Shortly after the United States and United Kingdom began their occupation of Iraq in April 2003, the two governments announced an ambitious program of reform. The occupiers’ governing authority, the Coalition Provisional Authority (CPA), declared that it would exercise ‘all executive, legislative and judicial authority’ in Iraq to advance ‘efforts to restore and establish national and local institutions for representative governance and facilitat[e] economic recovery and sustainable reconstruction and development’.Footnote 1 In the fourteen months that followed, the CPA pursued these goals with abandon, issuing decrees that fundamentally remade Iraq's legal, political, military, and economic institutions.Footnote 2 If one looked only at Iraq's written law, one might conclude that in just over a year the country had been transformed from an authoritarian one-party state with a centrally planned economy to a liberal democracy with one of the most permissive free-market systems in the world.
The Iraqi reforms posed a dilemma for international lawyers. On the one hand they presented a direct challenge to the traditional law of belligerent occupation. As codified in the 1907 Hague Regulations and the Fourth Geneva Convention of 1949, occupation law casts occupying powers as mere trustees who do not assume the legislative competence of ousted de jure sovereigns.Footnote 3 Article 43 of the Hague Regulations famously requires occupiers to respect the laws in force in the country ‘unless absolutely prevented’.Footnote 4 This ‘conservationist principle’, which limits the occupier's legislative authority, demarcates a critical boundary between occupation and annexation.Footnote 5 Occupiers enjoy limited legislative authority because they lack full sovereign rights over the territory, something annexation would presumably supply. But annexation is profoundly antithetical to contemporary international law.Footnote 6 In the traditional view, therefore, a ‘reformist occupation’ is a virtual oxymoron. As the UK Attorney General concluded on the eve of the Iraqi invasion, ‘wide-ranging reforms of governmental and administrative structures would not be lawful’.Footnote 7
This is not to say that an occupier is without legislative authority. That authority may be quite broad and permit a wide range of changes to local law, but the capacity to legislate must be commensurate with the temporary nature of occupation. Reforms deemed necessary to the tasks of occupation administration or to avoid the occupier enforcing local laws that infringe fundamental human rights are, in this view, necessary consequences of an occupier assuming temporary political authority. That is, they are essential to occupation being both effective and humane. There is even a plausible argument that an occupation lasting for many years (read Israel's) may require more extensive reforms since, as Adam Roberts observes, ‘[d]uring a long occupation, many practical problems may arise that do not admit of mere temporary solutions based on the idea of preserving the status quo ante’.Footnote 8 But reforms explicitly designed to outlast the occupation and address problems unrelated to the occupier's immediate security or governance concerns are not justifiable on ground that they are necessary for efficient administration during the occupation. Instead, as Nehal Bhuta puts it, transformative occupation derives its legitimacy ‘from the promise of the order to come’.Footnote 9 Such an occupation – again in the traditional view – appropriates legislative authority properly reserved for a locally chosen government that will follow the occupier's departure.Footnote 10
But the transformation of Iraq was not without normative roots.Footnote 11 The Iraq reforms seemed to resonate with many important trends in contemporary international law. They explicitly promoted human rights and democratic governance. They sought to bring political stability to a state emerging from armed conflict and years of authoritarian rule, a set of tasks now regularly undertaken by the United Nations and various regional organizations. They sought to open the Iraqi economy to foreign investment and greater integration within the global economy. They seemed consistent with the trajectory of occupation law more generally, which began as a means of securing the military objectives of European powers (though not when they operated outside Christian Europe) but moved in the post-World War II era to provide a ‘bill of rights’ to inhabitants of all occupied territories.Footnote 12 And the Iraqi reforms seemed to confirm the common observation that for most of its history the conservationist principle has been observed only in the breach, flouted and ignored so frequently that its prescriptions bear little relation to state practice of the past few decades.Footnote 13 According to this view, the CPA's blatant disregard for the conservationist principle simply affirmed its anachronism in an age when external reform of national institutions – particularly of the liberal democratic variety – is a common and widely accepted practice.Footnote 14
Each of these arguments figured prominently in the outpouring of scholarly commentary on occupation law in the years following the invasion of Iraq.Footnote 15 Professor Adam Roberts introduced the term ‘transformative occupation’ to describe Iraq and other occupations ‘whose stated purpose (whether or not actually achieved) is to change states that have failed, or have been under tyrannical rule’.Footnote 16 Much of the commentary recognized the novelty of transformative occupation: unlike earlier innovations such as the Fourth Geneva Convention that sought to protect the population of occupied states from depredations by occupiers, transformative occupation sought to legitimize rights-enhancing acts of legislative and constitutional change. Transformative occupation has thus been viewed as involving transformation away from repressive closed political systems and towards democratic systems that more closely adhere to international standards of governance and individual rights. From this perspective, the countervailing policies represented by the conservationist principle seemed weak indeed. Conserving existing law makes sense when the alternative is repression or even plunder by an occupier. But when the alternative is greater protection of human rights and the introduction of democratic politics, the principle appears regressive and even anachronistic.
These are compelling arguments. In this article I will suggest that they are almost all overstated, in some cases substantially so. The exaggerated nature of the conservationist principle's reported demise is critical in assessing the shifting normative expectations described by the principle's critics. Viewed in isolation, these claims might suggest that the conservationist principle survives with only a tenuous link to the policy assumptions, surrounding doctrines, and factual underpinnings of current international law. But I would like to argue that the conservationist principle retains a critical utility in contemporary law, indeed so critical that, I will argue, a presumption should exist against its demise or replacement by a doctrine of transformative occupation. The exaggerated nature of the criticism suggests that this presumption has not been rebutted in any convincing manner.
One certainly cannot deny that expectations for occupations have radically changed. Despite being outsiders to the territories, occupiers cannot be indifferent to the plight of citizens and so cannot treat their control as no more than an opportunity to bargain with the ousted regime over the terms of its return. The days of occupation as a means to secure the prerogatives of political elites are over, as they are for almost all aspects of armed conflict. But it is one thing to say that occupiers should refrain from neglecting or mistreating inhabitants. It is another to grant them licence to become agents of constitutional revolutions, which was effectively what occurred in Iraq. That task cannot be left to the unilateral discretion of a single state.
