1.1 Introduction

This chapter is concerned with the conditions when securitization is morally required. As such, this chapter does not make any claims about who has a duty to carry out securitization; consequently, it is silent on where specific actors’ duties to secure stem from (these issues are covered in subsequent chapters). Instead, it aims to locate the critical juncture when securitization is – from a moral point of view – not merely permitted (and thus optional) but obligatory (mandatory). Already, we can see from this that the critical juncture cannot depend simply on what actors think or believe they must do about a threat. Instead, the moral requirement to securitize is external to the beliefs of securitizing actors; that is, it must relate to the available evidence. I hold that the obligation to securitize arises from what has already been done to address the threat. In other words, the obligation to securitize rests on a prior obligation to politicize, by which I mean not simply the elevation of an issue into political forums where it is discussed but the situation when concrete if ordinary (non-exceptional) measures to address the issue are put in place. In more detail, I argue that securitization is morally obligatory only when other feasible and less harmful options have been tried once and demonstrably failed to satisfy just cause.Footnote ¹ Conversely, I hold that securitization is morally permissible before other less harmful options have been tried, namely when securitization is anticipated to have the best chances of achieving just cause compared to the chances of viable, less harmful, alternatives (Floyd, Reference Floyd2019a: chapter 5).

To make this argument, this chapter commences by explaining why in Just Securitization Theory (JST) obligation is bound to what I refer to as ameliorated moderate last resort (the situation when plausible and less harmful alternatives have been tried and have failed to meet the just cause), whereas permissibility is tied to reasonable chance of success, to wit when securitization is anticipated to be the ‘best option’ for addressing the just cause.

Overall, I show that last resort’s function regarding the obligation to securitize is comparable to just cause in the permissibility to securitize (cf. McMahan, Reference McMahan2005: 4 & 5),Footnote ² at least in so far as securitization cannot be required unless it is a last resort. In other words – and in a play on words – last resort is the must cause for the obligation to securitize.

It is important to note that JST’s initial just cause, as well as the other criteria concerning just initiation of securitization, is not replaced by must cause (bar chance of success); securitization cannot be obligatory in the absence of a just cause, or any of the other criteria of just initiation for securitization, because one cannot be morally required to perform an act or to do something that is morally impermissible.Footnote ³ Put differently:

Once I have laid this groundwork regarding last resort, I examine when last resort is satisfied regarding different threat types. To this end, this chapter works out plausible, less harmful, alternatives for a number of prominent and current threats and considers what factors influence for how long these ought to be tried before securitization is a last resort. The task of this chapter is not to do the impossible and to provide a blueprint that – for every possible threat scenario – reliably tells us when securitization is required. Instead, my aim is to tell users of JST how to go about establishing whether must cause is satisfied and on that basis when a relevant actor has a duty to securitize. The instructive nature of this chapter is an additional reason why all threat scenarios discussed refer to stylized and not real-world examples (cf. Introduction).

1.2 The Importance of Last ResortFootnote ⁴

Regardless of who has a moral duty to securitize and irrespective of the specific nature of the threat, what needs to be the case so that securitization is morally required (of relevant actors)? To answer this question, let us begin with the moral permissibility of securitization. In JST (Floyd, Reference Floyd2019a), securitization is morally permissible inter alia when it is judged ex ante to be the best option for satisfying just cause. To establish whether securitization is the best option, it is necessary to anticipate securitization’s chances of achieving the just cause and to compare these chances with those of plausible, less harmful, alternatives (notably politicization and purposeful inactionFootnote ⁵) at meeting the just cause. Only if securitization emerges as the best/most reasonable option is securitization permissible. In other words, securitization’s permissibility depends – in part – on the success condition.

That securitization is permissible when it is the best option is in line with our moral intuitions regarding self-defence. As Seth Lazar (Reference Lazar2012: 8) has argued: ‘Without taking a position on whether our moral theories should be evidence-relative, fact-relative, or both, it is clear that our primary interest when exploring the ethics of self-defence and war is in the evidence relative perspective, because the judgments it yields are much more in tune with our ordinary thinking about self-defense.’ Given that the obligation to securitize rests – like all duties – on finding a balance between the good done (the harm prevented) and the costs/harm incurred, the obligation to securitize could – provided all criteria are met – simply rest with the best option understood as having ex ante the best chances of success. To put this more simply, we might think that whenever all the permissibility conditions are met, it follows that the permissible option is also required – in short, that there is no gap between permissibility and requirement.

In the remainder of this section, I want to explain and defend why – where securitization is concerned – I do not support this inference. In other words, I shall explain why some securitizations that are morally justifiable are merely optional, while other securitizations are obligatory, even though any theory of the obligation to securitize must rest on a theory of permissibility. I will also explain why securitizations that are not morally required are – provided relevant criteria are satisfied – morally permissible. Pivotal to my reasoning is the logic of supererogation. A supererogatory act is one that is ‘good to do, but not wrong not to do’ (Heyd, Reference Heyd and Zalta2019: 9). By contrast, an obligatory act is right to do and wrong not to do.

Most security studies scholars I consulted on the question whether just cause is automatically a must cause agreed with me that there is, or should be, a split between the point when securitization is permissible and when it is obligatory. Although I too am of this view philosophically, the matter is not easy to justify. Hence, if securitization is the best option at point X to achieve just cause, then surely relevant actors at point X are also obligated to securitize (i.e., do the best thing). Or put another way, if there is no obligation at point X, then how can securitization be permissible at point X (cf. Oberman, Reference Oberman2015)?

