2.1.7 Preparing the constitution-makers

Most constitution-making exercises have included some form of education to prepare the constitution-makers for their role. Past efforts, however, have not always been well thought out or coordinated. Constitution-making bodies tend to be offered opportunities to travel or attend workshops organized by external groups rather than educational programs tailored to meet the constitution-makers’ specific needs.

It may be necessary to prepare the constitution-makers at several stages. For example, an overview of the rules of procedure and substantive constitutional issues may be needed at the start, with sessions about how to conduct civic education and public consultation efforts coming at a later stage. A process with several constitution-making bodies (e.g., a constitutional commission and then a constituent assembly) may require different educational programs for each.

As a result of conflict, many constitution-makers may have lost professional and educational opportunities, or they may not have experience in governance. Educational programs may therefore be welcomed and even requested. However, the offer of educational assistance can be a sensitive matter. Some constitution-makers may resent the suggestion that they do not know enough to do their job. Those who have served as parliamentarians in particular may feel that they have already mastered the issues—after all, they have been directly implementing the constitution. In the past, constitution-makers have even blocked attempts to create educational programs.

These barriers can potentially be overcome in the legal framework or rules of procedure by requiring all constitution-makers to attend officially organized educational programs. Sensitivity about the terminology used when organizing these programs can also be helpful. Terms used in the language of “education”—“study programs,” “orientation,” “initiation,” “induction,” “capacity-building,” “capacity enhancement,” and the like—can be viewed as condescending. The terms “seminars” and “workshops” are often considered the least offensive for those who feel they do not need an “education.” But for the sake of simplicity, in this section we use the word “education.”

The objectives

Constitution-makers can ideally be prepared to:

  • strategically plan for or manage the process (if this is required of members);
  • understand the structure of the process and any existing mandates or legislation governing the process, as well as rules of procedure and codes of conduct;
  • agree on principles that will guide their work and allow them to understand the mandate and principles set forth in the legal framework;
  • serve the national interest as well as their particular groups, and learn how to use the process to build consensus or even engage in nation-building;
  • communicate effectively with the public and other stakeholders;
  • carry out a coordinated and effective civic education and public consultation program; and
  • gain understanding about key substantive constitutional issues (such as human rights and decentralization).

The sections of this handbook on specific issues related to the process of constitution-making (such as rules of procedure and civic education) serve as a guide to the knowledge and skills that may be needed to carry out a task. This handbook does not discuss substantive constitutional issues (e.g., human rights, the judicial system), but we can provide some practical tips and points to consider for educating constitution-makers on these issues.

Developing a program for substantive constitutional issues

The content of an educational program on substantive constitutional issues could cover a wide range of topics. General topics that might be useful are:

  • the nature and purpose of constitutions generally;
  • the possible scope of constitutions and some of the arguments about what should and what should perhaps not go into a constitution;
  • how constitutions are used—legally, politically, and in other ways;
  • the language of constitutions—technical legal language and emotive language, and when these are used;
  • the importance of the structure of a constitution—how one part may interact with another;
  • the main elements of constitutions;
  • the main variations and options for designing key elements of the constitution, such as systems of government or separation of powers among branches of government;
  • international law and constitutions; and • how the existing constitution functions, including its strengths and weaknesses.

The purpose of educating constitution-makers about these issues is not to turn them into constitutional lawyers. Some have been elected or selected because they are representatives of marginalized groups or special interests. The aim is to help them translate their goals and aspirations for their constituents into constitutional terms. To participate effectively, they will need an understanding of how a constitution can—and sometimes cannot—advance the interests or groups they represent. The constitution-makers will also need a basic understanding of the wider issues, because they not only represent their particular groups and interests but should also reflect on the national interest. For example, women representatives are present not only to represent the interests of women but to bring their perspectives and knowledge to all issues.

Most constitution-makers understand some of the constitutional issues at stake, but likely only those issues that affect them and their own communities. Members of parliament may understand the electoral system and the lawmaking process well; fewer will understand the national financial system, or human rights. Few will understand how the security forces work and how they are—or could be better—controlled. Few people—probably not even most lawyers—will have read an actual constitution from beginning to end.

Many constitution-makers will find themselves having to educate their constituents or the wider public; unless they are well prepared it will probably be a poor teaching experience. The very exercise of constitution-making, and there having been conflict or disagreement about constitutional matters, are factors likely to generate new ideas for the constitution. But unless the constitution-makers are familiar with the existing constitution they will not be well placed to evaluate those new ideas.

Similarly, while constitution-makers should be expected to understand the language of a constitution and its key concepts, they should not be expected to acquire the technical skills of someone who drafts the constitution. They should understand their own roles as well as the roles of the various other actors in the process, including the legal drafters.

Given these considerations, at a minimum, educational programs on substance should enable the constitution-makers to:

  • participate fully in discussions on all issues and effectively draw on expert advice when they may require it, recognizing the limitations of their own knowledge;
  • explain basic constitutional concepts to the public;
  • understand suggestions that are offered for the draft constitution, and be able to ask questions of those who make submissions; and
  • read and understand drafts of the constitution—in part to ensure that decisions made on constitutional issues are accurately reflected in the language of the constitution.

Constitution-makers are sometimes divided into thematic committees that are responsible for weighing options on particular substantive areas, such as the judicial system. These committees may require more extensive education on their particular topics.

Approaches to learning

Expectations for educational programs vary from one culture to another. Some cultures will be more open to participatory methods. Usually, something more than a simple lecture is needed; many constitution-makers are not used to absorbing large quantities of information through reading or listening to lectures. Experience suggests that for learning about procedures, role- playing and other participatory methods are likely to be both acceptable and more effective. Studies on university teaching show that student attention flags after about twenty minutes. Lectures should be combined with techniques such as short discussions among small groups as well as question-and-answer sessions. These may help stimulate interest and attention.

Public consultation should be held with constitution-makers to learn what they want to know in order to do their jobs effectively. Different members of the constitution-making body may have different learning needs. Some members may be illiterate or not understand the working language being used.

Whoever designs and manages the educational program for the constitution-makers should look for local adult educational resources before requesting assistance from foreign experts. Some international organizations or embassies may have expertise on particular topics, but careful coordination will be needed to ensure that any externally sourced educational programs meet the needs of the process.

The international community also may offer study tours. In late 1947 the constitutional adviser to the Indian constituent assembly visited Washington, Ottawa, New York, Dublin, and London, and met a galaxy of distinguished judges, politicians, and scholars. As a result of those discussions he proposed a number of changes to the draft constitution that he had already prepared on the basis of the various committee proposals. This was a visit by a person with intimate knowledge of what was being proposed, a lawyer who was fully able to hold his own in discussions and benefit from them.

Many constitution-makers hope that they will be able to visit other countries to learn about their systems. Relatively recent examples have included:

  • visits by the members of the Fiji Constitution Review Commission to Malaysia, Mauritius, and South Africa (chosen because of the ethnic dimension in their politics and constitutional debates);
  • visits by officials and civil society leaders from the South Kordofan and Blue Nile states of Sudan to Indonesia to learn about the latter’s experience with secession (Timor-Leste) and autonomy (Aceh) because of the provision in the Sudan Comprehensive Peace Agreement about public consultation to be carried out by those Sudanese states over their future, in the light of the Southern Sudanese peace agreement;
  • trips organized by the Swiss Agency for Development and Cooperation for various Nepali political parties, government, academics, and members of civil society to Switzerland to look at the Swiss system, especially federalism;
  • tours by members of the Indonesian People’s Consultative Assembly involved in constitutional amendment to Thailand and South Korea to look at constitutional courts; and
  • visits by Ugandan constitution commissioners to the United States and various European and African countries.

Undoubtedly some of these visits have been of value. The members of the Reeves Commission in Fiji said, “We were able to form impressions that could not have been gathered from books or papers” (Fiji 1996: 61). But there is a big difference between a trip taken by members who are already knowledgeable about the situations in their own countries, and have some background in constitutional issues or political science, and one taken by nonexperts who may have been appointed for essentially political reasons.

The authors confess to a certain skepticism about the value of many of these tours. All too often they are viewed by the participants as chances to escape from the conditions at home, or as a rare opportunity for international travel or shopping expeditions. Ill-prepared participants may have no context into which to put what they see and hear. Many constitution-makers have been taken away on such tours when they are needed to perform key tasks in the constitution-making process at home. In Indonesia, a commentator said that although the ad hoc committee preparing a draft for one of the amendment processes for that country visited twenty-one other countries, this was something of a “picnic” because the committee members were not expected to prepare a detailed report.

