War

War can be necessary and proportionate only if it serves an end worth
all this death and destruction. Hence the importance of having a just
cause. And hence too the widespread belief that just causes are few
and far between. Indeed, traditional just war theory recognizes only
two kinds of justification for war: national defence (of one’s
own state or of an ally) and humanitarian intervention. What’s
more, humanitarian intervention is permissible only to avert the very
gravest of tragedies—“crimes that shock the moral
conscience of mankind” (Walzer 2006: 107).

Walzer argued that states’ claims to sovereignty and territorial
integrity are grounded in the human rights of their citizens, in three
ways. First, states ensure individual security. Rights to life and
liberty have value “only if they also have dimension”
(Walzer 2006: 58), which they derive from states’
borders—“within that world, men and women… are safe
from attack; once the lines are crossed, safety is gone” (Walzer
2006: 57). Second, states protect a common life, made by their
citizens over centuries of interaction. If the common life of a
political community is valued by its citizens, then it is worth
fighting for. Third, they have also formed a political association, an
organic social contract, whereby individuals have, over time and in
informal ways, conceded aspects of their liberty to the community, to
secure greater freedom for all.

These arguments for national defence are double-edged. They helped
explain why wars of national defence are permissible, but also make
justifying humanitarian intervention harder. One can in principle
successfully conclude a war in defence of oneself or one's allies without any lasting
damage to the political sovereignty or territorial integrity of any of
the contending parties. In Walzer’s view, humanitarian
interventions, in which one typically defends people against their own
state, necessarily undermine political sovereignty and territorial
integrity. So they must meet a higher burden of justification.

Walzer’s traditionalist stances on national defence and
humanitarian intervention met heavy criticism. Early sceptics (Doppelt
1978; Beitz 1980; Luban 1980a) challenged Walzer’s appeal to the
value of collective freedom, noting that in diverse political
communities freedom for the majority can mean oppression for the
minority (see also Caney 2006). In modern states, can we even speak of
a single common life? Even if we can, do wars really threaten it,
besides in extreme cases? And even if our common life and culture were
threatened, would their defence really justify killing innocent
people?

Critics also excoriated Walzer’s appeal to individual rights
(see especially Wasserstrom 1978; Luban 1980b). They questioned the
normative purchase of his metaphor of the organic social contract (if
hypothetical contracts aren’t worth the paper they’re not
written on, then what are metaphorical contracts worth?). They
challenged his claim that states guarantee individual security: most
obviously, when humanitarian intervention seems warranted, the state
is typically the greatest threat to its members.

David Rodin (2002) advanced the quintessentially reductivist critique
of Walzer, showing that his attempt to ground state defence in
individual defensive rights could not succeed. He popularized the
“bloodless invasion objection” to this argument for
national defensive rights. Suppose an unjustly aggressing army would
secure its objectives without loss of life if only the victim state
offers no resistance (the 2001 invasion of Afghanistan, and the 2003
invasion of Iraq, arguably meet this description, as might some of
Russia’s territorial expansions). If the right of national
defence is grounded in states’ members’ rights to
security, then in these cases there would be no right of national
defence, because their lives are at risk only if the victim state
fights back. And yet we typically do think it permissible to fight
against annexation and regime change.

By undermining the value of sovereignty, revisionists lowered the bar
against intervening militarily in other states. Often these arguments
were directly linked: some think that if states cannot protect the
security of their members, then they lack any rights to sovereignty
that a military intervention could undermine (Shue 1997).[11] Caney (2005) argues that military intervention could be permissible
were it to serve individual human rights better than non-intervention.
Others countenance so-called “redistributive wars”, fought
on behalf of the global poor to force rich states to address the
widespread violations of fundamental human rights caused by their
economic policies (Luban 1980b; Fabre 2012; Lippert-Rasmussen 2013;
Øverland 2013).