The case against separating out just cause (the point when securitization is permissible) and must cause (the point when securitization is obligatory) in the way I do is strengthened further by the fact that obligation at point X is merely a pro tanto obligation that can be overridden by concerns pertaining to the costs and risks to the securitizing actor. For example, regarding humanitarian intervention, we can say that all just states have pro tanto obligations to save people from grave harm; however, they have this obligation only if the cost (notably in terms of own soldiers sacrificed) is not prohibitive to their state’s own flourishing (see, e.g., McMahan, Reference McMahan, Davis, Keshen and McMahan2010: 57). In other words, separating out just and must cause is not only illogical but it is also unnecessary for developing a theory of the obligation to securitize.

The crux of the objection against separating out permissibility and obligation is this: if securitization is the best thing that can be done to satisfy just cause then surely – from a moral point of view and all other things being equal – it must also be done. In what follows I want to first defend my view that – in the absence of a must cause – just securitization is supererogatory.Footnote ⁶ And second, I what to explain what changes the balance so that hitherto optional securitization becomes mandatory.

Supererogationists hold the view that ‘supererogatory actions lie entirely and without qualification beyond the requirements of morality and that is the source of their unique value’ (Heyd, Reference Heyd and Zalta2019: 15). The existence of supererogation is contested (see Oberman discussed later in this chapter). The philosopher Nora Grigore (Reference Grigore2019) suggests that this contestation comes about because our understanding of morality is wedded to good acts being obligatory to avoid them from not being done. In short, in much of moral philosophy, the realization of the moral deed depends on obligation. However, to claim that all good acts are obligatory also brings problems. For one thing, it makes morality very demanding (Grigore, Reference Grigore2019: 1163). For another, it leaves no room for voluntarily good acts (Heyd, Reference Heyd and Zalta2019: 23–28). The demandingness point is ameliorated by the fact that moral obligations are pro tanto duties only that can – in many cases – still be overridden by costs to the agents (this will be addressed in Chapters 2–4). The fact that supererogationism leaves room for voluntarily good acts is key. Thus, I concur with supererogationists that there is intrinsic value in human beings having moral choice.Footnote ⁷ That is to say, there is moral value in some things being done voluntarily and actors having, within a certain range of options, a choice on how to respond.Footnote ⁸ Put differently, just cause + other criteria for just initiation of securitization give positive justifications (or deontic reasons) for securitization, but these reasons are not strong enough to override the value of autonomy (cf. Grigore, Reference Grigore2019: 1155). We can see that supererogation is also the reason why a securitization that actors are not required to do is still permitted, provided, of course, the criteria for just initiation of securitization are satisfied.Footnote ⁹

Turning now to the second part of my argument, it is time to examine what changes this balance so that relevant actors no longer have a choice regarding how to act on just cause. In what follows I shall argue that a qualitative shift in certainty renders deontic reasons for securitization stronger than the value of autonomy, making securitization mandatory. I shall further argue that the certainty that securitization is the best option increases relative to the amount and quality of the evidence collected in support of that view.

Recall that in JST securitization is morally permissible – inter alia – based on ex ante judgements regarding the success of securitization satisfying just cause in relation to other less harmful options. I contest that while these reasons are (provided other relevant criteria are met) strong enough to render securitization morally permissible, they are not strong enough to trump the value of autonomy. This changes when we have concrete evidence that other options do not work. To achieve sufficient certainty, it is thus important that alternatives are tried out, because trying out (not anticipation) is the method to obtain sufficient evidence and consequently to generate sufficiently strong deontic reasons that are able to override the pre-eminence of autonomy.Footnote ¹⁰ The premise here is that we can truly know (be sufficiently certain) that an action is, out of a set of viable options, the best one only when we have eliminated these other options by trying them first and they have failed to satisfy just cause. In rare cases, it is possible that actors possess sufficient evidence that alternatives are unworkable even without having tried them first. For example, if a state is targeted with ballistic missiles, its government must – capability permitting – respond with anti-ballistic missiles, as it is already too late to try, for example, diplomacy. On this view, in summary, what matters is the actor’s evidence about the likely success of their options. Trying out an option is the obvious way of getting that evidence. In exceptional cases, where an actor has evidence without trying out an option, trying out is not required.

This analysis gives way to the following guiding principle:

To be sure, while trying alternatives leaves securitization to emerge as the best option, it might still only be the best option among several worse alternatives. As Simon Caney (Reference Caney2005: 244) explains, the best option (in terms of the success of satisfying just cause) is not judged against some absolute standard but ‘“relative” to the other options available’.

In philosophy, the requirement to try alternative, less harmful, options (at least once) first (i.e., before war ensues) is known as the last resort (Frowe, Reference Frowe2011: 64). Last resort does more than to provide actors with greater evidence and thus increased certainty whether securitization is the best option. Importantly, by doing other things first, the need for securitization can decrease, because other options might go some way towards dealing with a threat. This is apposite for a number of reasons. First, the resources for securitization (including money and numbers of security practitioners) that most would-be securitizing actors have at their disposal are finite. Imagine a scenario whereby securitization is judged the best option in ten cases, but the designated securitizing actor has the resources to securitize only in five such cases. How should this actor deliberate between which case must be securitized, especially if all – or more than five cases – would not produce undue costs to the securitizing actor, thus requiring securitization in all or more than five cases, because securitization is anticipated to have the best chances of succeeding in satisfying just cause. We might say that in such cases they ought to do the best that they can and say securitize in five cases. Yet how are actors to deliberate which are the five most important cases given that in all cases securitization is expected to be the best option for satisfying just cause? Additional evidence is likely to assist in whittling down the number of cases that require securitizing, because once other options have been tried, we not only gain greater certainty regarding the best option but also the need for securitization goes down as threats are partially addressed by these other options, in some cases making securitization unnecessary.

