Constitution Making Under Occupation: The Politics of Imposed Revolution in Iraq

Our focus on democracy should not be presented to others as an imperial command … such a policy needs as its moral lodestone the traditional American value of prudence, not a neo-Trotskyite belief in a permanent revolution (even if it is a democratic rather than a proletarian one). The neoconservative insistence that the United States can be made safe only by making other nations accept American values is a recipe for provoking a clash of civilizations.

—R. Ellsworth and D. K. Simes, “Realism’s Shining Morality”

It is now commonly conceded that the United States, led on the ideological level by “neoconservative” intellectuals, tried to initiate and almost certainly failed to sustain a radical revolutionary project to remake the Islamic Middle East. Without reducing the cause of the war in Iraq to this one ideologically driven factor, few serious people would dare to deny that it was among its causes as well as part of its meaning. At the same time, the invasion of Iraq, the overthrow of its dictatorship, the occupation of the country for however long a period, and the initiation of a process of “regime change” have not generally been understood as a revolution or a revolutionary project.

We should not be surprised that it was a few conservatives opposed to all revolutions who were the first to properly decipher the revolutionary political character of what the United States has set in motion in Iraq.¹ Nevertheless, I myself have proposed the concept of “externally imposed revolution” from the very beginning² and would like to maintain it here against its main conceptual rivals, “nation building” and “regime change,” despite one reservation.³ Revolutions, as opposed to rebellions, have to involve a plausible process of establishing new regimes.⁴ They fully establish their legitimacy with the construction of the new, not merely the old’s destruction.⁵ The imposed revolution in Iraq has postulated, authentically or not, such a terminal point of legitimacy, namely constitutional democracy. However, a civil war or permanent revolutionary instability are antitheses of a political regime. By this standard, the imposed revolution in Iraq has not only failed so far and almost certainly will end up as a failure, but, because civil war and instability were always much more likely than a stable new regime, failure seemingly was preprogrammed.⁶ But failure was not a certainty. In my view, even the illegal war left open a small window of opportunity for a transition in Iraq toward some undetermined kind of constitutional democracy, and ultimately that is the possibility that validates the use of the term “revolutionary project.” That this revolution was to be externally imposed is part of the reason why the window of opportunity for success was so narrow.

I find the use of the term “nation building” overambitious and highly misleading. It has been used most absurdly for the American occupation of Germany and Japan,⁷ where, in the latter case, the identity of the nation as an “imagined community”⁸ was not in serious doubt, and, in the former one, a West German nation was never actually intended or created, as the events of 1989 and soon afterward showed clearly enough.⁹ In neither country was the meaning of “national community,” thankfully enough, open and available for the occupying power to create or mold. This state of affairs is implicitly conceded by some authors who use the term. According to Noah Feldman in What We Owe Iraq,

nation building in Germany and Japan aimed to transform former enemies into prosperous allies in the emerging new struggle … [knowing] these nations had the capacity for unity, organization, and productivity, we sought to make them over to move them into our column…. The objective was not to build democratic states for the benefit of their own citizens…. It was far less important that Germany and Japan be democratic than that they be capitalist and rich.

(7, my emphasis)

Aside from the fact that these lines do not give enough credit to the actual stress on democratization and liberalization during the two occupations,¹⁰ they concede by their very language that Germany and Japan were already nations before supposed nation building, with the latter term reserved for the development of capacities conducive to wealth, capitalism, and geopolitical adherence to the American side. Feldman then goes on to quickly concede that such are not our goals in Iraq. So why use the same and already misleading term “nation building”? Perhaps to create a “civic nation” with a viable identity out of the centrifugal main elements of Arab Shi’a, Arab Sunni, and Kurdish religious and secular populations hitherto held together (supposedly only) by a succession of authoritarian states? Feldman, in my view rightly, does not propose such an ambitious goal, which would justify his terminology: “with the Cold War behind us, the objective of nation building … must be to build stable, legitimate states whose own citizens will not seek to destroy us…. In short: the objective of nation building ought to be the creation of reasonably legitimate, reasonably liberal democracies” (8).

Terminology aside, what Feldman really seems to have in mind is either the “state building” of the social science literature¹¹ or “regime change” exactly as the neoconservatives have used this term.¹² The reasons become clearer upon examining Francis Fukuyama’s essay “Nation-Building and the Failure of Institutional Memory.” To be sure, while aggressively maintaining that “nations” can indeed be made by external powers, he does this on the basis of the single very questionable colonial case of India, and he entirely leaves out of his account the truly nation-building work of the movement led by the Indian National Congress. More importantly, he seems to concede (to unnamed European critics) the general inapplicability of the term, though he clings to it in the title of his essay.¹³ “Outside powers can succeed at negotiating and enforcing ceasefires between, say, rival ethnic groups; it is seldom that they can make these groups understand that they are part of a larger, nonethnic identity.” Indeed, according to him, what has occurred even in Germany and Japan was not “state building” (“state” apparently identified with “nation”!) but democratic relegitimation of government and the drafting of democratic constitutions. On the question of the state when distinguished from nation, Fukuyama is ambivalent. In another essay, he suggests nation building is a response to state failure, which leaves the American project in Iraq without a conceptual definition, because here, as he admits, state (suddenly clearly distinguished from nation) failure was caused, partly inadvertently and partly deliberately, by the invaders themselves.¹⁴ James Dobbins in a recent essay has no such difficulties with respect to Iraq, because, as he sees it, nation building is the proper response to both failed and rogue states.¹⁵ But his own bizarre definition of nation building (“the use of armed force in the aftermath of a conflict to underpin a transition to democracy”)¹⁶ is inconsistent with this stress on the state and is relevant only where state building is either not a problem or can be easily solved. Moreover, nation building is here almost synonymous with regime change, with the added and rather unnatural element of armed force built into the definition.¹⁷

The term “regime change” is formally less objectionable than “nation building” with respect to Iraq, despite its objectionable political uses. I will go further. Whoever the source of the theory may be, as it pertains to the desirability of democratic regimes from the point of view of the international order,¹⁸ it is for me not difficult to agree with the claim that relatively liberal and relatively democratic political systems are in themselves desirable everywhere and would reduce in the longer term the risk of war among states and the chances of the citizens of such states joining the enterprise of international terrorism. How this goal is to be achieved, however, is an entirely different matter (as Kant already well knew, rejecting “republican imposition”).¹⁹ Despite its rather new current implication, the term “regime change” evidently does not in itself suggest external force or even political rupture. The problem with the term as opposed to “nation building” is that it is too general and permissive and not specific and demanding. Almost a synonym of political transition, “regime change” points more accurately to the locus of change (to regime, that is, the form of government, rather than government in terms of incumbency on one side and state structure on the other),²⁰ but it does not reveal much about the modality of change, which here is the crucial question. To be sure, this is an advantage in the Iraqi case, with respect to nation building. While the latter term is almost nonsensical, empirically, when an external power is in the driver’s seat, the same is not true for regime change or political transition.

In a large and complex typology, Juan Linz and Alfred Stepan have isolated three ways in which defeat in war could play a major role in transitions from authoritarian to democratic forms of rule.²¹ Interestingly, however, a careful study of their options, based on preexisting regime types (totalitarian, sultanist, post-totalitarian, and authoritarian, with the first two allowing the same externally dominated transition path only) reveals that they may be thinking ultimately of only two types of cases. The first is when a dictatorship, its state and society, suffer total defeat in war and an external power is free to occupy and impose for a considerable period without much resistance. Germany and Japan could be considered examples of this phenomenon,²² even if, as I will later show, in neither case can we speak of absolute imposition. The second type is when a dictatorship suffers a military defeat that domestically discredits it and forces it to accept a process of internally steered and negotiated regime change with, or more usually without, some influence by the military victor. Here Greece in the 1970s and Argentina in the 1980s come to mind.²³ Neither type covers Iraq very well, because Iraqi society did not suffer total defeat, yet the military victor tried to assume total control over the transition process. Note that in all four examples, unlike Iraq, the dictatorship was the initiator of the hostilities it subsequently lost.²⁴ Perhaps too linked to examples of the past, Linz and Stepan do not take into account the possibility of an aggressive war against a dictatorship that may mobilize nationalist forces first on its behalf and later against the military victor. Even more importantly, the typology assumes that somehow the internal regime type and acts of the external power will be correlated in a harmonious way; in other words, it is strongly implied that a democratic external power (apparently, only they are relevant to democratic transitions!) that wins a war will choose the highly intrusive method only where there was totalitarianism or sultanism before, and not if the society has a sufficient level of organization, as under an authoritarian state, for example.²⁵ Two types of very possible “errors” are thereby disregarded: cognitively driven ones and normatively driven ones. In the first case, the victorious power and its experts may be simply wrong about the nature and politics of the society they defeat. In the second, they may not misunderstand the nature of the society but may wish to dominate it for whatever reasons, and they do not mind and may even wish to neutralize or suppress existing internal forces that were themselves not part of the dictatorship. We cannot exclude either or both of these possibilities, especially in the case of Iraq.²⁶

If regime change or political transition is to be kept as the general class concept under which America’s project in Iraq is to be understood, its specific modality has to be understood more clearly than it is possible in an analysis that draws its types from primarily historical cases. We need a more abstract scheme, and legal theory provides the answer. In the most persuasive comparative analysis, that of János Kis, reform is defined as continuity of both legitimate authority and legality, revolution as rupture in both dimensions, and coordinated transition or negotiated (regime) transition as rupture of legitimacy but generally with legal continuity (see table 1).²⁷ Legitimacy is understood here in the sociological sense of general or at least elite acceptance of the claims of the rulers to justify their rule, while legal continuity is understood, following Hans Kelsen,²⁸ as the limitation of change to a form that relies on a regime’s own rule of change. It is assumed by the scheme that after ruptures of legality and legitimacy a transition path (one of the three varieties, since reform involves no rupture) will involve the construction of a new legality, a new legitimacy, or both, in the case of a revolution. Legality, of course, means a legal order in the sense of the positivists and not necessarily the rule of law.²⁹ Especially since both legality and legitimacy can be matters of interpretation and indeed contestation, we have to keep in mind the ideal-typical character of this scheme, as we must for all such schemes. In reality, we may very well encounter borderline cases and mixed and contested types. While the scheme is applicable in principle to transitions to authoritarian rule and to counterrevolutions as well as revolutions, here our concern is exclusively with transitions from dictatorships toward more democratic forms of rule, at least in the minimal sense of Dahl’s polyarchy or near-polyarchy.³⁰ Finally, and most important here, the Kis’s four-part scheme can be expanded to eight if we differentiate among externally induced and internally generated versions of each path.

Given the large variety of external interventions possible, it is easy to postulate that there have been historical examples of each of these forms with strong outside influence. Even the negotiated transitions of eastern central Europe in 1989 and 1990 depended on the withdrawal of Soviet guarantees to ruling parties. In the twentieth century, it is very probable that the United States alone has strongly promoted all four types of change in Latin America, though admittedly not usually in a democratic direction. In the Linz and Stepan scheme already referred to, military defeat played a role in the transformations of Greece and Argentina (transformations corresponding to “coordinated transition” in Kis’s scheme), with comprehensive negotiations not playing the central role because of the various levels of collapse of the old regime forces. Thus the expansion of the scheme is both logically and historically justified.

Admittedly, while there are many situations where the idea of an externally influenced or imposed revolution may not seem very controversial, there are some problems with this notion when applied to an invasion and occupation of a defeated country. Charles Tilly rightly understands classical revolutions (following Trotsky) in terms of a doubling of governmental power and sovereignty between the old regime’s forces and new contenders, with the latter occupying sovereign power alone in the case of revolutionary victory.³¹ In the case of war, invasion, and occupation, an analogous process begins, but it occurs between a national government and a foreign power. However, except in the case of outright annexation, in the case of external invasion the doubling and subsequent resumption of unitary sovereignty over a territory seems to follow a different logic than in the case of internal revolution. Not every occupation following an invasion is revolutionary, but it does not require annexation to make something like a revolution. Some occupations can be seen as conservative. In fact, according to the Hague Convention of 1907, the occupying power was to be a placeholder for the absent governmental sovereign. Tilly’s revolutionary scenario would then be abrogated when the absent sovereign was restored. Accordingly, in Nehal Bhuta’s persuasive analysis, which uses Carl Schmitt’s relevant concepts,³² the occupation regime would have been a dictatorship, but a commissarial rather than a sovereign or a revolutionary one. A military occupation can be described as an externally imposed revolution only when it becomes transformative, instead of the classical occupatio bellica. Of course, annexation would be one form of transformative occupation that would do without new regime construction in the occupied country, since it would be incorporated into the already existing regime of the occupier. But the occupation government turning itself into the subject of new regime construction would be another no less revolutionary form and, as Bhuta rightly argues, it is the latter that was attempted in the case of Iraq.³³

Note: In this chapter, I argue that the typology is complicated depending on what is transformed, since a type of change can occur on the level of government, regime, and/or state.

Does the externally imposed aspect vitiate or increase the authoritarian potential of revolutions? The most basic issue has to do with the internal logic of revolution itself, from the legal point of view, and this is why the category is often left out altogether from typologies of transitions to democracy even though such a transition is exactly the goal of many revolutions.³⁴ The authoritarian consequences of revolutions are almost indisputable empirically,³⁵ and Hannah Arendt’s masterly analysis remains the best treatment of the elective affinity between revolution and dictatorship, even if in her book On Revolution³⁶ she produces an entirely exceptional and “successful” case that avoids this logic: the American Revolution. This thesis treats (following Carl Schmitt and one version of an old argument of the Abbe Sieyès) most revolutions as revolutionary dictatorships, with a part of the people (class, elite, party, or even one man) exercising constituent power in the state of nature, outside of all law and normative limitation, attempting to impose new rules on all others.³⁷ The argument is consistent with the typology used here, which keeps revolution in its framework for logical rather than empirical reasons, but it involves an asymmetry with other modalities of change: only here is a complete break in the forms of normative integration of political society. Revolution cannot be de jure anything but dictatorship, a point well understood by Lenin and Carl Schmitt.³⁸ Examining the nature of executive power during revolutionary breaks in legality reinforces the argument. Here the classical formula is that of a provisional or interim government that exercises either merely de facto powers or powers commissioned by the constituent assembly, powers that in either case tend to be unlimited by any separation of powers or checks and balances. Thus, extending Schmitt’s still unusual language, the sovereign dictatorship is reinforced by a more classical emergency or commissarial one.³⁹ These dictatorial forms involve actual or anticipated resistance and thus authoritarian preventive measures or countermeasures in a society with any complexity and conflict potential, especially given the utopian aspirations released by most revolutions. However democratic the goals, the revolutionary means inevitably wielded by minorities almost always tend to vitiate them.

