Conclusion

Writing this on February 7, 2008, almost nine months after finishing the sixth and final chapter, very little has changed with respect to Iraq’s constitutional conundrum. Finally, although many months too late and therefore technically in violation of the 2005 constitution, a Constitutional Review Committee was created. And if it ever submits amendments to the National Assembly, that too will be too late, in terms of the initial six-month deadline tied to initial government formation, which occurred in early 2006! Since no one cares about these legal niceties, it is much more important to point out that the likelihood of passing amendments acceptable to the Sunni parties in parliament is very small; the passing of such amendments by the National Assembly, which has already passed a law on regions in October 2006 incompatible with them, is even smaller; and ratification, given the possibility of a veto by three provinces, has almost no chance at all.¹ The constitution remains a source of fundamental and nearly irreconcilable division across Iraq’s large ethnoreligious political groupings. None of this has stopped the White House from rating progress on the “benchmark” of constitutional amendments as “satisfactory” nor prevented President Bush from repeatedly touting the new Iraqi constitution as one of the great achievements of the war and occupation. Nor has the U.S. Senate² been stopped from enacting a nonbinding resolution to the effect that Iraqis should adopt exactly the system that the benchmark of a new constitutional agreement was designed to somehow alter: the decentralized, partitioned “federation” of three entities on the model of Kurdistan, allowed but not required by the 2005 constitution, as the U.S. Senate seems to think.³

There has now been some learning on the American side. Toward the end of my last chapter, I wrote: “The amazing thing is that at least Secretary of State Rice and Ambassador Khalilzad, if not the Pentagon as well, in the end probably came to understand that they were guarding Iran’s prize and that many of the Sunnis they were shooting at were their geopolitical allies. But there is nothing they can do about this as long as they also want to stay in Iraq….” I was only wrong in writing the last phrase. General Petraeus and Ambassador L. Crocker have shown that they could do something, namely accept the overtures of the tribally based Sunni fundamentalist forces, who eventually grouped under the name of the Anbar Salvation (or Awakening) Council (Sahawa al Anbar), to join the Americans to fight Al Qaeda in Mesopotamia. These forces (who have undoubtedly killed numerous Americans) are also allied against the friends of Iran, and are thus enemies of the government that supposedly rules in Baghdad. It was this move, probably much more than the military surge, which has led to a perhaps temporary decline in the level of violence in Iraq in the second half of 2007. But the costs are also potentially very great. The United States has now helped to arm yet another militia that cannot be absorbed by an Iraqi government but is itself impotent to become that government or even the main force behind it. Its relation to older Sunni forces (the Iraqi Islamic Party, Sunni Endowment, Association of Muslim Scholars, etc.) being unclear, it is not even strong enough to fully control the Sunni-majority provinces.

This takes me back to the issue of constitutional learning. Using the older deals and amendments to the 2005 constitution are surely not going to help in reconciling the three main groups or bring the new Sunni extraparliamentary actor into the process. As late as February 2008, even a new de-Baathification law and a law dealing with oil resources have not been fully achieved. A deal on federalism is much more remote. Thus a new renegotiation of the procedures to deal with the problems would be required, especially because of the appearance of this powerful Sunni actor, which has more American support than any before. But once again, the actors should have learned not to approach that matter in the same old way, which led to disaster—to do so, according to Einstein, is the sure sign of insanity. For the American actors, “sanity” would mean that finally they understand that compelling the Iraqis without a timetable for leaving is impossible. Whether that lesson has been learned will now depend on the American electorate. The two Democratic candidates still standing have more or less the right approach even if so far they have paid little attention to what should happen in Iraq when we leave; the one Republican has the wrong approach. He wants to stay indefinitely, whatever happens in the political process. He thinks of Iraq as a war zone, and he would have the power to keep it one, indefinitely, if we let him.

As to the Iraqis, the lesson they need to learn is that in a strongly divided but culturally and socially interpenetrating country (whether Iraq or Israel) with differential resources, three or more sides heavily armed, and international allies, the desire to either impose a regime or to cleanly separate will lead to perpetual conflict. Only a historic compromise of all the major forces and the real sharing of power on many levels can lead to the way out. Whether this lesson has been learned in Iraq (or Israel) is doubtful.

