Responsible Veto Restraint: a Transitional Cosmopolitan Reform Measure for the Responsibility to Protect

This article examines reform to the ‘veto’ power held by the five permanent members of the United Nations Security Council. The responsibility to react to mass atrocity crimes under the Responsibility to Protect (R2P) lies predominantly in the hands of the Security Council, meaning that R2P and the veto are inseparable. Veto use can obstruct the Council from meeting its R2P, reflected by the ongoing crisis in Syria, over which 16 Council draft resolutions have been vetoed to date. This article applies a transitional cosmopolitan framework to offer an informal ‘Responsible Veto Restraint’ (rvr) recommendation for veto reform. This measure provides a more effective and feasible avenue for veto reform than the recommendations of the Accountability, Coherency, and Transparency Group’s Code of Conduct and the France-Mexico Joint initiative for veto restraint. rvr can help promote R2P action through the Security Council, offering an avenue for progress towards addressing the problem of atrocity crimes. global to protect (2020) 385-414

the Council's response to human rights violations.7 Similarly, Bellamy and Luck note that P5 disagreements often lead to Council ineffectiveness.8 Grover argues that not acting on R2P crises makes the Council complicit in atrocity crimes.9 Wyatt claims that 'the veto power continues to offer a constitutional mechanism whereby intervention can be precluded' despite breaches of international law and R2P violations.10 Coen notes an international legitimacy problem presented by the veto as those that possess this power can effectively determine which R2P situations warrant a response.11 Erskine believes that the decision not to act or inability to act in response to an R2P crisis (such as through veto use) equates to a failure of the Council to meet its moral responsibility.12 Relating to international response under R2P, Deitelhoff argues that while R2P's first and second pillars are uncontested, Pillar Three has suffered contestation in its implementation.13 Similarly, Welsh writes that while R2P's three pillars are meant to be taken as equal, there has been evidence of pillar sequencing and preferentialism, with Pillar Three being favoured the least by the international community.14 This contestation and distancing from Pillar Three is intrinsically tied to the use of the veto power, as veto use can effectively stall Pillar Three, undermining the international responsibility to respond to atrocity crimes. Approaching the issue from a legal perspective, Moustafa argues that the P5 have 'an obligation not to use the veto repeatedly in a way that prevents the Security Council from acting at all in the face of a breach of a peremptory norm' of international law.15 While this article takes a normative global responsibility to protect 12 (2020) 385-414 approach rather than a legal one, it agrees with Moustafa that veto restraint is necessary. If R2P is to promote progress in the area of human protection, then avenues of veto reform must be explored.
Yet veto reform is politically difficult to achieve and some preliminary feasibility constraints must be noted. First, veto reform challenges the vital interests of the P5 who have instrumental reasons to preserve their veto power. 16 The veto power ensures that the P5 members hold a permanent position of authority within the international community, and enables them to further their interests by blocking potentially damaging Security Council action. It seems highly improbable that any of the P5 members would agree to forfeit their veto rights entirely. The idea of abolishing the veto is therefore a politically untenable one and should not be pursued. A second constraint is that the veto is necessary for the management function of the P5 states towards international peace and security. The veto is not just a means to realise national interests of the P5, but exists to prevent great power conflict by guaranteeing their participation into the UN system.17 However, with the development of international norms of human protection the Council is now also expected to prevent mass atrocities as a matter of its vital function,18 and finds the legitimacy of the veto called into question where it obstructs Council action.19 Morris refers to this dilemma as the 'dual responsibility' of the P5.20 The growth in solidaristic values through norms such as R2P conflicts directly with the Council's traditional obligations to peace and security, meaning that meeting both these special responsibilities is immensely difficult in practice.21 For instance, authorising military action vis-à-vis R2P can disrupt a region and undermine the Council's traditional responsibility to international stability.22 To overcome this problem, veto reform must help facilitate action that contributes towards 16 Justin the Council meeting its cosmopolitan obligations under R2P, but simultaneously does not undermine its ability to manage threats to international peace and security. Linking on from this, a third constraint is that abolishing the veto may pave way for actions which are not aligned with the purpose of R2P. Without a check on the application of military force, states may be able to too easily authorise interventions that could potentially undermine human protection.23 This may stifle the potential for exploring other measures that are more appropriate for addressing an R2P case.24 This further supports the argument that the veto should not be entirely abolished. A final constraint to note here is the point that any formal amendment to the veto under the UN Charter is politically unfeasible. Article 108 of the Charter refers to the consensus required by the P5 states to adopt any Charter amendment.25 Webb goes as far to say that 'it would probably be easier to dissolve the UN than to amend the veto power under the Charter' .26 A formal commitment to limiting the veto power would be a legally binding decision that the P5 would be unable to easily reverse. This reduces the feasibility of a formal Charter amendment, making it appear that we need to focus on pursuing only non-formal amendments, at least in the short-term. This article applies a transitional cosmopolitan approach to make a new veto restraint proposal. In doing so, its aim is to promote incremental progress towards desirable veto practice. Building on suggestions in the Accountability, Coherency and Transparency (act) Group's Code of Conduct, and the France-Mexico initiative for veto restraint, the article calls for a 'Responsible Veto Restraint' (rvr) proposal. This is a recommendation for informal P5 veto restraint which differs from previous veto restraint proposals in several key aspects. rvr proposes that military measures, sanctions, and International Criminal Court (icc) referrals be separated from veto restraint, recommends the removal of subjective get-out clauses for the P5, and favours a more stringent trigger system for when veto restraint should become active. It is argued that by adopting these aspects, rvr can be a more effective and feasible measure for veto restraint than either the act Code or France-Mexico initiative.