On first sight, external imposition of a revolutionary logic tends to double the imposition and thus potentially the authoritarian consequences. The historical record seems to support this supposition. Revolutions as defined here and called by that name by the actors themselves can certainly be externally influenced, promoted, and even imposed, as we know from numerous examples in the twentieth century. From the East European to the Asian cases, these have not been democratic revolutions in terms of their outcome and generally their ideology as well. Since the states doing the promotion have been dictatorships and have sought to export their social-political system, as Stalin explained to Milovan Djilas during World War II,⁴⁰ the result surprised only some fellow travelers in the West. A large part of this record is thus irrelevant to our concern. Historically, however, the project to export democracy (republicanism) or at least constitutionalism by governments already organized according to these principles through violent external overthrow of authoritarian regimes is also not entirely new, if we can count the European old regimes among the latter. Yet the efforts of the French Revolution when still a republic to export its own political forms abroad more often than not culminated in military occupations, puppet regimes, or unstable revolutionary dictatorships.⁴¹ The Latin American revolutions also involved external force vis-à-vis local old-regime forces, for example in the Andean possessions of Spain, and again the success was ambiguous.⁴² In case of the United States, there have been many experiments in the violent overthrow of regimes, some of which at least were coupled with the export of democracy. Here too the success rate has been low, raised somewhat only by counting an outright colonial experiment such as the Philippines or a country with a long history of internal dictatorship, such as South Korea.⁴³

In the end, Germany and Japan, two very special cases, as I have repeatedly argued,⁴⁴ remain the best historical evidence that democracy can be exported and imposed through the violent external overthrow of dictatorships. Linz and Stepan seem have these cases in mind when they imply that external imposition can accomplish a democratic transition where no other option has a chance. They do not, however, consider the contrast with internal revolutions, which is instructive. While empirical evidence based on two cases (whose differences with Iraq I systematically present below) proves little, it could nevertheless be abstractly argued that external imposition has the advantage, in that an external occupier and monitor not only can remove the forces of the old regime but can also block the efforts of newly mobilized actors to impose nondemocratic solutions. External occupation may force such new movements and parties to work together and accept solutions that would not have been their first choice but that they can come to accept as “the only game in town.” For example, if a Shi’ite revolution had hypothetically overthrown Saddam, it would have been difficult to force victorious clerical leaders to accept any kind of power sharing and open competition with secular or Sunni elites. Even if the leaderships wished to compromise, they would generally not be able to restrain their victorious militants. In a revolution (legally a state of nature), to paraphrase Thucydides, the victor (inevitably a dictator, at least temporarily) takes what he can, and the defeated suffers what he must. With an American occupation, however, no indigenous force could be in a position to impose its own solutions unless the occupier so wished. (It is difficult to see this point now that both the interim constitution and the supposedly permanent one have been imposed on the Sunni part of the population, but this scenario would have been predetermined in an internal revolutionary scenario and was not in the externally imposed one.)

At the same time, in the case of external imposition there are fundamental problems with the legitimacy of a new transitional order even graver than in the case of an indigenous revolution.⁴⁵ The issue can be best explored and deepened if we consider another aspect of Arendt’s treatment of revolution: her analysis of the term in two phases, liberation and “constitution.”⁴⁶ According to her, a genuine revolution such as the American one would involve not only the destruction of an old regime, liberation, but also the project of the constitution, the construction of a relatively stable, new order that she identified (as did the founders of the United States) with the writing, enactment, and institutionalization of a written constitution. This stress, of course, is very welcome, since it highlights the importance of the topic of this book: constitution making. For the notion of revolution used here, I repeat, the bifurcation of the concept indicates not only breaks in legality and legitimacy, but also that the construction of new legality and legitimacy must be included. Given the at least minimally legitimating potential of legality (Weber’s rational legal legitimacy, after all) in negotiated transitions and the continuity of legitimacy in the other types of change, revolutions thus face the most serious problems of legitimacy among the types of change mentioned here. The doubling of revolution between liberation and constitution opens up a huge gap between the work of liberation (the overthrow of the old regime) and the work of constitution (the creation of the new one), in which legality is almost transparently based on facts alone and legitimacy is based on future promises made by an agent.⁴⁷ Confidence and trust in such an agent becomes all important; distrust and lack of confidence can lead to disaster, by reducing the size of the group or groups that can identify with the new regime. And whatever the advantages of an external agency otherwise, this mode of producing a revolution only increases the need and the difficulty of legitimation.

In indigenous revolutions, that portion of the people that plays a heroic and self-sacrificing role in the work of liberation has at least a claim to represent the people and their interests before the latter can express themselves through democratic channels. The reason why new elites in control of provisional governments, inevitable components of revolutionary transformations, can be accepted as legitimate for a relatively short period is that they have worked to liberate the country from a (generally) hated old regime. In theory, they have suffered what the people have suffered, or even more, and their hopes in part at least also coincide with the hopes of the rest. In an age of nationalism, it is easier to identify moreover with members of one’s own national groupings. In the case of an external imposition, however, it is almost impossible to distinguish liberation from occupation. Whatever the external liberator thinks of itself, it will have its own motives and interests, and these may not coincide with those of the country’s population. After defeat in war, unless a country is freed from an external occupier or its obvious puppets, the liberators will be seen by some and probably to most as occupiers if they stay long enough to do any good, to really help stabilize and frame the political competition. Any interim government they sponsor, especially if it has no major political credit of its own in the work of liberation, will suffer from this legitimacy problem, and the inevitable role of such governments in shaping more permanent arrangements will be the object of hypercritical scrutiny and suspicion. The occupying power—the Americans—can claim that they will let the Iraqi people gain control over their interim process of democratization before elections can be held, but they must then a priori identify “the people” and its plausible representatives. If any important group is excluded, it can claim to represent all dissatisfied parts of the population against the occupiers and their appointees. Whatever their legal status, that excluded group may find it as easy to speak in the name of “the people” as those groups who represent the occupiers—and to many, much more plausibly.⁴⁸

To be more precise in our understanding of the kind of revolution the Americans wrought in Iraq, we must analyze the term “liberation.” The concept has had an important place in Arendt’s theory of revolution and in American self-understandings of their actions in Iraq. More recently, an increasing number of commentators⁴⁹ have pointed to the supposed transformation of a genuine liberation to something universally regarded as occupation, especially its legal codification as such by UN Security Council Resolution 1483,⁵⁰ as the beginning of the disaster in Iraq.⁵¹ Are liberation and occupation incompatible? It depends, first of all, on the meaning of “liberation,” and second, on the nature of the occupation in question. There are at least two meanings of the term “liberation,” a less demanding and general one that narrowly focuses on the overthrow of an oppressor, and a more demanding one that in addition seems to imply and require that it should be an organized force (or forces) that is liberated and stands ready to take over the task of at least provisional governing.⁵² In the case of internal revolution, which Arendt analyzed, the two concepts reduce to one, because a revolution carried out by indigenous forces implies that at the very least they will be free or liberated to govern provisionally and to play the constituent role of building a new regime. The difference, however, is very real in the case of an external “liberator.” Here, for example, using the more demanding concept, liberation would have to entail the empowerment of a national government previously overthrown by a foreign occupier, drawn from elements including, for example, an underground army, resistance forces, or parties or movements capable of organizing political life. But what if none of these possibilities are available? According to the more general concept, it would still not be meaningless at all to consider a country liberated of a foreign oppressor or domestic dictatorship even if organized internal political force could not immediately take charge of the political process, because presumably its individuals could still be made free in their private capacity, and that could be the basis (depending on the nature of political arrangements) for the emergence of autonomous political forces. Thus it is obvious that neither concept would be compatible with the mere replacement of one oppressor by another. Neither a palace coup, nor the replacement of one colonial ruler by another, nor even the replacement of a dictator by a colonial ruler equally oppressive would count as liberation in either sense.⁵³ But what may be liberation to someone using the less demanding, more general concept would not be understood as such to another interpreter using the more demanding one. The latter may regard the very same state of affairs as a mere conquest of power followed by occupation, though not necessarily in a normatively negative sense.

The issue is further confused by the fact that when speaking of liberation in the general and less demanding sense the same events may represent, normatively speaking and in terms of fundamental interests and values, liberation to some and occupation to others. This issue does not come up in those liberations where a country is freed from the occupation of a hostile external power, but it was already an issue in a place such as my native Hungary, where the previous military occupation was by an allied power (Germany) and liberation was by an enemy power (the Soviet Union). In the actual case, Germany was a friendly power to some Hungarians and was hated by others, the Soviet Union was favored by very few and feared by many more, and probably most wanted to be occupied by neither at all. (I leave aside the question that in fact the Hungary of 1945 was more democratic until 1947 or 1948 at least than the Hungary before 1944, and thus the Soviet occupation could be represented as liberation in the more demanding sense from the point of view of many democratic, including non-Communist, forces.) The issue is even more complex—if at all possible—in the case of an indigenous but very oppressive dictatorship, say Iraq, overthrown by an aggressive external power seen by many groups as the enemy of the country’s national and perhaps religious aspirations. One must probably imagine that from a normative point of view, for most Iraqis (the Kurds being exceptional) the same events represented initially both liberation and occupation, which had to be welcomed and feared at the very same time. And it may very well be that it was the subsequent nature of the actual liberation/occupation that came to decide for most of them—probably very quickly—which was the way they were going to see the American presence: either as Wittgenstein’s rabbit (liberation) or hat (occupation).

How the Americans themselves answered a whole host of questions concerning the meaning of liberation was to play a very important role in Iraqi interpretations. There was no question that they regarded themselves as liberators and therefore legitimate. But their self-understanding concerning legitimacy (no more than regarding the legality of the invasion) did not automatically become the all-important view of others, without which legitimacy in the sociological sense would not be possible. Thus it very much mattered how the U.S. policymakers were going to answer one of the very first questions they faced: were they going to interpret liberation in the less or the more demanding sense? If it was to be in the more demanding sense, they needed to find the Iraqi version of the Afghan Northern Alliance⁵⁴ or, more absurdly, an Iraqi “de Gaulle” to whom they could hand provisional or interim power right away, so as not to become occupiers instead of liberators. There was, however, no Northern Alliance in Iraq in the sense of an armed force operating in the territory with some relative success against the government’s own forces, except the Kurdish minority, whose rule over Iraq would have been resented even more by the majority Arabs than would American rule. In 1991, when an Arab (or allied Arab-Kurdish) force on these lines could have emerged, internal uprisings in the bulk of the country were not supported by the United States because they would have brought power, supposedly, to the Shi’ite clerics allied with Iran. (The Northern Alliance was also friendly with Iran and Russia, but in the Gulf the stakes were much, much higher.) It seems that this argument was still alive in 2003, when the leader of SCIRI (the Supreme Council of the Islamic Revolution, then based in Tehran), the later tragically assassinated Muhammad Baqir al-Hakim, was publicly (for example, on PBS just before the invasion) asking for just such a role, and when this idea was still probably interpreted as handing Iraq to the allies of Iran. Power could also not be handed to Sunni generals, who also arguably possessed military forces within the country, because they were now understood to be incompetent, weak, and among the forces from which the country needed to be liberated. While handing power to the Shi’ite clergy was not possible, their second betrayal was also not possible. Thus earlier State Department ideas on this score were quashed.⁵⁵

Thus the question became whether there could be a “de Gaulle,” an exile leader such as Ahmed Chalabi or a set of leaders (Hakim; Jaffari, the leader of Da’wa; Talabani and Barzani, the two Kurdish leaders; and either or both Ayad Allawi and Chalabi; with one Sunni with no following, such as the aged A. Pachachi thrown in) to whom power could be handed in advance and who could form a provisional government immediately (or even before, in exile) upon overthrow of the Saddam regime. This would have been in fact the (externally facilitated version of the) classical revolutionary scenario, with the inevitable consequence that the provisional government would control the subsequent election of delegates to a constituent assembly.⁵⁶ Versions of this proposal were advanced before and during a conference of exiles in London in 2002 and before and after the actual invasion. But the analogy of Chalabi or even the gang of five, six, or seven exile leaders to Charles de Gaulle, who had a sizable army, control of some French colonies, an extensive internal underground network in France, and alliances with the powerful Communist underground, were always nothing short of ridiculous.⁵⁷ This time it was probably the State Department that torpedoed the plan to set up a provisional government in advance, with the president vetoing it in the end. The same drama, with the same six or seven claimants, repeated itself during the transition from Jay Garner’s proconsulship to L. Paul Bremer’s, with the same outcome.⁵⁸

Some interpreters argue that the fact that Jay Garner was given no plan by the Pentagon was meant to put him in the position where he would have no alternative but to hand over the reins of power very quickly to a provisional government led by Chalabi.⁵⁹ Whether or not that was actually or deliberately the case, “the plan not to plan” did not work. It is more reasonable to assume that at all points contradictory things were decided under a president with a strong will but with poor analytical ability to see the difference among alternatives.⁶⁰ The end result was that there was a decision, by design or default, for liberation in its less demanding sense. This meant a more or less extended occupation, of course, but we must be careful to note here that the occupation was still to be aggressively represented as liberation.

The issue is put by many very cynically, and I have no way to dismiss or to prove the claims involved. The argument runs that the Americans removed the old dictatorship in order to establish their own power position in the region. This could be securely done only by occupation, not by way of a provisional regime to which power would be transferred. The point then would be not that they intended to establish an occupation regime outright, since there is a lot of evidence that they feared it and understood the associated practical and symbolic problems—but that they feared all other alternatives much more.