I now come to what I myself have learned from this study—I dare to hope I learned quite a lot. What I did not learn about Iraq can also be discerned from the preceding chapters, and undoubtedly, because of my linguistic and geographical limitations, there is much under this heading for even sympathetic critics to find. I feel more secure regarding some conclusions I came to in the areas of methodology and theory. As to methodology, I learned something about the advantages and limits of an interpretive use of comparative methods. With respect to theory, I have now come to understand much better the superiority of the “postsovereign paradigm” of constitution making, on both the normative and functional levels.

I came to the topic of the book as a comparativist and a theorist, and indeed I had no other justification (intense political concern was only a motivation) for writing on Iraq but some previous work in these two areas. As a comparativist, I faced the problem that any historian could have predicted for almost any comparison, namely that no dictatorship was ended nor any constitution made in a manner like Iraq’s. Undeterred, in my first chapter I focused on a Germany/Japan comparison as a form of immanent critique of the conception of the occupiers themselves, who believed, with Charles Krauthammer, that these old cases proved that anywhere where the United States cared enough and put in sufficient resources it could impose a political order of its choice. Right away, however, here a causal form of comparative analysis tends to break down, because the now almost universally admitted insufficiency of U.S. troop involvement meant that arguably from the beginning the United States did not care enough after all (even if it should have or could have). I register this problem, but since my concern is constitution making, I note two other major differences (along with many other historical givens). The first has to do with the problem of imposition versus autonomy. In Germany, the American occupiers sponsored an almost entirely autonomous form of constitution making, while in Japan, where they imposed a constitution, they disguised their imposition behind the legality of the inherited system. In contrast to these, in Iraq they imposed and did not hide that state of affairs. The second difference had to do with the related problem of stateness. In Japan, the occupiers maintained the inherited state, whereas in Germany, where such a state was destroyed through the war itself, they allowed the rapid, bottom-up reconstruction of autonomous provincial “statehood.” Assume then that the occupying forces were insufficient in strength in Iraq. Could following either German or Japanese state-and constitution-making patterns (state destruction and an autonomous process or state continuity and an imposed process) have worked? I believe that we will never know, and this is the point where my analysis has to avoid all strong causal claims. The advantage of the interpretive, rather than causal, method that I use is that I am still able to say something of significance, namely that if the level of force did not itself doom the whole enterprise, not following either the German or Japanese pattern was in any case disastrous, even fatal, unless some other method was developed that could compensate for the missing autonomy (Germany) or legal continuity (Japan).

That is where the constitution-making method, the two-stage “post-sovereign” constitution making that was actually used, comes in. In principle, it could have had virtues that made up for much that was lacking in comparison to the German and Japanese models. At the same time, here the framework of comparison becomes very tenuous, because none of the previous examples of this constitution-making method—Spain, various central European states, and South Africa—involved external constitutional imposition of any kind. Similarly, regarding the issue of insufficient force, one could argue that this difference alone would have led to failure in the Iraqi case. Again, we will never know, because the method itself was adopted in a deviant manner, leading to pathologies. Some of these could be traced back to the factor of external imposition, but I believe that some were strongly contingent. My interpretive method focuses on what was contingent without being able to make strong causal claims for these factors.

I do not underemphasize the deterministic dimension. If the United States could be compared to the foreign sponsor of some of the state parties in central Europe in the late 1990s, nevertheless its role in pushing for a radical transition was unique. In its own (though never in Iraqi or international) eyes, this role gave it legitimacy and a moral mission that has no parallel in any relevant country, with respect to the agency that held the major means of violence during the transition. The deformations and pathologies that followed from this role were more like those the Soviet Union was guilty of when setting up its client states in the late 1940s (entirely legitimate, but only in its own eyes) than the timid role this imperial state played in the 1980s and 1990s. Again, we cannot know if this factor in itself doomed the process, but it is certainly plausible that it did.

But at the same time there was also contingency. The U.S. authorities were free, however they interpreted their own mission, to include all Iraqi forces in political bargaining, or at least many more political forces than they did, and to avoid making a fundamental state deal, subsequently almost etched in stone, with the Kurds alone. Elsewhere, from Hungary to South Africa, broad inclusion of political forces was the key to even partial legitimacy of the inevitably nondemocratic first stage of the process, and learning from these cases would have meant insisting on broad and deep inclusion. If the built-in differences of imposition and the external role did not lead to failure in itself, then the failure of inclusion certainly did. And it was not counterbalanced by subsequent belated attempts that I treat under the failure of sequencing, again in comparison with all the other cases.