In summary, I argue the following. One of the singular achievements of the post-Cold War era has been a move toward realizing the UN Charter's goal of multilateralizing armed conflict. In particular, reconstruction of post-conflict states is now a regular feature of Security Council resolutions, under Chapter VII of the Charter. In assuming the role of an agent for change in post-conflict societies, the Council has performed a critical sorting function: it has legitimized collective involvement in matters of governance and social relations that sit at the heart of states' domestic prerogatives, and it has delegitimized unilateral efforts to the same ends. Enshrining transformative occupation into doctrine threatens to reverse that hard-won legitimacy, for its only consequence would be to empower unilateral state occupiers.
Such a costly step, whatever its purported benefits, should come only after clear and convincing evidence that the conservationist principle has been thoroughly eroded. That appears lacking. While occupation law more generally has often been honoured in the breach, Iraq literally stands alone as the only example in the post-Cold War era of a liberal and democratic transformative occupation. While some commentators seek to include multilateral post-conflict missions on the list of transformative occupations – a move that would make the practice appear quite common and widely accepted – conflation of multilateral and unilateral transformative projects is highly problematic. The former arose in large part in order to delegitimize the latter. Finally, while the extra-territorial application of human rights obligations is now increasingly accepted, those instruments provide a mandate for legislative change only in a narrow set of circumstances. Those circumstances have yet to find their way before an international court or treaty body, none of which has found human rights instruments to provide a mandate for extra-territorial legislative action.
How transformation challenges occupation law: refining the problem
The challenge that transformative occupation presents to the conservationist principle is quite unlike the problems previously faced by occupation law. Much of the commentary on transformative occupation misses this disjunction, describing a steady erosion of occupation norms through the second half of the twentieth century, culminating in the wholesale transformation of Iraqi in 2003. Iraq plays such a prominent role in these analyses precisely because it is linked to this prior history.
The Fourth Geneva Convention (and its elaboration in Additional Protocol I) was the last major innovation in occupation law. Geneva law has famously been described as providing a ‘bill of rights’ for the occupied population.Footnote 17 After two world wars in which occupiers committed widespread brutalities and sought to bend local law to their own interests, the obvious priority for the treaty's drafters was to strengthen individual rights protections in the territories.Footnote 18 That it did. Seeking to remedy the inadequacies of Hague law, its numerous restraints assumed an occupier unconcerned, to say the least, with the welfare of individuals.Footnote 19 But transformative occupation does not involve an occupier who seeks to subordinate local institutions, citizens, and resources to its own self-interested goals. It rather assumes an occupier attempting to enhance individual rights, build and strengthen local institutions, and increase the occupied state's capacity to use and capitalize on its natural resources (or at least attempt to do these things). This is because reform of the territory is central to both the occupier's initial decision to use force (regulated by the jus ad bellum) and its conduct once it gains control over the territory (regulated by jus in bello). In other words, a transformative occupier does not seek a purely geostrategic objective in its use of force, one in which occupation appears as merely a phase of the armed conflict. Instead, transformation accomplishes the occupier's jus ad bellum objective. The traditional concern that an occupier may subordinate rights in the territory to broader military objectives is thus largely absent. Professor Roberts notes this changed rationale in his seminal article,
Put crudely, the traditional assumption of the laws of war is that bad (or potentially bad) occupants are occupying a good country (or at least one with a reasonable legal system that operates for the benefit of the inhabitants). In recent years, especially in some Western democratic states, various schools of thought have been based on the opposite idea, crudely summarized as good occupants occupying a bad country (or at least one with a bad system of government and laws).Footnote 20
Roberts is entirely correct that distinguishing ‘good’ from ‘bad’ occupiers is a crude and imprecise distinction, but one that is necessary nonetheless. A rapacious (‘bad’) occupation is inconsistent with much of existing Geneva law. A transformative (‘good’) occupation can easily be seen as almost entirely consistent with rights-based Geneva norms. First, a transformative occupier commits no violation of the rights-enhancing provisions that are the distinctive feature of the legal era ushered in by the Fourth Geneva Convention. Its objectives are instead rights-enhancing. Second, a thorough transformation of law and political institutions is future-oriented, seeking as its primary goal not the restraint of an occupier's own propensity to violate rights but the creation of new institutions that will permanently enhance local political culture after the occupation ends. If Geneva law is indeed focused on individuals rather than the state as such,Footnote 21 it should present no obstacle to consolidating institutionalization of human rights protections for the time after the occupation ends.
Third, transformation takes as its model not the self-interested acts of predatory occupiers but the multilateral post-conflict reconstruction missions that have proliferated in the last two decades.Footnote 22 These missions seek to avoid renewed conflict by creating an inclusive and pluralistic politics in the post-conflict state. Former antagonists can only co-exist peacefully if they can reasonably expect to find their views expressed in national policy. The only system capable of making such an assurance is pluralist democracy. The missions have thus advanced a broad range of majoritarian and counter-majoritarian initiatives designed to convert former military foes into political rivals. This view of ‘statebuilding as a foundation for peace-building’ is also broadly consistent with Geneva law.Footnote 23
Viewed in this light, the case for transformative occupation seems compelling. It implicates none of the prohibitions in Geneva law, instead seeking to further the values embodied in that law. One might conclude that, at worst, apart from the conservationist principle, contemporary occupation law simply does not address transformation. No other provision of the Fourth Geneva Convention prevents an occupier from enhancing human rights protections over their prior state. As for the conservationist principle, it appears out of touch with important trends in cognate areas of international law – in particular the proliferation of post-conflict reconstruction missions and the universalization of human rights protections. What's not to like about transformative occupation?
Challenging the argument for transformative occupation
Thus conceived, transformative occupation is seen as promoting many widely accepted goals of contemporary international law: the protection of human rights, democratic transitions, post-conflict reconstruction, and the use of force for humanitarian ends. But does state practice actually support such a complete (or almost complete) retreat from the conservationist principle? A closer examination suggests that it does not. The following sections examine three areas of practice often invoked to support a nascent acceptance of transformative occupation: the practice of state occupiers, multilateral post-conflict missions, and the extra-territorial application of human rights treaties.
State practice of occupiers
The flood of literature that followed the occupation of Iraq leaves the distinct impression that transformative occupations became common in the post-World War II era.Footnote 24 This is a highly misleading view. Few occupations in the post-World War II era have even attempted to bring about liberal democratic transformations. While the Iraq occupation certainly did so and produced a rich legal discourse in its wake, Iraq stands as an exceptional case in the universe of post-World War II occupations. The vast majority have been non-transformative and, when transformative, largely illiberal.
