Documents & images

Now available from Amazon.
Now available from Amazon.
Now available from Amazon.

Building Federal Subunits by Way of Referenda: Special Challenges for Iraq

By Reidar Visser (http://historiae.org)

9 June 2006

Parallel to the tortuous political process that is being recounted every day in media reports from Iraq, a much more low-key dialogue on the Iraqi constitution is taking place – under the leadership of the United Nations, and with a special focus on the revision of the constitution and the concomitant process of implementing federalism. The next conference related to this political track will be in the Spanish capital, Madrid.

There are indications that many international experts involved in UN work with Iraq consider Spain as an ideal model for Iraq. The UN Office for Constitutional Support as early as in June 2005 gave the Spanish constitution the place of honour on its website www.iraqiconstitution.org (now defunct). In subsequent dialogues with the Iraqi constitutional committee, international experts hired in as consultants repeatedly emphasised the advantages of a constitutional system in which different regions could achieve federal status with different tempi – precisely one of the principal distinguishing characteristics of the Spanish charter. Today, there are some conspicuous parallels between article 115 of the Iraqi constitution, on the formation of federal regions, and article 143 of the Spanish constitution.(1) Both stress the importance of local initiatives in the demarcation of the building blocks in a future federal system. Whereas a more traditional practice for federalising states is to merely convert existing administrative units into federal entities, these innovative constitutional mechanisms have been designed to unlock administrative legacies of the past and to open for fresh configurations: in the case of Spain, it allowed the country’s provinces to combine in new ways to form new regions (known as “autonomous communities”); in Iraq, the idea is that existing governorates can amalgamate into new entities in accordance with their own wishes.

By choosing this interesting and ultra-democratic path to a federal state model, Iraq has also exposed itself to special challenges and risks. In fact, in comparative perspective, it is highly unusual for federations to let their inhabitants have a say in the delineation of the subunits in the federal system. The majority of federations in the world are instead of an “evolutionary” type, which have materialised as unions of pre-existing political units (for instance Switzerland), or on the basis of imperial remnants from earlier ages (including such widely differing cases as Russia and Micronesia). A second group of federations are those that have been deliberately “designed”, often after a period of political upheaval and regime breakdown. Examples of this include post-war Germany, South Africa after apartheid and Ethiopia in the democratic era. Here too, the popular input to the process has been highly limited; most often “experts” or elite politicians have drawn up the maps, with the role of the public at large often reduced to voting “yes” or “no” to a complete constitutional package.

There are hybrids of these categories as well. Many Latin American federations, for instance, started on the basis of imperial remnants but witnessed shake-ups in their federal maps after periods of internal political upheaval (Venezuela) or because of special security arrangements implemented during times of war (Argentina, Brazil). There are also some “evolutionary” and “designed” federations that have incorporated certain more democratic features with regard to future administrative changes – as seen for instance in Brazil, Germany and Ethiopia. But in practice, the administrative changes that have been effected in accordance with these constitutional devices have been quite minimal.(2) And finally, there is the interesting case of Russia, which since 2005 has embarked on a process of rationalising its convoluted federal system, partly through special legislation, and theoretically with a considerable element of local initiative. But early reports suggest that in practice the democratic spirit of this process has been considerably tempered by the central government’s tight control of local governors – with whom the initiative for federal mergers technically rests.

Unions and imperial remnants: Evolutionary federations Hybrids with evolutionary and design elements

Top–down elite-controlled design in constitutional package or in gradual subdivision: Designed federations

Element of popular initiative in the demarcation of federal subunits Options for popular initiative in future administrative changes