Those who support or offer study tours seem to like them because they are high profile, please the participants, and redirect funds to the home country, as well as offering the rewarding opportunity for them to show off their own democratic wares.

Practical tips for ensuring that study tours meet the needs of the process include:

  • try to ensure that study visits are planned in consultation with those who plan the constitution- making events at home, so that a committee does not lose members at a crucial moment;
  • choose both the target country and the institutions to be visited carefully, to ensure that they really are relevant;
  • ensure that members who go are not taken just to please important political figures, but will benefit from the experience;
  • ensure that the language skills of the participants are sufficient (or that translations will be adequate);
  • provide preparation before the trip so that the participants know where they are going and why, and how the experience might relate to their own situations at home—they should be going with a purpose, be looking for certain information, and be able to ask questions;
  • hold a postvisit meeting to consolidate what the participants have learned;
  • brief those who will make presentations so that they understand something of the background of the people they will be meeting;
  • choose presenters carefully—it may be impressive to have the visitors meet the president or the chief justice, but they may not turn out to be the best resource persons; and
  • arrange that the participants contribute what they have learned to the wider constitution- making process.


In addition to face-to-face educational approaches, it may be useful to provide constitution- makers with essential information before the first sitting of the constitution-making body (especially a constituent assembly or other large, nonexpert body). This can come in the form of a handbook and include practical logistical information about transportation, accommodations, security practices, remuneration, and the like. It can also describe what resources, if any, such as researchers or Internet access, will be available to the members.

Documentation provided should include the legal mandate, the steps in the process, the time allotted to the process, any rules of procedure or codes of conduct that will govern the conduct of the members, and any other relevant materials. Ideally, background papers on basic constitutionalism, the key features of the existing constitution, a historical analysis of constitutionalism in the country, and an introduction to some of the key constitutional issues can also be provided. In Afghanistan [2004] the secretariat to the Constitutional Loya Jirga, a kind of constituent assembly, prepared these materials with assistance from a foreign adviser.

2.1.8 Guiding principles for the process

Agreeing on guiding principles: An increasingly important task

In developing a constitution-making process, it is important for the main groups involved to reach agreement on a set of principles intended to guide the process. These are often referred to as “guiding principles” (and sometimes as “constitutional principles,” or—in a few cases— “immutable principles”).

Such principles can address both how the process is to be conducted (for example, requiring that it should be consultative and participatory) and the content of the constitution that it is expected will result. Guiding principles tend to reflect key aspects of the historical context in which the particular process is taking place, and also the broader international norms, standards, and precedents outlined below.

Agreeing on guiding principles can have a number of important benefits, particularly in a situation of conflict or transition. Such principles can help ensure that the process is transparent and has legitimacy in the eyes of the public. All groups with interests in the process will likely be better prepared and more aware of how they can participate in and monitor the progress of the process. Agreed-upon principles regarding the contents of the constitution can be particularlyimportant in conflict, postconflict, and transitional situations where there has been a history of abuse by the state and other actors; principles can provide a shared vision of a better future. They can also provide assurance to minority groups facing the prospect of a loss of powers or privileges in a majority-dominated constitution-making process.

It is not always necessary to have guiding principles—for example, where constitution-making processes develop gradually, without the parties necessarily agreeing on the details. There are cases in which attempting to agree on principles in advance may prove to be divisive.

The nature of guiding principles

In most cases, guiding principles are terms of reference for the process as a whole. They are intended to provide guidance for the process without being unduly restrictive. They are usually regarded as political obligations rather than binding and enforceable legal principles. For that reason, they are usually expressed in general terms rather than in precise and directive language.

In a few cases, mainly involving processes intended to resolve serious conflict, the principles are extended to provide detailed provisions about the process and about what must be included in the constitution, with verification required before the proposed constitution can be brought into operation. In the process in South Africa [1996], for example, the main parties involved in negotiating the transition from the apartheid regime agreed to an elected constitutional assembly that would consult the people on the final constitution, and also to thirty-four constitutional principles giving detailed directions on the content of the new constitution to be developed by the constitutional assembly. They included the form and structures of government; relationships among national and subnational governments intended to ensure local autonomy; protection of the interests of significant minorities; protection of human rights; creation of independent public institutions; and the entrenchment of the constitution through amendment processes, which were to include roles for the provinces in the case of amendments affecting their interests. All of this was included in an interim constitution, which also created a constitutional court that had the task of verifying whether or not the thirty-four principles were reflected in the final constitution.

Principles in documents that establish constitution-making processes

Such principles are usually stated in the foundation documents for a process, such as the legislation establishing a constitutional commission or constituent assembly, the terms of reference for a parliamentary committee on constitutional development, or a peace agreement or interim constitution providing for a constitution-making process intended to contribute to conflict resolution. They may also be found (in varying degrees of detail) in international agreements that provide for constitution-making processes (as in the Bonn Agreement in relation to Afghanistan and the Paris Agreement for Cambodia) and in United Nations decisions and mission directives about processes in which it plays a significant role, for example in Namibia and Timor-Leste.

Guiding principles: A common phenomenon

There has been widespread interest in guiding constitutional principles since they were used in the South African process—so much so that some people have the impression that it was the first case in which they were used. But guiding principles of various kinds have been used in many processes, both before and since the South African experience. It is true, however, that they are seldom so rigorously enforceable as they were in South Africa. Earlier examples include the terms of reference that the British government set for the Indian independence constitution- making process that began late in 1946; the constitutional conferences involved in making some other decolonizing constitutions in the 1950s and 1960s; the terms of reference that Papua New Guinea’s colonial legislature provided to its constitutional planning committee in 1972; and Namibia’s “Constitutional Principles,” agreed to by Namibian freedom fighters (the South West Africa People’s Organization) and international community actors in 1982, seven years before the Namibian constituent assembly began meeting. The Indian process provided its own principles, in the form of the “Objectives Resolution” moved by Nehru early in the life of the constituent assembly. There have also been many examples of guiding principles being used since the South African case, including the Bougainville process [2004], the Burundi process [2005], the two Kenya processes [2005; 2010], and the Nepal process [ongoing] (wherein the interim constitution of 2007 provides for a number of important principles, though they are not clearly related to the constitution-making process, as opposed to functions of the state generally). The Bolivian constituent assembly included a committee on the “Vision of the Nation,” which developed principles for the rest of the process.

Sources of principles for constitution-making processes

The design and the operation of both national and subnational constitution-making processes are increasingly influenced, and even determined, by a wide range of norms, standards, and precedents (including precedents provided by other constitution-making processes), and inevitably they are also a source for guiding constitutional principles. human rights norms are having an increasing influence on the design of processes, in particular the emerging right to “democratic participation” based on the United Nations Declaration of Human Rights (1948, article 21) and the International Covenant on Civil and Political Rights (article 25), as well as various other United Nations conventions and declarations. Similarly, a number of United Nations and regional conventions and declarations on various aspects of human rights are having significant effects on the contents of constitutions, and so also on guiding principles intended to influence the process of making decisions on constitutions. There are also less formal sources of influence, such as international community concerns about good governance, independent accountability institutions, transitional justice, and so on—all of which have an impact on the framing of options for constitutions, and so on the content of guiding principles.

Who decides on the guiding principles?

Decisions on guiding principles can be made by a wide range of actors. In the case of decolonizing constitutions, the colonial government, or the local legislature established by the colonial government, has tended to frame terms of reference meant to keep processes within limits acceptable to the colonizers. In many postconflict situations, these principles are negotiated among the parties, and are included in peace agreements and—increasingly—in interim constitutions. In many instances where the international community is involved in a peace process, it plays a major part in the negotiation or even the determination of the principles. In other conflict and postconflict situations, the parties to the peace process agree on them. In still other cases, the government in power includes them in legislation establishing the process. Sometimes the government has consulted other parties with a view to ensuring that the process is as free and open as possible. In other cases, the government may be aiming to control or restrict what other parties can do during the process.

In many cases, however, those most interested in the development of guiding principles will not be involved in, or will not be controlling, the constitution-making process the principles are intended to guide. They often seek to use the principles to reduce the likelihood of their interests being adversely affected by choices made in the process. In conflict and postconflict situations, it will usually be important to have the widest possible range of interest groups involved in the process of agreeing on any guiding principles.

