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

These [occupation] authorities do not have the authority to appoint the members of the constitution writing council. There is no guarantee that this council will produce a constitution that responds to the paramount interests of the Iraqi people and expresses its national identity of which Islam and its noble social values are basic components. The [constitution writing] proposal is fundamentally unacceptable. There must be general elections in which each eligible Iraqi can choose his representative in a constituent assembly for writing the constitution. This is to be followed by a general referendum on the constitution approved by the constituent assembly. All believers must demand the realization of this important issue and participate in completing the task in the best manner.

—Ali al-Husaini al-Sistani, June 25, 2003¹

The model of constitution making used in Iraq was initially planned by no one; it was the result of the clash of major political forces. Thus it could be said to incorporate a compromise. Yet while there were important negotiations and even international mediation, ultimately the formula for the process and a good part of the contents of the Transitional Administrative Law were imposed by the occupying authority. However, there were two distinct dimensions to this imposition. The clash with the forces led by the Grand Ayatollah Sistani led to the compromise formula of a two-stage transition imposed (rather than in any serious way negotiated) by the Coalition Provisional Authority on the Interim Governing Council as a whole, and through them on Iraq itself. That process is the theme of the present chapter. There was no clash with the Kurdish Regional Government or the two parties that controlled it, but the CPA conducted serious negotiations with them—entirely outside the IGC, to which these parties indeed belonged—regarding the structure of the Iraqi state as it would appear in the interim constitution. The results were then imposed on the rest of the IGC when other important contents of the TAL were worked out. Chapter 4 will deal with this second process of imposition and bargaining, which also involved genuine give and take concerning American positions not assumed to be sufficiently fundamental (or deemed impossible to impose) to the CPA.² Interestingly, the second process to a significant extent vitiated the elements of compromise contained in the first, and this is why the Grand Ayatollah Sistani was confirmed in his determined antagonism to the TAL, which never would have emerged without his political movement in the first place. Thus these two chapters together deal with ultimately failed attempts to deal with the legitimacy crisis through constitution making.

From the outset, the Americans wanted to impose a constitution. The motives may have varied greatly, from a neoimperialist desire to get a friendly government to liberal democratic distrust in the political cultures and political forces they encountered.³ Whatever the motives, they could be justified by a logical self-assurance that since where there is no democracy the only way to begin is by imposition, thus it does not matter much whether the imposition is uni- or multilateral, internal or external, especially if the end result can be justified in terms of “universal” values. We cannot untangle the real motives, and the later testimonies of those involved (full of self-justification) are unreliable. For whatever it is worth, among these statements the idea that stands out is that only by American imposition could sufficient respect for rights, a degree of separation of religion and state, protection of minorities, and federalism be assured.⁴ A more skeptical interpretation would represent this as a public-relations attempt to retroactively justify the war by eventually being able to point to Iraq “as an example of enlightened democracy in the Arab world,” as manifested in constitutional documents.⁵ But once the question of not immediately transferring power to some Iraqi provisional government was decided, so was the question of constitutional imposition. In principle, it was possible just to impose enough to jumpstart full autonomy, but that would have had unpredictable consequences, which the Americans in Iraq were never prepared to live with and were warned not to do even by realist opponents of the war such as Scowcroft. That such a thing as “imposed constitutionalism” was possible was indicated by one and only one successful historical example, the occupation of Japan, and this is why I think the figure of MacArthur was so important for Bremer and his circle.⁶ It did not matter that partly because of historical givens and very soon because of their own grave acts of commission the Japanese situation was very different than the one in Iraq. What they learned from the occupation of Japan—but did not unlearn even under Sistani’s challenge—was that a determined group of American politicians and supposed experts could use whatever available political façade for making and imposing a constitution that could be, if done right, successfully presented and legitimated as a highly progressive and indigenous achievement.

Thus I simply do not accept at face value the continued insistence by Paul Bremer⁷ that in his scheme at all stages Iraqis were supposed to write their own constitution. When Bremer was confronted by the charge, apparently from Sistani, that Bremer planned to imitate MacArthur and impose on the Iraqis a constitution written by American experts, he gave a response quite literally worthy of MacArthur. Such a thing was furthest from his mind: “the Coalition has no intention of writing the constitution.”⁸ While it is impossible to say whether or not at all moments he and his advisors expected American experts (of Iraqi origin or not) to do the bulk of the drafting (as they actually were going to do in the case of the TAL), what they had in mind, indeed for the writing of the permanent constitution itself until Sistani’s victory on this point, was that officials appointed by the CPA directly or through the IGC but under the former’s sovereign authority would write and enact the basic document. Only what the CPA wanted to pass could be passed in such a model, whoever the literal drafters and their expert advisors were.⁹ They assumed, in other words, the legal authority to impose, and from the retrospect of the TAL we can infer that this intention referred to both the contents of the document and to the one-sided structure of authority.

Of course, the intention to impose made the adoption of the postsovereign paradigm discussed in the previous chapter in a fully legitimate form impossible, because that would have presupposed a many-sided rather than one-sided structure of authority and control. But before that choice could even come up, there were serious problems with a model of straight imposition. Unlike in Japan, where state and governmental continuity provided a ready-made formal process behind which the act of imposition could be veiled, in Iraq nothing like that was available. Thus a process had to be constructed rather than inherited—one that looked legitimate (that is, Iraqi) but that would not get out of hand. This turned out to be impossible for the initial constitutional project of the CPA.

Initially, there was much confusion about how policymakers wished to proceed, but it was definitely assumed that it would be an Iraqi body (“a conference”) selected by the CPA that would be formally in charge of the drafting.¹⁰ Once the IGC was formed, it was assumed that the selection process would operate through it or formally by it, but always under the CPA’s strict supervision. Any plan generated by the IGC or its Constitutional Preparatory Committee, as it is clear in Bremer’s memoirs, had to be thoroughly negotiated with and approved by the CPA. It may be true that Sistani’s first fatwa quoted above was initially entirely disregarded,¹¹ but in any case, under his pressure or not, the CPA’s plans developed in what could be claimed to be a more “democratic” direction. Bremer in particular later claimed he was impressed by the 1925 Iraqi Constitution having been ratified in a referendum, and he believed he was following that example when he linked the idea of a selected conference to popular ratification.¹² Quite amazingly, he does not seem to have realized even when writing his memoirs that the 1925 Constituent Assembly was elected and that this was probably one of Sistani’s models in the fatwa of June 25.¹³ It is possible that the idea of the referendum was a detached part of Sistani’s wider proposal. In any case, simply adding a referendum or a plebiscite to the CPA’s ideas was a characteristic attempt to shore up the democratic legitimacy of a constitution-making process that would be in the hands of the executive branch. It would have been the mark of Bonapartism, in quite a strict historical sense, especially because under conditions of foreign occupation the plebiscite in question, once a constitution was previously approved, would have to be regarded as highly constrained.¹⁴

That last suspicion is amply demonstrated by looking at Bremer’s seven-point program published in the Washington Post on September 8, 2003, under the revealing title “Iraq’s Path to Sovereignty.” Here the argument was repeated that in Iraq, elections presuppose a constitution, a new and permanent constitution. The steps relevant to constitution making were said to be (1) the creation of the IGC, (2) the creation of a Constitutional Preparatory Committee of the IGC, (3) the writing of the constitution by a process recommended by the Constitutional Preparatory Committee, and (4) popular ratification of the constitution.

The process was represented as the first time (perhaps Bremer was unaware of the 1925 constituent assembly back then) that “Iraq will have a permanent constitution written by and approved by the Iraqi people.” How the document was supposed to be written by “the” Iraqi people was unclear, given that there was no formula under point 3, and in any case a co-opted conference could not be said to be representative of the Iraqi people. However, what approval by “the” people meant was clearer. Steps 5 and 6, namely the election of an Iraqi government and the dissolution of the CPA along with the recovery of Iraqi sovereignty, could follow only if the popular response to step 4 was positive. The recovery of independence was linked to one and only one outcome of the plebiscite, thereby foreordaining the result whatever the voters thought of the constitution itself.