Second, a low threshold for the obligation to securitize can increase/trigger moral dilemmas – that is, ‘cases in which doing one thing we’re morally required to do rules out doing another thing we’re also morally required to do’ (Van der Vossen in Tesón and Van der Vossen, Reference Tesón and Van der Vossen2017: 217) – just states face. Such states are obliged to protect human rights and uphold civil liberties, but a requirement to securitize when it is anticipated to be the best option clashes with these other obligations because securitization often infringes these goods (cf. Floyd, Reference Floyd2019d).

While philosophers in particular might disagree with the differentiation between must cause and just cause, it seems to me that there is a precedent in the just war tradition for what I suggest here.Footnote ¹¹ Many just war scholars consider war permissible in line with Lazar’s earlier-mentioned view of evidence-relative certainty. However, it is also true that most just war scholars are not concerned with obligation. For the most part, the decision whether to go to war is considered the prerogative of states and not discussed further (cf. Oberman, Reference Oberman2015). The exception is formed by scholars theorizing humanitarian intervention. Most scholars writing on humanitarian intervention do so because they are in favour of the practice (an exception is Jackson, Reference Jackson2000; van der Vossen in Tesón and Van der Vossen, Reference Tesón and Van der Vossen2017) and because they think that able state actors have a duty to intervene when circumstances are right (see, for instance, Glanville, Reference Glanville, Gentry and Eckert2014). In my view, it is telling that many writers concerned with humanitarian intervention include as part of their substantive criteria the requirement of (if not by that name) ameliorated moderate last resort (e.g., ICISS, 2001: 36–37; Lango, Reference Lango2014: 140).Footnote ¹² It seems to me that while supporters of humanitarian intervention are concerned with the plight of people subject to grave human rights abuses, they are also wary of advocating the interference into state’s sovereignty. Consequently, for humanitarian intervention to be obligatory, we need sufficient certainty that it is the best thing to do. Sufficient certainty for the obligation to go to war for humanitarian reasons, however, rests with isolating war as actually the best option. This can be achieved only by trying other less harmful options (diplomacy, economic sanctions,Footnote ¹³ etc.) first.

While we can find a split between optional and mandatory wars in the wider just war literature (see, e.g., McMahan, Reference McMahan, Davis, Keshen and McMahan2010), some philosophers are challenging the view of optional wars or optional rescue killings. Helen Frowe (Reference Frowe2018), for example, has advanced the ‘Requirement Thesis’, which treats ‘lesser-evil justifications for harming for the sake of others as rescue cases. [It holds that] unless rescuing imposes a very serious psychological cost on the rescuer, of the sort that will seriously impede her capacity for flourishing, she is required to enact lesser-evil rescues’ (2018: 479). While Frowe’s formulation seems to challenge the distinction I draw between the permissibility and the obligation to securitize, consider that Frowe makes her case with reference to the well-known trolley dilemma in which a pedestrian is faced with the choice to divert a run-away trolley to save five people, yet whereby diverting the trolley will kill one innocent workman. This is important because it shows that Frowe does not allow for alternative options to killing. That is – in her scenario – last resort is satisfied in virtue of the fact that diverting the trolley is the only and hence actually best thing the pedestrian can do to save the five. None of this is at odds with my proposal. I hold that securitization is required only at the point when other – less harmful – things have been tried to divert the trolley, for example, driving a car onto the track in between the workman and the trolley, from which the driver is able to escape before impact.Footnote ¹⁴ And that when the car fails to stop the trolley by ramming it out of the way, I think securitization is obligatory.

A different critique is offered by Kieran Oberman. In a 2015 article, Oberman advances the Cost Principle, which holds that ‘those in power should not force people to render humanitarian assistance to those in dire need if the costs exceed that which people have a humanitarian obligation to bear’ (2015: 261). Put differently, he argues that it is morally impermissible for decision-makers to enact wars that pose unreasonable/supererogatory moral or financial costs on the executors of war (e.g., soldiers) and taxpayers. The flipside of this is that unless an otherwise just war incurs such costs, the war is not merely morally permissible but morally required. In other words, there is no such thing as an optional war. As will become apparent in subsequent chapters, I agree that the expected moral and financial costs to the securitizing actor play a major role in the actual obligation to securitize others (i.e., outsiders). In short, the moral costs and risks are duty-overriding or, as in the case of individual persons (see Chapter 3, Section 3.2), duty-voiding factors. This is the standard view in philosophy. McMahan (Reference McMahan, Davis, Keshen and McMahan2010: 57), for example, argues: ‘Among other things, while the question whether an act is permissible does not seem to depend on the cost to the agent of doing it, the question whether an act is morally required or obligatory does seem to depend on the costs.’ He makes this point by giving the following example:

Still, Oberman’s argument is a tough nut to crack because it requires a defence of government supererogation. Justin Weinberg’s (Reference Weinberg2011) work on the possibility of government supererogation is instructive here. It appears that one way to escape Oberman’s argument is to suggest that government supererogation is permissible, provided ‘through just procedures (say, some kind of deliberative democracy) we enact a policy which permits our government to perform such acts’ (Weinberg, Reference Weinberg2011: 275). Alas, for JST, this escape route does not work because the theory does not require legitimate authority or audience consent for just securitization (Floyd, Reference Floyd2019a: chapters 2 and 5). But this need not worry us. Weinberg makes it clear that invoking a procedural account of justice does not allow us to escape Oberman’s challenge because ‘outcomes of procedure matter, and those outcomes will be measured against a substantive account of justice’ (Weinberg, Reference Weinberg2011: 275–276). In other words, justice depends also on substantive forms of justice, not merely on just procedure.