Could other deviations from the pattern followed in other countries compensate for the negative effects of imposition and exclusion? There is only one such major deviation, the role of referenda, and its consequences were in many respects rather negative. No other country using the two-stage method felt obliged to give “the people” another channel beyond that of the voice of elected representatives. To me, a final ratificatory referendum may be a fallback into sovereign constitution making, and it was adopted in Iraq only because it was already contained both in the Bonapartist proposals of the Americans and the radical populist demands of the Grand Ayatollah Sistani, the two plans that had to be reconciled in the model adopted. More specifically, it was then used to hamstring the elected National Assembly, by tying it to the default position of the Transitional Administrative Law, which would be enforced by the possibility of three Kurdish provinces easily mustering the necessary two-thirds vote to veto a final constitution. Adding to the legitimacy problems of the authorization and negotiation of the TAL, such a restriction of the freely elected assembly meant, among other things, a violation of the UN-mediated compromise with Sistani. In the end, the National Assembly was reduced to less than even a rubber stamp, and the process returned to a kind of collective Bonapartism, in which the Kurdish and Shi’ite party leaders used the referendum to confirm their joint political diktat to the population. Certainly, with respect to the excluded, even the momentary passing hope that the three Sunni provinces could perhaps vote down the constitution did not compensate for their earlier exclusion.

Many lessons are implicit in these comparisons, but it is worthwhile to highlight them, especially within the context of the “rescue operation” I discussed in my preface. Given the multiple causes for failure just indicated, it is difficult to sort out the weight of each, and it is possible that some interpreters will blame in small or large part the constitution-making method itself, which I expressly want to avoid doing. My critics could take two different and opposite approaches. The first would deny that the method of constitution making matters at all and would claim (to the extent that admitting that this problem area is relevant at all) that it was the result (i.e., the constitution that was written) that doomed the prospects of a solution to Iraq’s problems. One interpretation of this argument is that pure imposition would have been fine as long as it produced the right constitutional result. I certainly agree that the TAL’s combination of confederal/federal/centralistic and consociational formulas was incoherent, and the solutions of the not fully completed final constitution were even worse. But I would deny that this was independent of the method of constitution making adopted; indeed, most of the untenable results followed from exclusionary bargaining and private dealings with the Kurds. The learning mechanisms I analyzed in terms of amendment and ratification rules were even more directly the results of the special Kurdish role in the process. A different and less deformed process would have produced different results.

Pure American imposition could have avoided all these substantive difficulties, but only at the cost of vastly exacerbating the already grave legitimacy problems. Here the burden of proof is on those who maintain the very implausible position that even the gravest legitimacy problems connected to constitution making do not matter as long as the result is good. Of course, if imposition were successfully hidden behind an apparently Iraqi process, then we would have had a better chance of acceptance. But then the constitution-making paradigm devolves into the Japanese example—but with the total destruction of old state institutions and the failure to form a provisional government, there were no Iraqi forms behind which to hide the imposition. What if they had not destroyed the Iraqi state? Forgetting the other vast differences with Japan, for the model to work, a plausible program would have required the immediate creation of an Iraqi government, relying on a more or less professionally intact ministry of justice, all controllable by the Americans. None of this was very likely given the divisions of Iraqi society and the different relationships of its groups to the invaders.

The second line of criticism would insist that the sovereign constitution-making method, a single-stage one such as that proposed by Sistani’s camp, could have produced more legitimate results and without the incoherence of the two Iraqi constitutions. There is little doubt that a higher degree of coherence would have been achieved, but only at the expense of legitimacy. Elections for the constituent assembly would then have been won, as they were later won, by a unified Shi’ite list of some kind. Despite some efforts, Sistani never managed to become the leader of a united Iraqi nationalist challenge to the occupation, perhaps because he expected the Americans to guarantee free elections against antidemocratic forces that, initially at least, were too strong. Whatever the reason, his side (assuming away its own internal divisions) would then have been in the position to impose a constitution on something like one-half of the Iraqi population, perhaps more if we also count secular Shi’ites. While one cannot exclude some compromises even in that situation, such a constitution would have been that of the bare majority and unrestricted by any prior rule or agreement. In such a situation, without prior constraining rules, a victorious party usually has difficulty controlling its own radical elements. If a constitution favoring two of the three major groups produced havoc, another favoring only one of the three would have probably fared worse—unless the whole effort turned against the occupation, and of that there was very little sign. And even in that last case, a civil war with the Kurds would have been impossible to avoid.

