Acceptability of Collateral Damage when Using Armed Drones: A Moral and Ethical Dilemma BY - Dr.S.Krishnan & Ms. Garima Dhaka

Acceptability of Collateral Damage when Using Armed Drones: A Moral and Ethical Dilemma
By
Dr.S.Krishnan
Associate Professor
Seedling School of Law and Governance
Jaipur National University, Jaipur
Ms. Garima Dhaka
Assistant Professor
Department of Law,
Vivekananda Global University, Jaipur
In 2001, The United States declared a War on Terror and invaded Afghanistan.1 Simultaneously, with wars in Afghanistan and Iraq, the media reported drone strikes in Pakistan, Afghanistan, Yemen and Somalia allegedly eliminating terrorists.2 In regions like Pakistan where no war is in progress but where it is no longer possible to speak of peacetime – we shall call them grey zones between war and peace – drone strikes against alleged terrorists take place on a regular basis, causing a great deal of collateral damage, including the death of innocent civilians.
The research question of this article is whether collateral damage – in the form of injuring or killing innocent civilians – caused by the use of armed drones, should be seen as morally permissible. First, this article will analyse the issues surrounding drone strikes causing collateral damage. Secondly, peacetime principles related to collateral damage will be discussed. Thirdly, the relevant principles of the “just war theory” concerning collateral damage will be analysed, and more specifically a doctrine relevant to the principle of discrimination, namely the doctrine of double effect. Finally, these principles will be applied to the ethical problems of collateral damage as they relate to drone strikes in the grey zone between war and peace.

Drone Strikes Causing Systematic Collateral Damage

At first sight, drones seem to have many advantages. They can be used to do the “3 Ds”: dull, dirty and dangerous work.3 The surveillance drone has the

1     Jeremy Sherlick, Greg Bruno, u.s. War in Afghanistan” Council on Foreign Relations
2     Cora Currier, “Everything we know so far about drone strikes” ProPublica (New York, 11 January 2013)  <https://www.propublica.org/article/everything-we-know-so-far-about-drone-strikes>.
3     Dean Irvine, “Doing military’s dangerous, dull and dirty work” cnn (Georgia, 16 February 2012) <http://edition.cnn.com/2012/02/15/business/singapore-airshow-drones/>.



possibility of staying in aloft unseen for long periods of time above the target, to observe it. Human Rights Watch states that high-resolution cameras can be used to guide missiles from armed drones toward any target desired.4 Nevertheless, in these grey zones between war and peace,5 many incidents with armed drones involving collateral damage take place.
 
According to Pakistan’s Bureau of Investigative Journalism, more than 400 drone strikes have taken place in Pakistan since 2004, with fewer than 4% of those killed having been identified as named members of al-Qaeda. According to other sources, the amount of killed terrorists due to drone strikes in Pakistan between 2004 and 2012 is estimated to be 2% of the total amount of casualties.6 The UK-based human rights group Reprieve calculated that it takes the equivalent of about 28 innocent lives to neutralize a single terrorist leader and often multiple drone strikes.7 In October 2015, a whistle-blower provided the American journal The Intercept with secret documents detailing the inner workings of us drone attacks in Afghanistan, Yemen and Somalia. The documents revealed that approximately 90 percent of the persons killed by drones in these regions were innocent civilians.8
While the media is an easy source for statistics, such should be treated with caution. People should pay close attention for organisations that use and manipulate statistics for their own political agenda. Besides, it is not always easy to find out the truth, which makes manipulation possible. In countries like Pakistan, there are no birth registers and few emergency services that record deaths; moreover, bodies are usually buried within 24 hours of death, well before an official research team arrives to investigate properly.9 However, much secrecy exists surrounding the official United States number of casualties

4     “Precisely wrong: Gaza civilians killed by Israeli drone-launched missiles” Human Rights Watch (2009) <http://www.hrw.org/sites/default/files/reports/iopt0609webwcover_0.pdf>.
5     The characteristics of these grey zones between war and peace are discussed in the fourth part of this article.
6     Jack Serle, “Drone strikes in Pakistan: Civilian casualties are inevitable, and the u.s. appears willing to accept them” The Bureau of Investigative Journalism (London, 16 October 2014)
7     “It takes 28 civilian lives to kill a single terrorist leader uk human rights group” rt (London, 25 November 2014) <https://www.rt.com/news/208527-reaping-lives-drone-strikes/>.
8     “The Drone Papers” The Intercept (New York City, 2015) <https://theintercept.com/ drone-papers/>.
9     Lode Vanoost, “Drones en terrorisme, zoek de verschillen” De Wereld Morgen (Brussel, 27 Sep- tember 2012) <http://www.dewereldmorgen.be/artikels/2012/09/27/drones-en-terrorisme

disclosed to the public. The American people may not be aware of the actual number of innocent civilians killed by drone strikes in American operations to date.10 Finally, the question should be asked who exactly should be considered innocent civilians, and who are just targets. By considering every man killed between 15 and 66 years old, who happens to be in the vicinity of a targeted terrorist, to also be a terrorist, you can drastically decrease the figures for collateral damage.11
 
How does one explain these enormous collateral damage statistics? In traditional wars, like the two World Wars, enemy soldiers usually wore uniforms with clear insignia, which made them easy to identify. In such situations, a soldier wearing a uniform lost his right not to be attacked. However, as more and more conflicts are being fought in cities where enemies do not usually wear uniforms, deliberately embedding themselves amongst civilian populations whereby they become human shields, distinguishing between those who can and cannot be attacked has become very difficult.12 This creates very difficult situations for the use of armed drones because their missiles can be directed to a precise location, but the missiles have an impact of several square meters.13 Consequently, in these situations, armed drones systematically cause collateral damage.
 
Risk reduction has always been an important component of warfare. With the invention of long-range weapons, such as artillery, fighter-bombers and the Tomahawk missile, the risks faced by the soldier on the battlefield have already been greatly reduced. However, drone warfare takes this an enormous step further, as military deployment near the battlefield is no longer an operational requirement. Drones can be launched and operated from safe places – even neighbouring countries – that are hundreds of kilometres away from their targets. For instance, the United States has been using an airfield in Saudi Arabia as a base from which to launch unmanned aerial vehicles for conducting
 

10         Jon Queally, “Senate Agrees: us People Can’t Know Overseas Drone Death Toll” Common- Dreams (Portland, 29 April 2014) <http://www.commondreams.org/about-us>.
11         Jo Becker, Scott Shane, “Secret ‘Kill List’ Proves a Test of Obama’s Principles and Will” The New York Times (New York City, 29 May 2012) <http://www.nytimes.com/2012/05/29/ world/obamas-leadership-in-war-on-al-qaeda.html?_r=0>.
12         Mark Thompson, “The Taliban’s Low-Tech Defense Against u.s. Drones” Time (Wash- ington, 29 April 2009) <http://content.time.com/time/nation/article/0,8599,1894370,00
13         Joakim Kasper, Oestergaard Balle, “About the Predator and Reaper” Aeroweb (Newtown, 27 June 2016) <http://www.fi-aeroweb.com/Defense/MQ-1-Predator-MQ-9-Reaper.html>.