There have been nineteen unilateral occupations since 1945, including the immediate post-World War II cases of Germany, Japan, and Austria.Footnote 25 This number does not include UN-sponsored nation-building missions, which are addressed separately below.Footnote 26 Only four of these nineteen cases involved comprehensive liberal democratic transformations and three of those four cases (Germany, Japan, and Austria) occurred over sixty years ago in the aftermath of World War II. There is only one case of comprehensive liberal reform in the post-Cold War era in which humanitarian concerns were a significant motive for warfare: that of the United States in Iraq. Six cases involved non-liberal transformations: Turkey in Cyprus, Indonesia in East Timor, Vietnam in Cambodia, the Soviet Union in Afghanistan, Iraq in Kuwait, and Israel in the Palestinian territories.Footnote 27 In the remaining nine cases, no meaningful transformation occurred.Footnote 28 One study of new constitutional orders promulgated under occupation – a related but not entirely identical phenomenon to transformative occupation – found that constitutional change is overwhelmingly the work of a few leading military powers, some democratic and some not.Footnote 29 The authors conclude that ‘occupation constitutions should be seen as a particular strategy of particular states, rather than a global phenomenon’.Footnote 30 In short, transformative occupation is very much the exception in a set of cases dominated by occupations that fit the traditional model comprehensively regulated by Hague and Geneva law.
Instead of examining the entire class of recent occupations, many authors focus only on the immediate post-World War II cases and two more recent high-profile cases: Israel in the Palestinian territories and the United States in Iraq.Footnote 31 Such a limited sample obviously cannot support the proposition that parties to the Hague and Geneva instruments have rejected the conservationist principle.Footnote 32 Even so, does this small subset of cases represent a unified legal phenomenon?
Germany, Austria, and Japan
Though following very different paths, the occupations of Germany, Japan, and Austria obviously involved liberal democratic transformations. But three factors counsel hesitation in marshalling these cases to argue for the conservationist principle's demise. First, whereas the three western allies used their post-World War II occupations to bring democratic change, the Soviet Union did not. Indeed, if one looks beyond Soviet actions in its East German zone of occupation to its actions elsewhere in eastern Europe, post-World War II practice appears anything but uniformly liberal and democratic.Footnote 33 In this sense, the status of occupiers' legislative powers seems to share the stagnant quality that marked so much of international practice in peace and security law during the Cold War.
Second, even assuming the Western occupations launched a challenge to the conservationist principle, that challenge was not taken up when the international community codified occupation law in the light of post-World War II practice. This occurred twice. First, during drafting of the Fourth Geneva Convention of 1949, the United States proposed language that would have effectively validated its actions in Germany and Japan and given occupiers virtually unlimited legislative discretion.Footnote 34 The proposal was opposed by the Soviets, as well as several Western states, and was not adopted.Footnote 35 Instead, Article 64 of the Convention reiterated the conservationist principle, though the rights-enhancing provisions elsewhere in the Convention effectively abrogated it for their specific purposes. The second codification was Additional Protocol I of 1977, which addressed a series of occupation law issues. None of its provisions purported to modify the conservationist principle.
Third, the arguments used in the late 1940s to justify the post-World War II occupations have not aged well. Post-World War II Allied lawyers faced a daunting task in justifying de-Nazification in Germany and Austria and the new constitution for Japan drafted by General MacArthur. All appeared blatantly inconsistent with Hague law, a conclusion they strove mightily to avoid. The justifications that these lawyers devised, while clever and even plausible at the time, are hardly replicable today.Footnote 36 Most relied on debellatio and related principles appurtenant to the right of conquest.Footnote 37 In a widely cited article, for example, Robert Jennings argued that, because the victorious allies could have annexed Germany, they were also ‘entitled to assume the rights of supreme authority unaccompanied by annexation; for the rights assumed by the Allies are coextensive with the rights comprised in annexation’.Footnote 38 These doctrines find scant support in contemporary international law, as they are wholly incompatible with the UN Charter's limitation on the aggressive use of force and affirmation of the sovereign equality of states. As a result, the post-World War II claims have dropped out of scholarly discussion and official justifications of state policy, which focus instead on the human rights-centred justifications examined here.Footnote 39 None of the post-World War II justifications are repeated in any of the now voluminous writings on transformative occupation. In short, the post-World War II occupations are separated by much doctrine and history from contemporary views of liberal democratic transformation. Claims that they support a legally unified conception of transformative occupation are unsustainable.
Israel in Palestine
The problem of legal anachronism is not present in the case of Israeli occupation of the Palestinian territories. Israel's direct administration lasted from its military victory in 1967 to its transfer of civilian authority to the Palestinians in the 1993 Oslo Accords.Footnote 40 It is certainly the most-discussed occupation of the modern era and is cited by virtually all commentaries on transformative occupation as supporting the erosion of the conservationist principle.Footnote 41 But it is also the most widely condemned: individual states and international organizations have loudly and consistently denounced the occupation itself, the extension of Israeli law to the territories, the proliferation of settlements, and a wide variety of practices that violate the human rights of individual Palestinians.Footnote 42 One puzzles over the attempt to claim Israeli actions as precedent for a liberal democratic model of transformation when Israel has been criticized for pursuing precisely the opposite path: violating Palestinians human rights and denying them the opportunity for democratic self-government that would come with full self-determination.Footnote 43 Such condemnations, like those of abuses in other illiberal occupations, find deep resonance in the negative prohibitions of existing occupation law and present no basis for discarding the conservationist principle in search of a new justificatory theory.