Thus, a method born out of a compromise between external imposition and internal populism would probably have been the right one, if practiced in the right way. There are strong theoretical reasons for this being the case, and paradoxically two missing elements of the Iraqi formula, “legitimacy” and “stateness,” have helped me understand why much better.

For some time I have stressed that where there is no democracy one cannot begin democratically.⁴ I have also tended to add that there are nevertheless legitimate and illegitimate beginnings, and I have treated legal continuity, revolutionary legitimacy, pluralistic inclusion, and other principles such as consensus and publicity from the point of view of an initial nondemocratic legitimacy.

This is relevant to populist democratic constitution making. Its overwhelming superiority is supposed to be because in this model the sovereign people, and no one but the sovereign people, gives itself a constitution. This position is based on unacceptable political mythology, incoherent originalism, and what has been called the metaphysics of presence.⁵ Both historical analysis and logical considerations reveal that the legal identity of the sovereign people, one capable of action only within representation, is determined by prior electoral and procedural rules that must be given to the “people” by elites who thereby constitute them as a people capable of action.⁶ Hans Kelsen already argued that “the people—from whom the constitution claims its origin—comes to legal existence first through the constitution.”⁷ This idea, as I see it, contains two: (1) some kind of nonpopular beginning of popular constitutions and (2) the legal, representative character of the people within constitutions. Regarding both, one needs to distinguish the “we” that speaks words like “we, the people” and the “we” in whose name these words are spoken.⁸ The first “we” that speaks (the “actor”) is not the “people” (one cannot begin democracy under nondemocracy democratically), and the second “we” (the “author”) never acts. The first “we” claims that it is a “representative” of the people and therefore has a right to act in its name.⁹ How can this act of representation—or equally substitution and arguably usurpation, given its element of arbitrariness—by a self-designated agent be made legitimate? Advocates of the classical democratic method of constitution making (who have often criticized liberal constitutionalists in terms of their result rather than process orientation) in the end answer in terms of the result: to the extent the constitution actually contributes, performatively, to the creation of its supposed author, the unified people.¹⁰ Leaving aside the irrelevance of that ideal to a divided society like Iraq—or even to the peaceful European Union—the formula resembles that of Joseph Weiler’s biblical “We will do, and hearken” (that is, impose a covenant first and understand the divine dispensation later), which was applied to a more appropriate model of European pluralistic culture of constitutional tolerance. Here too the initial process was redeemed by the result, and asking for its initially democratic credentials is equivalent to asking an unfree people to participate in the creation of freedom, forever privileging that unfree subject as the constitution’s author.¹¹ Thus Weiler goes so far as to make heteronomy not only inevitable but a virtue as well.

Nothing like a biblical dispensation from on high occurred in Europe, but Iraq shows the consequences of thinking along lines like Weiler’s. But the model of actor-author, somewhat relativized and further differentiated, very much helps to explain the power of the new model of constitution making. Here that model becomes reflexive, differentiating its components and indicating that a legitimate version has organizational requirements. What in the single-stage model is a single set of utterances made by one speaker(s) in the name of another subject that never speaks becomes two utterances, with two acting subjects, in two distinct stages, with increasing but never complete legitimacy. Unlike the speaker of the classical populist model, the speaker of neither stage here is able to fully identify itself with the popular sovereign. And yet their legitimacy can be greater. This is so fundamentally because of the changed character of the very beginning of the process. While the arbitrariness is still there in the first stage, it does not lead to an arbitrary and potentially self-serving imposition of initial rules. What can only be to an extent arbitrary is the choice of negotiating partners and decision rules, but sociological, historical, and moral criteria of inclusion and fairness exist for these, even if an old legal order has been disrupted. Inclusion may never be perfect, but as we have seen in Iraq one can do much better or much worse, at the very least. If the first stage is accomplished fairly, what was simply an actor in the classical populist model is now bifurcated into a subject of action (with its now receding arbitrary element) and authorized agent (authorized by all or most of the political organizations of the country). The speaker(s) who write an interim constitution speak at the same time directly in the name of political organizations, and they establish a process by which a democratically authorized set of speaker(s) can be elected. Still, it is the decision makers of the first stage that constitute the people as a legal entity capable of action. Broadly understood, they are representatives not only of political groups but of the population understood in terms of politically articulated segments. Evidently, they also constitute, performatively, the population in terms of these segments, by their action if not always by the rules they generate. They thus refer to the people in two possible senses, and this can lead to problems (for example, the freezing of consociationalism or power sharing) later on. Much depends on the extent to which they are genuinely superseded by the second set of speakers, both as authors who can legitimate an unchanged interim constitution and as actors creating a new one.