As I have said, the motives for imposed constitution making in Iraq, though convergent, were probably plural, complex, and difficult to reconstruct. It was a little different with the problem of justification. The CPA did not take the position of John Yoo,¹⁵ who argued in clear violation of the Hague and Geneva conventions that an occupying power had the right to make a constitution for the occupied. The constant stress, also there in the seven-point program, that it would be Iraqis that would make their own constitution indicates something different. As ambiguous as UN Security Council resolutions 1483 and 1511 may have been, and deliberately so, they seem to have pointed to a combined responsibility of the CPA and Iraqi actors to manage the political process of the transition. I am quite certain that the CPA’s lawyers believed that all of their various formulas, including the one eventually adopted, satisfied this rather vague demand, which does not deal adequately with the precise structure of authority, and indeed there was never a significant challenge of CPA constitution-making formulas from international sources. With respect to Iraqi actors, the argument had to be made repeatedly that an elected constituent assembly was not possible on empirical and even logical grounds. The empirical grounds, as weak as they were, had to do with the absence of an electoral law, electoral rolls, and census data, and insufficient time to organize such things. These arguments were in part spurious, because a single-country PR system like the one eventually chosen could be legislated in a day and requires no prior census data, only proofs of age and citizenship. Being a passive electoral system, its imposition would not be seen as trying to construct a particular outcome. Since this was very well known on all sides,¹⁶ the empirical arguments could only arouse the obvious suspicion that the United States feared the victory of those who were likely to win.

The arguments on logical grounds were somewhat better, but they pointed to a very different type of process than the one chosen. In terms of the legal regulation of all the processes needed for free elections, from private security to media access and various other rights, Iraq was under an old political regime in many respects, or an occupation regime, or their combination, and none of these was conducive for a democratic transition. To have elections, Bremer rightly and repeatedly argued, new laws, some of them constitutional laws, were needed, and thus for logical reasons free elections could not be expected to produce them. However, when this dilemma has arisen in other situations (for example, in the South African and some central European cases), the actors working for democratic transitions arranged inclusive round tables of all the important groups—and even the “talks about talks.” Even transitional rules have distributional consequences, and it creates enormous problems to impose them in a one-sided manner. That is precisely what the CPA intended to do, directly or through Iraqi proxies, and indeed initially for permanent rather than transitional rules. The logical argument was thus correctly employed against holding early elections, but it could not justify the alternative that was chosen.

The original perspective articulated by the fatwa of June 25, 2003, was maintained by Sistani with amazing consistency. While the exact formula advocated, a freely elected assembly plus referendum, seems closest to the French model first practiced in 1793,¹⁷ Sistani was obviously aware of the rhetorical power of advocating a democratic alternative against the Americans’ imposed model, who were evidently quite vulnerable to this type of “immanent criticism.”¹⁸ Nevertheless, it is highly likely that Sistani’s constitutional ideas come from his own tradition, even if that tradition itself absorbed classical European ideas of democratic constitution making.¹⁹ The tradition of the Iranian Constitutional Revolution undoubtedly strongly influenced him, as did some turn-of-the-last-century Shi’ite scholarship. But the former did not yet have a freely elected constitutional assembly,²⁰ and he follows the latter more in calling for limited government and accountability than for a council of guardians.²¹ He certainly has a more authentic claim to the heritage of the 1924 Iraqi Constituent Assembly than does Bremer. But this example was a complicated one for the Shi’a clergy to exploit, because it was during the 1923 elections that the leading Shi’ite mujtahids organized a boycott that led to the temporary expulsion of many of them from Iraq, the reduction of their political influence, and the overwhelming Sunni control of the Constituent Assembly (already guaranteed to some extent by the electoral rules).²²

It is often said that Sistani advocated a program calculated to bring to his forces the power of the majority. And indeed whatever plans he had for actual constitutional arrangements, a constitutional assembly elected before any provisional agreements, an interim constitution, and so on would only be under rules it chose to impose upon itself. If it had a narrow Shi’ite majority (and that is what the overall demographic distribution indicated), such a body could enact by simple majority simple majoritarian decision rules both for itself and the following national referendum. There would be no limits in principle with respect to religious and ethnic affairs, administrative centralization, and control and distribution of national resources that such a majority, however small, could not arrogate to itself. The intentions of Sistani and close supporters such as Hussain Shahristani, most likely relatively humane and tolerant, might not matter, because such a majority could fall under the sway of its most militant elements. Even before that happened, given Iraq’s strongly divided society, the minorities would reject the process, and instead of majority tyranny the more likely result would be constitutional and civil crisis. When Bremer and his colleagues referred to these issues in their attempts to block Sistani, they were not only being self-serving or demagogic. They were also articulating the views and interests of all non-Shi’ite groups in Iraq.²³ These perspectives were possibly reconcilable with Sistani’s positions, but he himself was not in a position to do the reconciling. However, it is also true that he showed no sign whatsoever of being aware of any constitution-making approaches that could unite constitutionalism and democracy and the rights of a majority with the needs of minorities.

However, the CPA was in the driver’s seat, not Sistani, and it was up to the CPA to offer formulas of reconciliation. Sistani’s opposition to imposed constitution making was clearly right in terms of what it opposed, and the mantle of democracy he assumed projected him for a while into a position of national leadership. However, because of his view that clerics should not play a direct political role, and because of the divided nature of Iraqi society, increasingly manifested in the Sunni insurrection, Sistani could not assume the national political stature that would have best fit his constitutional challenge. However, there is little in his later political statements to indicate that his view of Iraqi politics was highly exclusionary or narrowly majoritarian. Had he been offered a better deal than he eventually received, perhaps he would have been able to play a more integrative role. As things happened, he found himself outside, and in a peculiar way the moderate Shi’ites whom the Americans did in the end use to make the final deal were more hardline on the divisions of Iraqi society than Sistani might have been.

With a large popular movement behind him—but one with a relatively narrow ethnic-religious definition—it is quite amazing how far Sistani was able to deflect the Americans from their original goals. He stopped them from writing the permanent constitution and he received from them a commitment for a freely elected constitutional assembly. He got them to bring elections closer by about a year if not more. He got them to restore sovereignty, though only in a formal sense, a year and a half early. But his defeats are also significant, given how much more difficult he could have made matters for the occupation.²⁴ He had to accept elections much later than they could have been held. He had to accept the imposed TAL, even if he could block its UN Security Council authorization. He had to accept an interim government that would not be composed of the elements he wanted and that could not do its job effectively.

Aside from his own popular forces, his strength also depended on the ideological power of his message. If the occupation of Iraq was for the sake of democracy, he wanted that democracy to operate in the most fundamental political process of all: the making of rules for the political process. He quickly became a favorite of reporters and politicians from all over the world, even though he would not see them or give interviews.²⁵ But his limitations were obvious too: having taken a deeply and fundamentally political stance, he refused to enter politics, acting only through weak and often unreliable intermediaries. Of course, even these weak intermediaries were better than nothing, given his unwillingness to play a direct role. As things unfolded, it was unexpectedly the Constitutional Preparatory Committee rather than the main leaders of the IGC who proved receptive to his message. Most likely in this formally subordinate body, religious allegiance and expertise reinforced one another.

Without the large popular demonstrations in Najaf, Karbala, and, with the greatest effect, on five consecutive days in Baghdad, Sistani would not have achieved much.²⁶ His theoretical priority as the first among equals among the most senior clergy “worthy of emulation” (marja’ al- taqlid) or his leadership of the marja’iyya in Najaf²⁷ became really significant only when he demonstrated his power, and the constitutional issue gave him a great vehicle for this. It is possible that the crowds were moved by the combination of religious leadership and the discovery of his political attitude, reflecting that of the Shi’a masses, who disliked the occupation but wished to benefit from it if possible. Nonviolent public protest was the perfect means for this and was a better alternative for most than Moqtadah al- Sadr’s armed resistance. They could now demonstrate for something that was neither possible under Saddam nor given to them by the foreign occupiers: a freely elected constituent assembly. It was humiliating to have to be liberated by the Americans, but some of the shame could be redeemed if that liberation could be turned in quite a different direction than the liberator-occupiers themselves had chosen for Iraq. To the constitutional scholar, Sistani’s challenge was about democracy. To the Shi’ite masses, it was probably more about dignity. Regardless of what it was “about,” they marched and demonstrated in very great numbers, under Sistani’s picture and chanting Sistani’s slogans.