Other philosophers, equally unpersuaded by Walzer’s arguments,
nonetheless reject a substantively revisionist take on just cause. If
the individual self-defence-based view of jus ad bellum
cannot justify lethal defence against “lesser aggression”,
then we could follow Rodin (2014), and argue for radically
revisionist conclusions about just cause; or we could instead reject
the individual self-defence-based approach to justifying killing in
war (Emerton and Handfield 2014; Lazar 2014).

Some think we can solve the “problem of lesser aggression”
by invoking the importance of deterrence, as well as the impossibility
of knowing for sure that aggression will be bloodless (Fabre 2014).
Others think that we must take proper account of people’s
interest in having a democratically elected, or at least home-grown,
government, to justify national defence. On one popular account,
although no individual could permissibly kill to protect her own
“political interests”, when enough people are threatened,
their aggregated interests justify going to war (Hurka 2007; Frowe
2014). Counterintuitively, this means that more populous states have,
other things equal, more expansive rights of national defence.
However, perhaps states have a group right to national defence, which
requires only that a sufficient number of individuals have the
relevant political interests—any excess over the threshold is
morally irrelevant. Many already think about national
self-determination in this way: the population of the group seeking
independence has to be sufficiently large before we take their claim
seriously, but differences above that threshold matter much less
(Margalit and Raz 1990).

The revisionist take on humanitarian intervention might also have some
troubling results. If sovereignty and territorial integrity matter
little, then shouldn’t we use military force more often? As Kutz
(2014) has argued, revisionist views on national defence might license
the kind of military adventurism that went so badly wrong in Iraq,
where states have so little regard for sovereignty that they go to war
to improve the domestic political institutions of their
adversaries.

We can resolve this worry in one of two ways. First, recall just how
infrequently military intervention succeeds. Since it so often not
only fails, but actually makes things worse, we should use it only
when the ongoing crimes are so severe that we would take any risk to
try to stop them.

Second, perhaps the political interests underpinning the state’s
right to national defence are not simply interests in being part of an
ideal liberal democracy, but in being governed by, very broadly,
members of one’s own nation, or perhaps even an interest in
collective self-determination. This may take us back to Walzer’s
“romance of the nation-state”, but people clearly do care
about something like this. Unless we want to restrict rights of
national defence to liberal democracies alone (bearing in mind how few
of them there are in the world), we have to recognize that our
political interests are not all exclusively liberal-democratic.

What of redistributive wars? Too often arguments on this topic
artfully distinguish between just cause and other conditions of
jus ad bellum (Fabre 2012). Even when used by powerful states
against weak adversaries, military force is rarely a moral triumph. It
tends to cause more problems than it solves. Redistributive wars, as
fought on behalf of the “global poor” against the
“global rich”, would obviously fail to achieve their
objectives, indeed they would radically exacerbate the suffering of
those they aim to help. So they would be disproportionate, and cannot
satisfy the necessity constraint. The theoretical point that, in
principle, not only national defence and humanitarian intervention
could give just causes for war is sound. But this example is in
practice irrelevant (for a robust critique of redistributive
wars, see Benbaji 2014).

And yet, given the likely path of climate change, the future might see
resource wars grow in salience. As powerful states find themselves
lacking crucial resources, held by other states, we might find that
military attack is the best available means to secure these resources,
and save lives. Perhaps in some such circumstances resource wars could
be a realistic option.

3.2 Just Peace

The goods and bads relevant to ad bellum proportionality and
necessity extend far beyond the armistice. This is obvious, but has
recently received much-needed emphasis, both among philosophers and in
the broader public debate sparked by the conflicts in Iraq and
Afghanistan (Bass 2004; Coady 2008; May 2012). Achieving your just
cause is not enough. The aftermath of the war must also be
sufficiently tolerable if the war is to be proportionate, all things
considered. It is an open question how far into the future we have to
look to assess the morally relevant consequences of conflict.

3.3 Legitimate Authority

Historically, just war theory has been dominated by statists. Most
branches of the tradition have had some version of a
“legitimate”, “proper” or “right”
authority constraint, construed as a necessary condition for a war to
be ad bellum just.[12] In practice, this means that sovereigns and states have rights that
non-state actors lack. International law gives only states rights of
national defence and bestows “combatant rights” primarily
on the soldiers of states. Although Walzer said little about
legitimate authority, his arguments all assume that states have a
special moral standing that non-state actors lack.