Verifiable principles in conflict situations involving entrenched minorities

Sets of principles that are intended to be verified—sometimes called “immutable principles”— are not common. They are developed mainly in situations of deep conflict, such as in South Africa and Burundi, with the aim of providing assurances to a party (or parties) with deep concerns about their future (in relation to security, power, and similar issues). This is achieved by the immutable principles determining aspects of the contents of the ultimate constitution. As a result, they are often quite long and detailed, as in South Africa’s interim constitution in 1994, and in the Arusha Peace and Reconciliation Agreement for Burundi in 2000. The remarkably detailed principles in the Burundi case appeared in a protocol to the Peace and

Box 5. South Africa: The certification judgments

Many groups challenged particular aspects of the draft of the final constitution on the basis that they were inconsistent with the guiding principles for the final constitution stated in the interim constitution. According to the court, there were “47 advocates representing 29 political parties, organisations and individuals.” The case was fully argued, like any other litigation, and the court gave a reasoned 296-page decision in which it held that nine aspects of the draft constitution indeed did not meet the principles. Most important were the provisions on local government. The constitutional assembly had to revise these aspects, and when the constitution came back to the court it was certified as meeting the principles.

Reconciliation Agreement titled Democracy and Good Governance, which contained long and detailed “Constitutional Principles of the Post-Transition Constitution” that came close to being a complete constitution. In both cases, constitutional courts were given the task of verifying that the principles had been adhered to. A similar approach was used in Angola in 2010; the constitutional court certified that the constitution newly adopted by the parliament largely complied with principles laid down in the existing constitution.

An important reason for the detailed and verifiable principles in these cases was that they accommodated concerns of minority parties who had previously dominated the state and who could see that their interests were likely to be adversely affected by a new constitution produced by a process dominated by the previously excluded majority. By agreeing in advance on the detailed principles that the final constitution must adhere to, the minority groups were given strong assurances that their interests would be protected. That encouraged them to join rather than stay outside the process, reducing the risk that they would undermine the process. On the other hand, because the principles were agreed on by the parties to the conflict, and were intended to be verifiable (or enforceable), they were intended as limits on what could be decided in the course of the consultative and participatory constitution-making processes that followed the agreement on the principles.

In Burundi [2005], the draft constitution was withdrawn from the constitutional court when it failed to give a ruling in the limited time provided for by the Arusha agreement, and was instead submitted to popular judgment by referendum (the next step in the agreed-on process), where it was overwhelmingly approved. Perhaps because the principles in the agreement had come close to representing a complete constitution, the approval of the draft constitution by referendum was seen as providing sufficient verification that the principles had been adhered to.

While a careful and skilled analysis of a constitution such as the one in the South African certification case may be valuable, the results of court involvement are not necessarily beneficial. Six of the seven judges of the constitutional court of Angola had been appointed by President dos Santos; they largely approved the new constitution, which has been described as tailor-made for allowing dos Santos to remain in power.

Other cases of verifiable principles

In Namibia, the international community played a role in both making and verifying guiding principles, which were first agreed to in 1982 by the South West Africa People’s Organization, the Organization of African Unity, states adjacent to Namibia, and the United States, subsequently approved (indirectly) by the United Nations Security Council, and then adopted in 1989 by the Namibian constituent assembly. While no enforcement body was provided for in the principles, their having been approved by the United Nations was regarded as giving that body an enforcement role. In July 1990 the United Nations Secretary-General reported to the Security Council that the new constitution complied with the principles.

In Bougainville, Papua New Guinea, a 2001 constitutional settlement to a separatist conflict provided Bougainville with both the right to hold a referendum on independence, deferred for

ten to fifteen years, and a high degree of autonomy in the meantime. Autonomy included the power to make a subnational constitution providing for the structures and processes of a Bougainville government. It had to be made by a participatory process, include various democratic features, and adhere to “internationally accepted standards of good governance as they are applicable to and implemented in the circumstances of Bougainville and Papua New Guinea.” Concerns on the part of the Papua New Guinea government that the Bougainville constitution might go further than agreed or might unfairly favor particular groups were dealt with by specifying that the constitution would not come into operation until it was endorsed by Papua New Guinea, while at the same time limiting its grounds for refusing endorsement to a situation wherein it judged that the requirements of the constitutional settlement had not been met. To minimize the risk of endorsement being refused, the constitutional settlement required the Bougainville constitution-makers to consult extensively with the Papua New Guinea government in the process of developing the draft constitution.

There are other possible, less formal ways of trying to guarantee compliance with principles. They could be made the basis for the civic education of the public. They could—if brief enough—be printed on banners displayed in the venue of the constituent assembly. And they could be built into the oath of office of formally appointed or elected constitution-makers.

How are principles used?

Where principles are verifiable, there is a strong incentive for those involved in the process to take care at every stage that proposals for the constitution are taking full account of the principles. This will be important not just in the process of drafting the constitution, but also when proposals are being made during public participation efforts, or being debated in a constitutional commission, constituent assembly, or other deliberative body. In practice it may be important for a particular institution or body within the process to have responsibility for checking that the principles are being met. For example, in a constitutional commission or a constituent assembly, a particular committee could be given that task. In processes where the principles are more in the nature of political obligations or terms of reference, it will still usually be important, for the legitimacy of the process, that the principles are seen to be adhered to, and so a similar allocation of responsibilities for internal verification by those involved in the process will often be required.

Must guiding principles be followed?

In most cases guiding principles are expressed in general terms that are intended to guide rather than to restrict. Where a process is highly participatory, it would often be difficult to use guiding principles to limit choices being made as part of the process. The terms of the principles sometimes even highlight a dilemma that could arise if they were to be enforced. For example, the laws that have provided for Kenya’s two processes [2005; 2010] have required that the institutions established to undertake the processes give effect to generally laudable principles (such as accountability, accommodation of diversity, and respect for universal principles of human rights) while also ensuring “that the final outcome of the review process faithfully reflects the wishes of the people of Kenya,” with no guidance about handling situations where the people’s wishes might be contrary to the principles.


The experience of South Africa has perhaps imparted a rose-tinted hue to the question of guiding principles. While it is not necessarily a bad thing that principles be reached during a time when careful thought is possible, or that important political negotiations should shape the future constitution-making process, there is an antidemocratic quality to such principles. They are devised by those in a position of power at point A in the process in order to limit the possibilities of those who are making decisions at a later point. Those in power at point A may have their own interests in mind, and even if they are acting in the interests of the nation, their perspectives on those interests may be narrow. In Egypt [1971], President Sadat gave certain guidelines to the national assembly on the constitution, including the statement that 50 percent of the elected bodies were to comprise fellahin and workers. In Nigeria in 1988 the military president told the constituent assembly that there were certain “no-go areas”—including the federal system, the presidential system of government, and having no state religion.

Often there is no choice on the matter of principles. But if there is a possibility but not an inevitability of developing principles, the following points might be helpful:

  • principles may protect the interests of minorities who may not have much sympathy from the population at large;
  • discussion and decisions on principles may have value when developed by the constitution- makers themselves, to focus discussion on broad issues rather than getting prematurely into detail;
  • sometimes interest at a later stage focuses on “moral issues” such as abortion, and it may be helpful to have established early agreement on certain fundamentals;
  • similarly, political manipulation may come more to the fore at later stages and interest in fundamentals may diminish; and
  • it may sound antidemocratic to say so, but using principles to avoid what may turn out to be the dictatorship of the majority is not necessarily a bad thing—a major purpose of the modern constitution is precisely to limit that dictatorship once the constitution has been adopted.

On the other hand, the following may also be true:

  • principles may hold back a genuine process of reform, especially if they are determined by an outgoing regime;
  • developing principles may be time-consuming; and
  • to avoid the risks of the process being “straitjacketed,” the principles may have to be so general as not to be very helpful.

2.1.9 Interim constitutional arrangements

It has been estimated that one third of all constitutional design processes from 1975 to 2003 involved interim documents. If the meaning of the word “document” extends beyond constitutions, then the number of interim arrangements would be even larger, as some of them are based on understandings, treaties, or peace arrangements that affect the way state power is to be exercised, but are not constitutions. Several terms can be used to refer to what we call here “interim arrangements”: provisional, temporary, interim, and transitional constitutions. In many cases it is impossible to understand the design of a constitution-making process without knowledge of the interim arrangements, which are indeed part of its overall objectives, strategy, and design.