Part of the battle was fought in the streets, the other part in the IGC, especially its Constitutional Preparatory Committee. If these two bodies were meant to be a mere façade for the rule of the CPA, Sistani’s challenge temporarily put some life into them. It is highly amusing today to read Bremer’s cynical description of the plans to make it look like the IGC had created itself, by “immaculate conception,” and his indignant refutation of press charges concerning the lack of independence by members of the council.²⁸ The IGC was put together mostly by Bremer and his advisors, with some help from Sergio de Mello, the UN High Representative.²⁹ It is right to call the IGC exclusionary and lacking governing powers, but its members nevertheless were not puppets. Regarding representation of the country (“the most representative government Iraq ever had,” according to Talabani at the inauguration of the IGC), the representative character of the IGC was asymmetrical and incomplete but not for that reason entirely nonexistent. On a political level, Shi’ites and Kurds were represented in the twenty-one-person council by individuals with important ties to political or religious organizations in Iraq, the earlier exile status of some of the representatives notwithstanding. But as a result of the strict de-Baathification program, the five Sunni representatives had very weak ties to important constituencies.³⁰ Conversely, most important Sunni constituencies tainted by Baath, ex-Baath, post-Baath, and in general Arab nationalist associations were unrepresented, despite repeated efforts to the contrary.³¹ The Iraqi Islamic Party, a branch of the Sunni Islamic Brotherhood, was there, but it was not an important force in Iraq, at least not at that time. Significantly, the more radical Association of Muslim Scholars was absent. Regarding issues, the religious-ethnic structure and cleavages of Iraqi society were well represented—one might say too well. It was this representation that was to make the imposition of American-style separation of church and state impossible, but it was also the same representation that in the eyes of many Iraqis helped to harden the ethnic-religious divide of society.³² From a political point of view, it is clear that the IGC was grossly unrepresentative of a society that had a significant antioccupation segment from the beginning, one that became a quickly growing majority.³³ That opinion was politically absent from the council, whatever the private views of a few members. On the other hand, there was significant resentment within the IGC, whose key members expected to be part of a provisional government, to the way the occupation was being conducted and especially to Bremer’s policies. Here there were significant differences, however, since the Kurdish members could exercise their autonomy and leadership in their own provinces and had no problems with CPA interference on that score.

The constitutional issue was a good one for the IGC to use to make a small stand against the CPA, against which they were otherwise powerless. To begin with, this was one issue where they were assigned some power in Bremer’s formal scheme of things, power they could ideally employ for gain. Second, as a group, even the Shi’ite clerics among them could play the moderates to Sistani’s hardliners, and they indeed continued to assure Bremer they could moderate and channel the Grand Ayatollah’s demands.³⁴ This was not just a tactical matter. For some of them, even before they were appointed to the IGC, contradicting Sistani’s fatwa was almost impossible unless Sistani himself opened the door to compromise. Bremer should have realized this at the latest when he appointed the Shia cleric Mohammed Bahr al-Uloum to the council, and it is quite interesting that he then assured the candidate (who would certainly report to Sistani) that he would do nothing to contravene the fatwa.³⁵ Finally, they could hide behind their expert Constitutional Preparatory Committee, and it was that body that was to take the strongest stand against the Americans, with or without encouragement from IGC politicians. Expected by the Americans to produce a report choosing a co-opted conference to draft Iraq’s constitution—and a method of selecting it—that body, with a secular academic majority, voted (initially) twenty-four to zero to go with Sistani and elect a constituent assembly.³⁶ When this was not approved (by Bremer), the Constitutional Preparatory Committee simply deadlocked around three plans (election, selection by the IGC, and caucuses doing the selection). According to Bremer, this happened because twenty members he claimed favored his proposals for unelected caucuses could not overcome the pressure from Shi’a Islamists.³⁷ More likely, however, was that the majority supported elections and did not think that they could push through that option in the face of Bremer’s resistance.³⁸ But either way, the seven-point program that required that the IGC and its Constitutional Preparatory Committee produce a method for drafting a constitution came to an end without Sistani able to push through his alternative solution.

It is tempting to analyze this ending as a step toward a supposed compromise solution in terms of a simplistic four-actor model, where ultimately American moderates reached out to deal with Iraqi ones. In this scheme, Bremer would be the hardliner who stuck to his guns and kept trying various ways to ram through the original, top-down, seven-point program. For him, Sistani’s opposition seemed to be only about getting a Shi’ite majority at a “convention,” and that could be arranged in a selection process. But even he was slowly realizing that Sistani was not amenable to any version of that, though his idea that somehow the cleric “operated on a different rational plane than we Westerners” did not let him grasp why.³⁹ As the very reliable Chandrasekaran describes it, National Security Advisor Rice, her advisor Robert Blackwill, and Secretary Powell came to the conclusion that Bremer’s plan had shipwrecked but that waiting for free elections would extend the formal American occupation for an unexpectedly long period. Thus an entirely new approach was needed. In order to facilitate this, the Iraqi Stabilization Group was formed under Rice on October 6, 2003, bypassing the Pentagon, and it was that group that at the latest between October 27 and 30 told Bremer to go back to the drawing board. There was also pressure from the Pentagon for an early transfer of sovereignty to the IGC or a body appointed by it without a constitution, a position entirely unacceptable to Bremer.⁴⁰ Back in Baghdad, his advisors (again according to Chandrasekaran) had already prepared the alternative of promulgating an interim constitution, by the CPA and through the IGC, under which free elections could be held.⁴¹ This idea had the virtue of blocking a Pentagon proposal of transferring sovereignty to the IGC without a constitution, an idea Bremer adamantly and rightly opposed, but it required at the same time that the IGC accept it and this time get at least Sistani’s grudging acceptance. One carrot for the IGC, including its Shi’ite members, was that it would be the IGC itself that would write the interim constitution, with some expert help, not some new body created who knows how, producing the constitution according to who knows what procedure.⁴² Another carrot was that this CPA-by-way-of-IGC constitution writing was a way for IGC members to enter the interim government, since sovereignty would not be given to the IGC, even in an expanded version. The interim constitution would be called a “fundamental” or “basic” law or a “transitional administrative law” (the final term had not yet been decided on), so as to avoid a term with negative implications in Iraq⁴³ and most likely not revive the Sistani bloc’s earlier objections against imposed constitutions. When the plan was approved in Washington, the so-called November 15 Agreement was born.

1. The “Fundamental Law”

• To be drafted by the Governing Council, in close consultation with the CPA. Will be approved by both the GC and CPA, and will formally set forth the scope and structure of the sovereign Iraqi transitional administration.

• Elements of the “Fundamental Law”:

• Bill of rights, to include freedom of speech, legislature, religion; statement of equal rights of all Iraqis, regardless of gender, sect, and ethnicity; and guarantees of due process.

• Federal arrangement for Iraq, to include governorates and the separation and specification of powers to be exercised by central and local entities.

• Statement of the independence of the judiciary, and a mechanism for judicial review.

• Statement of civilian political control over Iraqi armed and security forces.

• Statement that Fundamental Law cannot be amended.

• An expiration date for Fundamental Law.

• Timetable for drafting of Iraq’s permanent constitution by a body directly elected by the Iraqi people; for ratifying the permanent constitution; and for holding elections under the new constitution.

• Drafting and approval of “Fundamental Law” to be complete by February 28, 2004.

2. Agreements with Coalition on Security

• To be agreed between the CPA and the GC.

• Security agreements to cover status of Coalition forces in Iraq, giving wide latitude to provide for the safety and security of the Iraqi people.

• Approval of bilateral agreements complete by the end of March 2004.

3. Selection of Transitional National Assembly

• Fundamental Law will specify the bodies of the national structure, and will ultimately spell out the process by which individuals will be selected for these bodies. However, certain guidelines must be agreed in advance.

• The transitional assembly will not be an expansion of the GC. The GC will have no formal role in selecting members of the assembly, and will dissolve upon the establishment and recognition of the transitional administration. Individual members of the GC will, however, be eligible to serve in the transitional assembly, if elected according to the process below.

• Election of members of the Transitional National Assembly will be conducted through a transparent, participatory, democratic process of caucuses in each of Iraq’s 18 governorates.

• In each governorate, the CPA will supervise a process by which an “Organizing Committee” of Iraqis will be formed. This Organizing Committee will include 5 individuals appointed by the Governing Council, 5 individuals appointed by the Provincial Council, and 1 individual appointed by the local council of the five largest cities within the governorate.

• The purpose of the Organizing Committee will be to convene a “Governorate Selection Caucus” of notables from around the governorate. To do so, it will solicit nominations from political parties, provincial/local councils, professional and civic associations, university faculties, tribal and religious groups. Nominees must meet the criteria set out for candidates in the Fundamental Law. To be selected as a member of the Governorate Selection Caucus, any nominee will need to be approved by an 11/15 majority of the Organizing Committee.

• Each Governorate Selection Caucus will elect representatives to represent the governorate in the new transitional assembly based on the governorate’s percentage of Iraq’s population.