In any case, the appearance of having undertaken the invasion and occupation of Iraq on behalf of imperialist goals had to be most strongly avoided. An occupation regime is inevitably a foreign dictatorship, and, significantly, on the eve of his departure L. Paul Bremer asked for and received from President Bush the plenitude of all powers for himself at the top of the Coalition Provisional Authority. Lakhdar Brahimi was quite right technically to subsequently call him the dictator of Iraq.⁶¹ But it should be noted that previously all attempts to establish a military occupation, with a U.S. general assuming the mantle of MacArthur, were strongly resisted.⁶² Bremer’s airs to openly play that role were greatly resented in Washington and by his own staff for, at the very least, public-relations reasons.⁶³ His was to be at all moments a dictatorship that had to be carefully disguised—a very difficult job indeed, given the arrogance and ignorance of the team that actually managed it on the American side.⁶⁴ A mere disguise, however, would not work in a country that had experienced both colonialism and neocolonialism unless measures were undertaken to indicate its good faith.

Continuing the train of this argument, the historical solution to the dilemma would have been to insist on that old staple of imperialist ideology, namely tutelary or pedagogic dictatorship, but fortunately that cannot work today. A foreign occupier would have to expropriate native authoritarian rulers, not on his own behalf but on behalf of future popular sovereignty and fundamental rights—the only two valid justifications in our liberal, democratic, postmetaphysical age. The claim would have been that the indigenous population (because of political culture, or because of having lived under an authoritarian regime for too long, or because of its economic conditions—take your pick) is not ready to exercise popular sovereignty and to respect the rights of individuals. Thus a long preparatory period under external referees and educators that will introduce gradually the new institutions would be necessary. This was still the idea under the famous League of Nations mandates (one of which created Iraq, more or less), but even then the timeframe for how long formal occupation was acceptable was already in the process of being greatly reduced.⁶⁵ Such a tutelary policy can combine various instruments of direct and indirect rule, as it in fact did under the British Mandate in Iraq, from which the Americans were to inherit, consciously or not, some administrative and even constitutional techniques.⁶⁶ For tutelary rule today, after the signing of the UN charter, decolonization, and the general taboo against colonialism, if the old arguments are to be used at all, one has to substitute international threats, as in the case of the cold war, or alternately restrict the functional area to which the claims are applied (for example, organizing elections) and, therefore, the acceptable duration of the formal occupation. All attempts to openly revive mandates and trusteeships under whatever name are clearly nonstarters⁶⁷ and fuel only the very plausible suspicion that the United States is involved in creating some kind of long-term protectorate in Iraq.⁶⁸ Thus Iraqis continually suspected a great number of occupation measures, some of which were obviously disastrous and others that could have been disastrous under different circumstances, as having to do with the secret desire of the United States and United Kingdom to extend the occupation in whatever legal form indefinitely.⁶⁹ Were they wrong? Feldman, a minor American participant, later (mid-2004) considered the dissolution of the Iraqi army to be a tremendous error. But afterward, according to him, there was no going back on the decision, and it became the “ethical” duty of U.S. forces to remain until a new Iraqi army under civilian control and capable of establishing public security could be created from the ground up. Facing this type of argument, it was easy to excuse the suspicion that (1) with militias taking over the task of local defense, the time period for forming a genuinely national army could be a very long one; and (2) the result with respect to the time factor could not have been merely an unintended consequence of well-intentioned but erroneous decisions.

The atmosphere of suspicion the Americans fostered further reduced their choices if they were going to in any sense keep alive the idea or the ideology, let us perhaps call it the latter, of liberation. The point was not only that they had to disguise dictatorship, but that they had to return this form to its classical essence: keep it short, and more technically, keep it commissarial: in other words, they had to avoid the usurpation of Iraqi sovereignty in its most fundamental, transformative dimension.⁷⁰

Of the relevant issues, it was the question of time that came to be best understood. If Patrick Cockburn (supported by important Iraqi perceptions) is right in saying that under whatever guise the period of the CPA represented an experiment in imperial direct rule, from the outset the project was doomed to as rapid a failure as Arnold Wilson’s attempt after World War I.⁷¹ While there were indeed advocates in the U.S. government for a program of “robust occupation” in Iraq, as in postwar Europe, the stronger side in government, this time the Pentagon, militated for a short occupation when it failed to have its Iraqi clients installed in the first place.⁷² This very much converged with Iraqi desires, and thus there was little chance that a program of long occupation could be formally adopted. While Bremer himself worked out a rather long period of CPA rule under a seven-point program that would have taken perhaps two years,⁷³ the time element was always the most vulnerable part of his mission, and in the end that element was radically curtailed. Of course, this did not mean that the United States was ready to leave Iraq altogether, after the end of a formal occupation regime. Iraqis were right to suspect that the status of some kind of protectorate was a serious possibility for them, perhaps on the model of arrangements with the British between 1932 and 1958.⁷⁴ It is almost certain that a future Iraqi government was projected that would negotiate extended troop presence and military bases, supposedly on its own initiative.⁷⁵ Also, a powerful U.S. embassy was going to take over many of the functions of the CPA. The models, to be sure, were supposedly Germany and Japan rather than neocolonial Iraq, but in any case they involved a very reduced period of formal occupation. Everything thus depended on how short a time a formal occupation regime really had to be restricted to and, even more, how it would be used.

Even as the duration of the occupation regime was in the process of being restricted under pressure, this regime had to solve very difficult tasks, taking into account two desiderata: effectiveness and legitimacy. The first had to do with generating outcomes acceptable from the point of view of (not always compatible) short- and long-term American interests, and this complicated task required that maximum powers would be exercised, given the shortness of the timeframe. If something had to be conceded, like the formation of a Governing Council or its Constitutional Committee, or surrendered, like the idea of making a permanent constitution, or devolved, like the role of picking the interim executive, or greatly accelerated, like the so-called transfer of sovereignty, the tendency always was to reduce the concessions to a façade behind which the CPA continued to exercise the real power—power that went beyond a commissarial dictatorship and toward a transformational or sovereign one.⁷⁶ All such moves, always transparent enough to the Iraqis, the external bloggers,⁷⁷ and eventually the international press, were utterly problematic from the point of view of legitimacy.

For legitimacy, the reduction of the time of the formal occupation mattered, but not enough. A process of legitimation built around the key words of “liberation” and “democracy” demanded other things than effectivity, namely the use of even the short time period for the gradual empowerment of Iraqi actors, understood as widely as possible. Either Iraqi actors had to be actually empowered or the actions of the American authorities had to be effectively represented as the basis of such future empowerment. To do this, one continually had to argue against democracy and participation (as that would have endangered effectivity) in the name of democracy. Authoritarian measures had to be given democratic window dressing or justified as the only possible road to democracy by pseudodemocratic public relations. Dealing with this contradiction is probably what generated the Bonapartist character of Bremer’s rule in Iraq, whose subterfuges did not lose all power just because they were transparent. Indeed, the category of Bonapartist dictatorship is applicable here in two ways, as the depiction of a foreign-imposed transformational regime⁷⁸ and of a dictatorship that uses democratic public relations and a façade of participation, representation, and consultation to hide its authoritarian practices.⁷⁹

Thus, behind the question of time and timing were three important substantive questions linked to legitimacy and effectivity and related to problems of government, constitution, and state. They all are functions of the ideology of liberation, which, like all genuine ideologies, had substantive dimensions that pointed beyond the ideological utilization itself. (1) If the problem initially was said to be one of finding a broad enough group of capable and representative leaders to form an Iraqi government, how could a relatively short formal occupation really and visibly stimulate the emergence of such a process, and to what extent could the duration of the occupation be legitimately extended in order to find such leaders? (2) If liberation (in the more demanding sense) could not rely on autonomous forces, could at least the process of making a constitution, of constituting a new regime, do so, and how long would one have to wait before that process could commence? (3) If finding a government and initiating a process of constitution making nevertheless extended the period of the occupation, could an occupation state be organized that would be able to guarantee security and essential services during this period without appearing unduly oppressive and visible to ordinary Iraqis? This third problem was vastly complicated by the American decision—a matter of choice rather than necessity—to dismantle the main pillars of the inherited Iraqi state: the army and the administration. Now the question became whether an occupation state could be used to rebuild the Iraqi state itself and whether such a task inevitably extended the period of the occupation, therefore possibly intensifying the resistance the occupation state was supposed to bring under control.

These three questions imply that the occupation regime could have been represented as liberation if it had been used in a timely and authentic manner to stimulate the emergence of autonomous forms of governmental participation, constitution making, and state building on the part of truly independent Iraqi actors. Significantly, these questions of legitimacy had to be answered both with respect to international law and UN Security Council resolutions, which had a bias against endless and unlimited occupational regimes, and to Iraqi opinion, which was even less permissive regarding foreign usurpation of sovereignty.

Below, I consider the problem of the duration of the occupation in relation to finding autonomous actors to form a government and write a constitution. Then I go on to discuss whether failures to promote autonomy could be ascribed to the absence of democratic intentions in the first place on the part of the imperial occupying power. After an excursus on historical comparisons and their lessons, mostly negative, I turn to an extended discussion of the occupation state and its failure. I end the chapter with a section on the relationship of state and constitution making.

Was the choice of formally turning the “liberation” of Iraq into an occupation, via UN SC Res. 1483 of May 22 (which formally referred to the CPA as “the Authority”), a serious error? Those who criticize the United States for not turning power over to a provisional government of exiles maintain this was the case.⁸⁰ Actually, a previous resolution (1472, of March 28) already referred to the “Occupying Power” and its duties, and 1483 went further only in explicitly placing the United States and the United Kingdom under the Hague and Geneva provisions for “belligerent occupation,” which involved important restrictions as well as entitlements (par. 4–5).⁸¹ Recognizing the CPA was merely the recognition of a fact, the fact of belligerent occupation, which before the CPA was administered by the U.S. military, according to international law. Especially given the particular generals involved, we now know that an explicitly military form of administration would probably have been smarter than Bremer’s, but there was no reason for anyone at the United Nations to suspect this at the time. The same resolution went on to stress “the right of the Iraqi people freely to determine their own political future” and the “resolve that the day when Iraqis govern themselves must come quickly.” It is a fair point that countries opposed to the war in Iraq should have held out for a statement that would have gotten a larger international role in establishing an inclusive process for building an Iraqi leadership, but in that case they may not have gotten the United States to verbally submit to the constraints of Hague and Geneva. In any case, they were much too weak and continually disorganized by the ambivalent but always pro-American position of the United Kingdom.

If, however, there is one UN Security Council resolution that should be strongly criticized, it is 1511, of October 16, 2003, which uncritically accepted the appointment of the Interim Governing Council as the organ (along with the ministries it supposedly controlled) that “embodies the sovereignty of Iraq” (par. 4) and its supposed control of the constitution-making process (par. 7).⁸² UN representatives knew very well, as I know from personal conversations, that the IGC was both unrepresentative and mere window dressing for the dictatorial powers of the CPA. By the time the resolution was written and passed, the IGC was in fact already a nonfunctioning organ, with the exception of its constitutional committee, which exercised its independence through acts of resistance to CPA plans.⁸³ At the same time, the American advisors to the ministries treated themselves as if they were (and they were in fact) the people in charge, much as their British predecessors did in the 1920s.⁸⁴ No UN Security Council resolution drafted by the Americans and passed by rather uninterested other members could gloss over the fact that the United States in establishing the IGC missed two major opportunities: (1) finding a process by which new, inclusive, internal Iraqi leadership could be generated, and (2) gradually transferring real responsibilities to an Iraqi leadership.

It is very doubtful that most Iraqis would have seen as liberation the empowerment of a group of (four to seven) exile leaders, or what was sometimes called the Iraqi Leadership Council or ILC, as the new government of Iraq upon the overthrow of Saddam.⁸⁵ The only (indirect) evidence produced for this claim is by Galbraith, who repeatedly points out that parties headed by five of the most likely candidates, the leaders of the two Shi’ite parties, SCIRI and Da’wa, the two Kurdish parties, and Ayad Allawi, were to eventually receive from a high of 90 percent (with a Sunni boycott; actually, the figure is 88 percent) to a low of 75 percent (without a boycott; actually 71 percent) of the votes in the two subsequent parliamentary elections.⁸⁶ Thus, the argument goes that they had support to begin with and it would have been politically if not electorally legitimate for them to rule Iraq in the name of the people. The argument is fallacious on its face, even assuming all the numbers to be correct—which they are not—because incumbents with patronage are in the position to gain visibility and control electoral outcomes to significant extents. Equally important is that with respect to the problem of Sunni exclusion, the appointment of such a group could have promoted the outbreak of an even more extensive civil war even earlier. But in fact we cannot determine whether being put in the position of genuine governmental power by the Americans would have helped or hurt SCIRI and Da’wa, who in the actual case rallied behind the Ayatollah al-Sistani’s subsequent challenge to the American occupiers and were certainly helped by that religious leader’s popularity. In other words, the vote total Galbraith recounts, not entirely accurately (especially because he excludes the very significant vote for the followers of Moqtadah al-Sadr in the second figure: they received 22 percent of the Shi’ite seats, four or five more than Allawi’s list), was in some important measure a function of a perceived oppositional status and not necessarily their support at the initial time nor what their support would have remained had they been appointed as the interim government. We certainly do not know how the Sadrist groupings, already strong in the second election, would have performed electorally had they been able to campaign against a government and Shi’ite rivals put into power by the Americans from the beginning. Finally, if Ahmad Chalabi had been the one leading such an interim government, as was first planned, it is worth noting that he was never successful in later political competition. Thus there is no reason to assume that he would have had any initial support (and there is a lot of evidence to the contrary), and we do not know how his lack of popularity would have affected his partners in an imposed government of exiles.⁸⁷

Thus the Americans and Bremer were probably smart not to simply put in power an exile government of their supporters and in overruling Jay Garner’s late effort (which may have been primarily a result of Pentagon manipulation and pressure) in that direction. But there were three other choices they should have made, but did not, that could have influenced the nature of their occupation regime. The first would have been to truly internationalize it. Evidently, the idea that an American occupation was self-interested was very strong in Iraq, and internationalization could have counteracted this.⁸⁸ The United Nations had a far better ability to negotiate with a wide variety of Iraqi actors. And there was a lot of local and cultural expertise available that the United States lacked but should have been able to draw on, and it could have been made available only with a genuine sharing of the control and responsibility. As it was, the Secretary-General’s special representative could only make a small contribution to the political process before his tragic death. It is also true that the United Nations had a negative reputation in some Iraq quarters, and this may have led to early conflicts. But given that (1) the United Nations did not invade Iraq, (2) it had no long-range plans for its domination, and (3) it had the ability to bring to the scene an international group of representatives that would by its composition reassure all Iraqi sides made the international body or its agencies far more preferable than an American representative to manage the postconflict political process.