Interim and incremental: Connected but different

Incremental reforms are different from interim arrangements. They are sometimes called “minimum reforms,” though this can be a misleading term in some instances of incremental reform, such as in Indonesia, where the changes made were far from minimal. Incremental reforms are contrasted with grand reform; unlike the latter, they are piecemeal, but, unlike interim arrangements, they are not necessarily a prelude to broader reforms. They are often the best reforms that are possible in difficult circumstances, and the hard decision is whether to reject them, recognizing that this may prevent further reforms, or to accept them in the hope that they will lead to broader reforms later, perhaps as a result of the dynamics of small but strategic early reforms. Chile, Hungary, and Indonesia offer significant examples of incremental processes that resulted in significant changes to the structure of the states in question. (For some arguments against a full-blown constitution-making process, to which incremental change might be an alternative, see part 2.1.2.)

Reasons for interim arrangements

When negotiating for peace, it is important that an agreement on a cease-fire be concluded speedily, but a long-term settlement, often seen as a new constitution, dealing with the underlying causes of conflict, would need considerably more time (as in South Africa). It may be premature to start negotiations on a new political order when myriad issues normally dealt with in a peace agreement have not been satisfactorily resolved (such as disarmament, demilitarization, exchange of prisoners, demobilization or integration of armies, and resettlement of the displaced). Depending on the sequence of events, if elections are to be held before the adoption of the final constitution, considerable time would be required to establish political, administrative, and security arrangements (as in Cambodia, Iraq, and Nepal), including decisions on who is entitled to vote—perhaps involving “lustration” (in this context, exclusion of those associated with past regimes), as in Iraq, or inclusion of communities hitherto excluded from citizenship and the franchise, as in Nepal.

A constitution-making process bedevilled by mistrust is unlikely to deliver a good, or any, constitution. Trust is a general condition for the success of any process, but it is particularly important when the negotiating parties have waged war against one another until recently. It has been argued that the difficulties in the Iraqi process arose because it was rushed; there was no time to build trust or develop a vision of the country.

It is desirable to use the “transitional” peaceful period to establish trust among the previous antagonists. At this stage a consociational approach may be important (where all key groups are represented), even if it is not the intention to retain this approach for the permanent constitution (as in Iraq, South Africa, and Sudan). Interim arrangements are sometimes a way to postpone difficult issues, which are hard to resolve in the aftermath of conflict but may be easier to tackle in the future with the goodwill that may have been established in the interim period.

The overall record of such arrangements is not impressive. But in many conflict and postconflict situations, some form of power sharing seems inevitable (as in Kenya in 2007, Cambodia, Iraq, South Africa, Sudan, and Zimbabwe). Consequently, more attention needs to be given to the modalities of the partnership arrangements. (In Fiji, Indo-Fijian parties declined the prime minister’s invitation to join the government, preferring to fight the next election without the “stigma” of participation in that government, before or after the 1997 constitution was agreed to. Perhaps unwisely, the basis of that constitution was a “government of national unity,” and a partnership between its two leading architects before the elections had laid a political foundation for it that did not hold when, contrary to their expectations, their parties performed poorly in the first election under the new constitution.)

A transitional constitution may be thought necessary to provide the legal framework for the running of the country when old institutions have collapsed or the old parties have disappeared or been greatly weakened (as in Somalia). In Afghanistan, for example, the arrangements were called “emergency interim arrangements” and justified due to “the unstable situation,” in view of the time it would take to re-create the state of Afghanistan. Crucial state institutions may have to be rehabilitated urgently while negotiations on a long-term settlement proceed (as in Afghanistan, Cambodia, and Iraq). Sometimes the old constitution may be considered unacceptable, for historical or ideological reasons, to one or more previously excluded groups, even if, through amendments, political understandings, or administrative measures, these groups could be given a share in governing the country (as in Nepal with the Maoists). In some countries, one or more key institutions involved in amendment procedures may have collapsed, making it impossible to amend the constitution, as can also happen if one key institution refuses to give consent to amendment. This might well have happened in Nepal, where the consent of the king, who had been sidelined by political parties, would have been necessary.

An important reason for interim arrangements is to provide a framework within which previous enemies can share state power, to facilitate the conclusion of the peace process, and to negotiate for a new constitution. They can promote stability in what could otherwise be a period of turmoil. Such arrangements can, however, give rise to their own problems, as the parties may

see them merely as a truce, before political “war” (competition for state power) starts again. This happened in Nepal and Sudan, where turmoil and suspicions continued well after the cease- fire and the peace agreement. Moreover, decision-making under interim power-sharing arrangements tends to be cumbersome and slow, and often requires consensus.

Another reason for interim arrangements is that in the immediate period following the cease- fire, previously warring parties dominate the political scene. Most of them have probably committed crimes against humanity, and may enjoy little public legitimacy. If the country were to move quickly to the permanent constitution, chances are that these warring parties would monopolize the process (as was the tendency in the Norwegian-sponsored process in Sri Lanka). It may therefore be better that the deals they make should be temporary, so that the process for the permanent constitution commences when conditions for greater public participation and the consideration of a wider range of social issues arise. On the other hand, it has been argued that the temporary nature of the arrangements makes it possible to find more imaginative solutions and cover issues that might not be dealt with in a permanent constitution.

A new use of transitional arrangements can be found in Kenya, Madagascar, and Zimbabwe following elections the results of which have been strongly contested. (The arrangements in these cases probably involve some modifications of the existing constitutions, especially concerning the structure of the executive; the ultimate aim may be a new constitution or a full return to the existing constitution. Kenya and Zimbabwe belong to the former category.)

The negative side of interim arrangements

Interim arrangements can be used to hijack the reform process. In Kenya in the mid-1990s, politicians used interim, and essentially minor, reform to forestall more fundamental change, taking the steam out of the civil society reform movement and holding up reform for at least a decade. Nepal in 1951 was under considerable pressure to democratize the political system through constitutional reform; the king promulgated an interim constitution as a prelude to reforms by a constituent assembly. Instead the interim arrangement lasted for eight years (and was changed to return to the king his old powers). It was not until 1990 that constitutional reform got back on track.

There are dangers in a long transition: momentum may dissipate, and agreement among the parties may disintegrate. If the focus is on the long-term democratic system, interim arrangements have to be dynamic, leading gradually to more inclusive forms and more accountability.

There is also a danger that if some reforms are instituted, the movement for radical change will lose momentum. The trick is to institute reforms whose logic is further reform, thus promoting the irreversibility of reform. (This could, for example, be done by inclusion of hitherto marginalized communities in the process.)

A transitional constitution may give opposing forces time to regroup and consolidate, which may put the objectives of peace at risk.

Interim constitutions have in some instances become the broad copy for the permanent constitution. In Iraq the earnest and prolonged nature of the negotiations on some issues in the Transitional Administration Law was such that the parties must have intended them to be the ultimate solutions (especially on issues such as federalism, language, and the future of the Kurdish forces, the peshmerga).

An interim constitution may exclude particular groups or issues. Both in the way it structures the process for the permanent constitution and if it becomes the model for the permanent constitution, this may have serious consequences for the country, and for the durability of the peace and the constitution.

Interim arrangements as road maps to a new constitution

To overcome the risk that change will lose momentum, it is essential that the interim arrangements include a road map to the new constitution. Indeed, some interim arrangements are largely about the road map, especially those negotiated under international auspices (as in Afghanistan, Cambodia, Kenya, and Sudan).

When the focus is the road map, it is not unusual to find more than one set of transitional provisions, sequentially. Both Afghanistan and Iraq demonstrate this: an initial bureaucratic procedure yields to a more consultative and representative administration, which has the basic responsibility of leading the nation to a new constitution or to elections for a constitution-making body.

Box 6. Negotiating interim arrangements in Nepal

After the People’s Movement of 2006 led to the end of the king’s absolute power, and in accordance with the agreement between the Maoists and seven “democratic” parties, decisions were made by consensus, and this is how the interim arrangements were decided. (The king, though nominally still present, was left out of the negotiations.) The initial decision on transitional arrangements (involving the removal of articles dealing with the monarchy and the recall of parliament with its immediate past members) was made nominally by seven parties, in reality by fewer. This happened in the face of Maoist opposition, but increasingly it was seen as a prelude to the next set of interim arrangements, in the making of which the Maoists would play a full role. Already the parties had conceded to the Maoists’ demands that the constitution would be made by a constituent assembly. Although an expert committee, under a much-respected Supreme Court judge, was appointed to draft the interim constitution, in practice the key decisions were made by nominees of the eight parties. Nearly five thousand submissions were made by the people, but there is little evidence that much heed was paid to them. The interim constitution was enacted nearly ten months after the recall of the ad hoc parliament.