• The Transitional National Assembly will be elected no later than May 31, 2004.

4. Restoration of Iraq’s Sovereignty

• Following the selection of members of the transitional assembly, it will meet to elect an executive branch, and to appoint ministers.

• By June 30, 2004, the new transitional administration will be recognized by the Coalition, and will assume full sovereign powers for governing Iraq. The CPA will dissolve.

5. Process for Adoption of Permanent Constitution

• The constitutional process and timeline will ultimately be included in the Fundamental Law, but need to be agreed in advance, as detailed below.

• A permanent constitution for Iraq will be prepared by a constitutional convention directly elected by the Iraqi people.

• Elections for the convention will be held no later than March 15, 2005.

• A draft of the constitution will be circulated for public comment and debate.

• A final draft of the constitution will be presented to the public, and a popular referendum will be held to ratify the constitution.

• Elections for a new Iraqi government will be held by December 31, 2005, at which point the Fundamental Law will expire and a new government will take power.

I first called the above “Agreement” a Bush-Sistani compromise, but in fact it was not exactly that. Clearly, on the American side important concessions were made. Bremer himself considered the move from a one-stage process of constitution making to a two-stage one and the abandonment of a longer occupation before free elections to be a significant diminution of his plans (“at the outer edge” of American interests in Iraq) and a reduction of the chances of success. Yet whatever the new plans were, they were the CPA’s plan in their own mind, their “plan B” so to speak, and Bremer makes fun of A. Pachachi, who presented them as his own to the IGC.⁴⁵ Things were different on the other side. First, as great as Sistani’s political role in producing the new approach in general was, his actual role in approving it was minimal and ambiguous. Bremer tells us that the new formula was checked with Sistani before it was taken to Washington, and that he approved.⁴⁶ We don’t know what was actually shown to Sistani, but whatever it was could not have been the final version. In the original version, an interim constitution would have been drafted by March 2004, elections for a Transitional National Assembly would have been held by the summer of 2004, and an Iraqi government would have been chosen by that assembly.⁴⁷ While it was not explicit, the interim constitution itself, despite some obvious problems of circularity, could have then been approved by the elected Transitional National Assembly. It would have been illogical, perhaps, to do this, but from the sociological point of view very useful nevertheless. (Sistani himself was to demand exactly such approval, as we will see.) Be that as it may, it was this plan more or less that was sent to Sistani for approval via Talabani, then president of the IGC. It is also clear that it was this plan he approved, if he approved anything, as reported by Abdel Mahdi on November 14, who himself found out about very significant changes in the plan only at that moment or on the next day.⁴⁸

The first version of the plan was changed in Washington on November 11 at an NSC meeting at the behest of Rice and Powell, who doubted the feasibility of elections by June. Late summer 2004 would be “a little too close to another election.” Bremer and his team then returned to the caucus option, in order to save the new framework.⁴⁹ While telling the NSC principals that Sistani was probably supportive of the plan according to Hakim and Jaffari (the leaders of SCIRI and Dawa) and that this was being double checked, neither he nor anyone else noticed (or said anything if they did notice) that this could have been the case only with respect to the version involving earlier elections.⁵⁰ This was obviously not some minor matter, however, but rather the very issue around which Sistani organized his whole challenge. The governance team in Baghdad was extremely unhappy when they found out about the change, Bremer admits.⁵¹

In fact, the altered final version was not checked with Sistani at all. Moreover, it was rammed through despite objections, after a discussion of a mere two and half hours, with four dissenting votes from Shi’ite representatives. The previous night, Bremer refused to describe details of the plan.⁵² Obviously, the arrangement was based on the support of the rest of the IGC, something that was not achieved for the previous seven-point program. However, as far as Sistani and his followers were concerned, the new plan had essentially been forced through the IGC. This was certainly a valid assessment, given the alterations, the short time period for discussion, the fact that the plan was leaked to the press before Sistani’s group had a chance to examine it, and the threats from Bremer that accompanied the whole charade.⁵³ So much for the Iraqi “talks about talks.”

Structurally, as already argued, there was a lot about the agreement that was indeed a compromise, allowed by the adoption of the two-stage model and the plan for two constitutions, one to be produced the American way, the other to be produced Sistani’s way. This compromise ultimately seemed to favor Sistani, because it was his constitution that was to be the final one. The question, however, was about how these two stages were to interact. In Bremer’s later recollection, “our objective would be to embed these points [basic principles of individual rights, federalism, and checks and balances] so thoroughly in the interim document that they would stand a chance of surviving into any permanent constitution.”⁵⁴ If successful, this view of the project would vitiate the compromise, because although there would be a freely elected constituent assembly, it might have nothing important left to do with respect to producing the final constitution. With different means perhaps than via the seven-point program, the project of imposed constitutionalism would have remained.

Before the new project could succeed, there was one big hurdle to overcome. On November 26, the Ayatollah Sistani denounced the November 15 Agreements and renewed his call for free elections:

First of all, the preparation of the Iraqi State (Basic) Law for the transitional period is being accomplished by the Interim Governing Council with the Occupation Authority. This process lacks legitimacy. Rather the [Basic Law] must be presented to the [elected] representatives of the Iraqi people for their approval. Second, the instrumentality envisaged in this plan for the election of the members of the transitional legislature does not guarantee the formation of an assembly that truly represents the Iraqi people. It must be changed to another process that would so guarantee, that is, to elections. In this way, the parliament would spring from the will of the Iraqis and would represent them in a just manner and would prevent any diminution of Islamic law.

He added: “Perhaps it would be possible to hold the elections on the basis of the ration cards and some other supplementary information.”⁵⁵

A closer look at the November 15 Agreement ultimately explains and in part justifies his rejection. Yet what was agreed upon was also to an extent Sistani’s product. The new fatwa insists on the version of the plan that was apparently first presented to him and adds only the requirement of the ratification of the TAL by the elected assembly. However, was this not a possible return to the one-stage model after all? At the same time, the highly ambiguous status of what was agreed upon—or rather imposed by the CPA—had to do with the simultaneous presence of the innovative two-stage model, which was the result of the clash and implicit compromise of two projects and the attempt to guarantee the original goals of one the projects, that of the occupying power imposing rather than genuinely negotiating the fundamental terms. These contradictions and Sistani’s (imagined and actual) response to some of them are worth examining.

The problem of authority arose because the CPA only made the November 15 Agreement (which was to be the foundation of the whole subsequent process) with its own creation and agent, the IGC—which was, legally speaking, an entity authorized entirely by itself.⁵⁶ In terms of international law and in particular UN SC Res. 1511, which recognized the CPA (point 1: temporarily exercising sovereignty) and the IGC (point 3: embodying sovereignty) as somehow jointly sovereign, this procedure could arguably be interpreted as legally valid. Its political legitimacy in Iraq was an entirely different matter. There was no question here in this externally imposed revolution, any more so than in other revolutions, of legal legitimacy in Weber’s sense. The legal order was dramatically ruptured. In indigenous revolutions, the instance that authorizes the drafting of Kelsen’s “first constitution” can typically rely on revolutionary, democratic, or charismatic forms of authority, but these were also unavailable to the parties to the November 15 Agreement. Finally, as I have shown in chapter 1, as a foreign occupier, the CPA’s claim of authority as the liberator of Iraq was always weak, and it became weaker every day the occupation continued without solving fundamental problems such as employment, public services, and especially physical security.