Even putting aside these condemnations, Israeli changes to law in the territories have not amounted to comprehensive transformation, let alone a liberal democratic one.Footnote 44 Most of the legislative changes (and they were substantialFootnote 45) can be understood as furthering the security and political interests of Israel rather than enhancing the quality of life in the territories.Footnote 46 The initial terms of each of the occupations – in the West Bank, Gaza, East Jerusalem, and the Golan Heights – were established in proclamations constituting military administrations for the territories. Each declared that the law in force would continue as long as it did not conflict with the order or subsequent proclamations.Footnote 47 The later changes can be grouped into four main areas. The first was the enactment of a substantial body of security-related decrees and the establishment of military courts to hear cases of their alleged violation.Footnote 48 Second was extending the jurisdiction of Israeli civilian courts to people, property, and events in the territories in which at least one party was an Israeli citizen.Footnote 49 Local courts continued to operate as before, except when claims were brought against the occupying authorities.Footnote 50 The extension of jurisdiction followed the growing interaction between Israelis and residents of the territories as the occupation lingered on and provided a means of resolving disputes in an Israeli forum under Israeli law.Footnote 51 Third was extending Israeli law and making Israeli courts available to the growing settler population in the territories.Footnote 52 The fourth set of changes involved a comprehensive effort to integrate the territories' economies with Israel's, described in the early days of the occupation as a ‘common market’.Footnote 53 These reforms were extensive and included the abolition of internal customs barriers and creation of a single external barrier for international trade; unification of indirect taxation; facilitation of free passage between the territories and Israel; introduction of Israeli currency into Gaza and the West Bank;Footnote 54 amendment of local banking laws to reflect Israeli practices;Footnote 55 and even replacement of the local traffic code.Footnote 56
While these measures were expansive and highly intrusive into life in the territories, it would be a category mistake to group them with reforms intended to enhance individual rights and autonomous political decision-making. With some notable exceptions, they were designed primarily to ease the task of administering the territories, promote the security interests of the Israeli Defense Forces, and ensure that Israelis living in or having contact with the territories continued to have recourse to Israeli laws and courts. They neither sought to facilitate a democratic transition nor to protect Palestinians from human rights violations perpetrated by the most powerful political actor in the territories, Israel itself. As a result, most of Israel's actions fit easily within the existing negative protections of the Fourth Geneva Convention.
Iraq 2003
The Iraq occupation was transformative beyond any case since 1945. Its explicit promotion of liberal democratic institutions and massive disregard for existing Iraqi law, as well as the US and UK's direct acknowledgment of their status as Occupying Powers, makes it the paradigmatic case in the current debate. One might argue that, for these reasons alone, it has advanced the legitimacy of transformative occupation.Footnote 57 But the legal import of the occupation of Iraq has become something of a Rorschach test. Those who saw a trend toward increasing acceptance of transformation view Iraq as confirming that trend; others who see it as violating existing law do not.Footnote 58 More fundamentally, some view occupation law as the critical normative framework for Iraq, while others believe that the Security Council authorized the reforms in Resolution 1483.Footnote 59 And some view that resolution as situated within the evolving nature of occupation law, while others view it as an act of legislative fiat that overrode but did not engage with occupation law as such.Footnote 60
Iraq, then, is at this point in history not so much a data point in the evolution of the conservationist principle as a blank canvas on which broader arguments about the legislative power of occupiers are applied. But even those who view Iraq as presaging a crisis for the conservationist principle acknowledge two limitations on its use as precedent for future transformations. The first is that the occupation was extraordinarily mismanaged.Footnote 61 The United States' unwillingness or inability to govern the country effectively will probably play an important role in assessing claims that unilateral transformations are essential to moving post-conflict states toward liberal democracy. The second is that to describe Iraq as having been fully transformed by its occupier is inaccurate. Critical aspects of the Iraq transformation took place after the occupation ended on 30 June 2004 and in response to Security Council demands for an early restoration of local sovereignty.Footnote 62 The first legislative elections, the drafting of a new constitution, the election of a new government pursuant to that constitution, and the Status of Forces Agreement providing for the withdrawal of all US combat troops by the end of 2011 all came after the CPA ceased to exist.Footnote 63 Of equal importance was that, as the occupation progressed, the United States increasingly sought assistance from the United Nations both to legitimize its actions in Iraq and to perform critical tasks it proved unable to carry out. Grant Harris has described an ‘invisible hand’ exerting pressure on occupiers to multilateralize their operations, and Iraq is his prime example.Footnote 64 The fourteen-month US occupation may present less a legacy of unilateral transformation than is often assumed.
Inspiration from multilateral missions
The second area of practice is multilateral, the mostly UN-authorized missions to post-conflict states that have proliferated since the early 1990s.Footnote 65 Much recent literature questioning the conservationist principle cites the UN missions in identifying a trend toward viewing the post-conflict period as a vital opportunity for reform.Footnote 66 And, indeed, reconstructing societies emerging from civil war has become a central task of international organizations. Many such missions have an explicitly transformative mandate. The clearest examples are the territorial administrations approved by the Security Council for Bosnia, Kosovo, East Timor, and Eastern Slavonia.Footnote 67 In many other cases the UN (with Security Council approval) has been integral in making transformation a part of winding down armed conflict: negotiating peace agreements that commit the parties to legal and political reform, crafting a role for the UN as monitor and facilitator of the reforms, providing expertise to implement the reforms, and pronouncing on their consistency with international standards. Examples include UNAMSIL in Sierra Leone,Footnote 68 UNOCI in Côte d'Ivoire,Footnote 69 ONUB in Burundi,Footnote 70 and MINUSTAH in Haiti.Footnote 71 In their most robust form, these changes seek to remake the political structures and cultures of states in which warring groups have been unable to co-exist in peaceful political competition. The model favours norms of tolerance, inclusion, and participation. The new institutions are familiar from Western democratic states, pairing majoritarian elections with robust protections of individual and minority group rights.Footnote 72
The parallels to transformative occupation are obvious. Both typically follow the end of major hostilities in armed conflicts. Both favour liberal democratic institutions implanted through wide-ranging legal reforms. Both pay little regard to the formalities of traditional sovereign exclusivity over national governance. Invoking a Chapter VII mandate (for the multilateral missions) or the prerogatives of de facto authority (for transformative occupiers), both focus on popular sovereignty rather than effective control as the critical fount of legitimate governmental authority.Footnote 73 Finally, both embody the hope – though it is often a hope tinged with scepticism and a sense that the initiatives are the best of a series of bad options – that these new structures will eventually socialize formerly warring factions into peaceful co-existence. The newly implanted democratic institutions are thus valued not only for their own sake but as means of avoiding renewed conflict.Footnote 74
Much thoughtful scholarship has explored a potential convergence between these two forms of post-conflict political transformation.Footnote 75 But to conflate multilateral missions with occupation by states acting without a Security Council mandate or authorization would ignore significant differences between the two. These differences flow from one critical fact: multilateral transformation is authorized by the UN Security Council under Chapter VII of the UN Charter and transformative occupation is not.