The scope of the interim arrangements depends on factors such as the anticipated length of time before the ultimate constitution is in place (the longer that time, the more detailed the interim arrangements must be); the feasibility of using the existing constitution for the time being; the discretion to be left to constitution-makers (the less the discretion, the longer the interim arrangements); and whether the interim arrangements are under the administration of the international community (in which case they will be brief—as in Cambodia and Timor-Leste).

The orientation and scope of the interim arrangements

It is clear, then, that the orientation and scope of interim arrangements depend on the context and strategies for establishing the new political order. Hence there is great variety in such arrangements. Some are brief, largely concerned with the road map, adjusting state institutions toward that objective. Some would even lack various institutions of government. Others are as detailed as a final constitution would probably be. The Nepal 1951 constitution had no provision for a legislature, but only for an advisory assembly for the king. The 2007 interim constitution of Nepal made no provision for elections. But the South African interim constitution of 1993 was complete and detailed.

Expert opinion seems to waver between those who favor a short and businesslike document with a bare minimum of content and those who argue that the interim constitution should offer considerable detail and be as democratic as the final constitution should be (in part to cultivate democratic practices and habits). The latter is not so easy given the difficulties in transition that we have mentioned. It might also introduce rigidity when flexibility is needed (particularly in volatile contexts, as Nepal discovered). It is perhaps more important to make the process itself inclusive and principled than to attempt to set forth democratic rules for the interim administration. But much depends on the context, and it is hard to be dogmatic about these matters.

Who negotiates the interim arrangements

As with other aspects of interim arrangements, there is great variety concerning their negotiation. In general they are negotiated, but in some instances they can be prescribed by one party when it is in general control of the state. The Ethiopian and Ugandan arrangements were the decisions of the governments that had captured state power. A unilateral decision is also the practice of military authorities on the execution of a coup d’etat.

In South Africa the interim arrangements were negotiated almost exclusively among several political parties. Most of the negotiations at this stage were held behind closed doors. If there were disagreements among the parties, decisions would be made by the two major protagonists, the African National Congress and the National Party, concurrence between which was described as “sufficient consensus.”

If the interim constitution is to be negotiated by internal forces, there is an obvious dilemma: if circumstances are not right for deciding on the final constitution, how much detail will it be possible to decide on—other than the actual process for producing the final constitution? For this reason, interim constitutions sometimes bear considerable resemblance to a previous constitution, perhaps with the most obviously offensive provisions removed. This was notably true in Nepal. If, on the other hand, full negotiations for the interim constitution are possible, won’t the same considerations effectively continue? And why should the final constitution be different? This was largely the case in South Africa [1996], where the detailed negotiations over the interim constitution produced provisions that were substantially reflected in the final constitution.

In recent years, the international community has played a key role in devising interim arrangements. This was true in Afghanistan [2004]. In Iraq [2005], major decisions were initially made by the United States, but the intervention of the United Nations was instrumental in reaching agreement on crucial aspects of the interim arrangements, including the road map. In the case of Cambodia [1993], major decisions were made in Paris at a conference that included key Cambodian groups, several interested states, and the United Nations. In both Sudan and Somalia Western states have played an important role.

On the whole, little space has been found for public participation by civil society. Where political parties are dominant, they may have influence on the negotiating parties (as in South Africa [1996]), but even then, the role is limited. In Nepal [ongoing process], groups that were excluded from decision-making were able to secure amendments to the arrangements after considerable agitation, accompanied by violence.

Forms of interim arrangements

There is considerable variety in the ways in which interim arrangements—meaning how affairs of the state are to be managed during the period when negotiations begin and the final settlement is implemented—have been organized. In many cases it is possible to use existing mechanisms (suitably modified, as in many transitions in Eastern Europe), while in others new arrangements may have to be created (which can consume time and energy). One factor is whether the new forces agitating for recognition can be accommodated within existing arrangements. Generalizations are hard, because much depends on the context.

A particular dilemma in structuring interim arrangements is whether to try to stick to the existing, even if discredited, constitution or adopt an interim constitution. The arrangements in South Africa [1996] illustrate several of the issues mentioned above. Although the African National Congress had fundamental moral and political objections to the apartheid constitution, it agreed to work within it for an initial phase. Its decision was motivated by at least two considerations. The first was to reassure the members of the white community that changes would not be abrupt and would not be imposed on them (since they were in charge of the amendment procedures). The second reason was to lay the foundation for the rule of law by

accepting the principle of legal continuity. (See part 2.1.6.) The “interim arrangements” dimension was part of the agreement among the parties engaged in negotiations that the government and the legislature would act in accordance with the instructions of an unofficial interparty executive council. (During this period the main legal pillars of the apartheid system were repealed by the apartheid legislature.) But even with this concession, the supporters of the African National Congress would not have accepted the extension of the apartheid constitutional and legal system. The initial interim period was therefore used to agree on new arrangements for the next phase. The new arrangements, in the form of an interim constitution, were fundamentally different from the apartheid constitution and were decisive in the move to a nonracial democratic system. They included elections to a constituent assembly, which changed the power configuration of South Africa. Interim arrangements played an important, constructive role in South Africa (in contrast, for example, to Sri Lanka, where few attempts at interim arrangements as defined here have been made other than the proposals by the Tamil Tigers— the Liberation Tigers of Tamil Eelam—which seemed designed to entrench their preferred system ahead of negotiations).

It may sometimes be possible to use the existing constitution either in slightly amended form (e.g., in Kenya in 2007) or without any amendment but with an understanding that authority under it would be exercised through joint decisions of competing groups (as in Hungary as it gradually moved away from communism). In Afghanistan an older constitution, deemed the most democratic of all previous constitutions, was adopted, but with so many modifications that it imposed little in the form of an effective framework for key decisions made during the interim constitution or on the road map.

Legality of interim arrangements

The question of the legality of the interim arrangements is likely to arise. Those opposed to the forces that come into power may challenge the legality of their actions. The safest course therefore might seem to be to operate through the existing constitution. As we have seen, often this is not possible. Others insist that new arrangements can be justified under the concept of “revolutionary legality.” (See box 4.) Unless revolutionary legality is bounded by clear principles and rules, it can easily degenerate into arbitrariness and even anarchy. The mandate of the people, which is frequently referred to these days, is too imprecise and its contents too contested to serve as the foundation for revolutionary legality.

If it is essential to establish interim arrangements outside the framework of the existing constitution, it is important for their legality that they be based on broad consensus. This consensus could perhaps be achieved through a roundtable with key groups.

Box 7. “Incremental” change in Chile

Chile is an example: after the military rule, General Pinochet lost a plebiscite in 1998 on his future as a presidential candidate, and the country moved gradually to a democratic system. Between 1989 and 2005 the constitution was amended seventeen times, until two commentators said that if the latest batch of reforms was adopted, “institutions will finally catch up with the democratic process” (Esteban and Vial 2005).

Managing the transition period

We now turn to how interim arrangements may deal with managing the period of transition until the new constitution is prepared. A key factor is whether the process is driven by local or external factors. If external, there are two possibilities: (a) the country is taken into international care and the United Nations or a regional organization takes over management of state affairs (as in Cambodia, Kosovo, and Timor-Leste), or (b) there is massive international involvement (as in Afghanistan, Bosnia-Herzegovina, Iraq, and Namibia). In the former case, power is restored to the country only after the dispute has been “resolved,” law and order established (including possible disarmament), a new constitution adopted, and elections held. In the second case the international group works closely with the national authorities (which are often of an “interim” nature) and keeps open the “seats of power” for a competitive electoral process. In Namibia [1990], for example, authority was vested in a South African administrator (sympathetic to the white-dominated faction) but his powers were exercised in close consultation with the United Nations representative, who was ultimately responsible to the Security Council through the Secretary-General. In Afghanistan [2004], the Bonn Agreement provided for an interim government that was to be endorsed by a partially elected Emergency Loya Jirga, and the United Nations was asked to provide assistance to it. In practice, the United Nations advised Hamid Karzai and provided a considerable measure of administrative support.

Cambodia’s [1996] and Iraq’s [2005] arrangements were in between complete external control and complete local control. In Cambodia, certain functions were discharged by the United Nations, particularly the organization of elections to the constituent assembly and the protection and promotion of human rights. The day-to-day administration was carried out in accordance with the decisions of the Cambodian cabinet, which consisted of various local political groups.