The speakers in the second stage are also representatives, this time of the citizens understood in terms of universal suffrage (who in turn are themselves only representatives, since not all citizens can or do vote, and the electorate represents the nonvoters as well). Their legitimacy can be higher than that of a classical constituent assembly, because the manner in which they were chosen was less arbitrary and more inclusive. Yet in the two-stage process, the rulemakers of the first stage, who structure the process as a whole, have no interest in mythologizing the second set of speakers as identical to the people and especially as direct embodiments of a constituent power outside of law. Nevertheless, in the second stage the representatives have the democratic authority to speak in the name of the people or the citizens.¹² Because they are neither identical to the people nor seen as such, their power can, should be, and always is to an extent limited. But one must be very careful with these limitations. This is so because the constitutional actors in the second stage are authorized, unlike those in the first, from a democratic point of view, or, if that begs the question, from a universalist and egalitarian one. What has become crystal clear in Iraq is that since the role of the first stage is to limit the more democratic one, the legitimacy of both stages is extremely important. Thus it is by no means irrelevant how one initiates that first stage. The element of heteronomy, if it is logically inevitable, must be as reduced and confined as possible. And the missing democratic legitimation, while it cannot be fully replaced, must be compensated for. Understanding the nonidentity of the actors—even of the democratic stage—with the people allows the first stage to partially define and limit them. But understanding the greater deficiency of the first stage requires that the second stage not be neutralized through either impossible learning mechanisms or executive usurpation. These are the lessons from the pathological case of Iraq, lessons that could not be clearly seen where the processes worked more or less properly.

Finally, state destruction and the failure of state (not nation!) rebuilding play extremely important roles in this book. There is no need to again rehearse the “no state, no constitution” thesis here. But I would like to again stress the priority of state building and rebuilding to constitution making and even make this argument entirely general. What I have in mind is the empirical regularity (though not entirely universal) that even in revolutions surviving institutions, organs, associations, and so on from the inherited state play important roles in constitution making even where the strict legal continuity of the state has been disrupted (in domestic if not international law).¹³ This phenomenon has to do with the necessity of political integration for political agency, which goes far beyond what Arendt recognized in the case of America, namely the role of small, inherited republics in the making of the big republic. Some pouvoir constitué is always part of the constituant,¹⁴ except perhaps, very tenuously, in the ideal limiting case of what Carré de Malberg and Kelsen called the “first” constitution.¹⁵ All previous constitutions contribute to the making of the following ones. Thus we can have a juridical theory also of revolutions, something that Carré de Malberg and Kelsen thought unlikely (though the latter came up with it in his rather implausible theory of international law).¹⁶ In terms of the model of actor and author, as several contemporary analysts now recognize, in some respects these two agents or agencies must be seen not only as two parts of the constituant but also as constitué: the actor by the old state and the author (who is both an actor and an author) by the new regime. But that constitution by the old order means the survival of some organ or agency as part of the state structure, organizational, political, or symbolic.¹⁷