The traditionalist, then, says it matters that the body fighting the
war have the appropriate authority to do so. Some think that authority
is grounded in the overall legitimacy of the state. Others think that
overall legitimacy is irrelevant—what matters is whether the
body fighting the war is authorized to do so by the polity that it
represents (Lazar forthcoming-b). Either way, states are much more
likely to satisfy the legitimate authority condition than non-state
actors.

Revisionists push back: relying on reductivist premises, they argue
that killing in war is justified by the protection of individual
rights, and our licence to defend our rights need not be mediated
through state institutions. Either we should disregard the legitimate
authority condition or we should see it as something that non-state
actors can, in fact, fulfil (Fabre 2008; Finlay 2010; Schwenkenbecher
2013).

Overall, state legitimacy definitely seems relevant for some questions
in war (Estlund 2007; Renzo 2013). But authorization is more
fundamental. Ideally, the body fighting the war should be authorized
to do so by the institutions of a constitutional democracy. Looser
forms of authorization are clearly possible; even a state that is not
overall legitimate might nonetheless be authorized by its polity to
fight wars of national defence.

Authorization of this kind matters to jus ad bellum in two
ways. First, fighting a war without authorization constitutes an
additional wrong, which has to be weighed against the goods that
fighting will bring about, and must pass the proportionality and
necessity tests. When a government involves its polity in a war, it
uses the resources of the community at large, as well as its name, and
exposes it to both moral and prudential risks (Lazar forthcoming-b).
Doing this unauthorized is obviously deeply morally problematic. Any
form of undemocratic decision-making by governments is objectionable;
taking decisions of this magnitude without the population’s
granting you the right to do so is especially wrong.

Second, authorization can allow the government to act on positive
reasons for fighting that would otherwise be unavailable. Consider the
claim that wars of national defence are in part justified by the
political interests of the citizens of the defending
state—interests, for example, in democratic participation or in
collective self-determination. A government may defend these
aggregated political interests only if it is authorized to do so.
Otherwise fighting would contravene the very interests in
self-determination that it is supposed to protect. But if it is
authorized, then that additional set of reasons supports fighting.

As a result, democratic states enjoy somewhat more expansive war
rights than non-democratic states and non-state movements. The latter
two groups cannot often claim the same degree of authorization as
democratic states. Although this might not vindicate the current bias
in international law towards states, it does suggest that it
corresponds to something more than the naked self-interest of the
framers of international law—which were, of course, states. This
obviously has significant implications for civil wars (see Parry
2016).

3.4 Proportionality

The central task of the proportionality constraint, recall, is to
identify reasons that tell in favour of fighting and those that tell
against it. Much of the latter task is reserved for the discussion of
jus in bello below, since it concerns weighing lives in
war.

Among the goods that help make a war proportionate, we have already
considered those in the just cause and others connected to just peace
and legitimate authority. Additionally, many also think that
proportionality can be swayed by reasonable partiality towards
one’s own state and co-citizens. Think back to the political
interests that help justify national defence. If we were wholly
impartial, then we should choose the course that will best realise
people’s political interests overall. So if fighting the
defensive war would undermine the political interests of the adversary
state’s citizens more than it would undermine our own, then we
should refuse to fight. But this is not how we typically think about
the permission to resort to war: we are typically entitled to be
somewhat partial towards the political interests of our
co-citizens.

Some propose further constraints on what goods can count towards the
proportionality of a war. McMahan and McKim (1993) argued that
benefits like economic progress cannot make an otherwise
disproportionate war proportionate. This is probably true in practice,
but perhaps not in principle—that would require a kind of
lexical priority between lives taken and economic benefits, and
lexical priorities are notoriously hard to defend. After all, economic
progress saves lives.