Thus, from the internal Iraqi point of view, the CPA-IGC agreement rested on brute force alone, on facts rather than norms. Given this state of affairs, it is striking that this agreement assigned almost all the crucial tasks, and not only ultimate authority, to the two contracting parties themselves. The task of creating the unamendable body of rules, referred to in the agreement as the “fundamental law” but elsewhere as the interim constitution or the “transitional administrative law,”⁵⁷ was given to the current IGC itself, “in close consultation with the CPA” (heading 1).⁵⁸ This was strictly speaking not a violation of UN SC Res. 1511, which called for the IGC’s Constitutional Preparatory Committee to prepare for a constitutional conference that would draft a permanent constitution. The IGC-drafted constitution would be interim or transitional and would not be called a constitution. However, the fact that it was to be unamendable by more representative bodies, including even the elected convention, was a remarkable postulate, given the low legitimacy its enactment (and the agreement itself) would have in Iraq. Other interim constitutions, for example in Hungary, South Africa, and now Nepal, have had relatively easy amendment rules.⁵⁹

Further important tasks were delegated to the selfsame agents, notwithstanding their low levels of authority and legitimacy in Iraq. Agreements concerning security—that is, the future role of U.S. military forces—would be agreed upon by the CPA and the IGC by March 2004, three months before the transfer of power (heading 2).⁶⁰ After a misleading introductory statement to the contrary, the IGC was given a crucial role along with the CPA in choosing an “Organizing Committee” in each “governorate,” which would select the caucus of notables of that province (“Governorate Selection Caucus[es]”). The latter would pick that province’s representatives to the Transitional National Assembly, to which the CPA would transfer sovereignty in June and that would elect a transitional executive (headings 3 and 4). Finally, the very procedures for the adoption of a permanent constitution by a freely elected (“no later than March 2005”) constitutional convention would be provided by the fundamental law produced by the CPA and the IGC by February 28 (heading 1). It is worth noting here one unusual aspect: the dates of the first and subsequent free elections (by December 31, 2005; heading 5) were already provided by the agreement, without waiting for the fundamental law that by this wording could bring the dates closer but not delay them.⁶¹

Those accustomed to coherent legal hierarchies immediately noticed the irksome problem with all this: instances of potentially higher political legitimacy—the supposedly representative caucuses, the transitional assembly, and the freely elected constitutional convention—are here authorized by an instance, or two of them jointly, with little legitimacy in Iraq. They were even forbidden to amend or revise the instance that would hold them in thrall, the fundamental law.⁶² The issue was not just an abstract one, and it may indeed have played a role in Sistani’s reservations. When at the intermediate stage of the process the Transitional Legislature and later the Transitional Executive would have been formed, these bodies could easily have considered themselves more legitimate than the process that created them and the original source of authorization, the will of Iraq’s conqueror. They could have been tempted to do what the Estates General did in France at the famous tennis courts: namely, sever themselves from their source of authorization, declare themselves the representatives of the Iraqi nation, and proceed to establish new rules for constitution making or even give Iraq a new constitution themselves. The Transitional Legislature was especially likely to do so, since it was, paradoxically, supposed to be sovereign and yet was denied any role in the constitution-making process. The fact that it was not even allowed to amend the interim constitution under which it was supposed to operate for a year and a half could easily have led to the rejection of the document as a whole.⁶³ But if the new, handpicked legislature subverted the planned procedure in any way, that would have compromised the concessions to Sistani, who would find it much more difficult to oppose an Iraqi legislature than the foreign CPA.⁶⁴ Given the projected timetable, this would have happened presumably only after the American elections, when Sistani’s probably effective call for massive Shi’a protests would no longer have worked on the occupation forces, which would be on hand to protect their clients.

It is also true that the constitutional convention, if and when elected, might also not have considered itself bound by the procedural rules of the interim constitution—or the constitutional decisions of the interim legislature, for that matter. But that would have happened only (if elections for a constitutional convention took place at all) at a later stage of the process, when new power relations may already have been frozen. Nevertheless, this danger also existed, especially if the legitimacy of previous stages of the process remained questionable.

Note that here the choice of terminology mattered a great deal. Sistani and his circle seemed to have in mind, at least by implication, a constituent assembly as in Iran and Iraq in the early twentieth century. Indeed, since there would be no other legislature, in their model the body to which sovereignty would be transferred after free elections could only be a classical, European-type sovereign constituent and legislative body with the plenitude of powers. The fact that Sistani’s fatwa called for a referendum to ratify the constitution confirmed that he was operating with the European radical democratic conception, where the only recourse against usurpation by the drafting assembly is an “appeal to the people.” The November 15 Agreement, however, spoke of a constitutional convention (heading 5).⁶⁵ In the American (and Latin American) tradition, “convention” denotes a legislative body with the one and only function of drafting and proposing a constitution, and it exists side by side with a normal legislature that continues to function as such. We have no reason to assume that the drafters of the agreement were ignorant of this distinction. They thus wished to devalue and possibly provide a way to control the freely elected body conceded to Sistani⁶⁶ by providing an interim legislature that would hold sovereign powers during the whole drafting process.⁶⁷ They did say, however, that the “Fundamental Law,” that is, the interim constitution, would expire only with the elections of the new Iraqi government under a new, permanent constitution. A convention in American theory is neither an organ that exercises sovereignty nor is it the foundation of the executive power. Thus the implication was that both the unelected interim legislature and the executive branch rooted in it would stay in power until then. If so, only two outcomes were imaginable: either the interim government would be able to control the constitutional convention, or it would not. In the former case, the concession to Sistani of democratic elections for the constitution-making body would turn out to be meaningless. But if the latter happened, Iraq would have a classical situation of dual power,⁶⁸ where only force can decide the issue, as in Russia in 1917, in 1918, and again in 1993.⁶⁹ The U.S. example of the Confederation Congress peacefully submitting to the extralegalities of the 1787 Convention is entirely unusual.⁷⁰ More common are the examples of Argentina under Perón and Venezuela under Chávez, where originally American-style conventions claimed and successfully asserted full sovereign powers. In Iraq, it is hard to say which assembly would have won this type of conflict, which could have become especially intense because of the probably quite different social bases of one body chosen in carefully regulated provincial caucuses and another by democratic elections. But it is understandable that Sistani would have wanted to avoid such a confrontation—in other words dual power with civil war—as its very possible outcome, no less than outright rule by forces he suspected.

No major player seems to have doubted in November 2003 that forces supported by Shi’a clerics could win free elections for a constitutional assembly under just about any democratic electoral rule.⁷¹ Sistani’s demand for such a procedure, however, came not only out of self-interest. He had good reasons to believe that Iraqi politics suffered from a dramatic legitimacy problem, and it was to his credit that he sought the answer in democratic legitimacy.⁷² In his view or that of his advisers, an interim constitution such as the one proposed by the November 15 Agreement failed this democratic test. They could not have been taken in by the shifting name “fundamental” or “provisional administrative” law, or even “Law of Administering the Iraqi State for the Transitional Period,” and they knew even if not immediately (who knows what they were first told or shown!) that an interim constitution was actually at issue.⁷³ It is possible that Bremer and the CPA made a huge mistake, as Peter Galbraith argues, in not projecting a very minimal document able to take the country to free elections for a constitutional assembly.⁷⁴ For that, perhaps the small amount of legitimacy they could draw on would have been enough. But if so, the mistake was natural, because as Galbraith very convincingly continues, “the Bush administration was desperate to leave its mark on Iraq’s constitution and the TAL was now its only chance.”⁷⁵ But when it came to a constitution under any name, Sistani’s position was that it could gain its validity only from a freely elected assembly. Moreover, his advisers had to understand, given Iraqi precedents, the high likelihood that an interim constitution would become a large part or even the whole of the permanent one. In the entirely reasonable view of Sistani and his advisers, such an outcome—here too subverting the apparent concession to their side—would be intolerable, especially in the case of an imposed interim constitution or its creation by an unelected body controlled by the Americans. Thus they retreated to the idea that, whoever drafted the interim constitution, at the very least in order to be valid it had to be ratified by a freely elected assembly. The first version of Bremer’s “plan B” was thus marginally acceptable for them, but only marginally, because ratification is not the same as drafting. But when the United States backed away from a freely elected transitional assembly, the interim constitution too became unacceptable.

Nevertheless, the idea that an interim document can gain its validity from an assembly elected under it is circular and therefore procedurally flawed, and there are important substantive considerations against it. Procedurally, free elections can be held only under some framework of rules that are not exhausted by an electoral law. If democratic legitimacy means electoral legitimacy, as it did for Sistani and his followers, this framework must come before there are elections. There can be no purely electoral beginning to a democracy.⁷⁶ Moreover, conditions of free electoral competition require basic civil and political rights and rights of access to all relevant media. These arrangements need enforcement as well. When one begins to produce interim “organic rules” that contain these, however, there is a tendency to produce something like a detailed interim constitution. So let us assume that an interim legislature is elected under an interim constitution and proceeds not to ratify it, a possibility if Sistani’s demands were met. Either the country still remains under more or less the same interim arrangements and the point becomes irrelevant, or, more likely, the interim legislature is forced to become a constituent assembly, which is what Sistani wanted to begin with for the first freely elected legislature. The two-stage process would collapse into a single stage, without the assembly being limited in any way.

It is worth noting that the idea of legitimating the interim constitution through a broad social agreement of the main political forces, reaching well beyond the IGC, was just as foreign apparently to Sistani as to his American opponents. Neither side had a reliable proposal for generating legitimacy for an interim settlement that would have binding force (and possibly enforcement, probably international) with respect to the subsequent stages of the constitution-making process. While Sistani was right in insisting on the original formula offered to him involving an elected transitional assembly, his new proposal still missed the essence of the two-stage formula he helped produce.