While in current circumstances achieving veto reform is politically difficult, a tempered form of veto restraint can help promote R2P's implementation. In France-Mexico initiative, determining that they are currently lacking in the criteria of effectiveness and feasibility. The third section seeks to build on these reforms, offering nuanced changes to produce the rvr proposal, more aligned with the transitional cosmopolitan approach. The final section examines the case-study of Syria to demonstrate the transitional cosmopolitan value of rvr.

A Transitional Cosmopolitan Approach to Veto Reform
To successfully achieve P5 veto restraint will require compromise on normative ideals in order to promote a measure which is satisfactory to demands for a stronger application of R2P, as well as competing state interests. One approach aimed at achieving such a compromise is that of transitional cosmopolitanism. A transitional cosmopolitan approach takes cosmopolitan ideals -such as obligations towards cross-border fundamental rights (evident in the concept of R2P) -as valuable tools only when they can be meaningfully applied to the reality of lived-in situations. It is an approach which accepts that promoting some normative progress through tempered actions is better than making no progress at all, and that this is better than calling for overly ambitious and unrealisable goals that are likely to be outrightly rejected. Transitional cosmopolitanism is a relatively new and developing school of thought. There are several authors among the cosmopolitan literature who can reasonably be identified as promoting a transitional approach to the application of cosmopolitan principles. Brown and Hobbs argue that cosmopolitical change takes time to achieve, with non-ideal progress towards a cosmopolitan condition still representing at least some progress, and that this can open the potential for further iterative advancement towards a more desirable cosmopolitan endpoint.28 Beardsworth et al. highlight a common charge against cosmopolitan thought that it is too idealistic and disconnected from political reality.29 Consequently, they argue that cosmopolitanism needs to better identify avenues for normative progress that are able to work within ongoing and evolving political dynamics.30 Similarly, Valentini argues that when we apply cosmopolitan principles, we need to be prepared to make 'strategic 28 Garrett She is also open to the potential for iterative progress, that is, the idea that some progress can be made even without universal consensus, and that over time this progress will continue as consensus increases.35 In a similar vein, Habermas views the development of international law as a steppingstone towards cosmopolitan constitutionalisation at the global level.36 He views the UN as part of this developing process, with the UN Charter's explicit links to human rights, prohibition on state violence, the relation of sovereignty to the (expanded) goals of international peace, and universal membership as a precedent for international law.37 Notably though, Habermas does not view transitional processes as linear, appreciating instead that backsliding is possible. This highlights the need to make continuous normative evaluations about contemporary circumstances, in order to identify avenues for cosmopolitan advancement, but also to minimise normative losses wherever possible. Whatever transitional cosmopolitanism is, it holds the following basic tenets. First, its purpose is to help foster the conditions through which we can strengthen the pursuit of human protection, but while working among the practical constraints in the way of achieving ideal progress. Second, a transitional cosmopolitanism accepts that imperfect normative progress still represents progress. Third, transitional cosmopolitanism promotes the idea that tempered progress can open up the possibility of further iterative steps and normative gains.

Effectiveness and Feasibility Tests
To develop the transitional cosmopolitan approach further, it is argued here that central to a transitional cosmopolitan approach to reform should be the twin criteria of effectiveness and feasibility of progress.
Regarding the first criterion, effectiveness is taken here as a three-prong concept which is in essence about whether a measure should be adopted. First, effectiveness is understood here to contain a pragmatic element: that is, a reform measure must contribute towards overcoming the problem which it seeks to address. As Pattison argues, to satisfy a pragmatic demand requires 'effectiveness in the purported measures which are being promoted' .38 Note that this aspect of effectiveness is not an overly demanding stipulation in that a given reform must have supreme transformational power, but merely that a reform measure must be able to contribute at least some positive change. Second, effectiveness refers here to the ability of a reform action to help move us in the direction of desirable normative progress. This is where transitional cosmopolitanism differs from a purely pragmatic or instrumentalist approach. A pragmatic solution simply focuses on overcoming a given problem, without making underlying normative stipulations. However, a lack of predetermined moral demands via a purely pragmatic approach can potentially permit actions which may violate valuable moral tenets, thus undermining the very purpose of pursuing reform.39 In contrast, effectiveness is understood here through a cosmopolitan lens, stipulating that we cannot determine something effective if it were to undermine global commitments to upholding fundamental human rights. Third, effectiveness is also measured here as the ability of a reform measure to open up avenues through which we may be able to attain longer-term progress. This attempts to consider the iterative potential of a given reform measure.