And this is not only because of the impossibility of beginning ex nihilo in revolutions and the need to rely on some inherited structures, institutions, organizational patterns, or groups to integrate society. The role of continuity is obviously even greater under reform, regime change, and transformations from above, the types discussed in chapter 1 along with revolution. In any significant large-scale transformation, the specter of revolution plays a major a role, and, whatever their historical reality, revolutions can in principle and in our imagination challenge the state structure and the regime. Great structural reforms are undertaken to preempt revolutions, and, likewise, negotiations of political power with important oppositional forces are undertaken to work out the parameters of regime change. In the case of the latter, both the structure of the regime and political power over the state is likely to change. All the participants need guarantees: the opposition must be confident that the existing power will not use state resources against them during the process, and the governmental forces need assurance that new incumbents will not later use the state in a repressive way. Agreement concerning constitutional rules is not enough; it is even more important that state structures be so organized that forces of violence, material resources, and population groups be so distributed and governed as to protect all negotiating partners from worst-case scenarios. Where the repressive role of the state (its administrative and military organs) with respect to some population groups plays a major role in the demise of an old regime, renegotiating the state structure so that it cannot happen again under new management is of prime importance. But to some extent, relevant questions such as the disposition of the militias of political organizations must be dealt with in all negotiations. It also follows that all those capable of materially affecting the relevant questions through the actual or potential use of violence of their own become the most important members of these negotiations, and their exclusion tends to make state bargaining at the very least difficult if not generally futile.

Once again we should be able to see a central advantage of the two-stage process of constitution making. From the point of view of the state-regime distinction, the great lesson of Iraq is that the first stage is necessarily the locus of the state bargain and the second stage can shift more in the direction of regime construction. Of course, the distinction is analytical only, and both types of issues are generally dealt with in both stages, with the added proviso that the first stage must contain both state and regime rules for the transitional period. But logically at least, the part of the first stage that should be negotiated among the main political actors controlling, or capable of controlling, the means of violence is the one that has some claims of being enshrined against the democratic will of the electorate as represented in a constituent assembly. In political life, the ethics of responsibility requires that we do not try to treat as equal those who cannot be in fact reduced to equality. However, making the distinction between state structure and regime should help in reducing the number of areas where this undemocratic element is given some of its due. Since this was not done in Iraq, in addition to the confederal state structures conceded to the Kurds, they were also granted consociational regime structures, which led to an entirely unjust arrangement from the point of view of Arab Iraqis, who can rightly say, “they have separated off their quasi-state where we have no say, but they are in a position to deadlock our political and constitutional development.”

Again, it will never happen anywhere that the first stage, even if dedicated to peacemaking, can be entirely kept away from constitutional or regime-building areas. But from the point of view of the new model as understood through the lens of Iraq, we can understand the centrality of state making in other first stages. In South Africa, for example, the interim constitution involved making important deals regarding the military forces of both sides and about the inherited administration, police powers, and federal arrangements. Not all issues regarding the state structure were enshrined in the thirty-four principles, but when they were, as in the case of federalism, the Constitutional Court wound up using its extraordinary powers even against the new dispensation of the freely elected constitutional assembly. In unitary states with more homogenous societies, I would grant that state rebuilding is very implicit, at best during the first stage of negotiations. But even in Hungary, for example, there was a contentious issue, which had to be resolved, regarding the Communist Party’s militia. More importantly, consensus about all state-related issues guarantees the continuity of the transition, which can then concentrate on regime change exclusively.

I realize now (especially after a friendly suggestion by Nehal Bhuta) that I have been concerned with state continuity even before Iraq, under the heading of legal continuity.¹⁸ Not being a follower of Kelsen, who has influenced me in a number of ways, I do not consider the two to be the same. But I have always argued that legal continuity from authoritarian to rule-of-law states has a fictional aspect. Amendment rules, for example, which were never the real rules of system change, are suddenly used for real, masking actual ruptures. If there is state continuity, then legal discontinuity, even acts of illegality (as in America in 1787, where the individual states supplied what was continuous), may be of relatively little consequence for constitutional stability. But without state continuity, legal continuity is impossible. It may be therefore true that the deeper continuity that really mattered in a Japan or a Hungary or a South Africa was state continuity, and it is another lesson of Iraq that without state continuity the stabilization of a democratic revolution or transition becomes extremely difficult. In the end, we may only have the German case to indicate that such a thing is possible, but that was under extraordinary and perhaps unrepeatable circumstances. The theoretical problem of Iraq was whether the two-stage, postsovereign method of constitution making could initiate a second major instance of democratic transformation in the context of state collapse. The challenge should never have been there to take up: first, because Iraq should not have been invaded, and second, because its state should not have been destroyed. Given that this challenge was taken up, we will never know whether the results could have been positive, because of the remarkable misjudgments and policy failures this book has documented.