While the procedural justice escape route from Oberman’s challenge then does not work, a division into different types of justice helps us to realize that government supererogatory acts are not necessarily ‘all-things-considered wrong’ (Weinberg, Reference Weinberg2011: 271). Many people might consider such acts right because they ‘maximise the good’ (Weinberg, Reference Weinberg2011: 278). Some scholars would consider risking our own soldiers’ lives to defend third parties virtuous (Kaldor, Reference Kaldor2012: 138–139; cf. Chapter 2, Section 2.3). For many, ‘Morality cannot be just a matter of considering your own needs and those of your nearest and dearest’ (Mulgan, Reference Mulgan2001: 135).

If we can accept that supererogatory acts are not necessarily wrong and hence impermissible, why then are they not also obligatory? Weinberg argues that whether or not an act is in fact required depends on the ‘theories of justice and goodness’ individuals accept (2011: 271). He suggests a test, whereby we must ask of the act in question whether it would ‘be unjust to fail to perform the act’ (Weinberg, Reference Weinberg2011: 271). Given that most people distinguish between permissibility and obligation, I believe that most reasonable people would accept that it would not be unjust for a state to fail to encounter harmful costs to itself and its people in order to save persons in other countries (indeed this is why Oberman’s account is so persuasive). In short, we can now say that a government’s act of supererogation aimed at satisfying a need for security of third parties is justified (here: permissible) because it benefits the insecure;Footnote ¹⁵ however, because it also produces significant costs to inter alia the executors of securitization, it is not required.

Another argument against Oberman’s argument can be extrapolated from David Runciman’s work on the moral agency of states. Runciman argues that the only way to make sense of states as moral agents at all is to consider them corporate as opposed to collective entities. Notably ‘a collective entity is nothing more than the sum of its parts, whereas a corporate entity is somehow separate from these […]’ (Runciman, Reference Runciman and Erskine2003: 41). According to Runciman it is only when considering the state as a corporate entity that we can make sense of the fact that ‘states can be acted for by their representatives’ (ibid, 41). Indeed, ‘states need representatives to act for them’ (ibid, 41). Although Runciman does not say this, it seems to me that it would be possible to argue that when representatives perform supererogatory acts that harm the people within the state, they act on behalf of the state as a separate corporate entity from the people, and that it is within their right to do so. In support of this argument also consider Ned Dobos and C.A.J. Coady’s (2014) observation (here with regards the role of the electorate in representative democracies) that ‘we empower our representatives to make decisions for us, and we do not feel that the government must always consult with us and make whatever decision is most popular’ (Dobos and Coady, Reference Dobos, Coady and Scheid2014: 93). Indeed, a third objection to Oberman’s position can be found in Dobos and Coady’s piece. They argue that since the electorate’s preferences are unlikely to be indisputably for or against intervention the public cannot meaningfully exercise its veto ‘in which case, we see no reason to think the decision is not covered by the authority vested in a representative government’ (Dobos and Coady, Reference Dobos, Coady and Scheid2014: 94).

To summarize the argument of this section, in line with our moral intuitions regarding self-defence I hold that when our concern is with the permissibility to securitize, then evidence-relative relative judgements regarding the success of securitization render sufficient certainty. Given inter alia the cost to securitizing actors and global stability, however, this same level of certainty is insufficient for the obligation to securitize. By insisting that other, viable and less harmful options are tried before securitization is obligatory, we increase certainty that securitization is the best thing to do to address the threat. Moreover, by trying other less harmful options first, the need for securitization in some cases is likely to decrease as some of these options might deal with the threat already.

1.3 The Nature and Origin of Threats in JST

The most important principle of JST is the just cause. However, ‘most important’ here does not mean that the satisfaction of just cause alone is sufficient for permitting securitization, instead, its special role arises from the fact that none of the other principles of just securitization can be satisfied unless just cause is (McMahan, Reference McMahan2005). In JST, just cause consists of two interrelated principles. The first is the just reason. The second is the just referent object of securitization. I have focused on the just referent object already in the Introduction; in this section, my concern is with the just reason. Notably, just reason is not about what kind of things may be saved by means of self-or other-securitization, but about when (i.e., under what circumstances) this can be done. I hold that securitization can be morally justifiable only if it concerns a real threat. Recall from the introduction that in JST security threats are socially and politically constructed, but that some security threats refer to real threats whereas others do not. Furthermore, and as also previously mentioned, I understand objective in the evidence-relative not the fact-relative sense. Fact-relativity refers to the situation ‘in which people know all of the relevant, reason-giving facts’ (Parfit, Reference Parfit2011: 162–163), while evidence-relativity refers to the situation when the available evidence suggests decisive reasons that the beliefs people hold about a given situation are true (Parfit, Reference Parfit2011: 150). This choice is justified because: fact-relative knowledge can, if at all, be obtained only with hindsight (cf. Herington, Reference Herington2013: 67). Moreover, the evidence gathered must approximate the actual facts, notably the failure to do so renders actors culpable. In life in general, all forward-looking decisions and choices are made in line with fact approximating evidence, indeed unless this is generally considered permissible, we could act only ever after the event. For example, governments plan and develop infrastructure years ahead in line with population projections for any given area, not after radical change in population numbers.