The second criterion proffered here for a transitional cosmopolitan approach to reform is that of feasibility. This refers to the practical viability of whether a measure could be attained. Reform that would be politically untenable by violating the vital interests of states would not be satisfactory to a transitional cosmopolitan framework as it would not offer scope for progress. Feasibility includes considerations for whether there is evidence to suggest some underlying progress, and a potential pathway towards achieving the measure. As Lawford-Smith argues, feasibility also includes a consideration 38 Pattison, The Alternatives to War, p. 21. 39 For instance, purely utilitarian arguments may permit solutions which are insensitive to deontological commitments to upholding fundamental human rights. global responsibility to protect 12 (2020) 385-414 for whether the potential pathway towards progress can actually lead us to the goal.40 To know whether these criteria of effectiveness and feasibility can be met we require a set of tests to judge them. Here, Gilabert and Lawford-Smith's framework for assessing the feasibility of a political reform measure is in-part adopted for the application of transitional cosmopolitanism.41 This process does more than merely assess feasibility, however, for the three tests serve to determine both the effectiveness and feasibility of a given measure and allows one to see how the two concepts must operate together to satisfy a transitional cosmopolitan approach.
The first test of this framework for effectiveness and feasibility is to determine whether effective progress can be made through adopting the reform measure. This concerns whether the problem can be at least partially overcome, or if avenues for future progress can be made by adopting the measure. If the reform were to fail this test, then it will possess no normative value as it could not promote positive change. The second test requires assessing whether the costs of bringing about the measure would be too morally hazardous. This requires questioning whether there is potential for unwanted side-effects that would damage the cosmopolitan commitment to uphold fundamental human rights. If a reform measure appears to possess such moral hazards, this will damage both its feasibility by way of its normative appeal, and its potential effectiveness as a measure for transitional cosmopolitan change. The third and final test is one of access: to determine if the route to the measure is practically possible. This is a test for whether there is possibility that the measure would be successfully adopted by the relevant actors. This requires one to weigh up the various soft constraints that may exist to determine the feasibility of the reform being adopted.
For a transitional cosmopolitan approach to be satisfied, both the criteria of effectiveness and feasibility must be met, and therefore all three tests satisfied. One may observe here that there may be a conflict between the criteria of effectiveness and feasibility when applied to the analysis of a particular measure, and hence that there may be a tension over whether it is appropriate to give weight to one test over another. This worry is likely to concern whether either effectiveness or feasibility should be given priority in situations where they clash. For instance, such a clash may occur over a question of whether it is appropriate to temper the transformational power of a given reform measure to aid in making the measure more feasible. On the one hand, such a sacrifice seems essential for finding a transitional solution. On the other, if effectiveness is sacrificed to the point that it would greatly undermine the possibility of making normative progress towards a cosmopolitical standard, then we cannot accept the measure just because its feasibility is greatly enhanced. The upshot is that a transitional cosmopolitan approach requires all three tests to be given due consideration and that no single test can be sacrificed entirely for the benefit of another. The application of the three tests to a given reform measure should be context specific to the measure in question. That is, when analysing the effectiveness and feasibility of a given measure, the contextual factors must be taken into consideration. These factors will differ from case to case, and are likely to affect how much transformative power a reform measure can appropriately attempt to generate when feasibility constraints are factored in. It is therefore not possible to say exactly how a potential conflict between the criteria is to be resolved a priori. The criteria of effectiveness and feasibility are merely to aid transitional cosmopolitanism as an approach aimed at generating a 'middle position' towards progress. In doing so, the aim of the criteria is to help in assessing the practical relevance of any given reform measure. The middle position is a compromise on ideal progress, acknowledging that some progress is possible, but also that there are practical constraints which make attaining the ideal considerably unlikely in the present context.

Evaluating the act Code and France-Mexico Initiative
Let us now use this transitional cosmopolitan framework to evaluate the act Code and France-Mexico initiative as the two most prominent veto restraint recommendations to date. Should these measures appear satisfactory to this transitional cosmopolitan approach, then this would suggest that they possess effectiveness and feasibility and should therefore continue to be promoted as the strongest recommendations for veto restraint. However, should they fail to satisfy this transitional cosmopolitan approach, it would seem that an alternative veto restraint measure needs to be promoted if progress is to be made. Launched in July 2015, the act Code calls on UN member states to 'pledge to support timely and decisive action by the unsc aimed at preventing or ending the commission of genocide, crimes against humanity or war crimes' . The act Code and France-Mexico initiative have some merit from the transitional cosmopolitan perspective. Both are aimed at restricting veto use in cases of R2P concern, meaning they are theoretically aligned with the cosmopolitan demand to promote action in the name of human protection. Further, as both are informal methods, the lack of Charter amendment required offers greater promise for their acceptance than a formal recommendation would. In contrast, however, Adediran argues that 'a Code of Conduct, just like Article 108, requires the acquiescence of all permanent members… [and therefore] it is not immediately clear how it considerably differs from the cumbersome demands of the Charter regarding Council reform' .48 While it is true that even an informal restraint measure would still require the support of all P5 states, this does not make such a recommendation worthless. For one, voluntary restraint offers a more pragmatic route because it would not force the P5 into a long-term binding commitment. Rather, it would give an opportunity to see how voluntary restraint operates in practice in the knowledge that should it serve to grossly undermine their interests or the function of the Security Council, it would be possible to revert back to the previously accepted application of the veto. The support that the act Code and France-Mexico initiative have garnered in a short lifespan is promising, reflecting the fact that veto restraint 'is not an idle hope' .49 Advocacy for veto restraint does seem to be putting increased pressure on the P5 members as they understand that veto restraint is a prominent issue for UN members. For Luck, veto restraint measures 'have generated renewed interest in prevention, protection, and R2P; [and] have forced members of the Security Council to reflect (at least a bit) on their Charter responsibilities' .50 The fact that veto restraint initiatives are directly linked to R2P is also important, for it shows that there is appetite among the international community for a more consistent application of R2P.