surveillance and combat missions in Yemen.14 Moreover, people are no longer sent on the ground anymore in these dangerous states to gather intelligence. The physical risks when using armed drones are thus reduced to zero, leading to the transfer of risk to innocent civilians.
Unreliable and incomplete intelligence is often the cause of incidents with high collateral damage.15 The difficulty of gathering intelligence depends, amongst other things, on the kind of war or conflict that is occurring. Analysing whether a person is a member of a terrorist organisation can be an enormous challenge for intelligence services, considering the organisation’s deliberate intermingling between civilians without uniforms. However, in many cases, a lack of boots on the ground means that there are simply no operatives physically present to gather intelligence. Intelligence specialists indicate that it would be better if surveillance drones were not the only sources of intelligence when it comes to verifying a target because intelligence gathered by these means remains subject to a high degree of uncertainty. Generally, these specialists pay a lot more attention to Human Intelligence (HUMINT), which will usually mean that the information collected is significantly more reliable.16 In Afghanistan, for example, in regions where troops and intelligence agents are indeed in place, ground commanders can bring important context

14         Micah Zenko and Emma Welsh, “Mapping the launch pads for Obama’s secret wars.” Foreign Policy ( Washington, 29 May 2012) <http://foreignpolicy.com/2012/05/29/where-the
15         Spencer Ackerman, “41 men targeted but 1,147 people killed: us drone strikes the facts on the ground” The Guardian (London, 24 November 2014) <http://www.theguardian
See also: us lacks intelligence to continue waging indiscriminate drone warfare in Yemen” rt (30 January 2015) <http://www.rt.com/usa/227715-us-yemen-intelligence-drones/>.
Matt Schiavenza, “Drones and the Myth of Precision Civilian casualties are inevitable, and the u.s. appears willing to accept them.” The Atlantic (Washington, 24 April 2015)
16         Marina Petrova, “The Drone Paradigm: Surgical Precision or Intelligence Mishaps” Future Foreign Policy (London, 3 June 2015) <http://www.futureforeignpolicy.com/ drone-paradigm-surgical-precision-intelligence-mishaps/>.
See also: Warren Strobel and Mark Hosenball, “Hostage locations difficult to track – and may be getting harder” Reuters (London, 24 April 2015) <http://www.reuters.com/ article/us-usa-security-hostages-intelligence-idUSKBN0NF0B320150424>.
“Former High-Level nsa Official: Drone Strikes by Metadata Alone ‘Undisciplined Slaughter’” Washingtonsblog (11 February 2014) <http://www.washingtonsblog.com/2014/ 02/former-high-level-nsa-official-drone-strikes-metadata-alone-undisciplined-slaughter

to pictures gathered from the sky. When an intelligence analyst has a question, he or she can make a call to a colleague on the ground for clarification. An intelligence analyst says: Without a commander on the ground, that puts that responsibility on us to be able to take every piece of data and make it make sense to the supporting commander. That is a harder challenge to do certain targets in that environment”.17 Bellamy describes a famous example of an operation where collateral damage occurred because of unreliable intelligence: the bombing of a wedding in Afghanistan in 2002. Initially, the United States indicated that the gathering in question was a reunion of Taliban members; however, following scrutiny by the media, the government was forced to admit that it was actually a wedding party and that the strike was authorized on the basis of faulty intelligence supplied by Afghans.18 Invariably, many operations depend on unreliable intelligence.19
 
Another consequence of such dependence on intelligence coming solely from surveillance drones is that the intelligence is incomplete: Because analysing images is a time-consuming activity, there is insufficient time to arrive at a complete intelligence picture. People present on the ground often need less time to gain insight to the situation. One example of such incomplete intelligence is a drone strike that resulted in collateral damage in Afghanistan in 2010. In this case, the intelligence indicated that the proposed target was a civilian convoy, but this information was overlooked by the drone operators, who were at the time under pressure and overloaded with information.20 A second example of incomplete intelligence is more recent. A United States
 

17         Marcus Weisgerber, “A Look Inside a Secret us Air Force Intelligence Center” Defense One (Washington, 18 November 2014) <http://www.defenseone.com/technology/2014/11/ look-inside-secret-us-air-force-intelligence-center/99347/>.
18         Alex J Bellamy, “Is the war on terror just?” (2005) 19(3) International Relations 275.
See also: Luke Harding, “No us apology over wedding bombing” The Guardian (London, 3 July 2002) <http://www.theguardian.com/world/2002/jul/03/afghanistan.lukeharding>.
19         Spencer Ackerman, “41 men targeted but 1,147 people killed: us drone strikes the facts on the ground” The Guardian 24 November 2014) (http://www.theguardian
See also: us lacks intelligence to continue waging indiscriminate drone warfare in Yemen” rt (30 January 2015) <http://www.rt.com/usa/227715-us-yemen-intelligence-drones/>.
Matt Schiavenza, “Drones and the Myth of Precision Civilian casualties are inevi- table, and the u.s. appears willing to accept them.” The Atlantic (Washington, 24 April 2015) <http://www.theatlantic.com/international/archive/2015/04/drones-and-the-myth
20         Jean-Baptiste Jeangène Vilmer, “Légalité et légitimité des drones armés” (2013) 3 Poli- tique étrangère 119 <http://politique-etrangere.com/wp-content/uploads/2013/10/ article_Jeangène-Vilmer_PE-3-2013.pdf>.

drone killed an American and an Italian hostage in an attack on an al Qaeda compound in Pakistan in January 2015. According to U.S. officials and lawmakers, the intelligence that underpinned the drone strike turned out to have been tragically incomplete.21 These are only three out of many cases.22
 

Peacetime and Collateral Damage

Principles Relating to Collateral Damage during Peacetime

The use of violence during peacetime is only possible in exceptional circum- stances, to protect oneself or a third party. Three conditions should be respected during defence actions: the action must be necessary, proportionate and targeted against an imminent threat.23 Suppose that the police have to end a hostage situation by neutralizing a hostage-taker. In this case, in order to neutralize the hostage-taker, the police will need to enter a room full of innocent civilians. When using force during peacetime, officers must carefully take into account the consequences of their actions, including any potential risks to civilians. The right to life is too valuable for that during peacetime. The police can sometimes use alternatives to stop the criminals, like rubber bullets. This may reduce the likelihood of collateral damage occurring. In very rare cases, collateral damage may be acceptable, such as in a national security context where an airplane that has been hijacked by terrorists is heading toward a populated area and is shot down.24
 