In Iraq the administration was at first completely under the authority of the United States, represented by Paul Bremer, an appointee of the United States president, under the general authority of the Coalition Provisional Authority, which consisted of the United States and the United Kingdom. In July 2003, Bremer appointed twenty-five Iraqis to the Iraqi Governing Council to assist him in this task. The council appointed a council of ministers and a constitutional preparatory committee. Bremer’s initiative was intended to speed up the process for the adoption of the final constitution, as Iraqi politicians were reluctant to move fast, considering that they needed time for public consultation and to establish trust among themselves.

Bremer’s plan was torpedoed by the Grand Ayatollah Ali al-Sistani, the most influential Shia cleric in Iraq, who insisted that only an elected body should draft the constitution. The interim constitution (the Transitional Administration Law) was drafted by a committee of the Iraqi Governing Council and adopted by it. An interim government (consisting mostly of members of the former council) was set up; it took responsibility for elections and the operations of the constituent assembly. The constitution was adopted largely in accordance with the time limit, and fresh elections produced a new government and parliament.

The transitional phase was dominated by the concerns of the United States, and the Transitional Administration Law was in part a document negotiated between the Iraqis and the United States, touching on matters of special economic and political interest to the latter.

Internally managed interim arrangements

Since the start and the progress of negotiations often have a dynamic effect on the relations among the parties, it is not unusual that interim arrangements tend to be modified over a period of time. South Africa provides a good illustration. It passed through two distinct stages of interim arrangements. At first it continued with the system and government set up under the old (and disputed) constitution. Nelson Mandela was anxious to maintain legal continuity. But the old system was infused with a decision-making process in which all key parties to the negotiations participated. An executive committee of these parties was set up under the peace process. The cabinet agreed to exercise the powers of the government and, to the extent necessary, of the legislature, in accordance with the advice of the committee. Using this mechanism, some apartheid laws were repealed, and considerable progress was made toward a new constitutional settlement. The second stage was reached with the settlement on an interim constitution (which was adopted under the old constitution) and the holding, under the interim constitution, of the country’s first nonracial general elections. During this second stage, there was a government of national unity in which all major political groups were represented in the cabinet. A principal mandate of the newly elected parliament was to draft and adopt the final constitution. On the dissolution of parliament, elections were held under the new constitution—and a government was formed to usher in the end of the transitional period.

2.1 Tasks—starting a process

The tasks grouped in this part are all performed early in a constitution-making process, and are mostly concerned with questions of design or interim arrangements. Almost all involve issues about how to carry out a constitution-making or review process, or how to handle the transition period until there is a new constitution.

We begin, however, with the logical first task: “Do we need a new constitution?” That requires some consideration of where the country is now, in constitutional terms.

2.1.1 The constitutional starting point

A country that is contemplating a constitution-making process may:

  • currently have no constitution at all;
  • have no acceptable constitution;
  • have a functioning constitution, but one that is expected to be replaced by a new constitution; or
  • contemplate only amendment of an existing document.

Having no constitution at all is a rare situation, but it can occur if a new country is carved out of an existing one, or if a number of existing countries decide to form a new, perhaps federal, state. Much more common is the situation in which conflict or radical political change has made the existing constitution unacceptable. Usually it is the institutions, the distribution of power, and the access to resources that are unacceptable, but sometimes even the existing document cannot be tolerated—perhaps because of who made it—even if the new institutions may not differ much from the old. Sometimes the existing situation is so unworkable or so unacceptable that constitution-making has to take place in two stages: first, an interim constitution is prepared; then, through processes established by the interim constitution, the final constitution is created.

The following examples indicate the variety of starting points and incentives that have affected constitution-making processes:

  • Timor-Leste—was a totally new country carved out of Indonesia; for a while it operated on the basis of United Nations regulations, but it needed a constitution.
  • South Africa—had a fully functioning constitution, but it was based on a racist rejection of any rights for the majority of the population; an interim constitution was adopted as the result of negotiations between the old regime and representatives of the majority. This was passed into law through the processes of the old constitution, and under its processes the final constitution was prepared.
  • Afghanistan—had been controlled by the Taliban, who ruled in compliance with their view of Sharia (though they did say they used an existing constitution with its un-Islamic elements removed). After the Taliban were driven from power, the only constitution that seemed acceptable to the United States and the transitional Afghan leaders was that of 1964; shorn of its royalist elements, it was adopted as the interim constitution.
  • Switzerland—had a constitution dating from 1874 that had been amended 140 times. Changing it was challenging, but it no longer reflected many accepted principles, including human rights; it was decided that a new document was needed.

2.1.10 Starting over when a process has “failed”

Failure—in the sense of not leading to a new constitution—seems to be the fate of perhaps half of all constitution-making processes in the world. Failure—in the sense of not producing a workable constitution—probably occurs in another significant proportion. And in others the process may be a failure in the eyes of some because it did not produce the constitution that they wanted. But a process that has not led to a new constitution is not necessarily a failed process. Much may have been achieved—even a realization that the country can live with its current constitution.

Beyond these rather general statements, we note elsewhere (see part 2.3.14) that failure may occur for different reasons, in different ways, and at different stages. Here we explore some of the factors, and the strategies, that may be relevant when deciding whether or how to try to remedy a “failure.”

An important warning: things are never the same at two different points in a country’s history. The people involved will have changed. The state of politics, peace and security issues, and the economy will affect the attitude of the people toward the need for a new constitution. It is likely to be particularly important who is in power. Sometimes pressure for a new constitution is really pressure to get rid of a certain ruler or a certain generation of politicians. When this is done, there may be, at least for a while, much less concern with constitutional change. And if the new rulers are the people who used to press for constitutional change, they may have discovered a fresh enthusiasm for the once-despised constitution—now that it has put them, and maintains them, in power. Finally, the earlier, even if “failed,” process will have had an impact: it will have shaped what people know about constitutions, and their hopes and expectations. They may be less interested in participating because of cynicism flowing from the failure; they may be more knowledgeable—or think they are more knowledgeable; they may be even more desperate for and committed to change; they may be so keen to have change that they overlook flaws in the proposed new constitution.

Back to square one

After a significant lapse of time, the failure of a process may be largely irrelevant—and the issues for those who have an interest in moving, or an obligation to move, toward a new constitution are essentially the same as if the country were starting constitution-making from scratch.

Even if the lapse of time is not so great, a country may decide that the best lesson from the past is to try completely afresh—either by the same sort of process or by a different process. In other words, no effort is made to salvage anything of the failed process.

Giving up

At the opposite end of the spectrum would be a decision to abandon the effort to change the constitution. Assuming that the country is not abandoning the aspiration of constitutionalism, the option would be to try to make the existing constitution work. As noted in part 2.1.2, this may well be a sensible option in many cases. No constitution is ever fully implemented, and the history of many countries would have been quite different if the government and the people had tried seriously to make what they had (in constitutional terms) work. Deciding to give up the search is unlikely to quiet demands for a new constitution unless the people have been involved in the decision or at least have acquiesced in it, and a serious and visible effort is made to use the institutions that the existing constitution presents.

What about the “no constitution” countries?

A particularly difficult variant of this situation will face a country that “does not have a constitution.” In reality, there are virtually no countries like this. Even a new country such as Timor-Leste could have adopted a set of institutions based on the constitution of Indonesia. There are people who would suggest that Somalia is putting its priorities in the wrong order by trying to proceed with constitution-making when even the extent of the government’s control over the capital city is contested. Either that country could build on local institutions and gradually build up a state, or it could use one of the earlier constitutions (or make its interim charter more permanent). The last course—relying for an extended time on an interim constitution—has been used in various countries, including Nepal and Sudan.

Gradual change

Another possible response is to give up the idea of a formal process of constitutional revision and to try to change the existing document incrementally, in the hope that over a period of time the country might be able to move toward a document that works and is legitimate in the eyes of the people. The Israeli parliament adopted a series of laws that laid down the constitutional framework, anticipating that this would ultimately lead to a unified constitution.

There are questions about such processes: is the result coherent? Is the process transparent and participatory? And some existing constitutions make change so difficult as to be impossible.

Starting from where you left off

Other countries have tried to retain what are seen as the gains of the “failed” process while moving ahead to translate them into a new constitution.