It is important to note that in JST objective existential threats are not tantamount to lethal threats to persons. This is for two reasons. First, a focus on lethal threats only would deny that some threats (notably some infectious diseases) have severe physically or mentally disabling effects, leaving people unable to be and function as humans should, but are not lethal. Second, regarding security, many relevant threats are not – in the first instance at least – lethal threats to people, but threats to non-human referents (states, political regimes, eco-systems etc.) instead. Such threats may be harmful to humans; indeed, I hold that securitization is proportionate only when the demise of a referent object as we know it (e.g., a state) is also sufficiently harmful to people, which in turn is the case when human well-being is severely undermined (i.e., to the point that people are objectively existentially threatened), but such threats are not necessarily lethal. Moreover, in the same way as humans do not have to die for a threat to be classed as existential, non-human referents do not have to be threatened with their actual disappearance to be existentially threatened. An established liberal democracy, for example, is existentially threatened not only when it is at risk at been invaded and annexed by a belligerent neighbouring state, but also when, for example, domestic forces seek to transform it into a dictatorship. In both cases the state as we know it will cease to exist, even if its territory, population etc. remain.

For all these reasons then, in JST existential threats refer to threats to the essential properties, character or being of putative referent objects. As such the terminology of ‘objective existential threat’ refers to:

1. (1) Direct lethal threats, meaning a threat is lethal to persons.

2. (2) Indirect lethal threats whereby the demise, or the significant alteration of an entity A (e.g., the state, an eco-system), entails the risk of death of people (e.g., when an aquatic ecosystem that disproportionally supports life of people turns toxic and leads to a famine).

3. (3) Direct objective existential but non-lethal threat, that is, threats that negatively compromise human well-being to the extent that humans cannot live in the way humans should.

4. (4) Indirect non-lethal objective existential threats whereby a threat to an entity A causes an objective non-lethal threat to people.

Conceptually the origins of all threats can be grouped into three threat categories. I differentiate between agent-intended threats, which refers to threats that one agent intentionally levels at another actor, order, or entity. An agent-lacking threat, in turn, is a threat that does not originate from human agents, for example, a vector-borne disease. Finally, third are agent-caused threats. This refers to a threat that is a consequence of an agent’s behaviour, but unlike in agent-intended threats, it is not intended by the threatening agent. I differentiate between two sub-types of agent-caused threats: (1) those caused by the offending agent’s obliviousness, that is, when people do not realize that their (combined) actions are potentially threatening to other entities; (2) such threat can also be caused by harmful neglect that is, when relevant agents fail to protect against the foreseeableFootnote ¹⁶ harmful events/consequences.

This concludes my discussion of threat types and origins of threats. These differences will become relevant again when we consider the length of time for which alternatives must be tried, for now it is necessary to discuss securitization and other less harmful alternatives on the basis of hypotheticals.

1.4 Securitization and Its Alternatives: Hypothetical Examples for Different Threat Types

In JST securitization is tantamount to the exception. Exceptionalism refers to measures and activities that are not considered acceptable in normal circumstances by reasonable persons, usually because of the risks or harm involved. However, even when we know that securitization refers to the exception the concept (i.e., securitization) is as slippery to grasp as a bar of soap in the bath. Unlike war, which inevitably means some use of kinetic force,Footnote ¹⁷ securitization is a shape-shifter. Thus, while we are able to say what securitization as the exception involves in the abstract, the nature of extraordinary measures in specific cases of securitization takes different forms. Accordingly, the threat from migration is or can be solved doing different things than say the threat from climate change, or that from cyber-attacks. For this reason, it is important that this section must not only discuss plausible, less harmful alternatives to securitization; it must also give some idea what securitization against different threats might look like. Given that the argument here follows on from JST, the security measures that make up securitization must at least in principle be justifiable; notably, for example, a lethal response to a non-lethal security threat is unjust (Floyd, Reference Floyd2019a: chapter 6).

Before the start of the analysis, it seems to me that plausible alternatives to securitization fall chiefly into two broad categories: (1) purposeful inaction and (2) politicization.Footnote ¹⁸ Securitization means both heightened politicization (in a sense that the issue is hyper-politicized) but, confusingly, also de-politicization (in so far as issues are moved out of the democratic process and subject to executive power) (Hansen, Reference Hansen2012: 521; Buzan et al, Reference Buzan, Wæver and de Wilde1998: 29). Importantly, politicization too is a shape-shifter. Thus, while politicization broadly means dealing with issues through the normal (often democratic political process) the realm of political solutions is vast and varied. Notably illegal unchecked immigration could be addressed by foreign aid and foreign direct investment into immigrant states, while climate change can be addressed by states transforming into carbon-neutral economies. In other words, what concretely politicization means is issue-dependent. This said, at the most basic level it means that issues are addressed using existing legislation that is applicable in normal times (i.e., not emergency legislation), or new – but non-emergency – legislation, that is, laws that are fully compatible with ordinary conduct. Politicization also means that the issue in question is not addressed using the military and traditional security establishment. At the risk of sounding banal, we can recognize politicization when actors do what they usually do in response to a type of issue or problem.

In this section I discuss the nature of securitization on a number of prominent threats across the different threat types (i.e., agent-intended, agent-caused, and agent-lacking). For each case this is followed by a discussion of the viable, less harmful alternatives. Where agent-intended threats are concerned I focus on cyber-threats and non-state terrorism. For agent-caused but not intended threats I focus on illegal immigration and climate change. For agent-lacking threats I focus on vector-borne infectious disease and natural disasters. It is important to notice that this discussion does not include an assessment of whether the substantive criteria of the just initiation of securitization as specified in JST are satisfied. All threats discussed could satisfy those criteria, however, not all of the examples of security measures used meet the criteria of just conduct in securitization. I use them here purely to aid the reader’s understanding of what is at stake in securitization.

1.5 How Long May Politicization Be Tried before Securitization Is Mandatory?

Using the six cases discussed earlier I now want to contemplate what needs to be considered when we try and work out how long politicization may be tried before securitization becomes – in the actual common-sense meaning of the word– necessary and on that basis mandatory, because we now have sufficient certainty that securitization is indispensable for satisfying just cause.