Second, local elections and provincial elections could have been extensively used from the beginning to generate autonomous leaders, as both Garner and some American military commanders wished. It was amazing that an occupation authority seeking new, internal leaders would not allow such a process to go forward and would quash the results where they already had taken place.⁸⁹ And third, when it came to the staffing of the Interim Governing Council advising the CPA, where, after all, the cost of including people with little political experience would have been small, amazingly enough the body ended up representing still an expanded version of the exile grouping that the Americans vetoed as an interim government. The same leaders they did not allow to govern were in effect given the patronage over the formation of future governing structures (and negatively, as well, as in their control over exclusion through de-Baathification). Thus instead of instituting a gradually more inclusive political process, the CPA did the reverse. And there was little embarrassment or even consciousness of the contradictions involved. When asked to include more radical Shi’ites or Sunni Arab nationalists, Bremer’s favorite line became, preposterously enough, “it is a fundamental principle of democratic government that people do not shoot their way to power,” forgetting that he himself had done exactly that, and so had all those he allowed to sit on the Interim Governing Council.⁹⁰

Given the exclusionary nature of the IGC, would it nevertheless have been better to transfer real leadership functions to it instead of using it as window dressing? From the point of view of diminishing the neocolonial aspects of the occupation, it certainly would have been better. But in the few areas where members of the IGC took charge, disasters were not avoided—quite the contrary. For example, de-Baathification policy, in the hands of Ahmad Chalabi, turned out to be an even more destructive purge of Iraqi professional life than it would have been had Paul Bremer managed it. There is little question of the contribution made by this policy to the polarization of Iraqi society on ethnic grounds and almost certainly to fueling the Sunni-based insurrection. The fundamental problem thus was in putting together a protogovernmental organ, with or without powers, something that would have been avoided had either the United Nations been in charge or had the process proceeded from the ground up, on a power-sharing basis.

In one respect the IGC did have power, namely regarding appointments, and this power was preserved until the so-called transfer of sovereignty⁹¹ and its own dissolution. In the end, the United States was not entirely able to avoid devolving some power to a legitimate instance, the United Nations, in generating Iraqi leaders to lead transitional governments. Sergio de Mello had a minor role in helping to make the powerless IGC more representative. But he failed to secure any representation for secular Sunni or Arab nationalist politicians. When it came to picking an interim government to which sovereignty would be transferred, in the face of the Grand Ayatollah Sistani’s objections, the United States was forced to formally defer to UN Special Representative L. Brahimi, who initially wanted to bypass the IGC altogether. We can now say that Brahimi failed to control this process, and while the United States (Ron Blackwill as much as Bremer) got the very candidates it wanted to run the interim government, the IGC got most of its important members in key positions as well. With the Chalabi-led de-Baathification process playing a major role, very few new leaders were generated in a process⁹² that preserved at most a paper-thin international legitimacy in the light of Brahimi’s own comments.⁹³ In any case, the choice of A. Allawi as prime minister suggested perhaps a future strategy of greater reconciliation among Iraqi groups than the one pursued by Bremer, and initially many Iraqis may have taken the restoration of sovereignty at face value, as did the legalistic UN Security Council. In any case, the interim government was to stay in power for only six months, and everything depended on the rules according to which elections and the process of the constitutional assembly were going to operate.

Politically, establishing a process of legitimate constitution making was less visibly important than gradually empowering an internally based, inclusive Iraqi governing structure. Theoretically, the reverse relationship holds, because a constitution was supposed to regulate the whole postoccupation period. International law seems to reflect this theoretical emphasis. While it contains few rules, largely indirect, for an occupying power’s organization of (generally military) executive powers, the Hague and Geneva Conventions are indeed concerned with legislation under occupation that can have long-term effects on the destiny of a population. Article 43 of the 1907 Hague Convention puts the matter thus: “the authority of the legitimate power having actually passed into the hands of the occupant, the latter shall take all steps in his power to re-establish and insure, as far as possible, public order and safety [i.e., civil life], while respecting, unless absolutely prevented, the laws in force in the country.”⁹⁴ This rule, subsequently confirmed by the Fourth Geneva Convention (art. 54–56, 64) of 1949, has been interpreted both in terms of an almost absolute prohibition⁹⁵ and as highly, indeed absurdly, permissively,⁹⁶ but it is almost impossible to reconcile it with a project of regime change⁹⁷ or transformative occupation⁹⁸ carried on by the occupying power, when the very purpose of occupatio bellica was that the occupation should be temporary and should allow the return of a territory to its own sovereign authority.

It is another matter that subsequent occupations violated these rules more often than they obeyed them, and that given contemporary standards of human rights and in relationship to some authoritarian regimes whose territory is occupied we are dealing with a hopeless case of underregulation.⁹⁹ In the case of Nazi Germany, the problem was solved by the revival by Kelsen and others of the old category of debellatio (subjugation), which made belligerent occupation irrelevant when a state was completely destroyed and when a territory could not be returned to its original sovereignty.¹⁰⁰ In such a case, the occupying power has full sovereignty, including all legislative and constituent powers, Kelsen argued, though the makers of the Grundgesetz disagreed.¹⁰¹ On first appearance, Iraq seems to fit Kelsen’s case of debellatio, since its government and subsequently its state organs were completely destroyed, and in any case it would have been a better argument for the U.S. assumption of constituent powers than the one made by John Yoo, except for two factors. First, as Jean Cohen has recently showed, since Geneva IV, in conjunction with the UN Charter and its right of self-determination and reinforced by the 1970 Declaration of Friendly Relations, has transferred its concept of protected sovereignty under occupations from government to people, debellatio has become an irrelevant concept as long as a population within a territory is still capable of self-determination.¹⁰² And second, as already mentioned, the United States and United Kingdom have willingly placed themselves under the law of belligerent occupation, with its restrictions (UN SC Res. 1483, par. 4–5).

The voluntary submission to Hague and Geneva can be interpreted in two ways, one intended and one logically inevitable, even if not intentional. Clearly by March 2003, U.S. policymakers realized that they were not going to immediately transfer power to an Iraqi provisional government—that there was not going to be liberation in the more demanding sense, in other words—and that they needed international legitimacy for their occupation. This was especially important because the 1970 Declaration of Friendly Relations by the UN General Assembly implied that occupations resulting from illegal wars, that is, wars neither of self-defense nor ordered by the UN Security Council according to chapter VII, were themselves illegal.¹⁰³ Of course, the United States and United Kingdom held that their war was legal according to a rather contorted argument,¹⁰⁴ but this view was a distinctly minority one in the international political and legal communities. Legalizing the occupation thus was necessary in its own right and could even be considered as a post facto confirmation of the legality of the war itself,¹⁰⁵ unless the Security Council’s action was to be construed as violating the Declaration of Friendly Relations.

At the same time, the unintended consequence of the action was to put the United States under the new post–Geneva IV interpretation of the law of occupation stressing popular rather than state or governmental sovereignty. After all, in Kelsen’s terms, with respect to governmental sovereignty, Iraq according to the Hague Convention would have been a case of debellatio, as there was no absent governmental sovereign whose rights could be protected.¹⁰⁶ The law of occupation is meaningful in Iraq only if we understand the sovereign whose rights survive even though suspended as the popular rather than the governmental sovereign. This means that the United States now legally committed itself to a form of occupation that would be compatible with liberation in the less demanding sense at least, and specifically to a process in which, regardless of who controls executive powers in the beginning, “the constituent power” (and in general most legislative powers with long-term effects) would be exercised by the Iraqi people themselves. Or, alternately, the United States allowed a resolution to pass that would legalize its occupation (and perhaps put to rest the question of the war’s illegality as well), but at the cost of pointing to usurpation if American authorities themselves established a sovereign rather than commissarial dictatorship or a transformative rather than belligerent occupation.¹⁰⁷

Paragraph 4 of 1483 “calls upon the Authority [the occupying powers] … to promote the welfare of the Iraqi people through … the creation of conditions in which the Iraqi people can freely determine their own political future,” and this was immediately interpreted by some as implying the creation of a constitution that would be a framework for elections.¹⁰⁸ The same lines, however, could also be interpreted as referring to the creation of conditions under which the Iraqi people could “freely determine” their future constitution, and the phrase “consistent with the Charter of the UN and other relevant international law” in the same paragraph and the requirement to “comply fully” with Geneva 1949 and Hague 1907 imply this latter reading.¹⁰⁹ Indeed, a way of making the two regulations consistent with problems of occupation of countries with previous regimes whose legality cannot be left untouched is again to focus on the stress on popular sovereignty in Geneva.¹¹⁰ In that case, the normatively valid core of the law of occupation remains the idea that if there is to be the creation of a new legal and constitutional system in a country under foreign occupation, the decisions concerning how such a system is to be created belong to agents that can legitimately speak in the name of the country’s own population and not to the occupying power or its agents. This idea converges with the common sense of democratic theory. We assume that democracy is a system that emerges from the autonomous activity of its (potential) citizens and is never, or rarely, the gift of political elites, who would seek to preserve undemocratic advantages hidden in new institutions.¹¹¹ The theory of democratic legitimacy can further clarify the matter. Any set of modern democratic institutions have distributional consequences and imply the rule of political elites through mechanisms of representation. We therefore assume that strong legitimation requirements have to be satisfied during the beginning stages of a new democracy, requirements that do not require the mythological attribution of purely democratic constituent power as the source of the new regime but that must go well beyond the idea that elites (especially foreign ones) have imposed it because they had the power to do so.¹¹²

Properly interpreted, then, UN SC Res. 1483 throws light on the two conditions under which an occupation regime can remain “liberation” in the less demanding sense: speed and autonomy. UN SC Res. 1483 insists on a speedy termination of the occupation and on the planting of the seeds of an autonomous Iraqi administration during it (par. 9). The separation between liberation and constitution in the concept of revolution allows the posing of the same two conditions: the temporal space between liberation and constitution making should be reduced, with the latter process organized autonomously and with the occupying power withdrawing, to the benefit of an indigenous process of self-determination. However, the issue is related to the previous one of forming an interim government. Arguably, unless a relatively inclusive Iraq government were formed, independent of the occupation authority and before the beginning of the constitution-making process, the latter could not be represented as a fundamental break with the logic of the occupation. Such a government, it was increasingly felt, could only come from elections, but elections were not possible without a new constitution.¹¹³ Thus the American authorities had to choose between two evils, even if they regarded them as such only on the public relations level: (1) the option that a constitution, despite international law, would be imposed by them, and (2) that they would empower what they regarded as a weak and unrepresentative interim governmental organ, the IGC, to be in charge of the process. As usual, they chose both options, but the second was meant to be only the Iraqi veil for the first: American imposition. That veil was greeted by UN SC Res. 1511 as “welcoming the decision of the Interim Governing Council of Iraq to form a preparatory constitutional committee to prepare for a constitutional conference that will draft a constitution … and urging it to complete the process quickly.” I believe it was 1511 rather than 1483 that helped to legitimate the process of imposed constitution making in Iraq—but not, as we will see, in the Iraqi eyes, which really counted. It is a great paradox of the recent constitutional history of Iraq that it was autonomous political and social action on behalf of popular (or populist) democratic constitution making that was to put an end to the open constitutional usurpation of the CPA, putting the process in a channel where it was arguably compatible with creatively interpreted international regulations.¹¹⁴ Yet once again it was to become a question whether the formal process was not meant as window dressing for a deeper, exclusionary, and imposed one after all.¹¹⁵ This question was superseded by a probably more important one: whether constitution making could work at all in the context of state failure.

Arguably, the attempt to make occupation and liberation compatible at all was not succeeding on the level of having a sufficiently short occupation regime that would allow the emergence of related autonomous and legitimate processes of government formation and constitution making. There was a strong suspicion among Iraqis, even after the formal transfer of sovereignty, that neither process would be autonomous. In short, the American invasion was suspected of being an imperialist enterprise rather than, as it is treated here, an attempt to externally impose democracy. Cockburn for example, as already mentioned, not only repeatedly notes Iraqi perceptions that the American occupation was an imperialist one, but he himself is inclined to treat the enterprise and its choice of governing structures as classically colonialist. Which interpretation is right? Are they incompatible? Is their possible relationship a clue to why occupation could not be successfully represented as liberation, even in its less demanding sense?