However, the proposals of the act Code and the France-Mexico initiative are far from perfect, and their chance of adoption by all P5 states remains unlikely. To see why this is the case, let us examine these proposals through the transitional cosmopolitan tests for effectiveness and feasibility.

2.1
Test of Effective Progress Let us begin by assessing whether the measures would contribute progress to overcoming the veto problem. The issue here for the act Code and the France-Mexico initiative is the vagueness and subjectivity in the terms 'credible' and 'vital interests' employed by the proposals respectively. First, as Vilmer argues, credibility is highly subjective, 'what is credible to a state that supports a resolution is not credible to a state that does not' .51 In practice then, were the act Code adopted, current disagreements on draft resolutions between Security Council members would remain, and veto use would likely follow where disagreements exist. Second, as Morris and Wheeler argue, the differences inherent in the P5's interpretation of 'vital interests' could make implementing veto restraint very difficult in practice.52 This suggests that the France-Mexico initiative would also run afoul in practice. It should be noted that the criteria of credibility and vital interests are intended to make the proposals more politically feasible by providing P5 members with reassurance that they can still global responsibility to protect 12 (2020) 385-414 safeguard their interests. However, since this would seem to come entirely at expense of the effectiveness of the proposals by simply enforcing the status quo of a deadlocked Council, this is untenable for the transitional cosmopolitan approach.

2.2
Test of Moral Hazards Next, to determine whether these measures can be deemed effective and feasible requires us to determine whether they would introduce damaging moral hazards. Here, the act Code and France-Mexico initiative appear to fall foul of similar arguments against abolishing the veto entirely. Vilmer, for instance, claims there is a risk of passing resolutions that are 'too strong' , potentially doing more harm than good.53 Again, this relates to the problem of too readily authorising potentially damaging military interventions before other measures are fully explored.54 Adediran argues that veto restraint 'advocates for the commitment of states to a specific course of action' .55 This logic seems questionable though, since veto restraint would not itself advocate any particular course of action. However, Adediran is right to caution against overzealously promoting military action which veto restraint could potentially assist. The act Code and France-Mexico initiative would also seem to fail on the second transitional cosmopolitan test then, due to the potential of introducing this moral hazard. This weakens both their effectiveness as measures aimed at promoting human protection, as well as their feasibility of adoption.

2.3
Test of Practical Potential Finally, relating to feasibility constraints, the primary obstacle here is the political will of the P5 to adopt restraint. Of course, this is not the case for all the P5. One of these measures comes from France as a P5 member, while the United Kingdom is also a signatory to both proposals. The UK has also stated, '[w]e cannot envisage the circumstances in which the United Kingdom would use its veto to block an appropriate response to a mass atrocity' .56 Morris and Wheeler have doubted the importance of France's and the UK's support for reform, arguing that it is effectively meaningless without support from all P5 members.57 Vilmer suggests that, despite its rhetorical support for constraint, 53 Vilmer the UK remains sceptical that it may be open to abuse.58 Adediran also argues that France's and the UK's support for veto restraint is a tacit acknowledgement of their own diminishing power in international politics.59 This suggests that their support may be an attempt to maintain their own legitimacy as international managers, rather than as a result of their normative commitment to veto reform. However, it does seem important that veto reform has some support among P5 members. If all were opposed to restraint, this would make it seem even less feasible. Further, when supplemented with additional pressure from a wider group of UN members and civil society groups, support from some P5 states may help grow a normative consensus in favour of reform, which other P5 members may struggle to ignore in the face of mounting political costs. Indeed, if veto restraint proposals garnered the support of the near-entire UN membership and some P5 members continued to ignore such calls, it would seem intuitive to suggest that the Security Council's legitimacy would be greatly called into question as the medium through which to manage international security.
What is particularly problematic for the feasibility of these restraint measures is the position held by the United States, Russia, and China. All three have thus far rejected calls for restraining their right to veto. The US has noted, 'we remain opposed to any expansion or alteration of the veto' ,60 despite its repeated calls for decisive action in the face of mass human rights violations. China has emphasised that the Security Council should be the one to lead the direction of Council reform and remains unconvinced by calls for veto restraint.61 The Russian position notes that 'any steps to improve working methods must be taken only by the Security Council itself' .62 As Vilmer notes of Russia, the veto power remains one of its few attributes of international power.63 It is therefore no surprise that Russia opposes reform. Further, both Russia and China have seen their traditional opposition to the use of force reinforced in the wake of the 2011 Libya intervention.64 The connection between veto restraint and the increased possibility of military force exercised under R2P Pillar Three remains a problem for veto reform, making it unlikely that either State would accept these two restraint proposals while this connection remains. In sum, these recommendations appear unfeasible as they are too demanding for those states reticent towards veto restraint. By failing the three tests, the act Code and the France-Mexico initiative fail to satisfy the transitional cosmopolitan criteria of effectiveness and feasibility. However, the fatalistic interpretation of this problem must be rejected if we are to follow the transitional cosmopolitan line that it is possible to make some incremental progress towards desirable veto practice.