St. Kitts-Nevis












South Africa








Even the Spanish parallel, on which international experts apparently pin their hopes as a model for Iraq, may not be as instructive as many think. Whereas theoretically, the Spanish people were to have a decisive influence in structuring their new federal world after the authoritarian rule of General Franco, there were important limits to the supposedly unshackled articulation of the popular will when the new federal system was established in Spain between 1978 and 1983. On paper, it all looked beautiful: a genuine grass-roots process, starting with the municipalities. But in fact, many features of the new federal Spain were rather abruptly agreed on by an elite of politicians in the spring and summer of 1981, after political crises had threatened to reverse the democratisation of the country. And more importantly, in the end, the process was to a considerable extent influenced if not dictated by administrative divisions of the past. Surely, Spanish politicians quibble over some of the new subdivisions that have emerged after 1978 (and the “genuineness” of their historical roots), but there is no escaping the fact that a comparison of the “new Spain” with administrative maps of earlier times reveals some striking similarities and shows that the basic features of today’s federal map, if not all the details, are older than is often thought – a majority of today’s regions were in fact candidates for regional autonomy also in the 1930s, before the Spanish civil war.(3) In other words, in delineating its federal structure, Spain was aided by history to a larger extent than the voluntarist and blank-slate spirit of its much-lauded constitutional clauses on regionalism would seem to suggest.(4) Even some of the Spanish regions commonly dismissed as “artificial” have in fact quite substantial historical roots – as can be seen for instance in the case of Cantabria.

Historical congruence in Spain: To the left, regions and provinces of Spain, from Hugh Thomas: The Spanish Civil War (1961). To the right, the regions of contemporary Spain

These two aspects of the Spanish experience – the impact of a modicum of elite consensus, and the considerable role played by historical legacies in shaping the current federal system – are of crucial significance to the current Iraqi debate. Achieving elite cooperation has been a problem in Iraq, so can the Iraqis at least expect to find a helping hand in history? The answer to this must be “no”. Iraq does not share Spain’s degree of correspondence between past administrative structures and contemporary regionalist ambitions. True, for much of the twentieth century, there was in Iraq a relatively fixed system of 18 administrative units, but the idea of simply transforming those units into federal states has repeatedly been dismissed by the Iraqis themselves. And the earlier administrative legacy of the area that makes up Iraq similarly does not offer obvious blueprints that correspond to today’s discussions about Iraqi federalism. To the extent that there was a degree of stability, it rested in the three Ottoman provinces of Basra, Baghdad and Mosul, although the two first ones were often ruled as a single entity. But today, only a minority among Iraq’s Shiites – principally those of the far south in the triangle of Basra, Nasiriyya and Amara – are interested in re-establishing the borders of the old Ottoman vilayet of Basra. Others are discussing a large-scale, sectarian unit extending from the Gulf to Baghdad (for which there is no historical precedence), whereas there are also many Shiites who see no need for federalism at all. In the north there is similar incongruence between the long-established Mosul province of the past and the ethnically defined Kurdish regional entity that has emerged since the 1990s. But in this case there is at least a certain degree of continuity with regard to the Kurdish demand for self-rule, which in one form or another dates back almost until the establishment of Iraq after the First World War. Thus, only the northern part of Iraq exhibits a relatively well-defined regionalism comparable to those localisms that form the backbone of today’s Spain.(5)

Historical incongruence in Iraq: Ethno-religious map of Iraq, with the old Ottoman provincial borders for Mosul, Baghdad and Basra indicated in red. The boundary line between Basra and Baghdad ran further to the south than is commonly thought and effectively bisected the Shiite population; this is often misrepresented in modern-day sketch maps purporting to show the administrative divisions of the Ottoman era.

The lack of congruence between past and present regionalisms – as well as rivalry between competing contemporary federal visions – are likely to be brought into play during the implementation of federalism in Iraq with a force quite unlike anything experienced by Spain, simply because of the great differences with regard to historical regional stability. A key problem in this regard is the salient article 115 of the Iraqi constitution – which includes no mechanisms for dealing with a multitude of conflicting federal visions competing for the same areas. One tenth of the “concerned” population, or a third of governorate council members in the “concerned” governorates may demand a referendum for achieving federal status for “their” areas. But this could create a chaotic situation in places like Basra, where there may well be disagreement on what exactly constitutes the “concerned area”. A third of the governorate council may wish to transform their area into a single-governorate oil-rich region; another third may prefer amalgamation with neighbouring Maysan and Dhi Qar in a slightly bigger entity; and the remaining members may favour a bigger single Shiite entity and unity with the Shiite spiritual heartland around Najaf and Karbala. On top of this, even if public opinion should crystallise around one particular vision, the current constitution does not contain any provisions that will prevent discontents for mounting new challenges, through new referenda.