Some goods lack weight in ad bellum proportionality, not
because they are lexically inferior to other values at stake, but
because they are conditional in particular ways. Soldiers have
conditional obligations to fulfil their roles, grounded in their
contracts, oaths, and their co-citizens’ legitimate
expectations. That carrying out an operation fulfils my oath gives me
a reason to perform that operation, which has to be weighed in the
proportionality calculation (Lazar 2015b). But these reasons cannot
contribute to ad bellum proportionality in the same way,
because they are conditional on the war as a whole being fought.
Political leaders cannot plausibly say: “were it not for all the
oaths that would be fulfilled by fighting, this war would be
disproportionate”. This is because fighting counts as fulfilling
those oaths only if the political leader decides to take her armed
forces to war.

Another reason to differentiate between proportionality ad
bellum and in bello is that the relevant comparators
change for the two kinds of assessment. In a loose sense, we determine
proportionality by asking whether some option is better than doing
nothing. The comparator for assessing the war as a whole, then, is
not fighting at all, ending the war as a whole. That option
is not available when considering particular actions within the
war—one can only decide whether or not to perform this
particular action.

3.5 Last Resort (Necessity)

Are pre-emptive wars, fought in anticipation of an imminent enemy
attack, permissible? What of preventive wars, in which the assault
occurs prior to the enemy having any realistic plan of attack (see, in
general, Shue and Rodin 2007)? Neoconservatives have recently argued,
superficially plausibly, that the criterion of last resort can be
satisfied long before the enemy finally launches an attack (see
President 2002). The right answer here is boringly familiar. In
principle, of course this is possible. But, in practice, we almost
always overestimate the likelihood of success from military means and
overlook the unintended consequences of our actions. International law
must therefore retain its restrictions, to deter the kind of
overzealous implementation of the last-resort principle that we saw in
the 2003 invasion of Iraq (Buchanan and Keohane 2004; Luban 2004).

Another frequently discussed question: what does the
“last” in last resort really mean? The idea is simple, and
is identical to in bello necessity. Going to war must be
compared with the alternative available strategies for dealing with
the enemy (which also includes the various ways in which we could
submit). Going to war is literally a last resort when no other
available means has any prospect of averting the threat. But our
circumstances are not often this straitened. Other options always have
some chance of success. So if you have a diplomatic
alternative to war, which is less harmful than going to war, and is at
least as likely to avert the threat, then going to war is not a last
resort. If the diplomatic alternative is less harmful, as well as less
likely to avert the threat, then the question is whether the reduction
in expected harm is great enough for us to be required to accept the
reduction in likelihood of averting the threat. If not, then war is
your last resort.[13]

4.1 Walzer and his Critics

The traditionalist jus in bello, as reflected in
international law, holds that conduct in war must satisfy three
principles:

1. Discrimination: Targeting noncombatants is impermissible.[14]
2. Proportionality: Collaterally harming noncombatants (that is,
   harming them foreseeably, but unintendedly) is permissible only if the
   harms are proportionate to the goals the attack is intended to achieve.[15]
3. Necessity: Collaterally harming noncombatants is permissible only
   if, in the pursuit of one’s military objectives, the least
   harmful means feasible are chosen.[16]

These principles divide the possible victims of war into two classes:
combatants and noncombatants. They place no constraints on killing combatants.[17] But—outside of “supreme emergencies”, rare
circumstances in which intentionally killing noncombatants is
necessary to avert an unconscionable threat—noncombatants may be
killed only unintendedly and, even then, only if the harm they suffer
is necessary and proportionate to the intended goals of the attack.[18] Obviously, then, much hangs on what makes one a combatant. This entry adopts a
conservative definition. Combatants are (most) members of the
organized armed forces of a group that is at war, as well as others
who directly participate in hostilities or have a continuous combat
function (for discussion, see Haque 2017). Noncombatants are
not combatants. There are, of course, many hard cases, especially in
asymmetric wars, but they are not considered here. “Soldier” is used interchangeably with “combatant” and
“civilian” interchangeably with
“noncombatant”.