First, approval by the Security Council allows for debate over the wisdom and objectives of the transformation. This prior vetting provides a critical legitimacy for the resulting missions. If Thomas Franck is correct that normative legitimacy emerges from an inclusive and transparent process of law-making, then consideration in the Security Council is critical.Footnote 76 An inclusive legislative process is more likely to produce inclusive implementation of transformative measures.Footnote 77 Here the international reaction to Iraq provides an object lesson. Facing an occupation that had become a fait accompli, following a war they had almost all opposed, members of the Security Council in 2003 provided only tepid support for the US-led reforms.Footnote 78 Equally, multilateral planning for an occupation arguably produces better outcomes. Again Iraq is illustrative: US planning for the post-war period was notoriously inept and insensitive to local concerns.Footnote 79 It is inconceivable, for example, that a UN-led mission would have disbanded the entire Iraqi army as one of its first actions. Equally, it seems unlikely that the free-market fundamentalism driving so many of the CPA's economic reforms would have survived Security Council consideration. While the UN has made its share of errors in post-conflict states, it has amassed a large repertoire of practice and at least makes a conscious effort to learn from its mistakes.Footnote 80
Second, unilateral intrusions into states' domestic orders are severely limited by contemporary international law. Eliding the distinction between unilateral and multilateral transformation acts requires overlooking these limitations, or at least minimizing their significance. Transformation is often described as seeking goals that are widely shared in the international community, and in a limited sense this is correct. Multilateral instruments have removed human rights and electoral matters from the realm of exclusive domestic jurisdiction. And, as previously noted, reforms in both areas are increasingly seen in particular as critical means of pacifying post-conflict states. But the scope of these normative goals cannot be divorced from the circumstances of their creation. Human rights treaties are carefully negotiated to accommodate states' divergent interests and to provide for limited enforcement mechanisms through national legal institutions and international bodies.Footnote 81 UN Security Council involvement in post-conflict reconstruction is subject to critical procedural limitations, principally the veto. More generally, all forms of state responsibility are specifically denoted as not licensing responses that violate the UN Charter or other serious international legal obligations of the injured state.Footnote 82 Normative objectives, in other words, are inseparable from the limited and collective means of their implementation. Human rights and democracy norms have never been understood (at least outside certain powerful states) as a licence for unilateral action.
The reason is clear: states do not trust each other's judgement, in the absence of collective oversight, to determine when and how these norms should be implemented. They fear pretexts. The temptations for political meddling are simply too great. That a prohibition on unilateral intervention is the necessary flip side of robust collective articulation of norms is made clear by the myriad non-intervention norms specifically addressed to unilateral action. The Friendly Relations Declaration, the Nicaragua case, the severe limitations on counter-measures, and robust principles of sovereign immunity are but the highlights of a post-World War II legal order that strongly favours the collective over the unilateral.Footnote 83
Are there reasons to exclude transformative occupation from these limiting principles? Not based on the desirability of the goals it seeks, for their collective authors thought them no less desirable but refused to license their unilateral implementation. Not based on the absence of other means of achieving reform, since the Security Council has been authorizing missions to post-conflict states for more than two decades. Not based on the claim that humanitarian law – the primary regime governing occupation – displays a greater tolerance of unilateral acts than human rights law. The usurpation of legislative authority contemplated by transformative occupation is specifically excluded by the law of occupation in its critical distinction between a de facto occupier and a de jure sovereign.Footnote 84 To elide this distinction and allow the former to effectively become the latter would grant occupiers the sole prerogative they lack by virtue of their temporary, de facto assertion of authority.Footnote 85 Transformative occupation is unique only in that it presents a propitious opportunity to alter domestic institutions. The defender of those institutions – the ousted regime – has ceased to exist. But to claim that improper unilateral acts can be easily accomplished is not much of an argument either.
Third, assimilating collective and unilateral transformations threatens to permeate the jus ad bellum/jus in bello divide. A critical function of a Chapter VII mandate for reform during occupation is to override or displace the conservationist principle. The Security Council exercises legislative authority to render lawful a practice that was (in my view) unlawful under the normally applicable jus in bello.Footnote 86 But, in so doing, the Security Council does not also legitimize the intervention that enabled foreign forces to occupy the territory. That would have required a separate authorization, one the Security Council (as in the case of Iraq) may have been unwilling to give,Footnote 87 and one that would implicate an entirely different body of law: the jus ad bellum regulating the initial use of force.
The question is whether the barrier between these two regimes can be maintained when a unilateral occupier, invoking multilateral missions as precedent, transforms the law of the territory it occupies. Answering this question requires understanding the incentives created by the possibility of transformative occupation. Imagine an intervention to halt rampant human rights abuses. To ensure that they do not reoccur once foreign troops leave, the interveners create liberal democratic institutions during the subsequent occupation. What kind of occupier is likely to engage in such actions? Non-democratic occupiers are highly unlikely either to intervene on humanitarian grounds or to engage in democratic reforms. Democratic occupiers may intervene for reasons unrelated to human rights and democracy, but in those cases their likelihood of undertaking a transformative occupation is uncertain. Panama and Grenada are cases of intervention for non-humanitarian reasons that were not followed by transformative occupations, while Iraq and the US/Allied wars against Germany and Japan in World War II are cases where non-humanitarian entry into conflict was followed by transformative occupation. The only scenario in which transformative occupation is almost certain to be attempted – to a greater or lesser degree in each case – is when a liberal democratic state intervenes for humanitarian reasons. For such a power to intervene because human rights abuses have become intolerable but then to leave the abusing government in place and depart would be self-defeating. And, indeed, one common argument made by supporters of transformative occupation is that contemporary warfare has come to be dominated by a transformative objective.Footnote 88 As Grant Harris writes, ‘the motivation for much of modern warfare is precisely to reshape occupied territory and intervene in the lives of its inhabitants’.Footnote 89 The most likely scenario for transformative occupation becoming a jus in bello question, therefore, is one in which it is also an asserted justification under the jus ad bellum.