Looking at our six cases again I want to begin this discussion by focusing on the differences among the six cases. I have picked the examples deliberately so that two of the threat scenarios discussed refer to agent-intended threats (Election and Jihad Terror), two scenarios involve agent-caused but not intended threats (Immigration collapse, Climate disaster) and two threat scenarios are agent-lacking (Dengue threat and Drought disaster).Footnote ³¹ Beyond that, two threats (Jihad terror and Dengue threat) are directly lethal to persons, two threats (Drought disaster and climate disaster) are indirectly lethal and the remaining two threats are non-lethal (Election and Immigration collapseFootnote ³²) (see Table 1.1).

These distinctions are potentially important; thus, it seems reasonable to argue that the relative lethality of threats renders securitization more urgent vis-à-vis securitization against non-lethal threats. Suggesting, in turn, that political solutions may be tried for a longer period where non-lethal threats are concerned and for a comparatively shorter period of time where lethal threats are concerned. With a view to our examples earlier, it would mean that the satisfaction of must cause (measured from the onset of politicization) takes necessarily longer in Election than Dengue threat, and longer in Immigration collapse than Drought disaster.

A related factor that quite plausibly influences the relative length of time for which alternatives may be tried could be the number of people existentially threatened. It seems plausible to hold that a threat that existentially threatens say 1,000 people needs to be dealt with quicker by means of securitization than one that affects 100 people. Thus, while here both threats are lethal, we might say that the threat bringing with it greater numbers of mortality is comparatively more lethal than a direct lethal threat with a small number of death (infectious disease versus terrorism comes to mind here). In this same vein we might also distinguish between direct lethal threats and indirect lethal threat, arguing for quicker securitization against direct lethal threats, on the grounds that the initial referent could be saved by other, less harmful, means before anyone dies.

On the face of it this lethality-qua-time reasoning seems convincing. Because the harm prevented by securitization is greater it seems true that lethal threats warrant a quicker response than non-lethal threats. And perhaps it is also true that threats leading to comparatively higher numbers of death must be prioritized. However, as soon as we consider that the lethality-qua-time logic suggests that the borderless threat of Climate disaster requires securitization comparably quicker than Drought Disaster (recall that our fictitious Millville ‘only’ has 500.000 inhabitants), doubts should set in. Our moral intuitions tell us that it does seem wrong that the commencement of safe-making by securitization of the 500.000 must necessarily take longer than that of >500.000. But why do such doubts creep in? For one thing, because 500.000 seems a big number, certainly compared to say 500. However, all of this is relative. Thus 500 is a big number compared to 50, while 50 is a big number concerned to just 5 deaths prevented by securitization. Yet we might also say that 500 people saved from being severely harmed because of securitization is ‘better’ than 15 deaths prevented by securitization. This last point is important; it disproves the idea that the harm prevented by securitization is always greater when lethal threats are concerned. In other words, it shows that lethality does not automatically warrant a quicker response to non-lethal threats in all cases, because the harm that securitization prevents is not always greater for those kinds of threat.

Instead of going on to spell out cases when this is so, recall that when we are discussing length of time for which politicization may be tried, we have already accepted that securitization is permissible. Notably, permissibility includes macro-proportionality that is, we have – at this point – settled that securitization and the harm that it will cause innocent bystanders and threateners is a proportionate response, and also that the threat securitization prevents is sufficiently harmful in its effects to warrant causing such harm. It seems to me that once we have passed this threshold and deemed securitization morally permissible (as we have, for the purposes of argument, in all six hypothetical cases) we are on a level playing field regarding threats and the harm that they cause. In short, once securitization is deemed morally permissible, the actual effects of individual threats (i.e., lethal, or not) do not matter, because we know that securitization is proportionate.Footnote ³³ One exception here pertains to cases where would-be securitizing actors have insufficient means to successfully securitize against two threats, which affect an equal number of people. Ceteris paribus in such cases it seems that the more harmful (usually the lethal threat but depending on numbers not categorically) ought to be given priority.

With lethality out of the way, another issue that needs investigating is whether must cause is satisfied more quickly when threats are agent-lacking that is, when there is no threatener. This argument stems from the observation that where agents are at the source of the threat (either by causing the threat intentionally or unintentionally) there is potentially always room for dialogue. That is, there is space in which would-be securitizing actors can warn agents at the source of the threat of what will happen unless these agents change their threatening behaviour. As such they can and perhaps should try and engage in conflict resolution as part of politicization. In agent-lacking threats, in turn, dialogue is impossible (we cannot reason with vectors), and it is therefore reasonable to suggest that the length of time for which alternatives need to be tried is comparatively shorter.

Looking at the stylized examples earlier, however, we can see that it is not true that dialogue and discussion with threateners is always an option. Notably with regards to lone wolf jihadi terrorists acting on the behalf of or in allegiance with, for example, IS this is impossible. The secret security services usually do not know who the people are (or rather that they will carry out attacks) before they have done so. While talking to Muslim communities (e.g., counter terrorism police and de-radicalisation specialists speaking at schools with majority Muslim pupils) can back-fire. Thus, by being identified as a possible offender, feelings of alienation can set in, which make people more susceptible to becoming radicalized (cf. Abbas, Reference Abbas2019 on PREVENT, see also Breen-Smyth, Reference Breen-Smyth2014 on the unintended consequences of counter-terrorism).