Reforming the Reform Proposals: Responsible Veto Restraint
In this section an alternative veto restraint proposal is offered which seeks to build on the act Code and France-Mexico initiative while rectifying some of their deficiencies. It proposes a 'Responsible Veto Restraint' (rvr) recommendation which seeks to better align with the transitional cosmopolitan approach and subsequently offer a more effective and feasible avenue for veto reform. First, to clarify, both the act Code and France-Mexico initiative relate to R2P in the context of generating better responses to cases of mass atrocity crimes. This theme is also central to rvr as a transitional cosmopolitan reform measure. rvr is a recommendation for veto restraint aimed at enabling the Security Council to pass resolutions pursuing R2P action. Second, the rvr recommendation adopts an altered trigger system to that of the France-Mexico initiative. rvr would not require the unga as the channel for triggering rvr. Instead, a majority of UN member states would be required present a joint declaration that acknowledged a case as one of manifest R2P failure to the UN Secretary-General directly. This would call on the Secretary-General to determine whether the matter was a genuine R2P crisis. If the Secretary-General were to determine that this was indeed the case, then rvr would come into action. Third, rvr does not include any subjective get-out clauses for the P5 to continue a status quo application of the veto. rvr therefore rejects the notion of 'vital interests' from the France-Mexico initiative and 'credible resolutions' from the act Code. Fourth, rvr adopts the France-Mexico initiative's focus on the veto power of the P5 members, and rejects the act Code's focus on the whole Security Council membership. Fifth, and perhaps most interestingly, rvr would not apply to votes on the application of sanctions, icc referrals, or the authorisation of military force. rvr instead applies to any Pillar Two action, and only Pillar Three action precluding sanctions, icc referral, and military force.

3.1
Test of Effective Progress Let us now assess whether these recommendations have value by once again utilising the transitional cosmopolitan tests for effectiveness and feasibility. Regarding whether the measure can offer positive progress: first, rvr can help to facilitate R2P-based action in the Security Council. Excluding sanctions, icc referral, and military measures from veto restraint should, in theory, assuage fears that the concept would be open to abuse. This would hopefully make it more likely that the P5 members would employ veto restraint over measures short of the use of force. Moreover, precluding sanctions, icc referral, and military force would make redundant any argument that said measures could have a directly destabilising effect on the target state, or that veto restraint would automatically lead to such measures. It should therefore prevent a P5 member from being able to cloud self-interested veto use over an R2P crisis within the reasoning of preserving regional stability or maintaining international peace. As Morris has similarly argued when suggesting a removal of coercive elements from R2P more generally, it can 'deprive those who cite such fears as a cloak for ulterior reasons' .65 Pattison's analysis on the alternatives to war66 is complementary to the theory of rvr. Pattison argues that the use of military force should always be the 'presumed last resort' in response to atrocity crimes.67 He argues that alternatives to war can always make at least some contribution towards protecting fundamental human rights, without the unwanted side-effects that military actions bring.68 rvr is compatible with this approach as it can help to promote: declaratory statements and condemnations of violence to demonstrate that impunity will not be granted; humanitarian access to provide vital resources to those in need; fact-finding missions to bring greater accountability and knowledge of a crisis; mediation efforts to broker peace; and provide counter-narratives to atrocity violence to promote the norm that atrocity crimes are unacceptable. While these measures are certainly not guaranteed to end cases of mass atrocities, they can at least promote action that helps ameliorate suffering and contributes towards meeting cosmopolitan demands. For example, rhetorical condemnation of actors responsible for mass crimes can sometimes have efficacy by raising the political costs of breaching a norm's rules.69 Condemnations also promote norms of human protection and reject impunity for 65 Justin global responsibility to protect 12 (2020) 385-414 mass crimes, potentially influencing the potential for future transgressions elsewhere.70 It could also be added to this that condemnation can serve as ample warning to transgressors that more robust action may be considered in future should they not quickly cease their actions. rvr would make it easier to adopt such measures, without imposing a looming spectre that they are merely the 'thin-end of the wedge' before coercive military action. Further, it is important to highlight that nothing in the proposals here calls for the entire removal of consideration for the use of sanctions, icc referral, or military force. The use of these measures in R2P situations would remain available to the Security Council on a case-by-case basis as part of the R2P toolkit. Some may argue that only applying veto restraint to measures short of sanctions, icc referral, and military force would make rvr too weak. Ultimately, however, watering down its scope is a necessary step to make veto restraint tenable. The desired outcome being to avoid complete Security Council deadlock over major R2P crises of the future, and as a consequence serve as an effective incremental step towards a more consistent application of R2P that can help stymie atrocity crimes. Furthermore, were the P5 to accept rvr in this form, it would at least send a powerful normative message that the Council is devoted to fulfilling its R2P, and would be an acknowledgement of the need to limit veto usage. This may provide the potential for longer-term and incremental progress towards achieving consistent veto restraint by acting as an iterative mechanism through which the P5 become socialised into a new practice of veto use. In this sense, rvr attempts to push practice in a favourable direction over time. The hope is that through the gradual socialisation of P5 members into a practice of veto restraint, more robust measures will become more feasible as trust is built between P5 members that veto restraint does not equate to automatically authorising the use of coercive military force or regime change. While opposition to military force may remain unmoved, rvr can still, over time, aid in pushing practice towards other robust measures short of coercive military force. The point is that excluding these measures now is necessary to offer the tentative first transitional steps towards progress in the long term, and that in doing so, at least some positive progress can be offered immediately. Perhaps in the long term, sanctions and icc referral could be dropped as part of rvr, or any potential formal Charter amendment to the veto power not include them. But for now, they are necessarily excluded from rvr for enabling short-term and feasible progress.