Both traditionalist just war theory and international law explicitly
license fighting in accordance with these constraints, regardless of
one’s objectives. In other words, they endorse:

Combatant Equality: Soldiers who satisfy Discrimination,
Proportionality, and Necessity fight permissibly, regardless of what
they are fighting for. [19]

We discuss Proportionality and Necessity below; for now let us
concentrate on Michael Walzer’s influential argument for
Discrimination and Combatant Equality, which has proved very
controversial.

Individual human beings enjoy fundamental rights to life and liberty,
which prohibit others from harming them in certain ways. Since
fighting wars obviously involves depriving others of life and liberty,
according to Walzer, it can be permissible only if each of the victims
has, “through some act of his own … surrendered or lost
his rights” (Walzer 2006: 135). He then claims that,
“simply by fighting”, all combatants “have lost
their title to life and liberty” (Walzer 2006: 136). First,
merely by posing a threat to me, a person alienates himself from me,
and from our common humanity, and so himself becomes a legitimate
target of lethal force (Walzer 2006: 142). Second, by participating in
the armed forces, a combatant has “allowed himself to be made
into a dangerous man” (Walzer 2006: 145), and thus surrendered
his rights. By contrast, noncombatants are “men and women with
rights, and… they cannot be used for some military purpose,
even if it is a legitimate purpose” (Walzer 2006: 137). This
introduces the concept of liability into the debate, which we need to
define carefully. On most accounts, that a person is liable to be
killed means that she is not wronged by being killed. Often this is
understood, as it was in Walzer, in terms of rights: everyone starts
out with a right to life, but that right can be forfeited or lost,
such that one can be killed without that right being violated or
infringed. Walzer and his critics all agreed that killing a person
intentionally is permissible only if either she has lost the
protection of her right to life, or if the good achieved thereby is
very great indeed, enough that, though she is wronged, it is not all
things considered wrong to kill her. Her right is permissibly
infringed. Walzer and his critics believe that such cases are very
rare in war, arising only when the alternative to intentionally
violating people’s right to life is an imminent catastrophe on
the order of Nazi victory in Europe (this is an example of a supreme emergency).

These simple building blocks give us both Discrimination and Combatant
Equality—the former, because noncombatants, in virtue of
retaining their rights, are not legitimate objects of attack; the
latter, because all combatants lose their rights, regardless of what
they are fighting for: hence, as long as they attack only enemy
combatants, they fight legitimately, because they do not violate
anyone’s rights.

These arguments have faced withering criticism. The simplest objection
against Combatant Equality brings it into conflict with
Proportionality (McMahan 1994; Rodin 2002; Hurka 2005). Unintended
noncombatant deaths are permissible only if proportionate to the
military objective sought. This means the objective is worth that much
innocent suffering. But military objectives are merely means to an
end. Their worth depends on how valuable the end is. How many innocent
deaths would be proportionate to Al Shabab’s successfully
gaining control of Mogadishu now or to Iraq’s capturing Kuwaiti
territory and oil reserves in 1991? In each case the answer is
obvious: none.

Proportionality is about weighing the evil inflicted against the evil
averted (Lee 2012). But the military success of unjust combatants does
not avert evil, it is itself evil. Evil intentionally inflicted can
only add to, not counterbalance, unintended evils. Combatant Equality
cannot be true.

Other arguments against Combatant Equality focus on Walzer’s
account of how one loses the right to life. They typically start by
accepting his premise that permissible killing in war does not violate
the rights of the victims against being killed, at least for
intentional killing.[20] This contrasts with the view that sometimes people’s rights to
life can be overridden, so war can be permissible despite infringing
people’s rights. Walzer’s critics then show that his
account of how we lose our right to life is simply not plausible.
Merely posing a threat to others—even a lethal threat—is
not sufficient to warrant the loss of one’s fundamental rights,
because sometimes one threatens others’ lives for very good
reasons (McMahan 1994). The soldiers of the Kurdish Peshmerga,
heroically fighting to rescue Yazidis from ISIL’s genocidal
attacks, do not thereby lose their rights not to be killed by their
adversaries. Posing threats to others in the pursuit of a just aim,
where those others are actively trying to thwart that just aim, cannot
void or vitiate one’s fundamental natural rights against being
harmed by those very people. The consent-based argument is equally
implausible as a general defence for Combatant Equality. Unjust
combatants have something to gain from waiving their rights against
lethal attack, if doing so causes just combatants to effect the same
waiver. And on most views, many unjust combatants have nothing to
lose, since by participating in an unjust war they have at least
weakened if not lost those rights already. Just combatants, by
contrast, have something to lose, and nothing to gain. So why would
combatants fighting for a just cause consent to be harmed by their
adversaries, in the pursuit of an unjust end?