Lessons from the Kenyan process—which are likely to be true of other countries, too—include:

  • old, contentious issues that are of concern to politicians (mainly about power) have not been resolved, and they threaten the new process as they did the old;
  • agreeing on what issues are “contentious” is hard, because additional issues constantly emerge;
  • errors of design that were made in the previous process, including some deliberately created to defeat the old process, remain to affect the new; and
  • people, including members of the press, have remarkably short memories about the details of the constitution (hailing as “new” provisions that have appeared in several drafts).

Box 8. Options for “starting over” in Kenya

This has been the approach taken by Kenya since the rejection of the government- mutilated draft in a referendum in 2005. Various suggestions were considered for restarting the process. These included different combinations of a constituent assembly, a panel of constitutional experts, and a referendum (the last being required by a court decision); or a multisectoral forum plus a committee of experts and a referendum; or electing the next parliament to act also as a constituent assembly; or an interim constitution, a constituent assembly, and a referendum; or abandoning the attempt to produce a whole new constitution. Eventually a committee of “experts”—six local and three foreign—was mandated to prepare a draft drawing on the various phases of the failed process. (See the case study on Kenya, appendix A.7.)

The “political will” question

The mechanics of restarting are one thing; generating the will to start again is another. It is important to understand why a first process failed, to rekindle interest and recapture momentum if a major review process is to work the second time around. In Kenya an agreement effectively forced on the country by the international community after an outbreak of ethnic violence compelled fresh movement on the constitution.

On not rushing to change the new document

Some people will not be satisfied with the new constitution; some will have lost some aspect of the argument, and others will have lost power. And some people who were spoilers during the process will continue to raise points (which are not necessarily their real objections to the new document). Once a country has a new constitution, it should, ideally, move on. Implementation will itself be a big task, and continued debate about the contents will only hold up implementation. There is a risk that constitutional amendment will replace the old constitution-making debate. Clearly, essential amendments—such as if some provision turns out to be quite unworkable—ought to be made, but in principle the effort in the first years of a constitution’s existence should be toward making it work as it was drafted.

Box 9. Who should participate—and how? (Sometimes known as “actor mapping”)

When a constitution-making process is begun, it is important to identify all the sections of society that need to be involved—to create a sort of picture of society, with all its divisions and institutions, to ensure that the constitution-making is a truly national event and everyone has a voice.

Certain groups often dominate—including, in postconflict societies, the parties to the conflict, who often seem to think that taking up arms gives them an exclusive right to participate.

  • women, either because they are generally disregarded in society or because organizers ignore the issues of culture and role that make it hard for women to participate in the same ways as men do;
  • ethnic or religious minorities;
  • marginalized caste or ethnic groups, because they are excluded from meetings, live in remote areas, belong to small language groups, or do not understand constitutional issues;
  • noncitizens—even if they are long-term residents;
  • the elderly, who may have to stay at home; and
  • immigrants, who, even if citizens, may be victims of exclusion.

Especially in a society emerging from conflict, there may be no institutions that really represent the people.

Parliament, political parties, and local government may have collapsed, or they may simply be ineffective or unrepresentative. (Such issues may have been the cause of the conflict.) Parties and formal institutions that do exist should not be ignored; their cooperation may be essential for the success of the project.

Formal organizations should be identified, including trade unions and farmers’ associations; civil society networks and organizations; chambers of commerce; professional associations (e.g., teachers, nurses, and lawyers); and bodies representing “traditional authorities.” But in most societies there will be many other organizations: informal sector workers, squatters, victims of conflict, savings groups. Many of these may be largely invisible, especially to foreigners, but even to nationals focused on the capital city. In many countries, churches, temples, and mosques may be the principal organizations in communities’ lives, and the local schools may also be an important focus of life and locus of communication.

Formal groups are not the only way of thinking about the people and how they may be involved. Many people will not be organized at all, but they have equal rights to be involved. Women especially may not be organized; persons with disabilities may be concealed; marginalized communities may not be linked into the national structures.

Groups are important for helping to get a sense of the population and its divisions. But it should not be assumed that people will want to be involved only through organizations to which they are affiliated. A person may have interests other than that of being a farmer (represented by the farmers’ association), a woman (represented by the women’s self-help association), or a Christian (represented by a church). That person may want to be involved as a person directly. This section of the book addresses these questions of how people can participate and have their voices heard. Ensuring the participation of all key groups, and even of those who may not be formally associated with groups, may also promote greater transparency and ownership of the process.

2.1.2 Deciding if a process is needed

No constitution is perfect, but this does not mean that a country needs a whole new constitution. Sometimes it may be a mistake to rush into a commitment to such a new constitution. People may insist on this precisely because they have not analyzed what is wrong with the existing constitution. “Let’s start from scratch” is a way of avoiding, at least for the time being, such detailed analyses. Before embarking on a major exercise of constitutional design, the question should perhaps be asked: “Is our journey really necessary?”

A problem with the government does not necessarily demand a new constitution. The people of the Philippines have resorted on several occasions to “people power” to remove governments, but though there have been initiatives to implement a different constitutional system, it is not assumed that immediate change is needed. The Kriegler report on postelection violence in Kenya observed that:

it is important that Kenyans honestly assess all the activities related to the 2007 generalelections so as to distinguish between those that can be attributed to anomalies, failures, and malpractices traceable to gaps or provisions in the constitution and laws of Kenya from those that can be attributed to a bad culture encompassing impunity, disrespect for the rule of law, and institutional incompetence (Kriegler Commission 2008).

There are arguments against constitution-making, including:

  • Expense. The costs of constitution-making in Africa have been estimated, in United States dollars, as the equivalent of $30 million for South Africa, $10 million for Uganda, $6 million for Ethiopia, and $4.5 million for Eritrea (or between 15 cents and $1.50 per person in the country).
  • Divisiveness. Constitution-making may be a great nation-building event, but if the wounds are too recent, or the process is not handled with extreme delicacy, the process may give rise to renewed or new conflicts.
  • Risk of failure. A majority of constitution-making processes may be said to have failed, in the sense that they have not led to the enactment of a new constitution.
  • Constitutions should have some permanency. A constitution that is changed frequently is not really a constitution at all, for it does not guide or regulate the affairs of government. Making a constitution work is not easy; it does not work unless politicians, citizens, courts, and other institutions take it seriously and take steps to make it work. A belief that problems can be solved by the mere adoption of a new constitution is a delusion.
  • None of this is intended to suggest that major constitution-making exercises are futile, but it is important to consider whether making a new constitution is necessary, or necessary immediately, or whether a more modest, incremental approach should be taken.

More modest enterprises

In some countries it may be enough to change a fundamental problem with the constitution, leaving the bulk of it unchanged. It might be more practical to have a simple process of review by a small group of experts given a limited task over a limited time, with the opportunity for public consultation, rather than a full-fledged process, which in some countries can be expensive, time-consuming, and even divisive.

There is a relationship between the complexity of the changes anticipated and the elaborate nature of the process that is set up. On the one hand, naturally, the more fundamental the changes, the more public input there should be. On the other hand, if an elaborate process is set up, it is quite likely that far-reaching proposals will be made, even if the initial mandate is limited. The original French constituent assembly was given the task of voting money for the king, but it seized the moment and became the government and the collective author of the constitution. Similar things have happened in West Africa, where national constitutional conferences have introduced changes that were more radical than had been anticipated.

Is a more modest approach feasible? Even in “no constitution” situations, it may not immediately be necessary to embark on a major constitution-making process; limited, temporary arrangements may be possible. Israel provides an interesting example. It became a separate country in 1948 and planned to hold a constituent assembly. But it was immediately invaded and, since there was a feeling that 1948–49 was not the right moment because of the potential for disagreement (mainly over the connection between religion and the state), it abandoned the constituent assembly idea and, over the years, has enacted a constitution in bits. (It simply used the United Kingdom’s pattern of government as its basic framework.) Various efforts to produce agreement on a single constitutional document have not borne fruit. In Chile and in Indonesia, major efforts at constitutional reform have not been successful, but over a period of years various changes have been enacted to move each country away from autocracy.

The constitutional moment?

It is sometimes suggested that certain situations make it more likely that a country will be able to prepare and adopt a new constitution. Some people feel that a crisis, or the perception that there is a crisis, is a prerequisite—indeed, that a country at peace with itself will rarely be able to make a new constitution. The “crisis” argument holds that a sense that something serious will happen if there is no constitution creates an impetus for parties whose rivalry might otherwise prove an obstacle to agreement to work together. A sense of shared excitement about the future may serve a similar purpose, though it is rare for all sections of society to share that excitement, as the recent experience of Bolivia shows.