Second, due to its specificity, a trigger system for when rvr becomes active appears to be a mechanism more aligned with the transitional cosmopolitan 70 Pattison, The Alternatives to War, pp. 92-98. demand for effective progress than a vague notion such as 'facts on the ground' in the act Code. A codified trigger system is less open to interpretation and political manoeuvring. However, as noted shortly, the trigger system in the France-Mexico initiative is troublesome. Hence, a slightly altered trigger system has been proposed here. Third, removing the concepts of 'credible' or 'vital interests' as subjective get-out clauses for the P5 attempts to make the recommendation here more effective. This seeks to make the proposal more aligned with the transitional cosmopolitan call for effective normative progress by attempting to make P5 members more accountable to their mandate under R2P.

3.2
Test of Moral Hazards Next, regarding the second transitional cosmopolitan test concerning moral hazards: first, more than simply easing P5 tensions, separating sanctions, icc referral, and military measures from veto restraint also serves to ease the tension between the Security Council's dual responsibility. rvr would not bind the P5 over morally hazardous resolutions that threaten international peace and security through the potentially destabilising effect of those measures. In this regard, the veto would stay true to its original normative purpose. In particular, precluding military measures from the rvr measure makes it more cautious than previous veto restraint proposals as it would make the Council unlikely to consider the use of military force as a first resort. As the potential for veto would remain over such a proposal, the Council would seem forced to discuss the alternatives to military measures as a first resort. Here it is again worth drawing on Pattison's analysis. Pattison argues that war is known to cause unwanted death and destruction, often involves doing harm, and that there is also uncertainty about its potential effectiveness in any given situation. Precluding military means should enhance rvr as an effective mechanism by avoiding the moral hazards associated with war that may undermine normative progress. Further, this helps contribute to rvr's feasibility, as it should help assuage fears that the measure could lead to unwanted side effects.
Second, the trigger system proposed in this rvr recommendation seeks to address a potential moral hazard. Increasing the number of states required for raising the issue with the Secretary-General to a majority of UN members makes the process more democratic than the France-Mexico initiative trigger system, while also helping to alleviate the potential for abuse by states attempting to overzealously advocate cases as requiring an R2P response.
Requiring only a small number of states to raise a case may lead to an overabundance of cases being brought forward for Secretary-General action that global responsibility to protect 12 (2020) 385-414 may undermine attention to other cases which are more genuine manifest R2P failures. In contrast, by having a more stringent trigger system, the recommendation made here is more cautious and less prone to abuse. Not requiring UN members to utilise the unga as the channel for activating rvr does potentially undermine the democratic credibility of the trigger process. However, this is necessary for enacting rvr in a timely and decisive manner consistent with the states' R2P commitment.71 Third, by applying only to the P5 members, another point rvr addresses is the potential for abuse if states are compelled to accept any resolution pertaining to an R2P issue. Adediran argues that adopting a code of conduct for veto use would undermine the contribution of the Security Council's nonpermanent members, as their votes would become predetermined.72 This would seem the case were the act Code's focus on the whole Security Council membership adopted, as the Code would compel all Council members to vote in favour of a given resolution. However, this argument does not apply to the France-Mexico initiative, which only calls on the P5 states to suspend their veto power rather than all Council members to avoid voting against a resolution. By adopting this aspect of the France-Mexico initiative, rvr can allay the tension highlighted by Adediran, promoting a form of veto reform that would still require a draft resolution to obtain a favourable vote of nine from 15 Council members. What must be emphasised here is that the purpose of veto restraint is not to create scenarios where any R2P-based action is automatically accepted, but to improve the deliberation process of the Security Council when voting on a resolution. By making it so that a draft resolution could no longer be scuppered by one member, the Council's voting process would be made significantly more democratic and able to promote action which the majority of the Council deems to be appropriate, rather than granting excessive power in the hands of five states. This would prevent a draft resolution being blocked by any one state, but also would remove fears that the Council would be bound to accept any given R2P draft resolution. This should serve to alleviate fears of potential abuse. It should also be noted here that the P5 members could still continue to voice their dissent towards a draft resolution by abstention from voting. global responsibility to protect 12 (2020) 385-414 international community.76 Where the Council fails to live up to its responsibilities under R2P, its own authority comes into question. If the P5 are seen to be in favour of progress towards Security Council reform, this may help to strengthen the acceptance of their authority from the wider UN membership. Linked to this, veto restraint should be viewed as a way to strengthen the veto power and not to undermine it. As Vilmer notes, 'repeated vetoes like in the case of Syria undermine the authority and eventually the centrality of the Security Council, restricting the veto gives it its full international effectiveness' .77 If the veto is to remain a meaningful tool for international management, then it requires acceptance by UN members. Reform may help to strengthen this acceptance, meaning that veto reform is likely in the long-term interests of the P5. Further, adopting