See also: WJ Editors, “In New Military, Data Overload Can Be Deadly” Small Wars Jour- nal (Maryland, 17 January 2011) <http://smallwarsjournal.com/blog/in-new-military-data
21         Adam Entous, Damian Paletta, Felicia Schwartz, “American, Italian Hostages Killed in cia Drone Strike in January” The Wall Street Journal (New York City, 23 April 2015)
See also: Spencer Ackerman, “41 men targeted but 1,147 people killed: us drone strikes the facts on the ground” The Guardian (London, 24 November 2014) <http://www
22         “It takes 28 civilian lives to kill a single terrorist leader uk human rights group” rt (Lon- don, 25 November 2014) <https://www.rt.com/news/208527-reaping-lives-drone-strikes/>. See also: “Only 84 of 2,379 us drone attacks victims in Pakistan confirmed Al-Qaeda militants report” rt (London, 18 October 2014) <http://www.rt.com/news/197100-usa
23         David Rodin, War and Self-Defense (11th edn, Clarendon Press 2002) 40.
24         Ben Jones and John M. Parrish, “Drones and dirty hands” 14 <http://wpsa.research.pdx

Police officers are sometimes exposed to a high level of threat. How much personal risk can and should a police officer take during peacetime? Police officers, in contrast to soldiers participating in hostilities, normally retain their right not to be killed. When police officers exceptionally and voluntarily risk their lives in order to protect the lives of civilians, this type of risk-taking must rather be considered a “supererogatory” act.
In the aftermath of the terrorist attacks in France (2015) and Belgium (2016), many countries introduced a state of emergency, whereby certain human rights were curtailed and police were given more power. The question arises whether the police may cause collateral damage out of necessity when con- fronted with people using very unjust methods, like terrorists. The proportionality principle states with regards to using violence: the more important your objective, the more harm you may cause. However, this is not in agreement with the principle that forbids any killing of innocent civilians intentionally during peacetime. In the Bataclan, the concert hall targeted during the Paris Attacks of November 2015, people were killed by the terrorists quickly and at random. Dynamic responses from police services were needed: when innocent people are dying, swift intervention is needed in order to save as many lives as possible. There is no time to get a clear picture of the situation or to try to negotiate. What matters most at that moment of intervention is that collateral damage is kept to a minimum and the agents involved are not ex- posed to unreasonable levels of risk. The doctrine of double effect is applicable here: You have the intention to kill the terrorists, who are not afraid of killing people, while entering the building with the side-effect of an unintentional but readily foreseeable killing of innocent civilians.25 We must emphasize that, during peacetime, this doctrine can only be used in these kinds of exceptional circumstances.

Importance of Alternatives

On the one hand, the lives of civilians should never be endangered. On the other, police officers normally should not have to risk their lives either. Consequently, dangerous situations like hostage situations are best avoided if it is possible. For example, intelligence gathering in order to arrest the suspect per- son can prevent hostage situations. A recent idea in the fight against terrorism is the pentito for terrorists, which offers terrorists the possibility of receiving a new identity in exchange for crucial information about a terrorist network.26

25         The doctrine of double effect is explained in greater detail in the third part.
26         Mark Eeckhaut, “Salah Abdeslam had de perfecte spijtoptant kunnen zijn” De Stan- daard (Groot-Bijgaarden, 20 February 2016) <http://www.standaard.be/cnt/dmf20160219

This could lead to faster dismantling of terrorist organisations. Interesting is a project ongoing in the U.S. where the government is currently trying to prevent school shootings. The idea is that since school shooters are predictable, they are also detectable, which means that people should keep an eye out for potential warning signs and be vigilant.27 Local good will is crucial when using this method. Beyond the standard methods employed by the country’s intelligence services, the government is trying to make citizens aware and to encourage them to watch out for potential threats, since it knows that the police cannot detect everything that is suspicious on their own. This approach also seems to work, at least in part, for detecting terrorists. Unfortunately, our societies are characterized by a lack of social control. It is not always possible to monitor everyone’s habits. When the hostage-taking can no longer be avoided, the alternative to lethal action is negotiations. These could be helpful once the criminal action is going on, in order to try to convince the hostage-taker to end the hostage situation.
 

War and Collateral Damage

Introduction to the Just War Theory

The just war theory has existed for more than 2000 years and has been developed by many thinkers, including ancient Greek philosophers and Christian theologians. The theory fits in a category with many other philosophies on warfare. In short, pacifism holds that war should always be avoided. Militarism holds that war is inherently good and can even improve cohesion in society. Realists believe that war should be fought for one’s own purpose, when one’s own interests are at stake. On the contrary, the just war theory prefers peaceful relations between states, but admits that sometimes the use of force is justified; for instance, to defend against an aggressor who has no just cause to invade your country.28 The American philosopher Michael Walzer reactualized this theory in the seventies with his book Just and Unjust Wars. The main objective of the just war theory is to offer guidance for dealing with actual wars and conflicts. The theory, which has an important normative component, speculates about when states are justified in waging war jus ad bellum and
 
 
 

27         “Preventing School Shootings” <http://preventingschoolshootings.com/>.
28         Carl Ceulemans, Over oorlog en ethiek: de traditie van de rechtvaardige oorlog in theorie en praktijk (1th edn, Garant 2011) 24–29.

in what ways states should behave during the prosecution of a war – jus in bello.29
Prior to the First and Second World Wars, national sovereignty was very important in international politics. States themselves decided when to wage war and when not to. The devastating nature of both World Wars made it abundantly clear that the Westphalian system, in which sovereign states enjoyed a quasi-limitless “compétence de guerre”, was no longer tenable. From that time onward, the right to resort to military violence was seriously curtailed. Article 51 of the United Nations Charter stated that only force used in self-defence against an ongoing attack was permissible. During this time, the concept of jus ad bellum was introduced in international law.30
The utility of jus in bello or fighting limited wars is twofold. It concerns limiting the total amount of suffering, but also keeping the possibility open for states to normalize their post-conflict relations. To be avoided is frustrating a party, because he thinks the end result seems unjust, or because the military action was unnecessarily cruel and in contradiction with the rules of war.31 As a consequence, one of the main principles of jus in bello is the moral equality of combatants, which gives all of the different parties in the war the same rights and duties, regardless the justness of their cause to go to war.32
In the last couple of years the jus post bellum has been developed. Here, the theory speculates about the post-war period and principles applicable to it.33 This article will henceforth discuss jus in bello and the issues related to collateral damage.
 

The Just War Theory and Collateral Damage

The Non-absolute Discrimination Principle

Before determining whether the just war theory permits collateral damage, it would be just as good to consider whether the use of an armed drone is itself in conflict with the principles of jus in bello and more specifically, the
 
 

29             Michael Walzer, Just and Unjust Wars: a moral argument with historical illustrations
(11th edn, Basic Books 1977) 21.
30             Ceulemans (n 28) 122–123.
31         Walzer, Just and Unjust Wars (n 29) 132–133.
32         Walzer, Just and Unjust Wars (n 29) 36–37.
33         Brian Orend, “Jus post bellum: The perspective of a just-war theorist.” (2007) 20(3) Leiden Journal of International Law 571.

discrimination principle, which states: once soldiers put on their uniforms, they can be attacked at any time by the other party. Unlike soldiers, innocent civilians have done nothing wrong to deserve losing their right to not be at- tacked.34 According to the just war theory, weapons or methods that are mala in se are “bad in themselves”, meaning that they should never be used on the battlefield due to their non-discriminatory and destructive effects; examples of this include genocide, rape and the use of biological, nuclear and chemical weapons or mines.35 Other examples are cluster munitions and barrel bombs. None of these effects are applicable to armed drones. As previously stated, drones can be used as very precise weapons, an armed drone is thus not necessarily a mala in se weapon. Even during peacetime it does not seem impossible that the police will at some point use armed drones to neutralise threats. Suppose you use a drone to save the life of an innocent person in a situation where the police are unable to be physically present. The weapon in itself does not change the ethics of the situation, assuming it reaches its target without resulting in collateral damage.
 