Herein lies the danger. Though unilateral humanitarian intervention is the subject of intense and ongoing debate, international lawyers are virtually unanimous that it contravenes the jus ad bellum.Footnote 90 If this is correct, how can that prohibition on initiating armed conflict to secure human rights be maintained if an intervener knows that, when it succeeds in occupying a territory, the jus in bello will allow it to enact the very reforms that prompted the intervention? Where transformation is the reason for intervention, in other words, the consequences of legitimizing transformation cannot realistically be confined to the jus in bello. Justification and purpose cannot be disentangled. As Adam Roberts acknowledges, ‘an element of artificiality marks the proposition that transformative goals may be acceptable, but only as a by-product of military action, not as its real justification’.Footnote 91 The danger is avoided when the Security Council must vote separately to legitimize intervention and then reform. But if the latter cases are taken as precedent for acts by unilateral occupiers, then Security Council authorization will have ceased to matter for jus in bello purposes. The incentives suggest that it will soon cease to matter for jus ad bellum purposes as well.
In sum, conflating unilateral and multilateral transformations creates a dangerously potent incentive for the jus in bello to legitimize violations of the jus ad bellum. Such a breach of the barrier between these two bodies of law admittedly occurs in the opposite direction from that usually warned against: the division is usually described as preventing claims of jus ad bellum legality from justifying violations of the jus in bello, such as claims of a just cause legitimizing abuse of detainees. But the problem here is arguably more insidious, for using jus in bello legality to justify initiation of an entire armed conflict may well have much farther-reaching consequences than legitimizing specific illegal acts in the course of armed conflict.
Extra-territorial application of human rights treaty obligations
The third part of the claim for emergence of transformative occupation is the extra-territorial application of an occupier's human rights treaty obligations. If an occupying power displaces the de jure government of a state that is bound by human rights treaties, it is argued, the occupier's authority within that state's territory should be subject to the same treaty norms. Otherwise, the population would be divested of the treaty's protection through no actions (or fault) of its own. Further, because human rights treaties require state parties to enact rights- protective legislation at home, the same obligations would apply when they assert governing authority through occupation.Footnote 92 The full panoply of treaty obligations effectively functions as a political blueprint for a human-rights-compliant society. Occupiers of illiberal states may thus be compelled to act as social engineers, creating the norms, procedures, and institutions of a treaty-compliant, liberal society. As Grant Harris notes, ‘there is no natural ending point for an occupant's obligations to protect human rights short of the creation of a functioning state with permanent institutions capable of doing so’.Footnote 93
The general proposition that states remain subject to at least some of their human rights treaty obligations during an occupation is now well established in international law.Footnote 94 But to move from this general principle to legitimizing transformative occupation requires a complex journey. In part this is because much of the jurisprudence on extra-territoriality – apart from that of the European Court of Human Rights – is quite superficial and provides little explanation of how a conflict between human rights (legislative obligations) and humanitarian law (prohibition on legislative changes) is to be resolved. Occupation sits uneasily on the border between armed conflict and peacetime governance. On the one hand, it presents an excellent opportunity to open closed societies to the robust debate and political competition provided by democratic institutions. On the other, the security environment in many occupations is so fragile that the military requires powers unthinkable during peacetime.Footnote 95 This duality creates a problem of legal categories. If occupation is viewed primarily as a stage of armed conflict, governance can become a vehicle for human rights protection only when doing so is compatible with the occupier's ability to maintain security. Occupation law tempers its protection of human rights with significant deference to considerations of military necessity.Footnote 96 But the opposite view – regarding occupation as equivalent to peacetime governance – would subject occupiers to human rights treaties that make no concessions to military necessity.
The options for reconciling these two regimes are unsatisfying. Deference to military necessity can obviously be retained by viewing occupation solely as a matter of humanitarian law – perhaps as the lex specialis – but this would negate the now-established application of human rights treaties to occupied territories. Alternatively, those treaties could serve as the sole relevant legal framework, but human rights treaties make no accommodation for military necessity since they were designed for peacetime governance.Footnote 97 Making no allowance for the military exigencies of occupation would be unworkable and probably unacceptable to most states. While derogation clauses obviously exist in human rights treaties, to invoke them in order to permit necessary security operations would effectively neuter the greater protections offered by human rights treaties and render the protections offered by the two regimes substantially similar.Footnote 98 The same would be true of a ‘flexible’ interpretation of the treaties to limit the scope of rights when they affect security concerns. The more nuanced approach of the International Court of Justice (ICJ)'s Nuclear Weapons opinion, which views humanitarian law as the lex specialis to be applied in the event of conflict with particular human rights norms,Footnote 99 promises a more fact-specific analysis but would similarly privilege military necessity over rights protection.
Even if one could overcome this doctrinal incoherence and occupiers were deemed fully bound by their human rights treaty obligations, legislative change in the occupied territory would only be required in a narrow set of circumstances. And those are circumstances in which the occupier's justification for acting without a Security Council mandate – that is, acting pursuant to occupation law as opposed to Chapter VII authorization – is at its weakest. A conflict between human rights treaty obligations and the conservationist principle would only arise if an occupier could not comply with the treaty without undertaking legislative action.Footnote 100 Human rights treaties contain three distinct forms of obligation: states parties must themselves refrain from violating protected rights; they must ensure that others within their jurisdiction refrain from such violations; and they must act affirmatively to ensure that right-protective procedures exist, remedies are provided, and victims are compensated.Footnote 101 I argue that only the third form presents an obvious conflict with the conservationist principle.
First and foremost, states parties must themselves refrain from violating protected rights. This is a purely negative obligation. Occupiers may comply by issuing orders to their military forces and civilian administrators that rights not be violated. Such inaction is hardly inconsistent with the conservationist principle. No new legislation is needed in order for occupiers themselves to act with restraint.
One might respond that an occupier will frequently govern through existing laws and institutions. If those violate human rights treaties in and of themselves (as opposed to the way in which they were applied by the ousted regime), then legislative change would become essential to compliance with the treaties. But this argument will apply only to a rather small category of laws that require government officials to violate human rights. If the law simply allows such violations – for example, by permitting discrimination against classes of citizens, permitting police to use abusive practices, permitting courts to truncate due process, or permitting governmental censorship, and so forth – the source of any violation would be an official's choice to violate rights, not a legal compulsion to do so. Ensuring that local administrators do not continue to exploit the law's permissiveness to violate rights would be a matter of the occupier's issuing clear instructions for the laws’ implementation.