Even if imperialist motivations were made part of the overall explanation of why Iraq happened the way it did, they can explain neither that an outright American military intervention occurred under such unpromising circumstances nor the actual course of the invasion, occupation, and especially the disastrous choices the occupiers were to make.¹¹⁶ Imperialism would have been compatible with a realistic attempt to contain Iraq and control the Persian Gulf (as Scowcroft argued in effect)¹¹⁷ or, after an invasion, with the establishment of indirect forms of rule based on the existing Iraqi state, a trajectory the U.S. State Department experts had long prepared for. Only within the larger framework of an externally imposed revolution, imagined, as I will argue below in a particularly radical manner, would imperialism take the form that it did, even if nonimperialists (Cockburn) or nonrevolutionaries (Scowcroft) could quickly foresee where it would necessarily lead. Revolutions, however, need ideologies or ideological motivations. Since, especially in our age, this ideology could not be an imperialist one, “democracy” (flying high in the period of the 1990s), authentically or not, entered the breach.¹¹⁸

During my many classes and lectures on this and related topics, students and others ask whether the idea of democratizing Iraq was ever seriously intended by the United States or the Bush administration and whether or not the real motives had to do with long-term imperial control of the Persian Gulf region and its resources. On one level, such a question involves pure metaphysics, postulating a collective actor, the United States, its administration, or, worse, U.S. Imperialism or U.S. Capital, actors that do not exist.¹¹⁹ Such actors are the vector sums of complex decisions made by many individual and group actors, and we rarely have access to all of their internal disagreements, conflicts, and decision-making processes, despite the growing literature on the subject. What can tentatively be said is that the decision to invade Iraq was the sum of several conflicting perspectives, of which the economic interest expressed in the imperialism thesis¹²⁰ was certainly one, and of which a democratization lobby that imagined a transformation of the Middle East in our image was probably another. We tend to personalize the first with the names of Cheney and Rumsfeld, the second with the neoconservative ideologists. It is now known that hard realists such as Henry Kissinger were ready to join them using a version of the deterrence argument, because the United States after being attacked by “radical Islam” had to reply on a significantly large scale, and Afghanistan was not enough.¹²¹ There is also the considerable likelihood that people concerned with the president’s electoral fortunes and internal powers were ready to use an external political adventure to enhance these. Certainly all subsequent moves in Iraq through the 2004 election track very well (as Iraqi actors continually noted) with the electoral needs of the president. Finally, we should not neglect the desire for absolute security, which was so well articulated in the “one percent” doctrine David Suskind has reconstructed for us, though given the absurdity of this idea¹²² it is hard to say whether for a Cheney or a Rumsfeld this argument was motive or merely rationalization.¹²³

Initially then, the role of democratic regime change was probably just one—and probably not the most important—of the motivations that brought a diverse group of actors together to support the war. Two factors have changed that. One was the failure of the older realism in its “two-pillars” policy to manage American influence in the Persian Gulf. Paradoxically, this became clearer as the older realists around the first President Bush, who were certainly concerned with our power position in the oil-rich region, and their academic supporters came to strongly oppose the plans for the war.¹²⁴ Their opponents could now charge that it was the earlier realism, based on the Nixon doctrine and its proxies, that established Saddam as the second pillar of American policy, which led to the invasion of Kuwait, gravely threatening the much more important first pillar, Saudi Arabia, the center of the region.¹²⁵ Dictators were unstable, unreliable, and dangerous.¹²⁶ But while we could opt for direct intervention instead of using surrogates, we could not adopt a policy of long-term direct or at least obtrusive presence rejected by the Nixon doctrine, as it would be openly imperialist, would not be supported by the American electorate, and would have required a standing military force the United States no longer possessed. It was within this intellectual context that the neoconservative argument for democratic regime change was adopted by other factions in the administration, not necessarily because they were fully convinced but because they had no other alternative if direct presence and dictatorial proxies were both rejected. It could be that the idea always was to leave sufficient force behind to make sure that the new democracies behaved. Or, on the basis of bad historical analogies born of the cold war, it was perhaps assumed that countries owing their democratization to the United States would remain out of gratefulness on the American side. Democratic governments overthrown by the United States itself were conveniently forgotten, or it was assumed that a process of democratization controlled from the outset and monitored later on could avoid such unfavorable outcomes as Mossadegh’s or Allende’s governments. It was also certainly assumed that managing the democratic transition and making financial contributions to the right parties could significantly contribute to the outcome of producing democracies that turned out right. The historical problems with actual democracies could have been explained away by reference to undesirable worlds where the United States was not the sole superpower and where small countries could gain significant support elsewhere.

The second factor was the total, dramatic, and utterly embarrassing failure of the WMD justification for the war—the primary justification.¹²⁷ For a considerable period afterward, democratic regime change was the only justification that remained standing, and thus it had to be pursued much more seriously than some of the architects of the war may have initially intended. It took the state failure (described below) as the justification to shift to defining Iraq as being the primary front in the War on Terror, the location where Al Qaeda had to be defeated. That justification was also present from the beginning, but it was based on even more elaborate disinformation and perversion of intelligence than the WMD hysteria. For a time, this justification could not be articulated without calling attention to the fact that it was the war itself that brought foreign terrorists to Iraq, and that it was the destruction of the state that established such a powerful base for them—if indeed their position is as strong as the propaganda would have it. For a considerable period, then, democratization and democratic regime change, with all its paraphernalia of elections, constitution making, referenda, governments of national unity, and so on, had to be taken utterly seriously.

At the same time, the defeated realists¹²⁸ were quick to remind the winners of the debate, the warmakers in the name of democracy, that not just any democracy would do in Iraq. The United States could not have expended all the “blood and treasure” it had in Iraq to merely deliver the country horribile dictu to the friends of Iran, the Shi’ites, who just happened to be the country’s demographic majority. You have to have a democracy that turns out right, one that puts and keeps in power friends of the United States. It is a mistake to rush into elections before that result can be ascertained. On the most sophisticated level, this argument logically meant that the United States should not have invaded Iraq at all in the name of regime change, just as Bush, Scowcroft, and Baker did not go to Baghdad in 1991. This issue, however, was now irrelevant. Slightly more practically, Scowcroft implied that the U.S. government would be still better off with some kind of authoritarian regime where one would have to manage only a small group and not whole electorates or political parties. This advice was impractical, ran counter to the spirit of the times, and was not in fact openly offered. What Scowcroft did suggest was to delay the process and to extend the occupation, advice that was still probably very impractical, because it could produce a nationalist backlash against the Americans, who would appear as successors to the British imperialists, perhaps resembling the events of 1920. There was indeed not only a Sunni insurrection to worry about but the possibility of one based in Shi’ite strata as well. At all costs, fighting such a two-front war had to be avoided. Thus the interpretation of the advice ultimately reduced to what the CPA was actually trying to do, namely to work toward a relatively quick transition to “democracy” while retaining sufficient control to ensure that this new regime turned out right from the American point of view.¹²⁹ Such a democracy, with its implied postulate of certainty, is not really democracy;¹³⁰ it is merely an empire’s democracy¹³¹ or a new Bonapartism.¹³² It is quite another question whether such a system can be stabilized after an aggressive war, in a country with multiple lines of political cleavage. In Iraq, an empire’s democracy continually hovered between open imposition and outright failure of the whole governing process. In retrospect, it is tough to believe that anyone could have believed that such a democracy was possible under the given conditions. Misunderstood historical analogies certainly helped to spread the misconception.

From the very beginning, those in charge of the American invasion and occupation of Iraq were willing to take seriously only the German and Japanese historical examples, which were, however, very poorly understood. James Dobbins notes four reasons for this highly anachronistic interest: the success of these efforts, their very large scale, the absence of salient controversies regarding them, and the fact that the Department of Defense rather than the State Department or USAID was in charge, as opposed to more recent and controversial efforts.¹³³ From a more theoretical perspective, the historical precedents of Germany and Japan also offered the promise that the legitimation problems of an externally imposed revolution were in principle soluble. What was most certainly not noticed was that these very problems were handled in significantly different ways in the two countries considered as models. External imposition can indeed be seen as two very different things: (1) imposing a fully democratic solution from the outside, or (2) merely removing the impediments to genuinely internal solutions. According to Carl J. Friedrich, the process of the democratization of Germany should be understood as the second of these types: the restoration of democracy.¹³⁴ In the words of the best American historian of this process: “German political life reached ‘point zero’ in May 1945. But there were still latent political traditions of pre-Hitler Germany on which a reconstruction of the body politic could fall back. There were also scores of political and intellectual leaders and thousands of faithful followers who had somehow weathered the totalitarian storm and were rallying now to rebuild society and state.”¹³⁵

Both parts of this statement are important. Unlike the Japanese state, the German state collapsed both as a territorial and as an organizational entity. As we have seen, Hans Kelsen, followed by other international lawyers, applied the term debellatio to Germany, implying the total absence of sovereignty and therefore the inapplicability of the Hague rules for belligerent occupation. It is also true that the reconstruction of democracy was on the whole an autonomous German process. For both reasons, therefore, it may be more fitting to choose the Japanese example as a lesson for Iraq, because here the work of imposition was much more drastic and democracy more obviously created than merely restored.¹³⁶ Note, however, that in Japan radical rupture was also avoided to a far greater extent than in Germany. The Americans made a determined attempt not to smash the internal capacity of the Japanese state. The Japanese cabinet and Diet were not removed but instead had to follow the directives of the American military rather than the directives of the dominant Japanese military groups, as they did in the years 1930 through 1945.¹³⁷ Most importantly, General MacArthur was strongly sympathetic to the Japanese government’s desire to allow the emperor, representing the unity of the state, to remain in place. The retention of the emperor was key to maintaining the coherence and loyalty of the police and the bureaucracy. Finally, even in the midst of imposition, the important fiction of legal continuity was maintained: the new constitution of 1947 was formally passed as a mere amendment to the 1889 Meiji constitution.¹³⁸

The problem of legitimation in Germany, where the state was completely destroyed, was handled through the restoration of democracy through largely indigenous processes; in Japan, where democratization was imposed, the solution relied on the preservation of the inherited state along with a species of traditional legitimacy. This difference should not hide the fact that the initial conditions of political transformation in Japan and Germany in 1945 were far more similar to each other than either was to contemporary Iraq.¹³⁹

(1) Germany and Japan were under the externally overthrown dictatorship for twelve and fifteen years, respectively; Iraq had been a dictatorship for forty years or more.

(2) Both Germany and Japan were completely defeated in the greatest war in history, with societies capable of little organized resistance; in Iraq, only the government was defeated.

(3) Germany and Japan were constitutional regimes before the 1930s; what occurred after 1945 was first and foremost a restoration of those constitutional regimes. Iraq has never known constitutional, rule-of-law government.¹⁴⁰ The skills, traditions, memories, institutions, and, most of all, the professionals to build a constitutional state were available in both Germany and Japan but not in Iraq.

(4) Germany and Japan are ethnically homogeneous. Japan is also religiously homogenous, and the violent Catholic-Protestant conflicts in Germany ended in the seventeenth century. The ethnic and religious situation in Iraq is obviously different, to say the very least. Here the ascriptive cleavage structure is three-dimensional and highly antagonistic (Arab-Kurd, Shi’ite-Sunni, secular-religious—with six of the eight possible combinations in existence, even if not always salient).

(5) Germany and Japan were liberated or occupied as the result of the aggression of the latter two countries; Iraq was liberated or occupied in a war in which the United States was, like Napoleon’s France two hundred years before, the aggressor.¹⁴¹

(6) German and Japanese elites and the Western, mainly American agents of democratization had a common enemy, whose support was relatively weak within their countries: the Soviet Union. Iraqi elites, in common with other Arab/Muslim leaders and opinion makers, have a major enemy that is a friend of the United States: Israel.

(7) The external boundaries of Japan were relatively secure; those of Germany could be threatened only by a new world war. The boundaries of Iraq, on the other hand, are porous and threatened by a plurality of states that have important allies within Iraq. Iraqi democracy is a threat to all of these states, some of them allies of the United States and therefore difficult to restrain by threats of force.

Looking at the matter from the point of view of international politics indicates some differences between Germany and Japan, which, however, further indicate their common distance from the situation of Iraq. The political transitions of Japan and Germany, because of the more complex occupation structure of the latter, took place in different international constellations, as Japanese writers were the first to point out.¹⁴² Japanese constitution making occurred at a time when the United States was unchallenged in the Pacific, before the victory of the Chinese Communists, and before the outbreak of the cold war. The Japanese were not (yet, or mainly) thought of as potential allies, a fact explaining their relatively rough treatment compared to Germany. In (West) Germany, the making of the Grundgesetz was an intrinsic part of the cold war division (of the country and Europe), and it was clear that the new Federal Republic would have to be an important ally on the very frontier between West and East. It was, moreover, important to demonstrate the superiority of the American-supported autonomous democracy to the Soviet-imposed pseudodemocracy next door. Given this train of thought, we can see that in an epoch of unbalanced American dominance, Iraq is seen as its terrain rather than its partner. There is no other sociopolitical model around that would require the demonstration of the superiority of our model. Moreover, the outcome of a genuinely democratic process in Iraq is likely to be very different from the point of view of American interests than in Germany in 1948 or Japan in 1946 and 1947. Given the political forces in Iraq and their international alliances, the United States is, from the outset, more ambivalent about the Iraqi democracy it supposedly promotes than about the two earlier examples.

Thus, when compared to Iraq, the German and Japanese models lead to a strongly negative balance sheet consisting of two major components. First, the given internal and external conditions of the three countries during the beginning of their occupations were so different as to discourage almost any attempt to apply German and Japanese lessons to Iraq. The earlier cases do not justify at all the optimistic scenarios concerning the later one. But, second, if we insist on learning something nevertheless—and after the invasion there was little choice in the matter—the lesson should have been simple enough. The only imaginable basis for initiating an autonomous political process in Iraq was either ensuring the continuity of the central state, as in Japan, or abetting the rapid and thoroughgoing organization of elected local government, as in Germany. Because of the far less favorable starting point in Iraq, most likely a combination of these desiderata would have been important if at all possible.¹⁴³ Adhering to them would most likely not have been sufficient, but disregarding either and especially both would make the chances of disaster very high.