The CPA drafters, on the other hand, were right to propose an interim constitution, something that never occurred to Sistani’s side, which did not fully understand this instrument in its contemporary form. Even the choice of a nonsovereign convention bound by preexisting interim rules had a lot to recommend it. Here I come to three substantive desiderata that may be compromised if free elections were held without prior, legitimate agreements on interim arrangements: the rights of minorities, federalism, and (relatively consensual) procedures for the making of the permanent constitution. In terms of substance, Bremer was right to worry about these desiderata. But he chose the wrong and I think strategically ineffective procedure to guarantee them.

Of course, neither Sistani’s sovereign constituent assembly nor a freely elected ratifying assembly that turned down an interim constitution could be bound in these three or any other substantive respects. But the problem was that neither could the freely elected constitutional convention of the November 15 Agreement, if it chose not to be bound by instances of weaker or entirely missing legitimacy. Whatever the interim constitution would enact, such an assembly, functioning through majority rule, could indeed legitimately if not legally reject establishing Western-type protections for minorities, possibly strong federalism, and very likely an independent, strong, and secular constitutional judiciary needed to enforce these rights. It is easy to speculate about the disasters that would result, which in the political and military sense may not differ much from the consequences of the clash of jurisdictions by assemblies also implied by the CPA proposal. Whatever Sistani and his circle wanted, once they had an overwhelming majority they might not have been able to control their own most radical tendencies unless self-binding occurred previously under conditions deemed legitimate.⁷⁷

Let me repeat the paradox. Both the choice of a nonsovereign convention and of interim minority and regional protections along with an independent judiciary, all present in the November 15 Agreement, were intelligent and innovative. The crucial point, however, remains that neither the procedural nor substantive parts of the agreement could stand up to the claims of a breakaway transitional legislature or constitutional convention. In other words, an interim constitution resting on little more than American fiat was open to repudiation from the first moment that either a transitional legislature of a freely elected constitutional convention or assembly were formed. This meant that it could not be used in a manner that would risk such repudiation by the electoral majority. For the Allawi government, which was to operate under the TAL, this was to mean the loss of the freedom of action needed to deal with the Sunni insurgency.

As we know from previous negotiated transitions, the alternative to the imposition of interim arrangements was not immediate free elections but a genuine and comprehensive historical compromise on interim arrangements. By excluding not only remnants of the Ba’ath (perhaps justified, as far as top echelons are concerned) but also all Arab nationalist parties and Sunni and Shi’ite radicals from the IGC and therefore the process leading to the agreements, Paul Bremer deprived the outcome of even this type of pluralistic rather than democratic legitimacy.⁷⁸

What Sistani found in the November 15 Agreement was in any case not a program of historical compromise but rather a top-down model for creating a transitional legislature and, much more importantly, an executive created by it, both based on co-optation and, potentially, clientelism.⁷⁹ The agreement’s most obvious flaw was indeed its formula for constructing an interim legislature that would be the repository of political sovereignty for a year and a half. Here journalistic attention has focused mainly on Sistani’s demand for a freely elected body, rather than the formula he opposed. It was one eminently deserving strong opposition. The caucuses the CPA and the IGC had in mind evidently could not resemble the direct democratic procedures of, say, Iowa—open caucuses in which citizens (or members of each party) can participate if they wish. The Iraqi provincial caucuses were meant to be closed meetings of notables rather than open ones of citizens. The last thing the Americans could have permitted is the voluntary participation of those most intensely concerned, perforce the most militant. Moreover, the provincial and local councils that would receive a role in organizing them were all picked by American military commanders, who still provided protection, and their independence was questionable at the very least.⁸⁰ What would have been especially problematic, however, was the ability of the current IGC and CPA to control participation. The IGC did receive, despite inconsistent denials, a formal role in selecting members of the caucuses, and, because of the three-stage process, its members could easily wind up being picked, as specifically permitted. In each governorate, the IGC was to pick five individuals to serve on an “organizing committee”; five others were to be picked by current provincial authorities, and five more by the authorities of the province’s five largest cities. Note that the provincial and city authorities were clients of the CPA. In each province, the organizing committee of fifteen would then pick the notables for the Governorate Selection Caucus, and the latter would elect the province’s representatives in proportion to its population.⁸¹ In case the method was not foolproof, the interim constitution (produced by the IGC and the CPA) would regulate eligibility on these three levels and undoubtedly exclude unstated categories from participating in any of them.⁸²

There was thus every reason to believe (even if some members of the IGC did not, at least according to outside testimony) that if the proposed procedure had been followed, the interim legislature would have been composed of friends, clients, and even members of the IGC. Nothing would have stopped such a legislature in that case from reappointing the council, its leading figures, and the ministers and state secretaries they had already named to every important executive position! The post–November 15 call of a few IGC members to preserve their own body contrary to the agreement arose either from an excess of caution or, more likely, a transparent ploy to deflect from the expected state of affairs.⁸³

The Americans (along with the part of the IGC closest to them) continued to adamantly oppose either free elections in June or postponing the transfer of sovereignty. The November 15 Agreement put off elections until March 2005.

But why did they not just accept the Sistani demand for free elections? The U.S. administration’s and CPA’s stated reasons have to do with a lack of time, relevant laws, and adequate voter rolls. Time was mostly a function of the law and the voter rolls, as far as they were concerned. Admittedly, it would have taken time to negotiate electoral, party, and media laws, but it would have taken no more time to impose them than it did to write the November 15 Agreement. Moreover, the United Nations could have been asked to write laws at least for the first election, as it eventually was, avoiding another American imposition. It has further been demonstrated in November by Iraqi Ministry of Planning officials that a census could have been completed by the summer of 2004.⁸⁴ Other plans had been offered as well. A single-district (the whole country) PR (the rule eventually adopted for the first elections) would not even have required a prior census of the population. So the CPA argument then shifted toward security considerations. It was alleged that the insurgency in the Sunni heartland in particular would depress participation there, and this could lead to an unfair result, in the case of a single-district PR (presumably favoring the Shi’a, who needed no extra help with their probable 60 percent of the voters). But a territorially based PR electoral system (the rule adopted for the second elections), with representatives assigned according to population and not participation, would not have had this unfair result. If elections were by district, as in the United States or United Kingdom, low turnout in some districts would not affect their representation. To be sure, both systems would have required a rough prior census for accurate apportionment of seats, but as was demonstrated in 2006, this was possible.

Thus, it was difficult to avoid the impression that the Americans opposed free elections now because a dramatic Shi’a victory that summer would have made mincemeat of the American justification for the intervention in Iraq, in an American presidential election year. It would indeed have been a tough sell to the American electorate that all of the country’s sacrifices and illegalities had the purpose of putting yet another Islamic government in power, the friends of Iran no less. Four of six justifications listed by Bremer for opposing Sistani’s demand, jotted down supposedly in late November 2003, all focus on the fear of a Shi’a victory.⁸⁵ The most organized group, the Shi’a Islamists, would win, and in Sunni districts, the winners would be perhaps the Baath in some incarnation. Giving in would encourage further radical Shi’ite demands, who would now appear superior to the CPA and the IGC. The Sunnis (to whom he was not reaching out anyway) could not be brought into the process if the Shi’a were seen as dominant.⁸⁶ What all this amounts to is that the CPA was clearly aiming at democracy with a certain outcome, excluding the possibility that one of America’s many enemies come to power in Iraq, and most American commentators are not disturbed unduly by this contradicto in adiecto. Democracies, however, are systems where any party (here including the Americans) can lose elections.⁸⁷ Postponing elections just because one’s side would lose them is certainly the wrong way to begin a democracy. It was never very clear how postponement could do anything else than delay the inevitable result.

Nevertheless, even if the U.S. administration’s and CPA’s rationalizations and actual reasons for opposing free elections before the transfer of sovereignty were highly questionable, nevertheless free elections too soon could have been a bad idea.⁸⁸ This was the view of knowledgeable UN officials (though not de Mello, in the beginning). If there was a serious reason to oppose early elections for a constituent assembly, it had to do with ensuring ample time to organize public discussions and participation regarding the meaning of a constitution and the choices involved in creating one, and to establish fundamental rights, including minority and regional rights, which could fall by the wayside in a majoritarian constitution-making effort.⁸⁹ The issue was not, in my view, primarily that of time but rather of the fundamental difference between two- and one-stage constitution-making processes. A two-stage process would be more time consuming, of course, but it need not have been as delayed as in the November 15 Agreement’s formulation. However, it would have had to allow for social negotiation and compromise concerning the fundamental rights of individuals, minorities, and regions before the majority got its say. Moreover, a two-stage process could have involved enforceable limitations on the powers of a constitutional convention, which would be impossible in a one-stage variant involving a sovereign constituent assembly. But this could not be achieved by the illegitimate formula of an imposed interim constitution and co-opted interim assembly.