According to the just war theory, the fact that the level of risk is diminished for only one of the parties involved in a conflict due to the fact that they are using drones – does not necessarily mean that this is immoral, as long as the use of these drones is consistent with the principles of jus in bello and the possibility remains for stable situation afterwards. However, Walzer states that risk diminishment and risk transfers are not compatible. An ethically responsible person, a soldier being no exception to this rule, ought not to save his own life if it implies killing innocent civilians. Soldiers must ensure that civilians are not affected by the atrocities of war, even if the soldiers have to take additional risks.36 According to Bellamy, who agrees with Walzer, when non-combatants are only protected as long as the lives of combatants are not in danger, this is a morally dubious position because it gives more weight to the lives of combatants than to non-combatants.37
Suppose a party has intentionally killed the civilians of their opponent, should the opponent then be morally permitted to react by also killing their adversary’s civilians? The motivation could be to anticipate this kind of cruel behaviour in the future. This is often the reason for reprisals. Walzer states: “The helplessness of the victims rules them out as objects of military attack, and their non-involvement in criminal activity rules them out as objects of

34             Exceptions exist like civilians actively participating in hostilities. 35            Vilmer (n 20) 120.
36        Walzer, Just and Unjust Wars (n 29) 155–156. 37              Bellamy (n 18) 289–290.

retributive violence.”38 Walzer is convinced that neither civilians nor prisoners of war may become the target of a reprisal. All reprisals against civilians should be morally condemned. It would be better to use alternative measures, for example negotiations or military action against enemy soldiers. Gross states: “If civilian immunity is to have any meaning whatsoever, then ordinary civilians cannot be made to pay for the crimes of others, regardless of the actions of one’s state (or quasi-state)”.39 Gross does not forbid reprisals completely, as long as the right to life is not violated: for example, property can be damaged.40 According to the just war theory, there is a situation in which it may be permissible to intentionally kill innocent civilians. Walzer states that the deliberate killing of the innocent is murder, however, in some extreme situations, commanders may have to commit murder or order others to do so. “In such situations, these individuals will have killed unjustly, perhaps for the sake of justice itself; nevertheless, justice itself requires that their actions be condemned.”41 Walzer refers to this kind of situation as a “supreme emergency”.42 It is an exceptional situation that forces an individual to break a moral rule, as also discussed while considering the peacetime principles regarding collateral damage. Jones and Parrish talk about the presence of moral conflict between two real yet incompatible moral values or obligations, where practically no available action permits the agent to avoid violating a deeply held moral principle.43 There is another way to justify actions that imply the killing of innocent civilians, which will be further elaborated in the next part.
 

War and the Doctrine of Double Effect

The Doctrine of Double Effect according to Walzer

In the 13th century, Thomas Aquinas introduced the doctrine of double effect to justify self-defence: by defending yourself, the purpose is to save your own life, so killing the unjust threat is an unintentional foreseen side-effect.44 This was a reaction against Augustine’s reasoning, which stated that self-defence

38         Walzer, Just and Unjust Wars (n 24) 213–214.
39         Michael L. Gross, “Killing civilians intentionally: double effect, reprisal, and necessity in the Middle East.” (2005) 120(4), Political Science Quarterly 569.
40             Gross (n 39) 570.
41         Walzer, Just and Unjust Wars (n 29) 323.
42         Walzer, Just and Unjust Wars (n 29) 323.
43         Ben Jones and John M. Parrish (n24) 5.
44         Jonathan Spelman, "The morality of killing in self-defense: A Christian perspective”, (1993) 8 Trans. by Thomas Williams (Indianapolis: Hackett) 2.

was not allowed, because it was just an act of self-love.45 The doctrine of double effect also found its way into the just war theory. It is possible that innocent civilians can be implied in the attack as well, although indirectly. Indirect or unintentional collateral damage can be unforeseeable, for example in an accident, but can also be foreseen. Accidental and unintentional collateral damage can have as its causes, for example, meteorological changes (e.g., wind speed), technical problems, or human error. Unintentional but foreseen collateral damage is, for example, an attack on a factory where it is known that innocent people of the neighbourhood are also likely to be killed. Here, Walzer refers to the doctrine of double effect. According to this doctrine, it is sometimes permissible for a negative effect to occur as a side effect of trying to do something positive (hence there is a “double effect”). In this respect, four conditions need to be fulfilled:
(1)                    The act must be good in itself or at least neutral, which means, for our purposes, that it is a legitimate act of war; (2) Its direct effects must be morally acceptable, e.g. the destruction of military supplies or killing of enemy soldiers; (3) The intention of the actor must be good, meaning that his aim is to achieve the acceptable effect, while not employing any evil effects as means to this end; (4) The good effect must be sufficiently good enough in order to compensate for permitting the evil effect to occur.46
Or, otherwise stated, the collateral damage must be proportionate to the military necessity.
 
However, Walzer thinks respecting these 4 conditions is still inadequate. He prefers to use the term “double intention” instead of double effect, reformulating the third condition as follows: “the intention of the actor must be good, meaning that his aim is to achieve the acceptable effect without using any evil effects as means to this end, and, being aware of the evil involved, seeks to minimize it and accept whatever the personal costs of doing so may be”.47 Not instrumentalising people is not enough, you also need to limit the killing of non-instrumentalised people as much as possible, even by taking one’s own risks if necessary.
What degree of risk “personal costs” should an attacker take in order to minimize, as much as possible, the risk of collateral damage? Walzer speaks of a “price to pay” in order to keep the number of innocent civilian casualties as

45         Spelman (n 52) 2.
46         Walzer, Just and Unjust Wars (n 29) 153.
47         Walzer, Just and Unjust Wars (n 29) 155.

low as possible. In such situations, Walzer argues that citizens need to be given “due care”. He criticizes for example the 1999 NATO air campaign in Kosovo be- cause it disproportionately harmed innocent civilians and their property. He argues that ground troops should have been used in place of air strikes so as to reduce the number of civilian deaths, even if this would have meant increased risks for NATO soldiers.48 However, what should be the limits of this “price to pay”? According to Walzer, the willingness to accept additional risk should be limited by considerations of military effectiveness. It would be “illogical to take additional risks when this would doom the mission to failure”; this is when the “price to pay” stops.49 But, in such situations, should the mission then continue or be aborted? A lot will depend on military necessity. One solution would be to abort the mission or to try to find an alternative; however, military operations will often continue despite collateral damage when military objectives are deemed high priority.
 