Second, states parties must ensure that others in territory they control do not violate rights. In the Congo/Uganda case the ICJ described this as an obligation ‘to protect the inhabitants of the occupied territory against acts of violence, and not to tolerate such violence by any third party’.Footnote 102 In an occupied territory, such third parties can usefully be grouped into two categories. The first are those in a subordinate relationship to the occupier, such as a compliant local administration or an allied armed group. While restraining such third parties will certainly require more affirmative acts than when an occupier simply restrains itself, in essence the obligation is still to abstention from human rights violations: the third parties must not torture, rape, murder, or discriminate, for example. They might be restrained by direct orders, or more active follow-up by the occupier might be needed, such as ensuring that orders filter down through a chain of command and that obstinate officials are relieved of their duties. All such acts would meet the ICJ's requirement that third-party acts not be tolerated. If the relationship is truly one of superior–subordinate, there seems no reason why a change to existing law should be expected to bring third-party violators to heel, when directly ordering – and perhaps even compelling – their compliance has not.
The second category of third parties are those not in a subordinate relationship to the occupier. Indeed, they may be insurgents who directly target the occupier. Not to ‘tolerate’ violations by such independent or hostile groups would involve maintaining security in the territory, a basic obligation of occupation law.Footnote 103 Because non-subordinate third parties do not act through the apparatus of state power, their human rights violations would not involve the sort of official misconduct that might require new legislation. As in the Congo case itself, violations by these groups are likely to involve independent acts of violence that can only be deterred or ended through enhanced security measures.Footnote 104 New legislation would do little to address these groups’ violations.
Only a third form of treaty obligation might require legislative action. In addition to the negative obligation to refrain from abuse, human rights treaties also require states to act affirmatively in order to prevent violations, investigate abuses, punish violators, and ensure justice to victims.Footnote 105 If existing law does not create these institutionalized monitoring and enforcement mechanisms, states parties can only comply by creating them through legislation. Some rights by their nature require institutional mechanisms, such as democratic elections and fairness in prosecuting criminal defendants. Such rights involve entitlements to a particular institution or procedure. Other rights may require expanding institutions to meet previously unaddressed challenges, such as violence against women in the private sphere. The degree of institutional change needed to protected rights with important procedural components can be quite far-reaching.Footnote 106
This third category presents a clear conflict with the conservationist principle. The first thing to be said is that there is literally no support in existing jurisprudence on extra-territoriality for the proposition that occupiers must build institutions and reform laws. All of the extra-territoriality cases against occupying powers before the ICJ and the European Court of Human Rights have involved human rights abuses by the occupiers themselves.Footnote 107 Each involved acts that could have been halted or prevented without changing the laws in force. One can also observe that this category does not involve the worst human rights abuses in occupied territories. Few truly egregious acts – massacres, rapes, torture, plunder, mass expulsion, persecution of disfavoured groups – require legislative solutions. If history is any guide, these are almost always the acts of occupying forces themselves or their close allies. They fall into the first or second categories described above and the parties can stop these actions by restraining themselves. The Iraqi plunder of occupied Kuwait in 1990–1991, for example, could have been halted by orders given down the chain of command.
But even if the cases are few, the question remains whether this third category of affirmative obligations should be applied extra-territorially. Here the discussion must inevitably turn to the type of occupation that the international community chooses to validate. For, as the prior discussion suggests, only a limited class of reform-minded occupiers is likely to pursue broad legislative change. From recent occupation experience, two specific scenarios seem to implicate this category. The first is a prolonged occupation in which no plans are made for an immediate return to local control and the occupier assumes responsibility for the territory over the long term. The second is where an occupation is commenced for the specific purpose of legal and political reform, namely a transformative occupation. In both situations the occupier makes governance decisions based not only on the pragmatic necessities of the moment (including compliance with the individual rights provisions of Geneva law) but also to address perceived long-term problems in the territory. In the case of prolonged occupation, the problem is legal stasis over time. As the Israeli Supreme Court has observed, ‘[a] prolonged military occupation brings in its wake social, economic and commercial changes which oblige him [the occupier] to adapt the law to the changing needs of the population’.Footnote 108 In the case of transformative occupation the problem is the many maladies of a closed authoritarian society.
Should human rights treaties enable such occupations, whose long-term strategic objectives and reform agendas are inexorably intertwined? If one answers yes, one must be prepared to defend the proposition that human rights treaties can permit (or even compel) the de facto annexation of occupied territory.Footnote 109 Legislating reforms intended to be permanent is fundamentally incompatible with the temporary nature of occupation.Footnote 110 Such fundamental policy decisions are the prerogative of the indigenous sovereign, and to appropriate them is to assume the mantle of that sovereign. As Pictet observes, the ‘temporary, de facto’ nature of the foreign presence is ‘what distinguishes occupation from annexation’.Footnote 111 The ICJ's holding that the Israeli security wall could become ‘tantamount to de facto annexation’ because it would ‘create a fait accompli on the ground that could well become permanent’Footnote 112 invokes a similar idea: when an occupier assumes powers reserved to the de jure sovereign – in the Wall case the demarcation of boundaries – it effectively treats the territory as its own, even without its formal incorporation.
The multilateral option
The attraction of transformative occupation is twofold. First, it promises to move illiberal state institutions toward compliance with international standards of human rights and democratic governance. Second, it takes advantage of a unique moment in which transformation appears possible. No forces stand between a reform-minded occupier and necessary political change. The conservationist principle can well appear unnecessarily obstructionist and even anachronistic in rejecting both of these justifications as unconvincing. The criticisms above, focusing on the agent of change – a state acting unilaterally – do not address these points.
But this Manichean view of transformation – as either a de facto annexation or a critical opportunity to advance human rights – is itself an anachronism. It assumes that the agent of transformation will be a state, since only states are limited by the conservationist principle and would be the only agents of change liberated by its obsolescence.Footnote 113 But a third option exists: authorization by the Security Council under Chapter VII.Footnote 114 Few doubt that if the Security Council were to authorize an occupier to reform local laws and institutions its legislative authority would override the conservationist principle.Footnote 115 As a result, its actions would cease to be governed by occupation law, at least in regard to limits on the occupier's legislative authority. Instead, such occupation would be akin to the many missions that the Security Council has created over the past two decades with detailed mandates to advance human rights and political democracy in post-conflict states.