Walzer’s case for Combatant Equality rests on showing that just
combatants lose their rights to life. His critics have shown that his
arguments to this end fail. So Combatant Equality is false. But they
have shown more than this. Inspired by Walzer to look at the
conditions under which we lose our rights to life, his critics have
made theoretical advances that place other central tenets of jus
in bello in jeopardy. They argued, contra Walzer, that
posing a threat is not sufficient for liability to be killed (McMahan
1994, 2009). But they also showed that posing the threat oneself is
not necessary for liability either. This is more controversial, but
revisionists have long argued that liability is grounded, in war as
elsewhere, in one’s responsibility for contributing to a
wrongful threat. The US president, for example, is responsible for a
drone strike she orders, even though she does not fire the
weapon herself.

As many have noted, this argument undermines Discrimination (McMahan
1994; Arneson 2006; Fabre 2012; Frowe 2014). In many states,
noncombatants play an important role in the resort to military force.
In modern industrialized countries, as much as 25 per cent of the
population works in war-related industries (Downes 2006: 157–8;
see also Gross 2010: 159; Valentino et al. 2010: 351); we provide the
belligerents with crucial financial and other services; we support and
sustain the soldiers who do the fighting; we pay our taxes and in
democracies we vote. Our contributions to the state’s capacity
over time give it the strength and support to concentrate on war.[21] If the state’s war is unjust, then many noncombatants are
responsible for contributing to wrongful threats. If that is enough
for them to lose their rights to life, then they are permissible
targets.

McMahan (2011a) has sought to avert this troubling implication of his
arguments by contending that almost all noncombatants on the unjust
side (unjust noncombatants) are less responsible than all unjust
combatants. But this involves applying a double standard, talking up
the responsibility of combatants, while talking down that of
noncombatants, and mistakes a central element in his account of
liability to be killed. On his view, a person is liable to be killed
in self- or other-defence in virtue of being, of those able to bear an
unavoidable and indivisible harm, the one who is most responsible for
this situation coming about (McMahan 2002, 2005b). Even if
noncombatants are only minimally responsible for their
states’ unjust wars—that is, they are not blameworthy,
they merely voluntarily acted in a way that foreseeably contributed to
this result—on McMahan’s view this is enough to make them
liable to be killed, if doing so is necessary to save the lives of
wholly innocent combatants and noncombatants on the just side (see
especially McMahan 2009: 225).

One response is to reject this comparative account of how
responsibility determines liability, and argue for a non-comparative
approach, according to which one’s degree of responsibility must
be great enough to warrant such a severe derogation from one’s
fundamental rights. But if we do this, we must surely concede that
many combatants on the unjust side are not sufficiently responsible
for unjustified threats to be liable to be killed. Whether through
fear, disgust, principle or ineptitude, many combatants are wholly
ineffective in war, and contribute little or nothing to threats posed
by their side. The much-cited research of S. L. A. Marshall claimed
that only 15–25 per cent of Allied soldiers in the Second World
War who could have fired their weapons did so (Marshall 1978). Most
soldiers have a natural aversion to killing, which even intensive
psychological training may not overcome (Grossman 1995). Many
contribute no more to unjustified threats than do noncombatants. They
also lack the “mens rea” that might make
liability appropriate in the absence of a significant causal
contribution. They are not often blameworthy. The loss of their right
to life is not a fitting response to their conduct.