Based on this analysis, I would like to suggest that the time for which alternatives to securitizing may be tried before must cause is satisfied can – somewhat unfortunately – not be generated using a formula that takes account of lethality and source of the threat. With this dismissed as impossible, what, if anything, concrete can we say about the length of time for which alternatives may be tried? One concrete thing we can say is that it makes little sense to dismiss less harmful alternatives to securitizing, unless these alternatives have been tried for a period of time that is long enough allowing them to fail. And at the same time short enough though as not to unnecessarily endanger would-be referents by delaying securitization. Let us call this ideal time period: the sufficient-time gap. I will return to this more fully in a moment but let us consider firstly what failure means in each of our six cases.

Regarding Election, political alternatives to securitizing fail when they prove unable to secure the democratic process; in part this could be established using opinion polls that test for the belief in false facts. Politicization of Jihad terror fail when the number of (foiled) attacks (not casualties) does not decrease following the introduction of such measures. Alternatives to securitizing in Immigration collapse fail if the number of irregular migrants seeking to enter a given state does not markedly decrease. Political solutions to Climate disaster fail unless they manage to reduce carbon emissions, allowing the rise of global temperatures to stabilize. Political solutions to Dengue threat fail unless the number of new infections is decreasing in statistically relevant terms. Finally, in Drought disaster politicization fails when the water level in aquifers and dams has decreased at a rate incompatible with low water usage by all.

One thing we can observe from these examples is that the time that needs to elapse to establish whether politicization has failed differs widely depending on the nature of the threat. Regarding Dengue threat, for example, it should be possible to see relatively quickly (i.e., in a matter of days or weeks) whether politicization manages to generate a much-needed decrease in new infections, the same should be true of Drought disaster, in so far as relevant authorities should be able to tell – relatively quickly – whether water levels go down too rapidly. By contrast, at the extreme end, the sufficient-time gap for Climate disaster could span months, a year, 18 months or more. This is not only because vast amounts of data need to be collected from different states, but also because unless data is collected for a lengthy period (i.e., over the course of 12 or 18 months) to establish the effectiveness of measures. Moreover, unless a longish time window is allowed it is impossible to discount short-term changes to emissions reductions caused by e.g., economic factors such as temporary downturns, weather factors (i.e., less heating in summer months). In short, short-term data collection would generate an unreliable result.

The analysis of the six cases clearly shows that the sufficient-time gap can differ widely in distinct cases. In each case the politicizing actor needs to monitor whether the political solutions are making a sufficient positive change. In some cases – notably infectious diseases – this will be relatively easy as we rely on concrete data (i.e., medical statistics); in others – notably, terrorism – it will be much harder due to known unknowns (i.e., the fact that we can’t know for sure who else is out there planning to harm people). What then can be done in cases where, for example, states as politicizing actors-cum-securitizing actors disagree with most of the civil society that sufficient-time gap is closed and must cause satisfied?Footnote ³⁴

I want to suggest that this situation can be resolved with help of the reasonable person standard. Specifically, the would-be securitizing actor ought to defer the decision on whether must cause is satisfied to the reasonable person understood in Ripstein’s terms as the normatively justified person (1998: 8 c.f. introduction). Of course, the reasonable person is not a real person but instead a heuristic. In practice it would see the would-be securitizing actor publicly tot up the reasons and the evidence in favour of the satisfaction of must cause in order to determine whether this outweighs reasons for not securitizing. This process would enable the rendering of a plausible, non-biased decision regards sufficient-time gap and hence on whether must cause is satisfied. If reasons are miscounted, and a mistake is made in favour of the satisfaction of must cause, it should be remembered here that at this stage securitization is morally permissible.

The reasonable person standard is associated with common law. The practice of mandatory securitization can benefit from the methods of common-law in another way. In common law countries such as the UK, and unlike in civil law countries such as, for example, Germany, the doctrine of precedent applies. Concretely, ‘[t]his means that the judgment of each case can bind all subsequent cases depending on the seniority of the court (the court system has a hierarchical structure). As such case law becomes part of the law by either setting legal precedents where there is no legislation or interpreting legislation’ (Oxford Lib Guide, 2018). Quite aside from the question whether mandatory securitization should be codified into law; precedents can act as useful guides on how to proceed in similar situations. Moreover, precedents reduce the need for the reasonable person standards. To facilitate this method, any given state, regional body etc. should aim to document the threats, the means and nature of politicization and how long for this was tried. Such databases could then also be used to serve as valuable guides to settle disputed cases regarding the satisfaction of must cause.Footnote ³⁵

1.6 Conclusion

In this chapter, I have argued that securitization is not automatically obligatory when it is morally permissible, which inter alia is the case when securitization is ex ante judged to be the best way to satisfy just cause. Instead, I have suggested that securitization is a moral duty only when politicization has been tried and when this has failed to satisfy just cause, because only then do we have certainty that is sufficient to override the intrinsic moral value of autonomy. Importantly, the so determined ‘must cause’ does not replace just cause (notably securitization cannot be mandatory unless it is permissible, and permissibility depends inter alia on the satisfaction of just cause, proportionality etc). Rather, what it does do is to specify the critical juncture when circumstances are such that the threat no longer merely permits securitization but requires securitization.

The distinction between in Pattison’s words merely a ‘right’ to securitization and a ‘duty’ to do so (2010: 15ff), raises three outstanding issues in need of discussion in this chapter’s conclusion. First, are the criteria for the moral permissibility of self-securitization the same as they are for just other-securitization? Specifically, does the right to other-securitization not also depend on the – at least – tacit consent of the general population in cases where securitizing actors are states? Second, in JST the prospect of success is not gauged in terms of securitization’s absolute chances of satisfying just cause, but relative to the chances of the success of other less harmful options. In other words, securitization is permissible even if resistance appears futile. This begs the question whether executors of securitization can be required to partake in securitization that has only a slight chance of succeeding in satisfying just cause? Or is it the case that mandatory securitization requires a more stringent success condition than JST?