Interestingly, when making the case for delaying elections, the Americans did not focus on the problem of personal security. This may have been because the November 15 Agreement is itself particularly vulnerable on this question. Granted, the Fundamental Law was meant to incorporate a variety of equal freedoms, including political, minority, and due-process rights along with an independent judiciary to enforce them. A gaping hole in the agreement, however, was its lack of rights against the power that was likely to monopolize (or dominate) the means of state violence, namely the U.S. military authorities. The transference of sovereignty (illogically) had to do only with the dissolution of the CPA, the civilian arm of a coalition that is fundamentally a military dictatorship. The agreement (heading 2) spoke of security arrangements to be made between the CPA and the IGC that would give coalition forces “wide latitude” to provide for “safety and security” for the indefinite future. Sistani had good reason to believe that such an agreement could be made only by fully legitimate representatives of the Iraqi people, and undoubtedly he and his followers would not have considered arrangements made before the end of March binding. But the fact is that U.S. troops were likely to stay in Iraq for some time; they were engaged in counterinsurgency as well as police work, and therefore rights of assembly, association, press, and speech as well as due-process rights had to protect private and public actors from the American forces if they were to protect them at all.⁹⁰ No really free elections could be held unless the relevant rights were secured against all capable of repression—and that included the coalition forces, who held in November 2003 between eight thousand and fifteen thousand people, according to different estimates. None have been charged, and, given the likely presence of many noncombatants among them, astonishingly few have been released. The human-rights situation in Iraq was thus dismal, and it is difficult to see how the Fundamental Law planned would provide any remedies. While repression was then mainly directed at the Sunni militants, there was understandably much suspicion concerning this state of affairs on the Shi’a side as well. The November 15 Agreement, along with its projected addendum on security, did not recover Iraqi sovereignty in a meaningful way, and Sistani’s call for free elections and approval (or rejection) of all interim documents only by a freely elected assembly was thus also meant to be a more effective formula for its recovery.

Any illusion concerning the viability of the November 15 Agreement as it stood was dispelled a few days later, when on November 26 the Ayatollah al-Sistani himself denounced it and renewed his call for free elections. The new fatwa demanded two things, the creation of a transitional assembly by election and not by co-opted caucuses and the approval of what he recognized as an interim constitution behind the phrase “basic law” by such an elected assembly. Sistani, in other words, was not, at least explicitly, going to battle against the two-stage process that was in effect his unintended achievement, but rather, his fight was against an illegitimate process of drafting an interim constitution and the establishment of an illegitimate legislature. In both interrelated cases, his opposition was about democratic legitimacy. Nevertheless, we can assume that he counted on a Shi’ite majority in any freely elected legislature and thus, as the best-case scenario from the point of view of the two-stage process, the prior framing of an interim constitution that he at the very least could live with. After all, as in the case of all ratification, the process of drafting should be affected by knowing who will be the agent that will have to ratify the document. (As I already explained, there was also a worst-case scenario, which would arise when an elected assembly—now the repository of sovereignty—rather than the electorate was the agent of ratification, and it rejected the interim constitution. This would have meant the reemergence of a one-stage process dominated by a sovereign constituent assembly!)

How far ahead Sistani was thinking we do not know. How his opposition to the November 15 Agreement came to pass has been described by at least two authors, but I do not fully buy these accounts. According to Bremer as well as Chandrasekaran, it was the SCIRI leaders Adel Abdel Mahdi and Abdulazziz al-Hakim who went to Najaf to gain Sistani’s opposition to the scheme on November 25. Sistani obliged but surprised them by not simply denouncing the caucuses but also advocating instead an elected interim parliament that supposedly SCIRI did not want at that time.⁹¹ I am not sure why they would not have wanted such a thing, but the claim here is that they preferred the IGC simply creating a co-opted legislature directly, an idea Sistani did not support. What is hard to believe is that they expected Sistani to support that particular preference if they helped sell him the whole scheme with elections for a transitional assembly in the first place. Did they also think that he was only interested in a Shi’a majority, as perhaps arranged by the IGC, a solution that Bremer would have been glad to offer much earlier? I think rather that they knew Sistani was going to be furious about the change of plans and they wanted to rally to his side—a very smart political move.

According to Bremer, everyone was betrayed by Sistani, who “moved the goalposts” because he had already accepted that the interim government would not be selected by free elections.⁹² But this charge is nonsensical by Bremer’s own account. As I have shown, at the time when Sistani was first consulted about the new interim arrangements, the scheme still involved free elections for the interim parliament, and it was only afterward, on the demands of Washington—and reluctantly as far as the Governance Team in Baghdad was concerned—that Bremer and his close advisors gave up the idea. If the goalposts had been moved, the Americans had moved them. What that move allowed Sistani to do, admittedly, was rethink his attitude to the proposal, in particular concerning the enactment of the basic law. It may be that he assumed that the elected transitional legislature was supposed to have something to do with enacting it, though this would have been both a mistake and illogical on his part. Thus the dropping of the electoral formula would have been doubly wrong in his eyes. Moreover, even if the transparent cover of a basic or fundamental law for an interim constitution was not going to fool Sistani long, it may have fooled him for a brief moment. It soon became clear that this was not a matter of agreeing only on a few common principles, as Bremer warned Talabani right in the beginning, with the latter quickly adding the need to include federalism in the interim document. Thus everyone knew they were engaging in a subterfuge, and the responsibility was theirs and not Sistani’s when the real intentions became clear.⁹³ What the Americans gained thereby was a determined enemy of what came to be known as the TAL.

The most important thing about the November 15 Agreement for the Americans was the timeline. Formal exercise of sovereignty had to be restored by June 2004, in time to affect the U.S. elections.⁹⁴ As a result, two strategic tracks had to be followed roughly at the same time: (1) dealing with Sistani and (2) beginning work on the interim constitution, now to be called the Transitional Administrative Law. There was no time to do these things in order. As a result, the second battle with Sistani, which ended through mediation and which will be the last theme of this chapter, would not be the last one. Just as a new deal was fashioned with him concerning the problem of unelected caucuses and the timing of elections, an interim constitution was being drafted (as I will describe and analyze in the next chapter) that reaffirmed his suspicions and reinforced his doubts concerning the process of which he now became a significant outside actor.

The mediation in question had to be accomplished by another prestigious outside actor, one whom Sistani was willing to see, and only a high representative of the United Nations would obviously qualify for the role. The Secretary-General and his ambassador, L. Brahimi, who were chosen for the task, were opponents both of the war and the Saddam dictatorship, as well as of any neocolonial aspects of the occupation, including the authoritarian steering of the political process, but they were, as we have seen, also on record with the opinion that a transitional society, in particular Iraq, should not be rushed into new elections.⁹⁵ Brahimi, quite critical of the IGC, which had now disappointed Sistani as well, was thus perfect for the job of convincing the latter that early elections were not possible. Yet his mission would only be a success if he could also convince the Americans that in return for such a concession from Sistani, one that he actually attained, a more legitimate process of drafting the TAL than the in-house procedure concentrating on the old IGC and some additional American and Iraqi experts was needed.⁹⁶ Neither this concession nor a convincing process of picking the interim government was arrived at after several consultations.⁹⁷ In the end, the only thing Sistani got in return for accepting late elections was the dropping of the idea of a transitional legislature along with the caucuses he objected to. But this left the other problem, the legitimacy of the TAL, entirely open.

