating and floating ideas about safeguards that address pressing ethical concerns.  For example, it is common to hear calls for the introduction of oversight to drone campaigns.  Political scientists generally have a good sense of which proposed institutional arrangements might provide successful oversight because we are trained to consider issues like incentive compatibility.

Further, we’re likely to have knowledge of oversight institutions at work in other countries that might be emulated.  My own proposal is based on adaptation of the UK’s system of independent review for terrorism legislation.  I think it addresses the single most important ethical issue regarding drone strikes: we have no way of knowing whether or not the U.S. government is acting in accordance with the requirements of necessity, discrimination, and proportionality.  Inconsistent studies of post-strike damage have not settled the issue, and we can’t simply take the Obama administration at its word.  Instead, the government needs something beyond existing congressional review to demonstrate credibly to audiences at home and abroad that too many civilians are not dying compared to the threat posed by targets and to show that there is appropriate cause for deeming individuals targetable.

This oversight, which can ideally provide some indication when strikes begin to violate the requirement of proportionality, may be the key to preventing “endless war”: it might help us know when, if not already, campaigns have taken out so many targets that further killing cannot be justified.  Clinton Watts and Frank Cilluffo propose another tangible solution that has a chance of being acceptable both to government and human rights advocates.  Their idea is based on the modification of an existing American institution, the Foreign Intelligence Surveillance Act (FISA) court; it is covered in this post.  If you are aware of other proposals, please link them in the comments, and feel free to post your own ideas.

Slightly edited for readability (7:30pm)


6 Responses to How to Improve the Drones Debate

  1. Vladimir November 15, 2012 at 2:55 pm #

    I am not opposed to drone strikes including so called in “signature strikes” in principle. What I find astonishing is that these signature strikes employing the techniques of racial profiling to the use of lethal force are supported by an African-American president.

  2. Patrick Johnston November 16, 2012 at 8:39 am #

    Anoop Sarbahi and I have a working paper, which we’re currently revising, on the relationship between drone strikes and militant activities in FATA. It doesn’t get at recruitment issues directly, but it is relevant to many of the issues brought up in this post, and it has implications for thinking about the utility of this tactic, at least in the context of northwest Pakistan. Email me at ( if you’d like a copy.

  3. Peter Kaye November 16, 2012 at 7:43 pm #

    I consider drone strikes to be a very important weapon in the war against terrorism, particularly those attacking Taliban and Al Qaeda targets. In conflicts of this sort where conventional means of defeating the enemy are denied as a result of the difficulty of identifying visible targets within the general population, drone attacks, based entirely on intelligence are more than justified and reduce the possibility of widespread colateral damage amongst the general public. They are the only weapon the terrorists are really afraid of.

  4. Christopher Kam November 16, 2012 at 10:53 pm #

    “….drone attacks, based entirely on intelligence are more than justified and reduce the possibility of widespread colateral damage amongst the general public.”. Which country’s “general public” would that be?

  5. Christopher Kam November 17, 2012 at 1:11 am #

    I applaud the effort to think critically (i.e., logically) about what is an important and contentious issue. That said, let me point to a couple holes in the argument set out above and offer some additional perspectives on the situation. On proprtionality as a criteria with which to evaluate the moral basis for employing drone srikes: 1) It strikes me as pertinent that the administration has altered (or at least endorsed) a definition of an enemy combatant to include all males above a certain age in a given area; this clouds our ability to assess accurately the scale of collateral damage and with it to judge proportionality. 2) The efficiency of drone strikes (i.e., whether they kill 2 terrorist for every 1 new recruit they incite, or vice versa) is irrelevant to any moral judgement about the use of drones. We sometimes have recourse to efficiency as a moral standard as when there exists a Pareto efficient, and hence unambiguously welfare enhancing, solution. This is not one of those situations: x claims to be better off as a result of killing y; whether or not x is actually better off, y is clearly worse off. On the extrajudicial and illegal nature of drone strikes: The author does a nice job at pointing out that these actions could contravene various types of laws,, e.g., domestic, international, etc. I think the most troubling case in this regard was the assassination of al-Awlaki and his son. Both individuals were US citizens who were alleged to be linked to al Queda. The allegations were serious and the evidence pretty convincing (at least with respect to the father – yet even he was not alleged to have directly killed anybody ). Yet, surely the administration violated the 5th amendment when it assassinated two American citizens without a “a presentment or indictment of a Grand Jury.” To my mind stretches the “public danger” clause to a dangerous limit. (At the very least, the administration should have the canoes to argue their case in court to establish a clear legal precedent.) So, yes, I think there is a very dangerous lack of accountability here, and clearly those dangers extend even to American citizens. This is not welfare enhancing; quite the opposite, save for the folks in the administration who get to craft the disposition matrix, we are all (at least weakly) worse off . What’s the solution? I think there is a fairly obvious one: use existing criminal law (apply for extraditions, put the accused on trial, etc.). Would that not constrain the administration and limit to legal and constitutional remedies? Yes – and that would be exactly the point.

  6. Thomas November 17, 2012 at 12:48 pm #

    Great article, and finally some ideas on how to address the issue of drone warfare – which will, for sure, stay with us. One point I would like to add is the fact that critics often mention the “playstation” mentality. This, however, is in my opinion unfounded criticism as there is evidence that drone operators suffer from PTSD. I do get the feeling that some criticism is rather based on emotions rather than informed and rational opinions.