That so much energy would be expended in arguing to expand the unilateral authority of Occupying Powers is itself surprising. The goal of the UN peace and security regime has been to multilateralize all aspects of armed conflict.Footnote 116 In the post-Cold War era it has largely succeeded. According to two important datasets of armed conflict,Footnote 117 there were ten major inter-state armed conflicts between 1990 and 2010.Footnote 118 All but two of these were addressed in one form or another by the UN Security Council, which took actions ranging from authorizing intervention to supporting regional peace processes.Footnote 119 The Council's involvement has not been episodic but holistic, as it regularly addresses all aspects of armed conflict from inception to termination.Footnote 120 It seeks to mediate disputes that appear likely to result in armed conflict; authorizes responses to cross-border incursions; condemns violations of humanitarian law in the course of armed conflicts, including referring matters to the International Criminal Court; assists in negotiating ceasefires and eventual peace agreements; and, as has been noted, dispatches reconstruction missions to post-conflict states.Footnote 121 The UN (via Security Council action) has moved well beyond simply responding to aggression to deploy sophisticated strategies of prevention, mediation, reconciliation, reconstruction, and exit from conflict zones.Footnote 122
This move to multilateralism has been particularly evident at the post-conflict stage.Footnote 123 The UN has become the indispensable actor in such transitional operations, displaying an evident learning curve and increasing store of institutional expertise. To ensure the organization continues to learn from both its successes and its failures, in 2005 the Security Council created the Peacebuilding Commission, with a mandate to ‘bring together all relevant actors to marshal resources and to advise on and propose integrated strategies for post-conflict peacebuilding and recovery’, as well as to ‘focus attention on the reconstruction and institution-building efforts necessary for recovery from conflict’.Footnote 124 Reviewing the UN post-conflict record in 2006, Doyle and Sambanis concluded that the organization is most successful when it is involved in all aspects of a transition from conflict resolution to stable peacetime governance.Footnote 125 A greater role for the United Nations thus enhances the effectiveness of post-conflict reconstruction.
With the entire trajectory of international conflict management moving toward collective responses, and with the UN more experienced at reform than any national military (including that of the United States), an expansion of an occupier's prerogatives via occupation law seems a throwback to a time when collective options were either non-existent (before World War II) or non-functional (the Cold War era). Enhancing unilateralism is out of touch with the many areas in which international law in the post-World War II era has refused to license unilateral action, even when that action is described as essential to achieving humanitarian objectives. Examples include a general rejection of unilateral humanitarian intervention, a reluctance to endorse broad conceptions of universal jurisdiction, and refusal to recognize human-rights-inspired exceptions to sovereign immunity in national courts. It is also out of touch with the many ways in which the Security Council has integrated human rights and democracy concerns into its strategy for post-conflict states. Indeed, the Security Council has pursued these objectives to such an extent that some scholars argue that they are being introduced too quickly into politically fragile societies.Footnote 126 In the early twenty-first century, enhancing unilateral authority over domestic affairs is the real anachronism.
Of course, the multilateral option will not always be available. Vetoes by permanent members, a lack of political will, local resistance to reform, and a host of other factors may all render the Security Council incapable of approving reforms. As of the time of writing, the Security Council's inability to address the Syrian crisis in a meaningful way seems a case in point. But this observation does not necessarily provide affirmative support for the unilateral option, for the Security Council may be unavailable on fewer occasions than might be supposed. If an occupier has received Security Council approval for an initial use of force for the purpose of transformation, the Security Council is quite unlikely to refuse support for the transformation itself. Security Council refusal only seems likely where an occupier has either failed to seek approval at all for the intervention or approval was sought and refused.Footnote 127 In such cases, the Security Council would not be asked to approve reforms until after an occupation had become a fait accompli. At that point, with the occupier in place, many of the strongly held reasons why Security Council members might have opposed the initial intervention would lose their potency. The invaded state's sovereignty would have been compromised, the occupier would have gained a foothold in a region from which certain Security Council members might want it excluded, and the geopolitical consequences of the intervention, if any, would have occurred. None of those feared eventualities could be reversed by opposing a resolution sanctioning reform in the territory. Perhaps more importantly, the Security Council would be given an opportunity to dictate the parameters of the occupation. In short, the incentives in such circumstances would seem to be aligned against Security Council refusal to approve reforms. The limited possibility that the multilateral option might fail thus provides only weak support for radically expanding the unilateral option.
Seen in this light, the conservationist principle is not a proxy for international law abdicating responsibility for needed legal change in occupied territories. It is instead a limitation on unilateral self-help that finds much support elsewhere in contemporary doctrine.
Conclusions
The Iraq occupation spawned an important debate over the role of occupiers in territories in substantial need of legal reform. The central problem was not the one traditionally facing occupation law: how to protect inhabitants from a rapacious or brutal occupier that subordinates their wellbeing to its own political objectives. It was rather what to make of an occupier that seeks to enhance the inhabitants' wellbeing. Some commentators found little reason to prevent such an occupier from legislating new standards of human rights and democratic government now well established in international law. In particular, the conservationist principle that restrained such supposedly beneficial legislation was anachronistic. It was inconsistent with post-World War II occupation practice, failed to follow the lead of UN-authorized post-conflict reconstruction missions, and was incompatible with an occupier's extra-territorial human rights obligations.
Each of these grounds for declaring the conservationist principle an anachronism, however, turned out to be substantially overstated. Iraq was the only case in the era of Geneva law in which an occupier has legislated for an explicitly transformative purpose. UN-authorized missions, while seeking similar goals, differ in significant ways from unilateral occupations, which make their invocation as legal precedent highly problematic. And the extra-territorial application of human rights treaty obligations has been used only to restrain occupiers from committing or tolerating abuses. No international court or tribunal has found an occupier in violation of its treaty obligations for failing to legislate in foreign territory it controls.
None of these responses to the supposed emergence of transformative occupation is dispositive in the sense that it reveals transformation to be necessarily incompatible with contemporary occupation law or that state practice shows definitively negative reactions to particular cases of transformation. But the bar need not be set that high. It is sufficient for the purposes of the argument I have made that these claims are only moderately persuasive. That is because they carry a substantial burden of showing why the legislative powers of occupiers should be exempt from the general trend of multilateralizing all aspects of armed conflict. The arguments supporting transformative occupation fail to carry this burden.