rvr and improving the potential for R2P action can satisfy the direct interests of the P5 too: serving their normative interest in preventing and responding to mass atrocity crimes, as well as their instrumental interests such as stymying refugee flows and the growth of international terrorism, which are both directly linked to outbreaks of mass violence.78 It cannot be conclusively determined if the rvr measures recommended here could overcome the reservations of the three remaining P5 members currently opposed to veto reform. They could have an impact in the longer term though. For instance, they could help to assuage fears in certain sections of the international community -particularly non-P5 'brics' states and other medium powers -who have been troubled by the application of military force in Libya, but who are supportive of the tenets of R2P more generally. By doing so, the reform measures proposed here should theoretically be able to garner wider support, which as noted, would enhance the pressure on further P5 members for adopting veto restraint. The generally favourable attitude of the United States towards R2P and China's desire to be perceived as a responsible world leader79 offer potential openings for this political pressure to reap some success. For instance, Odgaard has argued that China's policy of 'coexistence' with liberal internationalist principles has meant that China is on occasion 76 Ian Hurd, willing to compromise on its non-interventionist preference.80 China's rise as a world power and desire to be perceived as a responsible leader81 perhaps offers hope that it may one day accept a tempered form of veto restraint. Moreover, by only applying to measures short of sanctions and military force, rvr would undermine arguments made by veto-casting members that international action over an atrocity case is a prelude to authorising damaging military force. As noted above this would prevent P5 members from being able to cloak self-interest in the name of normative reasoning, such as for preserving regional stability. With the safeguard against sanctions and military force offered by rvr, if P5 members continued to veto proposed non-military measures then international pressure would mount as it would be clear that vetoes were being cast simply to further the self-interest of the veto casting members. This pressure would raise the political costs of vetoing. As Bellamy notes, 'with sustained pressure, vetoes could be made so politically expensive that the permanent members would consider employing them only when absolutely necessary' .82 Whether desire to avoid this pressure would be enough to actively change P5 attitudes remains to be seen, and we can therefore only theorise as to whether this proposal can feasibly overcome soft constraints in the way of reform. Overall though, the tempered nature of the rvr recommendation would seem better placed to satisfy the transitional cosmopolitan test for overcoming soft constraints in the way of progress. Therefore, rvr offers a more feasible avenue for progress than the act Code or France-Mexico initiative.

Test of Practical Potential
The rvr recommendation better satisfies the transitional cosmopolitan tests for effectiveness and feasibility than the act Code and France-Mexico initiative. In doing so, rvr attempts to help in establishing the middle position for promoting progress, providing a recommendation which can help us achieve a more consistent implementation of R2P. rvr attempts to satisfy the transitional cosmopolitan approach by aligning the currently diverging positions of the P5 members vis-à-vis veto restraint, reducing the tensions in the Security Council's dual-responsibility, and promoting at least some R2P action that can contribute to R2P's stronger cosmopolitan application. International's Head of UN Office, described the vetoes as 'despicable and dangerous' as for many Syrians, humanitarian aid is 'the difference between having food to eat and starving' .90 It is for draft resolutions such as these that veto restraint is clearly morally necessitated and where the rvr mechanism could readily be applied.91 The example in this section focuses on the early part of the Syria crisis in 2011-12, specifically, on the first two Security Council draft resolutions that were put forward and subsequently vetoed by Russia and China. The contrast between these two draft proposals provides useful insight into the transitional cosmopolitan value of rvr. The case demonstrates that were rvr in place at the early part of the Syrian crisis, it would have been possible to pass R2P action in the Council that may have helped ameliorate the situation.
On 4 October 2011, the Security Council put to vote its first draft resolution for the Syrian crisis.92 The draft received the requisite nine votes in favour, with four abstentions. However, vetoes were cast by Russia and China, preventing the resolution from being adopted. The draft declared that there had been 'continued grave and systematic human rights violations and the use of force against civilians by the Syrian authorities' .93 The draft also called for access for latter element of the draft resolution may have made rvr's application more controversial due to the potential link with Venezuelan regime change. ruling out the use of force and making no mention of potential sanctions. Were rvr honoured then, the second draft resolution would have been passed. One might observe that despite its cautious wording, Russia and China did still veto the second draft resolution anyway. But the guarantee provided by rvr that the veto power would be maintained over any future consideration of sanctions or military force, combined with the potential political costs of forgoing their rvr commitment, may have been enough to influence Russia and China to avoid veto use. What this example demonstrates is that rvr can create the space for at least some R2P action in cases where P5 interests clash. This is likely to be action well short of what would be the ideal Security Council response, but this is still better than the complete failure to produce any action.
In opening up the space for action, rvr could help in the norm building of R2P, helping to alter Council behaviour over time as an understanding of the need to avoid veto use over less robust R2P actions is generated.