Morally relevant to this discussion could be whether civilians are intentionally used as human shields. Has action been taken to remove the innocent civilians from the neighbourhood of the military targets? In Vemork, Norway, during the Second World War, the heavy water plant had to be destroyed al- though it was surrounded by innocent civilians who were not evacuated.50 Hamas, a Palestinian Islamic organisation, used civilians as human shields in 2014 while firing rockets into Israel.51 If the civilians are used on purpose as a human shield in this case, Hamas carries the moral responsibility for these killings. In the case of being forced by its leaders, the innocent civilians cannot be blamed. Another option could be that the civilians have chosen to be there. In the case of the free choice, the civilians are morally responsible as well, and this would tend to plead in favour of authorising more collateral damage, unless they do not know that they are being used as human shields. However, women and children would probably never consent to become human shields. The risk to suppose that human shields will not be attacked is too high for that. There is also the danger that we do not know the real intentions of the

48         Michael Walzer, Arguing about War (1th edn, Yale University Press 2004) 99–104.
49         Walzer, Just and Unjust Wars (n 29) 156.
50         This destruction was vitally important to stop the production of heavy water, so as to delay the development of an atomic bomb by German scientists.Walzer, Just and Unjust Wars (n 29) 156.
51         Matthew Blake “Hamas admits it did use schools and hospitals in Gaza Strip as ‘hu- man shields’ to launch rocket attacks on Israel but claims it was ‘mistake’” Mail Online (12 September 2014) <http://www.dailymail.co.uk/news/article-2753176/Hamas-DID-use

innocent civilians. This could lead to slippery slopes. An enemy surrounding himself with innocent civilians is not an action without moral consequences, otherwise they can surround all military targets with innocent civilians un- punished. Then it could be asked why personal risks still should be taken. How- ever, the notion of voluntary human shields is rather unrealistic and forced human shields should not be punished for the immoral actions taken by their leaders. Consequently, the moral consequences can never be transferred to the innocent civilians.
Another morally relevant issue concerns whose civilians – the enemy’s or yours – will become the subject of the military actions. Suppose that an aggressor invades your country. You fight back and the opponent uses your own civilians as a human shield. Suppose that you push the enemy back to his country and he uses his own civilians as a human shield. You will try to protect your own people as much as possible, but should the same personal risks be taken to protect the enemy’s women and children, whom they have deliberately used as human shields? Human shields should always be protected, to the extent possible, even with personal risks, if necessary. Walzer states: “But the structure of rights stands independently of political allegiance; it establishes obligations that are owed, so to speak, to humanity itself and to particular human beings and not merely to one’s fellow citizens.”52
A last relevant moral issue concerns unintentional benefits due to unintended foreseen military action that can occur during a military action. The problem lies not with these unintended foreseen benefits themselves. “To rule out unintended harms that bring unintended side benefits would severely limit the scope of counterinsurgency warfare and conventional aerial bombardment.”53 However, as soon as military operations take place because of these expected intended advantages, this becomes morally problematic.

The Third Condition of the Doctrine of Double Effect according to Walzer, in Greater Detail

Epistemological Problems
Conditions one, two and four of the doctrine according to Walzer are verifiable: It concerns facts, even if it is not always easy in war to discover the truth. You can investigate whether the act and its consequences are legitimate, and whether the amount of collateral damage is proportionate to the purpose – although comparing between military advantages and collateral damage is,
 

52             Walzer, Just and Unjust Wars (n 29) 158. 53    Gross (n 39) 563, 564.

again, very difficult. It is possible, in theory, to verify the amount of collateral damage, although problems occur, like parties hiding or destroying evidence, or even cultural customs, such as burying the bodies very quickly after their death.
In contrast, what is more difficult to determine is whether “his aim is to achieve the acceptable effect, while not employing any evil effects as means to this end”, or in other words, to discover whether the killing of innocent civilians was intentional, or rather, unintended but foreseen. Intentions are more difficult to ascertain in war than, say, during a case of self-defence in a peace- time situation. In a peacetime context, the perpetrator of violence will normally be interrogated and arrested if necessary, and he will be forced to tell the truth, which is why he will mostly not so readily represent an intentional action as an unintended foreseeable act. In war, the different parties do not appear automatically before a court, so they do not automatically have to justify controversial actions.
Besides, it is often more difficult to determine whether an action was really militarily necessary, due to multiple complex factors, in contrast to the more clear personal motivation of self-defence. Nowadays, more and more conflicts are fought out in cities amongst civilians, where the right intentions are even more difficult to appreciate, and there is therefore more occasions to manipulate the truth. The question is whether we should still base ourselves on the unverifiable intentions of an agent when weighing the justness of collateral damage, because it does not necessarily concern self-defensive action, but an action taken out of a more complex military necessity. Gross remarks that it is not about “the agent’s subjective intentions or motives but about his plans and the means he uses to achieve them”54 However, plans can also be manipulated. Your plan could include the destruction of a military factory in a crowded neighbourhood but your real intention could be to hit the innocent civilians to lower morale. Gross answers that you could investigate the importance of the military objective and weigh this against the utility of the collateral damage and make suppositions based on these facts. When collateral damage was not necessary to destroy the military objective, this is a suspect situation. The killing of innocent civilians would then be deemed intentional.
A last epistemological problem is whether the agent is aware of the complete situation, which is complex. Does he know or should he have known that collateral damage would be caused if carrying out the attack?
 
 
 
 

54             Gross (n 42) 562.

Different Visions on the Third Condition of the Doctrine of Double Effect
When the doctrine of double effect is discussed in the literature, the paradigmatic cases given to distinguish between intentional acts and unintentional but foreseen acts are those of the strategic bomber and the terrorist bomber.55 They eventually both kill innocent civilians, but the terrorist bomber, who intentionally kills them, is morally condemned in the framework of the doctrine of double effect, while the strategic bomber, who unintentionally but foresee- ably kills innocent civilians, is morally spared. This point of view is not without its critics. According to the consequentialists, the result of both actions is the same: whatever the intention may be, innocent civilians die. The difference between intentional and unintentional but foreseen consequences is often very blurry: some say it is the same thing: the strategic bomber kills unintentionally but foreseeably, but if you know with enough certainty that your action may cause collateral damage, then some will state that your unintentional but foreseen action is also intentional. You also “use” civilians intentionally, or the actions are at least morally equal. Or, is it right to say that strategic bombers, who are said to be unintentional in causing collateral damage, are acting justly, while terrorist bombings, which are considered to intentionally kill innocent civilians, are unjust acts?
 
To answer this question, we will rely on the reasoning of Quinn and Mc- Mahan. In the first instance, Quinn asks whether a moral distinction should be made based on the character of the connection between what is intended and the resulting, foreseen harm: “If the connection is close enough, then the doctrine should treat the harm as if it were strictly intended.56 Quinn gives the example of the strategic bomber bombing an automotive factory: he would not target civilians, but the fact is that the factory is built in a crowded neighbourhood. He states: “So the kind of thing the bomber strictly intends immediately and invariably results in some innocent deaths.57 However, according to Quinn, the strategic bomber and the terrorist bomber would be morally condemned and that is not what our moral intuition tells us. Other moral differences should thus be found.58
 

55      Warren S. Quinn, “Actions, intentions, and consequences” (1989) 18(4) Philosophy and Public Affairs 336.Jeff McMahan, “Revising the doctrine of double effect” (1994) 11(2) Journal of Applied Philosophy 201.
56        Quinn (n 55) 338.
57        Quinn (n 55) 338.
58        Quinn (n 55) 338.