The danger is that situations like these could create chaos and chronic administrative instability. They could play right into the hands of those who wish to derail the entire process of establishing a new democracy in Iraq. It should be remembered in this regard that it took Spain no less than five years to create its federal map between 1978 and 1983; this despite the stabilising role of the monarchy, the relatively benign international climate, and not least a considerable degree of consensus on what the new map should look like. The Spanish debate on the relationship between the centre and the various autonomous units – with their different statuses and ambitions – is still going on today. Can Iraq afford a similar and probably even more time-consuming process? Might not this come at the expense of basic security and public services?

One possible remedy is to try to make article 115 of the constitution more robust, so that it can deal with a greater number of eventualities. This should be of interest to all Iraqis, regardless of their preferences for any particular federal scheme, simply as a precautionary measure against the danger that federalisation might create further political instability. Some help is in fact at hand in the Spanish constitution which Iraq has emulated in other areas. The Spanish charter establishes a five-year moratorium for any failed regionalist projects; this is a simple device which could provide at least a minimum of respite during the transitional period. Another feature of the Iraqi constitution worth looking at is the requirement for calling a referendum, especially the stipulation that a third of the members of a provincial council can make a referendum initiative. This seems rather overindulgent. In a political climate dominated by militias, a faction of that size is exceedingly easy to create. It will be open to manipulation, by Iraqi political parties from outside the region “concerned”, or even by foreigners. Tighter rules in this area will be needed to achieve stability; again the Spanish precedent may offer possible alternatives. Here the initiative for creating regions is vested in the municipalities, which due to their sheer numbers may be more difficult to manipulate.

But beyond the ongoing federal puzzle-solving activity, there is also a second and more radical option – that of accepting the historical dissimilarities between Spain and Iraq. That would mean sealing off all entrances to the federal labyrinth south of the Kurdish highlands. This could be done permanently or at least temporarily; in any case the goal would be to make a determined effort to focus energies on what most Iraqis demand most: security, jobs and general stability. In the absence of those basics, many Iraqis will continue to consider federalism debates as an abstract elite pastime with minimal relevance to their own situation. They will see it as a paradoxical political game which merely serves to postpone the rehabilitation of a country that has the potential to be one of the richest in the Middle East – if its politicians could only move on from political squabbles to actually governing the country. And while some Shiites increasingly refer to security as an argument for federalism, there is also the fear that the federal process could easily drag on and thereby deliver the opposite result. Before implementation, there would have to be constitutional revisions, legislation for implementing federalism and referenda – if all went according to plan, that is. Probably there would be some kind of federation in the end, but perhaps, as the Iraqi saying puts it, only “after the destruction of Basra”.(6) Over the past year, Iraq’s politicians have globetrotted and have been globetrotted extensively in search of the perfect federal model; it is to be hoped that after Madrid they will spare a few glances for their own history before putting the finishing touches to the planned revisions of the Iraqi constitution.

Suggestions for further reading:

Levy, Karl (ed.): Italian Regionalism: History, Identity and Politics (Oxford: Berg, 1996).

Moreno, Luis: The Federalization of Spain (London: Frank Cass, 2001).

Morrow, Jonathan: “Iraq’s Constitutional Process II: An Opportunity Lost”, United States Institute of Peace Special Report 155, 2005.

de Silva, K.M.: “Sri Lanka: Ethnic Conflict and the Search for a Durable Peace, 1978–1999”, Ethnic Studies Report vol. 17 no. 2, 1999.