Since the German and Japanese “gold standard” (as Galbraith puts it) for occupation had only limited value for Iraq, it is not surprising that analysts supporting the war have sought other examples. In a study performed by the Rand Corporation¹⁴⁴ led by Dobbins, which made the rounds of the inner circles of government,¹⁴⁵ in addition to Germany and Japan, the cases of Somalia, Haiti, Bosnia, Kosovo, and Afghanistan were added. Subsequently, Dobbins was to argue much more strongly for the irrelevance of the German and Japanese cases, which his group still took seriously, and for the salience of the ex-Yugoslav cases in particular, and the conclusions of the original study support this focus.¹⁴⁶ The recent cases here all belong to an epoch in which the United States as sole superpower is in principle capable of exerting its will in the world on an entirely new basis. But as the failed cases (Somalia, Haiti) testify, such power in itself does not lead to the desired result. Undiscouraged, neoconservatives or “democratic realists” have argued that failure is inevitable unless vital national interests and political principle both justify the necessarily very high expenditure of the required resources.¹⁴⁷ Conversely, the Dobbins group seems to demonstrate that where the expenditure of resources (military and financial) is sufficiently high, the task can be accomplished even where the initial conditions are much less favorable than in Germany and Japan, where the society is much less homogenous, and where it is capable of further military resistance (for example, Bosnia and Kosovo, though Iraq is depicted as a still more difficult case than these two). Admittedly, in Bosnia and Kosovo the resources came in large part from Europe, and, based on these experiences and also on the Japan/Germany comparison, Dobbins’s group favored international participation in shouldering the costs and in helping to compose the command structure. That would mean in Iraq that the United States, not willing to share the command, would alone have to shoulder the burdens. Based on Bosnian and Kosovo comparisons, the Rand report estimated for Iraq a minimum occupation force of 460,000 to 494,000 in 2003 and 258,000 to 526,000 in 2005 and an expense of thirty-six billion dollars over the first two years. (It is in this context that the report recommended the gradual dissolution of the established Iraqi army and the creation of an entirely new one.) But assuming the geopolitical stakes, it was assumed nevertheless that the costs could be managed and paid. Disregarding recommendations like this, at least with respect to the size of the U.S. military force (the funds committed were very high, though not properly utilized), could only be described as an effort of calculated inexperience or a calculated attempt to learn not to learn.¹⁴⁸

The assumption behind the Rand report’s conclusions is the same as the premise of Charles Krauthammer’s “democratic realism,” even if they do not marshal up questionable examples such as the long dictatorship of South Korea to support it: “the single most important factor in the success of nation building is seriousness.”¹⁴⁹ Everything thus depends on us; the other is a passive object for our experiment. The more difficult the case, the more we must do in terms of the commitment of resources. But what happens if our own actions make the problems worse, not better? More reflective conservatives tend to recall what happens when dictatorships dissolve in multiethnic and multireligious societies held together by authoritarian states. According to James Kurth, liberating a segment like Kosovo or Bosnia was thus less indicative of what was likely to happen in Iraq than was the dissolution of Yugoslavia as a whole or the breakup of the Soviet Union.¹⁵⁰

As suggestive as the violence of some of Kurth’s examples may be, given what is now happening in Iraq, it is not entirely obvious that they had to inevitably apply to this case. And there are two reasons for this. First, it is not axiomatic that the end of a dictatorial regime must mean the end of a state in a multiethnic but not federally organized society. All his examples were federal as well as multiethnic. While it is true that the Shi’ites have experienced Saddam’s rule as very oppressive, interpreters tend to forget the historical importance of Iraqi nationalism as opposed to Arab nationalism and the historical desire of the Shi’ites to participate in an Iraqi state rather than to destroy it.¹⁵¹ And second, it is not certain that an occupation state could not have been temporarily grafted onto the inherited state, preparing the ground for a new process of state rebuilding (whose burden would have to be carried by the constitution-making process).

For an occupation to be represented as liberation, it would have to above all function effectively; paradoxically, it would have to be an effective occupation, one providing public security and basic social services on a continuous basis. To most people, this was more important than even the questions of the occupation’s duration and whether that time was being used effectively to transfer governmental power and organize an autonomous constitutional process.¹⁵² And, clearly, a longer but more effective occupation regime would allow in principle at least a more lasting transference of power and the building of stable institutions than a shorter but disorganized occupation. It is undeniable that this point was clear to the founders of the CPA,¹⁵³ but they did not understand the important corollary: an effective occupation regime would have to rely on an effective occupation state. As we have seen already, historically this problem was solved in two different ways. Here briefly: an effective occupation state could be grafted onto the inherited state (Japan) or it could be a function of its own effective forces of coercion, administration, and economic support (Germany, Bosnia, and Kosovo), the latter requiring the presence of much larger occupation forces per capita.¹⁵⁴ The most remarkable thing about the occupation of Iraq was that it followed none of the earlier patterns: unwilling to use a large enough American force and unable to mobilize a large force of allies to join the enterprise, the U.S. occupation authority proceeded both to disband the Iraqi forces that could have been used to provide security and to eliminate or reorganize the inherited institutions on which the occupation state could have been grafted. State destruction without the hope of creating an occupation state to replace it: that was the formula that made the failure of almost everything else highly likely, though I will argue not entirely inevitable.

Here some definitions are in order. For the purposes of this chapter, I will consider the state as the Weberian one: “a compulsory political organization with continuous operations [within a given territorial area, whose] administrative staff successfully upholds the claim to the monopoly of the legitimate use of physical force in the enforcement of its order.”¹⁵⁵ When older German and French writers analyze the category in terms of population, territory, and a form of organization or “political organ” capable of producing a public power exercising coercion over them,¹⁵⁶ they are rightly adding only the element of membership, admittedly with the risk that “state” and “nation” may be wrongly conflated.¹⁵⁷ Their subtraction of legitimacy from the definition is only apparent, because it is contained in the idea of a public power. Nevertheless, I will leave open here whether a plausible claim of monopoly produces legitimacy or publicness or presupposes it.¹⁵⁸ Admittedly, other important features could be added for the modern state, such as its link to an order of positive law and legislation¹⁵⁹ and its role as an organ of communication in reflexively monitoring society.¹⁶⁰ Even if we distinguish between nation and state, as we should, it may be important to stress that a state’s people are themselves socially (and not just coercively) integrated in a variety of ways, and this entails important state symbols and rituals at the state’s end (communicated through a variety of socializing agencies) and entitlements on the members’ end (administered through a variety of administrative structures).¹⁶¹ It is, however, a mistake in my view to reduce the state either to the sum total of bureaucratic offices or to the territorial entity (or, for that matter, the people or the nation). The state must be seen as the operation of ultimately coercive political and bureaucratic structures on people (distinguishing between a people with some kind of collective “state identity” and all people, members and nonmembers both) within a territory.¹⁶² Here, in the given context, I will concentrate on these three dimensions: coercive organization with a (legitimate) monopoly, territory, and people, the last in the sense of members.

In distinction to the concept of the state, I will understand government here in the American manner, to refer to all the functional branches of political power, not only the executive.¹⁶³ From the point of view of international law, government is the body of individuals that “by virtue of the effective constitution of the state, represents the State in relation to other States.”¹⁶⁴ Government is the political center or political top of the administrative and coercive apparatus of the state; when differentiated, all three branches of governmental power have a relation of formal superiority to that apparatus. From the point of view of a formal action theory, government is the source of goals and objectives; administration is the instrument that carries these out. The reality, as organization theory has repeatedly shown, is far more complex, however: goals emanate from below as well as from above.¹⁶⁵ Regime is the institutional structure of state and government, including the rules and practices that govern the relations of governors and governed.¹⁶⁶ It is best understood in relation to the term constitution. Until the late eighteenth century, “constitution” meant the empirical structure of state and government, and, as in Aristotle’s Politics, all states had constitutions in this sense. In the late eighteenth century, primarily through the American and French Revolutions, the term “constitution” was transformed to mean the legal rules regulating the establishment and practice of political government.¹⁶⁷ Not only did the term now mean secondary legal rules, namely rules of rule making, identifying, adjudicating, and applying,¹⁶⁸ but when differentiated, there was also the implication that constitutional rules would be made in a special procedure, included in a single document—a written constitution—and protected from amendment by stringent rules.¹⁶⁹ However, as Dieter Grimm rightly notes, the old notion of constitution in the empirical sense survives, and, together with the legal practices, customs, and conventions indispensable for any actual constitutional tradition, these can be understood as a kind of “constitutional reality” that still influences the law.¹⁷⁰ The actual political order is thus governed by a combination of institutionalized legal, unwritten legal, and empirical regularities, and it is this combination we should refer to by the traditional term regime. This term denotes little more than Aristotle’s politeia, that is, one of the meanings of “constitution,” as long as we realize that today a regime will be a complex combination of formal and informal, normative and empirical elements, all however requiring some level of institutionalization.

The three terms “government,” “regime,” and “state” imply three categories of political change that are, more or less, in the order of their radicalism: coups d’état, revolution, and (what I will call) revolutionary state destruction, one of whose forms is a not necessarily very radical revolution-secession.¹⁷¹ Little needs to be said about the first. It is a form of illegal change in which merely the government is replaced, but the old rules of the political game remain in place despite a one-time violation of the key rule according to which the incumbents of government are supposed to be replaced. It is wrong to consider, as does Kelsen, such an event the equivalent of a revolution or even the demise of an old constitution (which would be discredited, though very possibly preserved). While in a very narrow technical sense one could argue that during a coup, with temporarily two contending armed forces, a state in the sense of a legal monopoly over the means of violence has ceased to exist, this is neither true in international law nor in terms of the important territorial, popular, or even administrative components of the state. Revolution is more complex. Generally it entails not only a replacement of the incumbents of government, but also a change in the form of government. Thus it contains a coup d’état, and sometimes it can look like a coup or begin with one. But more importantly, by attacking the regime as well as the government, revolution arguably smashes the old state structure as well. Such was the view and desire of Lenin before Russian conditions taught him the Tocquevillian insight of the partial survival of the old state and its bureaucracy even in radical revolutions.¹⁷² Because regime and constitution regulate both state and government, it is difficult to see how a regime change can leave a state more or less intact. At the very least, it will imply a reform of the state structure and organization. In great revolutions, there have been certainly important tendencies toward the disintegration of one or another dimension of the inherited state apparatus: in France, the bureaucracy though not the army; in Russia, both until they were reconstituted from largely old elements.¹⁷³ Theoretically, however, three distinctions can be made. Even a radical revolution can use important segments of the old state organization, such as the army and the police, to build its own state. Here the objection would be that a radical revolution needs to reorganize the mode of operation of even these organs of power. Such an effort can be relatively complete when a revolutionary force has a very large, trained military force and incipient political apparatus at its disposal, as did the Chinese Communists. The counterargument is on the whole more plausible empirically: radical revolutions in particular immediately need state power to suppress class-based forces of the old regime and to impose their very likely elite minoritarian conceptions about the reorganization of society. Generally, they may have the forces to take the commanding heights of power but not to administer, control, and defend the country as a whole. Second, regime transformation is indeed likely to weaken a polity with respect to its external environment, and thus the mobilization of inherited state structures on behalf of the external defense of the revolution is likely to be necessary. Even a radical revolution would wish to preserve the territorial integrity and continuity of the state along with its international recognition, and in most cases the new regime would also wish to be perceived as the legal heir of its predecessor. Finally, even in a radical revolution, the same population or those speaking in its name could play the dominant role in the building of new institutions. Thus Carl Schmitt argued that in a democratic age, during some “revolutions,” as in 1848 and 1870 in France, the (same) people retain their constituent power. From a legal point of view, he referred to these phenomena as “setting aside” or, more clearly put, “replacing” the constitution (Verfassungsbeseitigung; Verfassungswechsel), in contrast to its annihilation (Verfassungsvernichtung).¹⁷⁴ He called both types of events “revolutions,” but only the latter, the annihilation of the constitution, involved state and regime destruction in the terms of internal and possibly even international law.

The combination of preserving some administrative and coercive organs, territorial integrity, and (reference to) the same political community thus implies that a radical revolution can and most likely will assume a conservative posture with respect to the state. Within a two-part distinction between revolution and coup d’état alone, this insight in an age of bureaucratization could lead to Weber’s diagnosis concerning the passing of revolutions and their general replacement by coups, a diagnosis especially paradoxical at the dawn of the twentieth century.¹⁷⁵ Using my three-part distinction based on the differentiation of government, regime, and state, I should note only the state-strengthening character of revolutions not directed against the old state structure itself.¹⁷⁶

The survival of the inherited state in revolutions in turn implies the persistence of a “minimal constitution,”¹⁷⁷ which French thinkers have rightly interpreted in terms of the survival of political organs that can help organize the constituent process even in the midst of revolutions.¹⁷⁸ In any case, there is a clear distinction between a revolution that builds on the existing state and the type of radical change that destroys a preexisting state, for example through a revolution-secession. Here the preexisting territorial integrity and the political community are destroyed, and whatever remains of administrative and coercive apparatuses can exercise monopolies only over parts of the former state, not the whole.

Important consequences follow from these distinctions—especially the distinction between revolution and revolutionary state destruction—for the problems of constitution making, the potential for civil war, and for the specific nature of an externally imposed revolution. Regarding constitution making, revolution proper no longer needs to be understood as creating a tabula rasa or creating a legal state of nature; only revolutionary state destruction has this consequence. And even in this latter case, if the state destroyed in a secession-revolution is a federal state with organized units, the constitution-making process can rely on their “state” organizations to avoid a legal state of nature. This is what happened at the beginning of the American Civil War, but there are many recent examples as well. Thus, state destruction is the most radical from a constitutional point of view when it occurs in a previously centralized state. From the point of view of radicalism, this is the most radical form of revolution. It is, of course, also the most difficult context in which to carry out the second stage Arendt insisted on: the constitution of freedom, as she called it, or less dramatically, the institutionalization of a new regime.

Civil war is a danger in all three forms of change, including the coup d’état, where two military leaders, for example, can have equal forces to contest the seat of the established governmental power. It is a greater danger in revolutions, because the prospect of regime change with important distributional and symbolic consequences is likely to mobilize politicized social forces along all important cleavage lines. This is an additional reason why revolutionaries, whatever their initial intentions, are likely to avail themselves of and strengthen the forces of the old state, and generally their success depends on their ability to do this effectively. Thus while revolution may almost tautologically imply civil war (as Tilly rightly notes),¹⁷⁹ the actual duration and depth of such a conflict may be manageable. The case of revolutionary state destruction, including revolution-secession, is fundamentally different. Here the old apparatus will be broken up or split. There will be no force available to contain armed political claimants to power, or the fragmented old state forces themselves will become part of a contest for power. A revolution-secession can lead either to the strengthening of the original unit-states that become independent or to the creation of new, weaker or stronger, federal or confederal structures, as was the case for both sides in the American Civil War.¹⁸⁰ However, to the extent the state breakup or the secession is contested, the result can be only an extended civil conflict.