Nor does it matter that only the terrorist bomber wishes for the civilians to die, because for the terrorist bomber, they need not die, because appearing to be dead is enough to intimidate his opponent; yet, after the conflict the civilians may be seen to be alive once again. In contrast, the relevant moral difference is that the terror bomber needs the civilians to serve his goal. “What matters is that the effect serves the agent’s end precisely because it is an effect on civilians.”59 Quinn calls this “direct agency in the production of harm”.60 In case of the strategic bomber, the civilian’s involvement does not serve his goals. “Perhaps the strategic bomber cannot honestly say that his effect will be unintentional in any standard sense, or that he does not mean to kill them because the relation between what he intends and the resulting foreseen harm is very close, but he does not use them to attain his purpose”.61 Quinn calls this “indirect agency in the production of harm”.62
 
Quinn makes another distinction in “direct agency”. When the agent “benefits from the presence of the victims”, then exploiting them, as the terrorist bomber does, is deemed to be a form of direct opportunistic agency. When the agent “aims to remove an obstacle or difficulty that the victim presents”, this is called direct eliminative agency.63 An enemy using human shields is one example. Direct opportunistic agency is morally worse than direct eliminative agency, but the latter is also unethical because: “Someone who gets in your way presents a strategic problem, a causal obstacle whose removal will be a service to your goals.”64
Finally, according to Quinn, a moral operator should be introduced: moral rights. Direct agency is only to be morally condemned if independent moral rights are violated.65 For example, “non-combatants (even those on the wrong side) are not morally obligated to serve the right side by accepting the role of demoralizing civilian casualties,” so their moral rights are violated. “The victims are made to play a role in the service of the agent’s goal that is not morally required of them.”66 The difficulty when including moral operators while judging an action is to find a consensus about what exactly is a rights violation, and what is not.
 

59             Quinn (n 55) 342.
60             Quinn (n 55) 344.
61             Quinn (n 55) 342.
62             Quinn (n 55) 341–342.
63             Quinn (n 55) 344.
64             Quinn (n 55) 350.
65             Quinn (n 55) 349.
66             Quinn (n 55) 349.

A second person that has tried to develop the doctrine of double effect is Jeff McMahan. This is his description about what an unjust action in the framework of the doctrine of double effect should be: “An act is an instance of potentially harmful direct agency if it is intended to affect or involve a per- son P in the agent’s plans and if one of the following is among the effects that the act is strictly intended to have: [i] an effect on P that is itself bad for P or constitutes a harm, [ii] an effect on P that, while not itself a harm given the description under which it is strictly intended, is nevertheless such that the agent believes that there is an alternative description of it under which it is bad for or constitutes a harm to P, or [iii] an effect on P that, while not itself a harm given the description under which it is strictly intended, is nevertheless believed by the agent to be harmful to P because it is believed to be causally sufficient, in the circumstances, either for a significant harm to P or for a high probability of such a harm.”67
 
When applied to the self-defence case, any harm done toward the unjust threat is a “potentially harmful direct agency”. Indeed, it satisfies condition i: it concerns an effect on the attacker that is negative for the attacker. The application of criterion iii would also condemn self-defence: the effect created by the defender on the attacker is meant to involve the attacker, and although the harm itself is not intended, it is believed to be causally sufficient for a significant harm to the attacker. Criterion ii is also fulfilled, as there exists an alternative description to “save myself” under which it is bad for or constitutes a harm to P, like saying that “the defender attacks the attacker”.
 
McMahan thinks that the causal connection between one’s action and the resulting foreseen harm to the victim is morally relevant to the question of whether or not to condemn a moral action. However, McMahan adds other moral relevancies. McMahan states that the difference between the two bombers should lie in the fact that in case that there is a close relation between the agent’s action and the foreseen resulting harm, the strategic bomber has no plans to harm the innocent civilians in opposition to the terrorist bomber. A difference has to be made between “intending to involve a person” and “in- tending to do something that involves a person”. This is similar to Quinn’s condition of using or involving a person to serve the agent’s goals. Uniacke in this context proposes a “test of failure” to test intentionality: would the mission fail if the harmful effects were avoided?68 If the answer is yes, then the harm is unintentional but foreseen.

67        McMahan (n 55) 209–210.
68        Gross (n 39) 562.

McMahan counters the critics that his three criteria morally condemn self-defence, which is counterintuitive, by adding moral operators like Quinn did when introducing moral rights, called “nullifiers”. These are supplementary conditions to morally judge an action or not. Harmful direct agency is impermissible unless nullified by various moral considerations or moral operators such as moral guilt, moral duty, and the consent of the person who is being harmed.69 In the case of self-defence, the moral guilt of the attacker does indeed nullify the wrongness of the harmful direct agent that is the defender in this case. In the case of the strategic bomber and the terrorist bomber however, there are no nullifiers, and so it satisfies the criterion of harmful direct agency: “the persons who are harmed have no duty to serve as victims, have not consented to be victims, and are not liable to be treated as victims by virtue of being guilty or dangerous.”70 This risks morally condemning both cases. However, McMahan sees the difference with “in the agent’s plans”. In the rest of the article, we assume that the civilians have moral rights not to be killed and that they are innocent.
The Third Condition of the Doctrine of Double Effect and the Use of Armed Drones
Because we see no relevant moral difference between the two concerning causing collateral damage, the use of armed drones can be compared to the strategic bomber case, but with a more systematic character. When collateral damage is caused with a high probability, and systematically, then there is definitely a close relation between the agent’s actions and the resulting foreseen harm. However, that criterion alone is not enough to morally condemn the action.
 
The first question to be answered is whether the caused collateral damage is part of the agent’s plans. We may assume that the users of armed drones do not kill innocent civilians intentionally, although it is strange that, according to some governments, every male person between 15 and 66 years old in the proximity of a terrorist is also considered a terrorist. However, means and methods are also part of the plans and they should also be judged. If these armed drones systematically cause collateral damage, then it becomes a weapon mala in se. When a police officer has to kill someone in a mass of people during peace- time, then his weapon has to be precise enough to eliminate the threat. He could try to get closer to the target, or use rubber bullets. When the criminal

69         McMahan (n 55) 211.
70         David R. Mapel, “Revising the doctrine of double effect” (2001) 18(3) Journal of Applied Philosophy 268.

protects himself with five people around him, then the police officer cannot shoot, unless the suspect starts killing bystanders or starts threatening increasing numbers of people. A car driver who knows that his breaks are broken, and then unintentionally kills a person, is negligent and will also be punished by criminal law. His systematic, recidivist behaviour will be punished more severely. These are two examples of lack of precaution because of improper means or methods. This also counts for armed drones that are made improper weapons or weapons mala in se by the enemy because the terrorists often surround themselves with innocent civilians.
 