UN officials were entirely clear concerning both the actual shortcomings of the November 15 Agreement and the potential problem with the TAL’s legitimacy, and they did not hide their views.⁹⁸ As Jamal Benomar put it: “instead of a detailed interim constitution, it would have been wiser to develop a consensus among all Iraqi stakeholders before producing a general set of principles, consistent with international law, to guide the transition.”⁹⁹ But these same officials sincerely believed that elections should not be held early, and by the time they arrived in Iraq on February 6, 2003, they found it difficult to shape issues other than that one. Although they had ideas regarding how the overall process could have been significantly improved, they found it difficult to introduce them into the negotiations at this late stage. The TAL was already being written in a matter they did not like, but that could be altered—if at all—only in open conflict with both the IGC and the CPA, which the United Nations would not risk. Since such a conflictual perspective was no longer viable, the UN officials adjusted their priorities, and since what they objected to most in the November 15 Agreement was the caucuses choosing the interim government,¹⁰⁰ they concentrated their fire on this bete noir of Sistani, hoping to arrange a tradeoff. They succeeded, but this was to be their only success. Upon mediation, Sistani gave up on the early elections and Bremer let go of his caucuses.¹⁰¹ There would be no elections before the ones for the constituent assembly, and there would be no transitional legislature of any kind.¹⁰² But since now there was no plan at all for picking a transitional government, Brahimi converted a part of what he had brought with him, namely ideas for a national conference, a round table, or a suitably expanded and pluralized governing council, into options for selecting a transitional executive.¹⁰³ These plans would have been more suitable for negotiating the interim constitutional arrangements but perhaps also could have served in the more restricted capacity. In fact, as the UN report makes clear, in line with an aspect of Sistani’s last fatwa, either a national conference or a round table could have amended and then approved the fundamental law, if it were given an amendment structure as opposed to the November 15 arrangements. But these ideas were all opposed by the IGC in either case, and given the uncertain support to any of the options by either Brahimi or Bremer,¹⁰⁴ none of them were picked, leaving the issue of choosing a transitional government to whom sovereignty would be transferred unsolved for the drafters of the interim constitution, who were not given the power to solve it. Among American policymakers, an option favored especially by R. Blackwill, namely that Brahimi should come back and in consultation do the job himself, vaguely came to be established as the only possible option if the IGC’s direct role was to be avoided.¹⁰⁵ The issue of the legitimacy of the fundamental law was placed on the back burner, even though UN officials predicted that the actual drafting (in part by the CPA) and enactment (by the CPA), arguably violating international law, in secrecy and without public consultations of any kind, was likely to give rise to a serious crisis in a “postdecolonization era.”¹⁰⁶

Before the Iraqi constitution-making process settled on what I would call a pathological or at least deficient version of the two-stage postsovereign model of constitution making, other possibilities were tried out, at least ideally. First came the model of top-down imposed constitution making, which was justified, if at all, on realistic, liberal, or formal-logical grounds. This model was an insult to the occupied country, neoimperialist even in its most sympathetic liberal guise, and reminiscent of the colonialist’s white man’s burden. No one would think it appropriate today for a European country, the United States, or even a U.S. state, and it was not appropriate for Iraq, where even in the 1920s the British used a far more democratic method, at least in appearance. The Ayatollah Sistani’s successful challenge to the implementation of the top-down model calls to mind Paine’s famous slogan: “The constitution of a country is not the act of its government, but of the people constituting a government.”¹⁰⁷ While enlightened monarchs, reforming dictators, and colonial powers have historically imposed constitutions on passive populations and weak political forces, the legitimacy and therefore stability of such an enterprise is highly doubtful. But what democratic procedures would today satisfy Paine’s maxim?

As I have repeatedly argued both here and elsewhere,¹⁰⁸ Sistani’s choice, the revolutionary democratic European formula involving a sovereign constituent assembly with the plenitude of powers and, supposedly, purely democratic beginnings, has become dangerous and undesirable today, especially in deeply divided societies. Such a democratic beginning under nondemocratic conditions is logically impossible, because free elections themselves require a prior, quasi-constitutional political framework. The choice of the first electoral rule cannot be the result of an election that presupposes it. A sovereign constituent assembly is possible but undesirable. It practices exactly what Carl Schmitt called sovereign dictatorship: nothing stops it from making itself permanent but its own good will or strong popular resistance.¹⁰⁹ In addition, a majority in a sovereign assembly cannot be stopped from or even faulted for adopting majoritarian decision rules. Such an assembly, not dependent on the votes of or agreements with minorities, may very well choose to adopt few or no protections for them. It has little reason for not adopting its first preference among constitutional models, and that preference may not be democratic in a broad sense of the term. Again, only self-restraint or external resistance could control a sovereign assembly’s majoritarian tendencies, but the first factor may be absent, and the second, if present, could lead to massive repression or civil war.¹¹⁰ We are accustomed to thinking that majority rule in a constitutional democracy needs to be limited by fundamental rights. This recognition needs to be extended to the constituent process, and it speaks against the sovereign constituent assemblies of the European revolutionary tradition.

Grudgingly, I have to admit, either for self-serving reasons or because of a genuine concern for the rights of minorities and women, the CPA group around Bremer saw these problems in Sistani’s model even if they had not studied comparative constitutional theory and history—as they evidently had not. When they were forced to abandon the imposed, liberal alternative to the democratic populist proposal and chose in its place the two-stage paradigm with the centerpiece of a interim constitution, they had an opportunity to substitute a legitimate, democratic model recently tried out in several countries for Sistani’s model, in essence one of two French revolutionary models. But they did not seem to know anything about the history and character of the model they were adopting. Thus they grafted on what they knew or were instinctively familiar with: the American model of a convention that admittedly also dispenses with a sovereign unlimited instance of constitution making. But that requires the coexistence of a constitutional convention specializing in constitution making alone with an ordinary legislature in which the executive is rooted and that deals with all other political tasks. In America, where the constitutional change was from a republic to a republic, legitimate legislatures inherited from the previous constitutional arrangement were available and could be used. In Iraq, this was not the case. Thus the CPA spent an inordinate amount of time in trying to put together a long-term interim legislature before Brahimi, unfamiliar with American constitutional assumptions, made them call off the quest.

In fact, in most places, and especially during transitions from dictatorships, the American formula of a nonsovereign constitutional convention has preconditions that may make this option either impractical or dangerous. Even where there is some kind of inherited legislature, as in Russia in 1993, the specter of dual democratic legitimacy and dual power may surface with explosive results. That was, as I tried to show, the danger inherent in the November 15 Agreement’s formula. Even if the drafters of this agreement had been more careful and ended the tenure of the Transitional Legislative Assembly the moment the constitutional convention was convened, they would have had no way to block either of these bodies from violating or changing the rules delivered by instances less legitimate than themselves.

It is important to note that the concession that eliminated the caucuses and along with them the Transitional Assembly also converted the constitutional convention of the American formula into a constitutional assembly of the European and indeed, the early twentieth-century Iranian, Iraqi, and Indian type—an assembly that doubled also as a regular parliament. Note, however, that the November 15 Agreement was not abandoned but rather altered in yet a new compromise. The survival of the interim constitution implied potentially, and in the TAL quite explicitly, that the constitutional assembly though freely elected would nevertheless not be a sovereign one, checked only as Sistani wanted by a subsequent referendum. It was now the contemporary two-stage postsovereign paradigm that was adopted, but since the interim constitution itself was still to be imposed, that paradigm was destined to be implemented in a pathological version. Here the first point of the Agreement on Political Process of November 15, by far the most important one, remained entirely unaltered.

The UN officials who had the comparative knowledge to understand all of the implications also had the final opportunity to straighten the matter out, and unlike the Americans (who should have had the motive but it seems lacked the knowledge) and Sistani (who may or may not have had the knowledge but lacked the motive),¹¹¹ they understood what needed to be done and were committed to doing it. But unlike Sistani and the CPA, Brahimi’s team lacked the power. Or, more precisely, they had a little power, given the fix in which the Americans found themselves, but the timing was wrong. As I will show in the next chapter, by the time they arrived in Iraq, on February 6, it was probably too late to seriously alter the formula and the timetables of November 15, which were linked to the American elections. Round tables and national conferences would have meant delays in producing the interim constitution, and to Bremer’s bosses in Washington, timing and speed became crucial. The choice was either to take a big stand and risk being able to do nothing or to concentrate on the things that could be certainly accomplished. Being diplomats, the UN officials chose the latter option. In my view, they should have tried the former, but I am no diplomat. In any case, marginal corrections and successful short-term mediation of the conflict between the CPA and Sistani could not significantly improve the process or do better than delay the next battle over the interim constitution. The process remained pathological, and when the interim constitution was actually crafted, its contradictions were to have serious consequences. Iraq was to miss its best if not yet its last chance for state rebuilding through constitution making.

FOR THE GOVERNING COUNCIL:	FOR THE COALITION PROVISIONAL AUTHORITY:
______________________	______________________
Jalal Talabani	L. Paul Bremer;
______________________
David Richmond