Finally, it is very important for us that an externally imposed revolution has a freer hand in choosing among options than a classical internal revolution. When Lenin faced the problem of actually destroying the existing state, he had to realize that if he followed through his original ideas he would then have to face a still continuing external war and counterrevolution without the support and with the enmity of those sectors of the inherited military and administration that were now potentially loyal to the new regime. An external agent of revolution such as the United States, in its various attempts to impose democracy, had a choice at least when it faced an inherited state organization in an occupied country. It could act like MacArthur in Japan and build on the existing state. But with its own vast military resources, it could afford to be more like Lenin than Lenin himself. To the extent that it saw the inherited state and its administrative and military structures as part and parcel of an old regime that would always reproduce that old regime, even if the formal governmental structures were replaced, the idea of destroying these structures could be entertained supposedly without the fear of a scenario in which social order collapses because of missing state inputs and the emergence of armed political actors who cannot be controlled without the presence of much stronger state coercive forces. An occupation state could in theory replace the inherited state while an entirely new state structure was being created ex nihilo. This would involve creating a new army and police force as well as a state administration. Even de facto fragmentation of the territory could be permitted, as long as the agents controlling the potential breakaway region could not leave the political process of state rebuilding. The constitution-making process would thus serve as the context where both a new territorial state structure and new regime rules would be negotiated. It would not occur in a state of nature, because the occupying power (under its own interpretation or misinterpretation of international law) would supply the constitutional minimum needed for the process to move forward.

For the American occupiers of Iraq, only the third of these consequences was initially relevant. They only saw the availability of a choice, not the potential problems. Thinking in terms of historical analogies alone and relying on the German and Japanese models, they may have imagined that state destruction was a matter of choice without grave consequences for the occupier and that occupation with or without the destruction of the occupied state can allow for effective constitution making without the danger of civil war. Misunderstanding the Rand report, which stressed the greater similarity of Bosnia and Kosovo to Iraq, Bremer seemed to consider the Japanese and German examples especially relevant models to follow.¹⁸¹ But the connection between the parts of these models seemed to have been entirely unclear to him and his colleagues, since bizarrely enough they adopted, as I will show, their initial constitution-making project from Japan, where the old state was preserved (and where its mechanisms could disguise their imposition), even though they were going to head the German way and destroy inherited state structures. Without a theory or any serious prior analysis—but with almost hysterical warnings from other American participants—the CPA, supported by the Pentagon, embarked on an experiment during 2003 and 2004 with consequences that are now almost universally recognized as devastating.¹⁸² This they did despite a fundamental built-in flaw they were clearly conscious of, one called to our attention by L. Paul Bremer himself,¹⁸³ that there were not going to be enough American forces to monopolize the forces of violence, and that was going to be a problem, as the complete breakdown of public order and security as the Saddam government collapsed in April and May 2003 showed plainly. We will never know if having had enough American forces would have made the experiment work.¹⁸⁴ It is possible that such a force would have only increased the resentment and size of the insurrection. But it is also possible that a crucial period of time would have been gained to enable implementation of the rest of the project. We also cannot even know if having enough American forces would have worked if there had been an attempt to build on the inherited Iraqi state. What is clear is that not having enough occupation forces meant that the experiment in state destruction should have been abandoned before it was started. Jay Garner and others who had been a bit longer in Iraq were right; the new arrival Bremer was not.

In any case, CPA orders 1 and 2 proceeded to destroy the Iraqi administration (meaning both the top levels and large parts of the civil service) and the army. It is right to treat the two decisions together as the destruction of the Iraqi state in its administrative capacity, as long as we realize that its territorial integrity was already severely compromised.¹⁸⁵ There is more than ample literature on these two interrelated decisions, and there is no need to go into them in great detail here.¹⁸⁶ As to the army, the argument that it was already disbanded is clearly fallacious in light of the testimony of American officers already involved in trying to reconstitute it. Generals Abizaid (the new regional commander, the only one with local knowledge) and McKiernan (the highly professional first commander of American forces in Iraq) and Col. Pat Hughes, Garner’s planning chief, were heavily involved in trying to create the core of a viable army from the old, including even the formation of a new general staff.¹⁸⁷ If the soldiers went home, they could have been recalled. Their officers were ready to reorganize them even taking into account the purging of their ranks that Americans would have in any case insisted on. The political risks involved in recalling the army were real, because the officer core was indeed a stronghold of Arab (Sunnicentered) nationalist power, but this could have been gradually altered by selective decommissioning and commissioning and by a new stress on an inclusive Iraqi nationalism (always a minority alternative).¹⁸⁸ That the survival of the army would have been politically resented by organized Kurdish and Shi’ite groups is certainly true,¹⁸⁹ but with all their expected benefits from the occupation these groups would not have violently resisted. Their capacity to do so was in any case doubtful. In the longer term, gradual reform, reorganization, and affirmative action for underrepresented groups could have dealt with the problems. Keeping the army and most of the old apparatus would have involved costs, but getting rid of them involved even greater costs. It was up to the occupying power to evaluate which cost was greater: keeping a more or less efficient state machinery in place or not and attacking the main symbols of the integrity of the Iraqi state or keeping in place a set of institutions the leaders of Shi’ite and Kurdish groups wished to see destroyed. It was not enough to weigh the costs (dissatisfaction of major groups, deceleration of the change) and forget or easily discount the huge potential cost on the other side (collapse of security, collapse of administration, collapse of symbols of stateness).¹⁹⁰

And it was not just a matter of a static situation at the time the choice was made, because the choice was to affect profoundly the developments in Iraq. The destruction of the army by CPA order 2, “Dissolution of Entities”¹⁹¹ (meaning army and all military formations, the ministry of defense and all security-related ministries, but not apparently the interior ministry), created ready recruits for the insurrection in the stratum most capable of fighting a long-term resistance against any new government. When the fired military men demonstrated against Bremer, they did so openly, some even mentioning suicide bombings, and there is no question that many were to join the insurrection, with devastating results.¹⁹² The CPA’s first two orders solved the most difficult problem for the insurrectionists—recruitment—for them.¹⁹³ Moreover, it has been estimated that dismissing about 7 percent of the workforce produced severe economic hardship for about 10 percent of the population as a whole.¹⁹⁴ Finally, having to organize a new army ex nihilo created a long-term security vacuum in which militias took over law-and-order functions. It was difficult to prevent them from then entering the new army in an organized fashion. Thus there would be no new national army at all except in name. The one historically anticolonial and statewide institution—the one remaining symbol of state sovereignty¹⁹⁵ that was considered the guarantee against the return of neocolonialism—was destroyed, and many Iraqis interpreted this as an act of deliberate imperialist policy intended to maintain the country’s long-term dependence on the current occupying power.¹⁹⁶

As to the administration of the country and the provision of professional services, de-Baathification (CPA order 1) as ordered by Bremer and as administered by Ahmad Chalabi had devastating consequences. Surprisingly, the army was not a center of Baath power and organization, but this was discovered only when it was dissolved. As a matter of fact, Saddam Hussein trusted neither the army nor in the last phases of his rule the Baath organization itself. Reduced in size, deprived of all ideological vigor, the late Baath was a tool for selecting “careerists and obliging technocrats, and providing a parallel bureaucracy to ensure the loyalty of the bureaucracy and armed forces.”¹⁹⁷ However, most members of the bureaucracy at the upper levels still had to show their loyalty by joining, as did professionals including teachers, doctors, and university professors wishing to practice in their fields. In his book My Year in Iraq, Bremer speaks of ridding “the Iraqi government of the small group of true believers at the top of the party and those who have committed crimes in its name.” CPA order 1 had to do with the top four levels of the party of two million, definitely far more people than the twenty thousand he estimates. Either deliberately or inadvertently he is confusing the figures and the Baath categories initially agreed upon in Washington with the much larger figures his own order entailed. Even more seriously, he glibly glosses over the important new distinction in his order, pointed out by Rajiv Chandrasekaran, between eliminating the top four categories of Baath from all jobs and eliminating all Baath members from the top three levels of government jobs. Based on a memo by Douglas Feith of the Pentagon, it was the two steps put together that generated the high numbers of firings.¹⁹⁸ Moreover, there must have been few true believers and even fewer actual criminals among them. On the other hand, the admitted “administrative inconvenience” involved was a pathetic understatement of the likely and actual results. Finally, government here meant civil-service employees in a country where the state was by far the largest employer, and “full Baath members” purged from the top layers of the management of every “ministry, affiliated corporation, and other government institutions, institutes, hospitals” would not be allowed government employment anywhere else. This meant teachers, doctors, managers, engineers, and so on. Ten to fifteen thousand teachers were fired, leaving only one or two teachers in some Sunni-dominated areas. Ministries lost the bulk of their staffs; Americans lost the partners they were already working with.¹⁹⁹ Some hospitals were in danger of being closed.²⁰⁰ One American official fired 1,700 university professors and staff, and felt justified in doing so by the German analogy,²⁰¹ though the official does not seem to realize how idiotic that analogy was. Some estimate that 35,000 employees of the bureaucracy, mostly Sunnis, lost their jobs overnight, with 65,000 targeted.²⁰² The significance of this number is that it hides the fact that the people involved were probably the most expert and most able to lead institutions, including state-owned companies, which were severely affected.²⁰³ But the most disastrous result may have been, according to a joint study of the inspectors general of the State Department and the Pentagon cited by T. Ricks, the complete decapitation and disorganization of the Iraqi police, which had been exempted from CPA order 2, which dissolved the security institutions along with the army.²⁰⁴

There is much evidence suggesting that the consequences of CPA order 1 were immediately far reaching, and Bremer’s attempt to put the blame for this on Chalabi’s de-Baathification commission in the IGC is only partly justified. On the other hand, Chalabi and his group’s influence seems to have been crucial in making the original decision.²⁰⁵ In the face of State Department opposition, it was Chalabi’s group who apparently convinced the Pentagon with entirely mistaken analogies such as “de-Nazification,”²⁰⁶ when the generally rejected idea of de-Communization in eastern and central Europe would have been the far better analogy given the respective lengths of Baath, Communist, and Nazi regimes and their types of relationships to society. It seems well established that it was supporters of Chalabi such as Feith who convinced Bremer of the idea of drastic de-Baathification, although he like all others continues to deflect the blame. There is some evidence that Z. Khalilzad, with whom Bremer was initially supposed to work, teaming with A. Allawi, fought for a very limited process of de-Baathification and sought to recruit the professional and expert strata for the new order. But Bremer made a special point of insisting that Khalilzad be recalled, and with Garner out, the Chalabi position became dominant.²⁰⁷ It is thus no use to put the blame on Chalabi’s later administration of the program. Moreover, interestingly, the de-Baathification order was announced to Chalabi and the ILC the same day they were told that power was not going to be handed over to them, as Garner promised. Bremer, who called his role that of first a good cop (before a bad one), seemed to believe, rightly, that de-Baathification would be the carrot that the seven leaders, especially Chalabi, needed in order to accept the stick of the CPA’s dictatorship.²⁰⁸ Bremer notes what he thinks is a contradiction, namely that Chalabi himself pointed out that some Baathists were forced to join. What in fact he was doing is indicating his objectivity and his suitability to be in charge of a process that would gain him influence over both those who replaced old employees and those Baathists who were after all allowed to stay. And indeed, to make the carrot even more tempting, Chalabi was put in charge of the de-Baathification commission, giving him immense potential power.

Some interpreters seem to feel that de-Baathification and the decommissioning of the Iraqi army would have worked less destructively if Bremer had a delivered on Garner’s promise and put the ILC in charge of the provisional government.²⁰⁹ Note that Garner’s plan, even for that case, involved only very minor de-Baathification, and he would have reconstituted the old army. But the idea that the people in power, who pushed for the most radical reconstruction of the Sunni-dominated state—without adequate Sunni representatives—would be forced to negotiate with Sunnis was a nonstarter unless they lost the protection of American military forces: if, in other words, the United States withdrew from Iraq or threatened to do so. Even then, it is more likely that they might have opted for the drastic measures they themselves, namely Shi’ite and Kurd exile groups, pushed for the hardest, even at the cost of launching a civil war immediately. With the United States in Iraq, they would have felt even more confident to pursue their maximalist goals of revolutionary restructuring, which, given the small size of the American force, would not have been able to prevent civil war. If the goal was to produce a negotiated settlement between the real forces in Iraq, and that included the Iraqi military as well as Sunni Arab nationalists, the best way to accomplish this would be for a neutral monitor to force parties into contexts of inclusive negotiation and power sharing. The United States was in the position to play this role, and it is not clear why in the long term its hostility to a given side had to be so great as to inhibit its playing it. What is certain is that with CPA orders 1 and 2 the possibility of becoming a neutral referee was gravely compromised if not lost altogether.

That the experiment in state destruction was pursued meant several interrelated things for the Iraqi state, some obvious, some less so.²¹⁰ Best treated around the dimensions of the control over violence, population, and territory, the first three are directly connected to the combined problem of the small size of the American forces and the dissolution of the Iraqi army. The fourth is connected to the dramatic decline of the Iraqi state as the producer of goods and services.

(1) Most obviously, the state in the sense of any imaginable monopoly over the means of violence was gone. There was not enough of an occupation force to reestablish it. The Iraqi army was gone; the police force was severely compromised. If there was to be any order, Iraqis had to rely on themselves, their tribes, their mosques, and the various militia that emerged from mixed political, religious, neighborhood, and tribal foundations. The decentralization of violence, however, only added to the problem itself. Because the relationship of each of the forms of coercive organization to the occupation was different, and because there were traditional grievances against one another, it was only a matter of time before antioccupation activities released grave sectarian conflicts as well. An insurrection based in the Sunni Arab part of the population always had the potential to produce a Shi’ite-Sunni civil war even without the deliberate provocations of the most radical Sunni Islamist factions. This potential was unavoidably released when most Shi’ite political groups, themselves no friends of the occupation, chose the strategy of coming to power not through open resistance but through elections the Americans in the end were forced to accept.