And finally, third, are the ‘jus in securitization’ rules the same for morally permitted and for morally required securitization? Or is it the case that laxer standards should apply to the just conduct in securitization when it is morally required?

The first question arises because of the clear differentiation between self- and other-securitization developed in this book, but merely mentioned in passing in my 2019 book The Morality of Security. As already touched upon in the discussion of Oberman’s work in Section 1.2, other-securitization can be at odds with public opinion and still justified. Consent is also not required for just self-securitization. As I have explained in my 2019 book, in state-led self-securitization, referent objects (most likely (members of) the public) do not have to consent (tacitly or overtly) to securitization. There are several reasons for this, but the main reason is that the inclusion of beneficiary consent seeks to prevent against agent-benefiting securitization.Footnote ³⁶ I hold that the possibility of agent-benefiting securitization is pre-empted by the substantive criteria of just securitization, notably right intention. Given that the general public within the securitizing state is not the primary beneficiary of other-securitization, the phenomenon acutely raises the issue whether or not other-securitization requires the consent of the securitizing state’s general population? Or, more realistically whether other-securitization should ‘reflect its citizens’ opinions in its decision making’, whose taxes and man-power are utilized to make it possible (Pattison, Reference Pattison2010: 134). To solve this puzzle let’s begin by asking when and why the general public might object to other-securitization. I can see three possible scenarios: (1) we don’t think that we should help; (2) we don’t care about other (far away) people, and (3) helping will negatively affect us, for example, the cost of other-securitization is too high. The first two fall foul of a general obligation towards others on the grounds of the moral equality of people, which imposes remedial responsibilities (positive duties) on capable actors towards all regardless of place and origin (cf. Chapter 2, Section 2.3). In short, the first two do not stand up at the bar of justice. The third objection is the weighty one. We shall see in Chapter 2, Section 2.5, that the obligation to securitize is overridden when the costs for the would-be securitizing actor are expected to be too high. Also recall McMahan’s bridge example from earlier (Section 1.2). Here the passer-by’s pro-tanto obligation to rescue the $1000 dollar bill is overridden if this can only be achieved at great costs to himself.Footnote ³⁷ The interesting question is whether the passer-by is morally permitted to rescue the $1000 dollar bill even if this means falling to his death? The answer is likely to depend on the view one has of the sanctity of life (cf. Section 3.2), as well as on the position the passer-by has in life (e.g., do they have caring duties, for example, for children). However, we believe that individual actors are morally permitted to incur great costs (provided they do so voluntarily) in performing moral acts, but that such acts are supererogatory never obligatory (McMahan, Reference McMahan, Davis, Keshen and McMahan2010: 58). But are governments morally permitted to perform supererogatory other-securitization without the consent of the public who would be the entities incurring the high costs? The answer is that permissibility depends also on the extent of the costs incurred (cf. Frowe, Reference Frowe2014a: 151). That is to say, permissibility does not depend on the view of the majority, but rather on proportionality, which includes the costs to people governments are responsible for (Shue, Reference Shue, Lazar and Frowe2018: 266). In other words, the criteria governing the moral permissibility of self and other-securitization are coterminous.

Turning now to our second question, can one be required to do a harmful act that cannot be expected to succeed in averting the threat? On the face of it this seems not only improbable, but it also raises the issue whether there is a mistake in JST, which does not specify an absolute success condition. Instead, here the prospect of success is judged relative to the success of less harmful options in satisfying just cause. Bluntly put, in JST out of a set of options, securitization simply must be the least hopeless one at achieving just cause. One reason for framing JST in this way is that the theory permits just resistance by the weak against unjust harm by the strong (cf. Floyd, Reference Floyd2022). One way to justify a less stringent success condition is to stress the honour of the victim. Daniel Statman puts this as follows: ‘Whenever victims of aggression are overwhelmed by an aggression but, nonetheless, find the courage to rise against him through some form of determined resistance, however hopeless, they are thereby reaffirming their honor’ (Statman, Reference Statman2008: 679).

To my mind this reasoning sits comfortably with the value of autonomy stressed earlier in the context of permissibility and supererogation (Chapter 1, Section 1.2). Where wars and securitizations by collectives (states etc.) are concerned there is however a problem with this logic. Helen Frowe has shown that honour-based accounts don’t work for collective defence (war) because collateral damage inflicted on innocent bystanders is morally permissible only if the war has a good prospect of success (Frowe, Reference Frowe2014a: 153–154). While honour-based justifications sit uneasily with, for instance, state-led securitizations, Frowe shows that futile violence is unjust not because it does not have an absolute chance of succeeding in satisfying just cause, but rather because such action is disproportionate.

This means that provided securitization is proportionate, a relative chance of success suffices for securitization’s permissibility and its requirement.

Finally, our third question: are the principles governing ‘just conduct in securitization’ the same for mandatory securitization as they are for merely morally permitted securitization? This question comes about because it seems reasonable to suggest that if securitization is a bona fide last resort then the rules for what kind of emergency measures may be used also are relaxed. While this seems intuitively right, it is to be rejected. It cannot be the case that simply because relevant actors have a duty to securitize that they may do so in any which way they see fit, including in complete disregard of basic human rights of those that are subject or object of securitization. If there is a moral disparity, it pertains to how excusable wrongful acts are only. For instance, while torture is morally wrong, its use in securitization might sometimes be excusable (i.e., under extreme duress). It is almost certainly more readily excusable when securitization was a last resort.