The second question to be answered is whether moral rights are violated or if nullifiers are present. In the cases of the strategic and the terrorist bomber, we have just concluded that indeed moral rights are violated and that no nullifiers are present. There are no morally relevant differences between these cases and drone strikes in relation to the innocent civilians. Consequently, based on this criterion, the drone strikes would be also morally condemned.
 
The third question concerns the second part of the third condition of Walzer’s doctrine of double effect: you “should try to minimize the evil and accept whatever the personal costs of doing so may be”. Is this condition satisfied in the case of the use of armed drones? In war, armed drones can be used to assist ground troops, so the physical risks taken are not always necessarily zero. If the risk of collateral damage is too high, alternatives can be investigated; for example ground troops can be used to carry out the operation. In the fourth part of this article, it will become clear that this is problematic in the grey zone between war and peace. It must be remarked that minimizing risks does not necessarily imply taking one’s own risks: you could try to negotiate with the enemy, or try to warn the civilians by spreading flyers from an airplane so that they have the chance to leave the neighbourhood of the military target. However, even supposing that the civilians are forewarned, they could still be forced to stay.
Must all military operations causing any collateral damage that does not satisfy these conditions really be morally forbidden? Some parties surely use human shields as protection in a systematic way, it is like they are wearing a protective vest. These enemies do this because they know that the armed drones are not capable of targeting just one person in a group of people, because of the enormous impact of the drones’ missile. And, if the armed drone is the only weapon available in the region, then there are no alternative interventions. The enemy then forces his opponent to effectively commit war crimes or to not react. As a consequence, the enemy can do what he wants when violating the discrimination principle, by surrounding himself with civilians.

However, the availability of other appropriate means or methods does not put us in this morally problematic situation. The enemy always used human shields, for example during sieges, or they use factories in a populous neighbourhood for military purposes. Human shields remind us of Quinn’s direct eliminative and direct opportunistic agency. Human shields are rather the object of direct eliminative agencies: they are an obstacle, and the agent has no purpose to exploit them. However, Quinn condemned both agencies, because even the removal of the obstacles will also “be a service to your goals”. Because of the physical proximity of ground troops, alternatives are present. And even if there is no better alternative to apply, the fact that alternatives exist already makes a moral difference.
 

Other Opinions on Collateral Damage

Revisionists of the just war theory disagree with the moral equality of combat- ants, arguing that unjust combatants should get fewer rights than just combatants once a war has started. What counts to them is the fact that to lose the right not to be attacked involves moral responsibility, regardless if you are a soldier or a civilian.71 However, McMahan provides nuances: in most cases, civilians will not be so responsible that they should be attacked. Often, less drastic measures are enough, like economic sanctions.72
 
Some of these revisionists state that just combatants should have more rights than is currently the case, like the right to kill non-combatants intentionally. “On this view, a combatant fighting a war of exceptional justness and importance might possess exceptional permissions to inflict high levels of collateral harm on non-combatants”.73 Compared to peacetime, just combatants are similar to the police, in the sense that they have many rights, and the unjust combatants and terrorists are similar to the criminal, in the sense that they have fewer rights. After all, the police officer fights for a just cause, namely to guarantee the security of society, while the criminal tries to destabilise the public order. However, a bank robber does not rob a bank according to any honour code, but he does not lose all his rights either. The guard cannot shoot the bank robber just like that. The most important difference however is the causing of collateral damage. As seen in the previous section, the police officer

71         Jeff McMahan, “The Ethics of Killing in War” (2004) 114 Ethics 719, 720.
72         Jeff McMahan (n 71) 729.
73         David Rodin, “Liability of Ordinary Soldiers for Crimes of Aggression” (2007) 6 The
Washington University Global Studies Law Review 591.

cannot cause collateral damage, even if his cause is obviously just, except in exceptional circumstances. In our view, it is a bridge too far to allow the intentional killing of innocent civilians in non-exceptional circumstances, however unjust the opponent is. If you kill them intentionally, then you are no better than a terrorist. When confronted with terrorism, the proportionality principle applies: because your central rights are in danger, including the right to life, you can react harder, but this must not conflict with the discrimination principle.
The Grey Zone between War Peace and Peace and Collateral Damage
Characteristics of the Violence Used in the Grey Zone between War and Peace
Walzer states: “However, in the places in between, such as in states that lose control of parts of their country or are wracked by civil war in which terrorists can set up camp, the situation has a different ‘feel’ because … it happens outside the moral and legal conventions of ordinary warfare”.74 Terrorists take advantage of situations where states are unable or unwilling to stop them. As a reaction, armed drones are used, often by other states, like the United States. It is not possible to speak of peacetime, because of the amount of violence, but no conventional war is going on at present either. At first sight, there seems to be a new “context” that we shall call the grey zone between war and peace. However, this grey zone has characteristics of both peace and war contexts, so it is a hybrid context.
 
Different from classic warfare but similar to law enforcement is the locality of the use of violence, because of the targeted character of armed drones. The purpose of the drone strikes is not to conquer a country, but to diminish or eliminate a terrorist threat with precise actions against one person or a small group. The use of violence is often less frequent than during war; one drone strike takes place and then it can be quiet again for a considerable time. To overpower or to exhaust the enemy by means of successive actions, like in war, is not the actual objective. Because of this controlled, local and less frequent use of violence, the danger of unpredictable consequences like those seen in war seems more limited.75 The drone-using state is capable of doing this for a time: the drone-attacks are relatively cheap and lives from your own side are not lost. It seems to have become an endless or eternal conflict between a clearly just and a clearly unjust party, namely those who fight terrorism, and the terrorists themselves. Many of these characteristics remind us of law enforcement.

74         Michael Walzer, “On fighting terrorism justly” (2007) 21(4) International Relations 480.
75         Michael Walzer, “Regime change and just war” (2006) (53)3 Dissent 106.

Indeed, there too, the used violence is intentional but controlled – it is local and less frequent, and is mostly focussed on one specific person (or a group of identified persons, or a behaviour) in order to neutralise their threat. There too, it is clear who is the just party, and who is the unjust party, in this end- less situation of law enforcement, namely the law enforcer and the suspect or criminal.
 