(2) The population of the country, another dimension of stateness, was increasingly fragmented on ethnic-religious lines. Perhaps there never was an Iraqi nation, or perhaps it was already severely compromised by Saddam’s treatment of the Kurds and the Shi’a. But there was nevertheless common membership in a polity that provided a large variety of social (educational, medical) and infrastructural (communication, transportation, electricity, water, sanitation) services and enforced a variety of duties (military service, primarily) on that basis. To be Iraqi meant at least a minimum entitlement²¹¹ to many things and a responsibility to perform certain duties that would apply across the board to different population segments. When the organization of violence took place on ethnic-religious grounds, penetrating the police and the new armed forces, this meant that people could expect state-type outputs only from “their own people,” and that the others, even when wearing Iraqi uniforms, had become deadly threats to them. It was not only a matter of security and protection but also the provision of local services. The general insecurity, disorganization of the state apparatus, fall in revenues, and open attacks by insurgents meant a collapse of what used to be central state services, whether social or infrastructural. Many of these would now be supplied by local governmental organs under the protection of militias, both organized increasingly on ethnic-religious lines. Public services still supplied by the central state, such as water and electricity, were distinguished by their scarcity and poor quality. There is some evidence, however, that in mixed areas including Baghdad even central state infrastructural services are now provided on ethnic bases,²¹² and the same is probably true for whatever is left of medical services and the increasingly segregated system of education. The trend toward ethnicization was also supported by the continued exclusion of organized Sunni forces from the political process and the establishment of ethnic-religious quotas whenever governmental structures were negotiated. Unsurprisingly, political parties when successful were also organized on religious-ethnic lines.

(3) The territorial integrity of the state was severely compromised from the outset. Admittedly, this was a function of the particular ending of the first Gulf War, which left the Kurds with an autonomous area. But the paucity of American forces meant that Kurdistan was unoccupied by the United States in 2003, the CPA’s control therefore could not be extended to all of Iraq’s territory, and thus the Kurds could enter constitutional negotiations as agents of a de facto independent territory that as far as they were concerned might not reenter Iraq. Failure to secure Turkish support for the war, which would have led to a Turkish occupation of Kurdistan,²¹³ and the unavailability of American forces meant that the Kurds would be allowed to create more and more facts on the ground and could be entirely recalcitrant when it came to constitutional negotiations. Had there been an Iraqi army, any threat reinforcing the bargaining position of Arab parts of Iraq would have been far more credible.

(4) Despite some economic liberalization already under Saddam, before the war the Iraqi state was both the major producer of services and of (state subsidized) goods, including the all-important nationalized oil industry (the basis of all subsidies).²¹⁴ Under the American occupation, the decline of the state in its economic functions was in part the result of the security problem (hence of the destruction of the army) and of the disorganization of the state apparatus (through de-Baathification) and in part the consequence of other conscious efforts on the part of the CPA during the period of formal occupation. With respect to the oil industry, exposed to both sabotage and corruption, the resulting decline in potential revenues (somewhat balanced by the rise in the price of the commodity) affected all other dimensions of the effort to restore state activity. Here the culprit was primarily the security situation, a situation involving obvious negative feedback: less oil, less money, less security, less oil. It is not clear, however, that given the apparent inside knowledge of the industry and its infrastructure by the neo-Baathist insurgency anything can be done to improve the situation short of a comprehensive political bargain.²¹⁵ Other state industries were victims of reorganization schemes that produced factory shutdowns, bankruptcies, and unemployment. Here de-Baathification played an important role in getting rid of expert managerial staffs, though there are no exact numbers indicating the extent. Even more damaging were reorganization schemes starting with a privatization effort that failed in the face of international legal prohibitions²¹⁶ and the fairly general incomprehension of Iraqis, who blamed their country’s lamentable economic situation on Saddam’s wars and not on a structure that previously produced a high level of (admittedly oil-related) welfare. When privatization failed, mostly because there were no buyers or even leasers for Iraqi firms under the given security conditions,²¹⁷ its apparently wholly amateurish American advocates shifted to a policy of “shrinkage” based on the elimination of subsidies that almost no Iraqi company was prepared to survive. Worse, a simultaneous elimination of all company debts and financial assets, which was supposed to help weaker companies with high debts that were still not in a position to resume production, in actuality deprived stronger companies with serious bank holdings of their ability to do so. Shrinkage there was, with most of the one hundred thousand workers of state firms left unemployed as a result; even now, plans involve rehiring only ten to fifteen thousand of them.²¹⁸ By the time the transfer of “sovereignty” brought these experiments to an end, the damage was done.

The four dimensions of state destruction certainly had independent motives, even if they might appear “rational” or “logical” in terms of some imperialist grand design to keep Iraq weak and dependent. The state apparatus and the symbols of statehood were victims of the revolutionary radicalism of the occupiers, goaded on, to be sure, by office seekers and Iraqi enemies of the older central state. Thus the state’s people fell apart both because of the occupier’s policies and because there were large groups that understood their identities in alternative and competing ways. The state’s territory fell apart because the occupation was powerless to reunite it, despite some intentions to that effect, for example, the toying with the idea of Turkish destruction of independent Kurdistan. The state in its economic capacity fell apart both because of the security situation and the ideological antistatist proclivities of the occupiers. All these dimensions were mutually reinforcing, obviously, in their negative consequences. But in all of them, the failure of the classically Weberian dimension of the monopoly or near monopoly of the means of large-scale violence has pride of place, and it is a dimension that reinforces the connection of all the others. Given the extreme tenuousness of the assumption in this context that the small American forces were sufficient to provide security, a tenuousness understood by the leaders of the CPA, their destruction of the Iraqi army and civil apparatus can only be described as wanton state destruction.

Given state destruction, admittedly, it becomes true after all to describe the American project, eventually, as both regime change and state building. A very important part of both these burdens had to be assumed by the constitution-making process. Not only would a new constitution have to establish a new structure of government, it would also have to reestablish the territorial integrity of an Iraqi state, the political organization that could maintain it, and presumably a locus of services, rights, and obligations that could again produce a minimum Iraqi collective identity. Such a burden is a difficult one for constitution makers.

Historically, all modern constitution making presupposes successful state making. It may be valid to argue, as did Raymond Carré de Malberg, that since all states have a constitution in at least the eighteenth-century sense or in the sense of Kelsen’s material constitution, the state and its first constitution are co-original.²¹⁹ The continuity of the state in revolutions thus implies, as already argued, the continuity of a minimal constitution. This does not change the fact that even revolutionary constitutions are made in most cases for a preexisting, even if already minimally constituted, state. State making in revolutions then means the building or rebuilding of the state under a new constitution but on the foundations of a preexisting state. In a case such as Iraq, which was a state before the American invasion and occupation, it is hard to see how public power and public security, and therefore the object for which constitutions would be made and that it is meant to reorganize and regulate, could come from any other arrangement of the polity than a state. This theoretical point has been persuasively argued by Dieter Grimm, who has further maintained, in the tradition of Schmitt’s Verfassungslehre, that the subject of constitution making as well as the object presupposes a state, in other words, unified public powers capable of being activated by a people or a nation as a constituent power.²²⁰ Accordingly, if a state did not exist before constitution making, or if it was destroyed, the logically prior act of state making would have to proceed simultaneously with constitution making. That this is possible is shown by the historical examples of the formation of federations and federal states where a new subject (state) is constituted from old subjects (states) in some kind of constitutional treaty.²²¹ Even in such contexts, the perspective advocated by Grimm can be sustained, because as we see from the example of the United States both in 1776 and 1787, the units that formed the confederation and federal state in constitution making themselves had well-organized states with close enough approximations of public monopolies of the legitimate means of violence. Here both the subject and object of constitution making were public powers. Nevertheless, I believe in any case that the notion that private powers cannot be the source of constitutional authority is belied by the role of juridically private conventions in the American states in the period 1775 to 1780 and the round tables in the recent experiences of the new democracies. In these cases, only new forms of public legitimacy had to be generated for private organizations, and any agreements emerging from private spheres had to be approved in public political procedures. Of course, Grimm would answer that even these experiments presuppose “stateness” or public powers with the requisite monopolies either existing before or constituted during the constitution-making process, at least as the object of constitution making.

However, what the American invasion and occupation in Iraq did was to introduce both serious doubts concerning the legitimacy of the “subject” of constitution making and equally serious uncertainty about the “object” for which a constitution would be made. This issue of legitimacy I already discussed, but I can briefly restate it here. The foreign occupying power has no right under international law to give the occupied country a constitution. By implication, neither do its Iraqi agents such as the IGC.²²² Private organizations, like political parties, not organized in a public body, also have no legal right to do so, but they could, as set against a foreign occupier, generate political legitimacy under conditions resembling the round tables of Poland, Hungary, and South Africa. I will return to this point in later chapters, but I will say here that this has not happened in Iraq because of the narrow, exclusionary bases of the constitutional negotiating process under the shadow and influence of foreign imposition.

The situation is even worse with the “object” for which constitutions are made: public power. The issue of unresolved state structure goes so far as to have consequences for organizational choices on almost all levels, making the writing of a constitution that could be taken seriously extremely difficult. The problem was both dedifferentiation and disorganization of the Iraqi state. The differentiation and independence of something like that state both from an external power and internal private powers with their own armies allows considerable doubt. The occupation forces, now under simply the name of “coalition forces” or “multinational forces” (MNF), as well as a large number of private military contractors (that is, mercenaries of a new type), remain a state within the “state” not subject in any way to Iraqi law or constitutional restraints. These could be written into constitutions in general but did not apply to the major forces of violence and incarceration in particular. With the de facto Kurdish entity, which acts like a full-fledged state within its territory—and at times elsewhere where its forces are engaged nominally under an Iraqi flag—and the assertion of various uncontrollable forms of local rule, all with their own military forces, the boundaries of state authority or authorities are extremely poorly defined even if we forget the foreign forces. From this point of view, the insurrection may be particularly ugly and violent, but legally it represents more or less the same problem as the other violent nonstate organizations that control people and territory and cannot themselves be controlled. The organizations that in the modern world are called upon to enforce “stateness”—the “national” army, the “public” administration, and the “national” police—lay in shambles because of deliberate American policy. As to the issue of security and the control of violence, it is clear that the enforcement of any constitution in the legal sense cannot focus or rely exclusively on public authorities but must also and to an equal extent take into account the very private powers (and their good will) that are themselves the greatest threats to legality. The issue, however, is not only that of violence and security. Public services, enforcement of contracts, and in most areas personal relations too are under the control of juridically private organizations, in the latter case religious organizations. In such a situation, Grimm’s question concerning the very meaning and possibility of a constitution makes a great deal of sense. One makes a constitution, but it establishes or regulates only a small part of political power in the country—and by no means the most important part.

Because Iraq was a state before the beginning of the American war to change its regime, in principle constitution making here could have followed the pattern established with the overthrow of old regimes, one relying on inherited state structures. Instead, there was a set of policies, coordinated or not, whose aim or at least consequence was to destroy the Iraqi state as well as the regime.²²³ As a result, fragments of the old state went underground. The very same Sunni elites that were being deprived of employment in the state organs were also denied all representation in the political process, and they were being given an underground army or armies, fueled by powerful new resentments, in a context where there were no effective military forces available to control them or even to prevent their access to the enormous amount of ordinance stored in various parts of the country. The result was not only the insurrection and the collapse of public security but the impossibility of dismantling militias and private military organizations. In short, the effort to temporarily replace the dismantled Iraqi state by the occupation state was strikingly unsuccessful: no one established any kind of monopoly over the legitimate means of violence except for the political authorities relying on the Peshmerga in Kurdistan—the Kurdish quasi-state that was neither dismantled nor occupied but strengthened in its autonomy, and whose strengthening was a sign of Iraq’s disintegration and of the difficulties of rebuilding its state.

To sum up then: the destruction of the state in Iraq placed tremendous burdens on the constitution makers of Iraq, whoever they were going to be, operating under whatever procedure. Their task was both the creation of the new regime and, logically prior to that, the rebuilding of the Iraqi state in its political and territorial integrity. Evidently, success would depend on other efforts, arguably even more important: the reconstruction of the forces of violence and administration. And all of these concerns were interconnected, because a unified military and administration was not possible without a general state bargain on the constitutional level, and a constitution could hardly be made for a country in civil war. Arguably, such a task was impossible, given both the bad faith of the Americans in ceding real autonomy and the security situation, which could be linked to state destruction. On the other hand, the emergence of autonomous, popular forces on the terrain of constitution making and the reluctant concession of the Americans to a stronger UN role in the process gave the constitution-making process at least a chance at becoming autonomous (Sistani’s role) and inclusive (Brahimi’s role). And if these forces had been successful in converting the process of constitution making into an autonomous, legitimate, national and inclusive effort, it was just possible that a significant part of the insurrection would have joined the political process at an early stage. Thus the security situation and incipient civil war that always threatened to reduce the political process to a mere sideshow could have been perhaps marginalized by a different political, constitutional process. This did not happen, because a convincing and broadly legitimate political process was subverted relatively soon. If anything, the failed constitution-making process became an additional source of conflict and violent confrontation.

As we now see, the margin of error on the part of constitutional negotiations dealing with state making is much smaller than it is in the case of constitution making in general. If, however, the state-making part fails, the possible “success” of the rest of the constitutional product becomes irrelevant. One cannot make a constitution for a failed or nonreconstituted state. I will argue that the model of constitution making adopted was actually suitable when abstractly considered for the double task of state reconstruction and regime creation. In the Iraqi case, compared to others, failure of a part of the process would have far more devastating consequences than elsewhere. To understand what could have been possible, I will now turn to a theoretical and comparative consideration of the specific model of constitution making adopted in Iraq.