The argument here is that had a resolution of condemnation and humanitarian access been passed in the early days of the conflict, it would have promoted human protection. As Gareth Evans argues; 'the case for a condemnatory statement was overwhelming and had that been supplemented by the kind of measures that were initially applied in Libya … Assad would certainly have been given cause for pause' .104 Action fostered by rvr may not itself be particularly robust, but it can nevertheless help in ameliorating R2P crises. In the Syria case, had the Security Council had been able to pass early action, then Assad's sense of impunity may have been reduced, potentially causing him to scale back his campaign of force and helping to stymy the conflict before it spilled into all-out civil war. The use of diplomatic pressure can have a meaningful effect. This is highlighted by Weiss who draws on the example of the pressure applied against Assad following his later use of chemical weapons. 105 This particular diplomatic response backed Assad into a corner where he was forced to comply, showing the effect diplomatic pressure can have when applied by the international community. At the very least, the second draft resolution would have permitted humanitarian access early on during the crisis which would have actively worked to save human lives. Such steps would have represented an attempt to meet the demands of cosmopolitan human protection. This would have satisfied the transitional cosmopolitan demand for at least some action in response to a clear manifest failing of R2P Pillar One; acknowledging that the whilst over the Syria case the ideal response was not possible, it can still be possible to work amongst practical constraints in order to promote some normatively desirable progress. Furthermore, as noted, rvr does not in any way close down the potential for more robust R2P action, rather, it would merely serve to make less controversial R2P action a stronger possibility.
Less optimistically, it may simply be the case that due to the deep complexities of the Syria case, the potential link of the second draft resolution with a call for regime change, or the instrumental ties between Russia and Assad, Russia and China would have vetoed the draft resolution anyway. Perhaps it is also just simply too much to expect that a P5 member will ever forego its instrumental interests to promote R2P action. A more optimistic caveat, however, could be that R2P crises of the future may not be quite so morally and politically complex as the Syria case. So long as there are not significant instrumental ties present, and rather only more normative opposition to the use of force more generally, or only minor interests at stake, rvr can still help to promote at least some R2P action in response to the crises of the future. For instance, where a P5 member is sceptical of the interests of other Security Council members, but only has minor interests at stake itself, it would seem fair to suggest that they would be conducive to at least accepting action short of sanctions and military force. Over time, as trust is rebuilt between the P5, they may even become conducive to accepting more robust action, so long as they have the assurance provided by rvr that the need for veto restraint in the face of R2P breaches does not equate to automatically authorising the use of coercive military force. rvr can help in rebuilding the relationship between P5 members which has been so heavily tarnished by the events in Libya and Syria. If honoured by Russia and China, it can show the P3 members that it is possible to act through the Council on R2P matters. And if the P3 members operate through the Council and avoid unilateral action, they can show Russia and China that they can be trusted to honour the purpose of the Council, and the veto power. Without rvr or some other form of veto restraint measure, the Council seems destined to fail to meet its R2P for any future crisis where competing P5 interests are at stake.

Conclusion
This article has argued for the adoption of an informal 'Responsible Veto Restraint' measure. This nuanced approach to altering recent act Code and France-Mexico initiative proposals for veto restraint presents an avenue for global responsibility to protect 12 (2020) 385-414 veto reform that is transitional, appreciative of the purpose of the veto, and in theory, able to help in overcoming Security Council inaction over contentious R2P cases. While practical constraints -mainly the opposition of some of the veto wielding states themselves -make veto reform difficult, the transitional cosmopolitan view is that these constraints should not be interpreted as making reform an inherent impossibility. It may be possible to achieve a degree of veto restraint that can aid R2P in its application. However, the transitional cosmopolitan line requires that if we are to achieve this, we must temper expectations over what reform can help us achieve in the present, while fostering the avenues through which further progress can be made. rvr, satisfying the requirements of the transitional cosmopolitan tests for effectiveness and feasibility for reform, attempts to help in establishing this transitional middle position. This cannot aid in promoting action that would be perfect from a cosmopolitan view, but at the very least it can signal the potential for progress and act as a step in the normatively desirable direction offered by a cosmopolitan approach to human protection. The hard truth is that imposing veto restraint remains immensely difficult with the attitudes of some P5 members remaining less than favourable. Perhaps the case is that striking a transitional cosmopolitan balance for veto reform is too difficult: in attempting to make reform more feasible we perhaps only undermine its potential effectiveness and vice versa. Finding the balance between the two is immensely challenging, but certainly not something we should give up on. While even a mild proposal like the one offered here has issues, it has attempted to strike a balance by offering a transitional solution that attempts to satisfy the concerns of P5 members that could still contribute towards meeting the goals of R2P.
The success of veto reform will rest on continued campaigning in the UN, wider advocacy groups, and particularly on the attitudes of the P5 members themselves. Achieving reform will certainly not be an easy task. Yet efforts must continue to be made if the Security Council is to be able to live up to its responsibility to curb mass atrocities. It may be too premature to declare the Council unfit for realising R2P, yet if it continues to fail to provide meaningful action in the face of mass atrocities then the international community may have no choice but to consider alternatives for implementing R2P action.