However, the intensity of the violence is extremely heavy, with the use of missiles carrying an enormous payload, often with many injured, dead, and indeed collateral damage. This reminds us of warfare, and the local population often perceives it as such.76 Conventional police forces do not seem to be equipped to deal with this level of threat. The violence used is cross-border, which implies a violation of the sovereignty of the countries who are unable or unwilling to help stop the terrorists. Because it is through rather invisible drones instead of airplanes crossing the borders, this border-crossing is often tolerated by other countries. The opponent is an unjust threat, but it is a different one than that seen during ordinary law enforcement action: it is more than “just” a criminal, it involves terrorists.77 Many definitions of terrorism exist. It is important to us that they do it for a public purpose, and the method used is to kill as many innocent civilians as possible, in order to create fear.78
Principles Related to Collateral Damage in the Grey Zone between War and Peace
The question arises now which principles should be applied when drone at- tacks take place, if they can be used after all in this non-war context. Should peacetime or wartime principles be applied, or something between the two? One approach could be to apply the same principles in this grey zone as would be applied during peacetime. However, this would ignore the character of the threat, which is now of a terrorist nature. Because the threat of terrorism focuses on innocent civilians, we could say that this threat is more serious than during war, where soldiers are normally only targeting other soldiers, and not civilians. In that sense, the necessity to stop the threat is on average more important here than it is during war.79 The allowed amount of collateral damage needs to be
 

76         Christian Enemark, “Armed drones and the ethics of war: military virtue in a post-heroic age” (2013) Routledge, 75.
77         It should be noted that terrorism is not inherent to the grey zone, it is also possible during peace and wartime.
78         There are also more moderate forms of terrorism. See: Walzer (n 29) 198–201.
79         The question is what is considered a necessity. In war, generally, necessity is what con- tributes to military victory. During peacetime, necessity will be considered in a more context-dependent way, case by case. On the other hand, during wartime, necessity needs

evaluated against this necessity, like during peacetime in exceptional circum- stances. If the amount of force a government is allowed to use in the grey zone is too restricted, terrorist groups may decide to maintain a sufficiently low level of violence in order to prevent the use of heavy weapons like armed drones from being used against them because no war is going on but enough violence is used to easily defend against the weapons and methods accompanied with law enforcement. Adapted principles seem necessary here, otherwise terrorists would also systematically use human shields, and then no reaction would be authorized because during peacetime, causing collateral damage is not allowed, even when not unintentional and unforeseen. However, in this grey zone, causing collateral damage should also be accompanied by the same conditions as during wartime. The improper weapons and means are, just like in war, part of the agent’s plans, even if he does not want to cause systematic collateral damage. Besides, unlike during a more conventional war, where drones assist ground troops who are exposed to a certain physical risk, armed drones are used here without any other physical presence in the region. When alternatives are not an option at all, when personal risks cannot be taken to minimise collateral damage, then this becomes morally problematic. We can conclude that the third condition of the doctrine of double effect according to Walzer is not satisfied. One reaction could be that this extreme situation of a clearly just and unjust party warrants the just party to act without risk exposure, similar to the police officer who normally should not be exposed to risks. We would respond that first of all, the main problem here is that there is no possibility at all to minimise collateral damage, taking personal risks is only necessary if there are no other alternatives. If there must be the possibility to cause collateral damage, so then a price must be paid. If the armed drone is the only option, because it is the only weapon in the region at that moment, it cannot be concluded that the discrimination principle should be mitigated, because then you are adapting moral reasoning to a technology. Users of armed drones in these grey zones have two options: making the armed drones a just weapon, meaning that they respect amongst other things the conditions of the doctrine of double effect while causing collateral damage, or forbidding their use entirely.
 
Drone Strikes in the Grey Zone: Alternatives to Minimise Collateral                Damage
Can unintentional foreseen collateral damage be minimised when using armed drones? If intelligence services were present on the ground, the amount of
 

sometimes also to be considered case by case: for example, when soldiers are attacking civilians.

potential collateral damage could be, on average, reduced. Walzer is clear about this: “Collecting information about the targeted individuals, their schedules, their whereabouts, their families and their neighbours, is critically important, and if it involves risk for agents in the fields, the risks must be accepted before the killing can be justified”.80 Again, taking personal risks is not always necessary but will very often be the consequence. For example, gathering HUMINT will involve personal risks: sending infiltrators, spies or installing bugs all imply physical risk exposure. However, is it reasonable to put boots on the ground in remote tribal areas in Pakistan? If the price is too high, it might be better to not carry out these drone strikes and focus on alternatives or to authorize drone strikes only when it is certain that no innocent bystanders are present. It also might be possible to try to limit causes with a less high cost in physical risk, for example solving the communication problems or a better management of the overload on information. In practice, the higher the physical cost of some- thing, the more it is difficult to change it.
So what other alternatives are possible? Non-lethal alternatives would be a solution; here, fewer people would be killed and collateral damage could be partly avoided. In the future, mini-drones much like a robot insect could sting the target with poison. Robot insects could sit somewhere close to an enemy and collect information about their intentions. It does not seem unrealistic to suppose that, eventually, drones will be capable of capturing alleged terrorists – perhaps through the use of mechanical arms – and transporting them to places where they can be arrested and guaranteed a fair trial. However, these technologies are still under development. The alternative of spreading flyers to warn the innocent civilians is not realistic: first, a drone cannot spread flyers and two, it is a drone’s purpose to stay quiet and surprise the enemy.
Today, we could ask whether enough communication has taken place with states that harbour terrorists. Do these governments willingly permit safe havens to exist or is it something out of their control? Do these governments willingly permit drone strikes or are they viewed as a violation of national sovereignty? In any case, cooperation with these governments would be ideal, since then national police or armed forces could be used to intervene as opposed to sending drones. Moreover, to what extent can a political solution be found to respond to the threat

80         Michael Walzer, “Targeted Killing and Drone Warfare” 11 Dissent Magazine (New York, Jan- uary 2013) <https://www.dissentmagazine.org/online_articles/targeted-killing-and-drone
81         David Cortright, “License to kill” (2012) cato Unbound <http://www.cato-unbound

of terrorism itself? This may be the only way of reaching a stable and peaceful outcome.81 In any case, the collateral damage associated with drone strikes leads to the recruitment of future terrorists. It leads to a vicious circle. When people see their family and friends die around them, this creates resentment.
 

Conclusions

The use of armed drones is not in itself necessarily problematic. In theory, drones have the capacity to neutralize their targets very precisely. However, recent drone strikes taking place in the so-called grey zone between war and peace, as is witnessed in countries like Pakistan and Yemen, have caused a great deal of collateral damage. Terrorists make the armed drones out to be mala in se weapons, i.e. weapons that systematically cause collateral damage, because they surround themselves with human shields just as though they were wearing protective vests. The enemy knows that their enemy has no alternatives present, other than killing with that drone, and they then take advantage of that situation.
 
The question is whether this reactive, systematically caused, collateral dam- age can be morally justified. Although some philosophers morally allow the intentional killing of innocent civilians as a function of the justness of the cause for which they went to war, we think that collateral damage must be reduced as much as possible. The doctrine of double effect states that is it possible to cause unintentional but foreseen collateral damage. This damage must not be in the agent’s plans. When the connection between your action and the foreseen resulting harm is clear, consequently, because of the systemic character of the collateral damage, the inappropriate weapons and methods used become part of the agent’s plans. In war, ground troops can normally be considered an alternative to eliminate the threat, and they might intervene physically and will expend maximal effort to protect civilians while taking risks themselves if needed to minimise collateral damage, but in this grey zone between war and peace, where drones are the only means available in the region, where the intelligence gathered is often incomplete and unreliable, you cannot take any alternative to minimise the collateral damage. If the drones’ users are not capable of stopping the violations of the conditions of the doctrine of double effect, they must find alternatives instead of drone strikes to stop the terrorist threat.