Visser, Reidar: Basra, the Failed Gulf State: Separatism and Nationalism in Southern Iraq (Berlin: Lit Verlag, 2005; New Brunswick: Transaction Publishers, 2006)


1. Article 143 of the Spanish constitution reads, “(1) In the exercise of the right to autonomy recognised in Article 2, bordering provinces with common historical, cultural, and economic characteristics, the island territories, and the provinces with a historical regional unity may accede to self-government and constitute themselves into autonomous communities in accordance with the provisions of that title and the respective statutes. (2) The initiative for the autonomous process belongs to all the interested deputations or to the pertinent inter-island body and to two-thirds of the municipalities whose population represents at least the majority of the electorate of each province or island. These requirements must be fulfilled within a period of six months from the first agreement adopted on the subject by one of the interested local corporations.” (The text is greatly inspired and in parts verbatim copied from articles 11 and 12 of the 1931 Spanish republican constitution and as such an interesting relict from a radical and fiercely anti-clerical document written at a time when “self-determination” was in vogue in international politics.) Article 115 of the new Iraqi constitution says, “One or more governorates shall have the right to organise into a region based on a request to be voted on in a referendum submitted in one of the following two methods: (a.) A request by one-third of the council members of each governorate intending to form a region. (b.) A request by one-tenth of the voters in each of the governorates intending to form a region.”

2. In Germany, for instance, the main change has been that in the south-western part of the country (Baden Württemberg) which was specifically arranged for in the special article 118 of the constitution. On the other hand, the “normal” route to administrative changes – as laid down in article 29 – has not generated any flurry of popular initiatives for changing the structure of the German federation. Italy, which is normally not considered a federation, is also interesting in this regard, because it theoretically allows for municipal, bottom–up initiatives in the redrawing of provincial boundaries. But as in the cases of the federations with similar arrangements in force, this has only had a limited impact on the administrative map of the country, which has remained quite stable in its overall structure throughout the post-war period.

3. Perhaps some of the most contested decisions concerned the “detachment” of La Rioja and Cantabria from the “historical” Castile, and the amalgamation of the latter with Leon. Similarly, parts of Murcia “defected” to New Castile.

4. It is noteworthy too that in some areas where the local right to initiative prescribed by the Spanish constitution was indeed followed to the letter, it caused problems. This was the case in Andalusia, which tried to follow the “quick track” to regional status specified in article 151 of the constitution – similar to article 143 but involving an even stricter requirement of local support for the regionalist initiative. In this case, one of the provinces involved, Almeria, at first rejected the regionalist bid in 1980 and the region was established only in 1982, after intervention by the central government and special negotiations.

5. Further carefulness must be exercised in comparisons between Spain and Iraq in this regard. In Spain there is a debate concerning the relationship between the “historical” regions, meaning those that claim to have existed for several centuries, and more recent ones. It seems reasonably clear that in Iraq’s constitution, Kurdistan is playing the role of what the Spaniards initially referred to as “historical nationalities” of Catalonia, Galicia and Basque Country, which were offered a fast-track route to federal status. But the parallel is imperfect: even most of the “new” Spanish regions (say, Extremadura) are in administrative terms older than the most well-established region of Iraq, the Kurdish one. Whereas the Spanish regions can at least be traced on maps of the past (though this did not always correspond to the primary political entities), federal regions floated in today’s Iraq like the “Region of the Centre and the South” (iqlim al-wast wa-al-janub, the big Shiite super-province) do not even have a historical name. (The same goes for partitionist nomenclature like “Sunnistan” and “Shiistan” which is employed by enthusiastic foreigners eager to see a tripartite division of Iraq). Empirical studies of the Iraqi situation before the twentieth century show that the most widespread regional name in use by the local inhabitants was in fact “Iraq” – often dismissed by modern-day analysts as a complete “invention” but a term which historically was in use at least from the Gulf north to Takrit. To the extent that there were genuine regions on a smaller scale than this, the concept of the “Middle Euphrates” (al-furat al-awsat), limited to the areas from Samawa to the Shiite holy cities, may have been the most recurrent one; at one point this also corresponded roughly to a distinctive political entity, the medieval Mazyadid emirate based in Hilla.

6. The Arabic expression is “ba‘da kharab al-basra”, often believed to originate from the Persian siege of Basra in the mid-1770s, when the promised support from Baghdad against the Persians failed to materialise until after the city had fallen. It is used more generally to refer to actions that are accomplished when it is too late for them to have material impact.

Copyright © 2006 historiae.org
This document may be freely reproduced as long as http://historiae.org is credited as the source.