Some countries have, however, made constitutions, or carried out major reviews, while at peace. Canada, Finland, and Switzerland are recent examples.

It is not always easy to predict whether the circumstances will be right for adopting a whole new constitution. This is perhaps particularly so with a major review, which may take some years, and which may result in a radically different situation than existed at the beginning. This was what happened to the Kenyan process—especially because there was an election partway through the process, and the incoming government was unenthusiastic about the proposed changes.

Not all constitutional moments are suitable for the adoption of improved constitutions. in 2010, the president of Sri Lanka capitalized on his military victory over the Tamil rebels to introduce sweeping changes to the constitution—some of which enhanced the power of his office and his personal power.

2.1.3 Starting a process: The law and the politics

Here we explore briefly the matter of how constitutional processes are started. There are legal questions, but also political ones—how does a group within a society get constitution-making on the national agenda?

When there is no constitution

In the rare case of existing states agreeing to form a union, and thus having no constitution that governs their new entity, they can either continue to operate separately or devise some temporary cooperation agreement until their new document comes into existence. Their agreement will probably set forth the process for making their new joint constitution.

Some recent examples of beginning with a constitutional “blank slate” have involved the international community—for example, Cambodia [1993], Timor-Leste [2002], Iraq [2005], and Somalia [ongoing process]. In Afghanistan in 2001, the 1964 constitution was revived on a temporary basis by the Bonn Agreement made among Afghan forces, with amendments because there was no functioning monarchy. In Iraq, the United States, which was an occupying power, essentially guided the drafting of a temporary constitution. In these situations there was no basis in existing national law for the constitution-making process, and no legal way to challenge it. After a coup, all or part of the constitution is usually suspended.

Table 3: Constitutional review timetable

Country Constitutional review
United States [1787] Congress can pass amendments by two-thirds vote; two-thirds of the state legislatures can call a constitutional convention
Portugal [1974] Normally there cannot be a review more frequently than every five years (but four-fifths of the legislature can vote to do reviews more frequently)
Papua New Guinea [1975] There must be a review after three years
Fiji [1997] There must be a review after seven years
Switzerland [1999] The parliament, one chamber of parliament, or the people can initiate changes
Kenya [2010] Any legislator can introduce amendments; one million citizens can initiate an amendment

When there is an existing constitution

An existing constitution will include a provision for constitutional amendment. Rarely will it specify the entire process of amendment; often it will say only how the final adoption of a change will take place.

Some constitutions talk only of “altering” or “amending” the constitution or provisions thereof. A few constitutions specifically mention the possibility of enacting a whole new constitution. If this is not envisaged, sometimes there is doubt about whether the amendment process could be used to introduce a whole new constitution. In 2008 the Kenyan constitution was altered precisely to provide a mechanism for introducing a completely new document.

Some constitutions have several different amending procedures for different provisions— perhaps providing for different majorities in parliament, or requiring a referendum for some changes. (Canada’s constitution is a complex example.) This is one reason for not having a procedure for adopting a whole new constitution.

A few constitutions provide for a constitutional review. They may specify who can introduce amendments—the government, members of the legislature, the people—and they may have timetable requirements to prevent frequent amendments, or even to ensure periodic review. Table 3 shows a few constitutional provisions about starting review processes.

Some constitutions have provisions that cannot be amended at all, while the Indian Supreme Court has held that some aspects of the constitution (such as federalism, republicanism, and secularism) are so basic that they cannot be changed.

Some examples of initiation

Table 4 summarizes how constitution-making processes were initiated in a few countries where existing laws could not provide a mechanism, or even a starting point.

Table 4: Constitutional review initiation

Bangladesh [1972] Fiji
[ongoing process]
What did existing law say? Declaration of existence of state, adopted by Israel itself: new constitution was to be adopted by an elected constituent assembly No law—because Bangladesh had been part of Pakistan Constitution (itself a postcoup document) must be reviewed within seven years, but no mechanism was provided for this Bonn Agreement— basic framework for process: constitutional commission and Constitutional Loya Jirga Constitution: amendments required two- thirds majority of each house (but second house was not recalled when parliament was reconvened)
What steps were followed? Assembly converted itself into an ordinary parliament and did not adopt a constitution; it used an institutional framework carried over from period of British mandate
Proclamation of independence declared the Bangladesh leader president with all powers— succeeded by war with Pakistan. After president released from detention by Pakistan he promulgated a provisional constitution and an order for a constituent assembly, and
the assembly drafted constitution
Government (under internal and external pressure) set up, in consultation with the opposition, a review commission, which produced a draft that went to parliament under the existing constitution Steps in Bonn Agreement were followed—though the president added (by decree) a constitution drafting committee to prepare a draft for the constitutional commission Interim constitution was drafted by agreement between political parties and passed by one house

The political dimensions of starting a process

The impetus for changing a constitution may come from within the government. This may be the result of the realization that the existing document has become unworkable—either because it has been changed so often or because circumstances have changed so that the constitution is no longer appropriate. Switzerland is an example of the former, and a committee was set up in the 1970s to look at the possibility of changing the constitution. Finland is an example of the latter; pressure for reform came from the parliamentarians, who wanted a greater role for parliament.

Unfortunately, pressure from within the government is as likely to be self-interested as it is to be focused on the national interest. The eighteenth amendment to the constitution of Sri Lanka was passed at the urging of the president, and removes limits on the number of terms the president may serve, places the power of appointment to many important state offices (including the election commission) in the hands of the president, and in other ways weakens democracy. It was introduced at the insistence of the president, certified as “urgent,” and passed in one day.

Governments are often reluctant to embark on reviewing the constitution under which they got into power, so constitution-making processes are often generated by civil society or “informal” political networks. The tactics that get a process moving might include constitutional action, legal action, intellectual action, academic action, and people’s initiatives. A combination of strategies is often needed.

Constitutional action might be a citizen initiative (specifically provided for by some constitutions—see table 5) or a general petition process under the constitution. Many constitutions leave the initiation of reviews to parliament.

Legal action is expensive, time-consuming, and restricted in its scope—but sometimes it is a court case that shows that constitutional change is needed, though this is likely to focus on some fairly narrow point rather than on the constitution as a whole. A court may explicitly criticize a constitution, or the outcome of a court case may show the constitution’s weaknesses.

By “intellectual action” we mean reasoned efforts to argue for a new constitution, and thus to convince people and the government that something must be done. Again taking the Kenyan example, during the 1990s a number of organizations produced drafts of new constitutions designed to show not only that the existing constitution was weak but that a workable alternative was possible.

Box 1. Colombia’s popular movement for reform

A particularly interesting case arose in Colombia [1991], where a popular movement developed in favor of setting up a constituent assembly (though constitutional reform was the responsibility of the congress under the existing constitution). It was proposed that this motion should be put to the people in a general election. But the law prohibited popular initiatives on the ballot paper; nonetheless, a ballot paper containing the issue was printed in the press, and five million people used that to vote in favor of the proposition. The president used a decree under emergency powers to propose the same question formally on the forthcoming presidential ballot. This had to go to the constitutional court, which approved the decree, relying on the sovereignty of the people. A total of 88 percent of the people voted for the constitutional assembly.

“Academic action” in the sense of the sort of writing that may critique a constitution in books and journals is unlikely to have much impact. But it may provide ammunition for activists, for lawyers taking legal action, and for the more conscientious members of the media.

“People’s initiatives” have included people organizing for constitutional change by starting an unofficial commission and even consulting the public about its views.

2.1.4 Design

All sorts of factors will have an impact on these questions of design; our concern here is not so much the design as such, but the design process.

An important element is the history and culture of the particular country. Some countries— even some regions—start from assumptions about how the process will be done. Latin American countries often use constituent assemblies. French West African countries have often used constitutional conferences. Nigeria generally uses a commission followed by an assembly of some sort—a model used also in East Africa.

When most countries approach the matter of designing a constitution-making process, they are in crisis—or at least in a situation of some tension within society. A government department is not making decisions about the process in a detached, technical way. It is likely that the details of the process will have to be negotiated among political parties, or between previously warring groups (or even still-warring groups), which may be a long, drawn-out process. On the other hand, there may be such pressure to start the process, and such a sense that something serious will happen if there is delay, that decisions about the process are made in a hurry. It is impossible to dictate the right way to design the process; that depends on individual countries and their circumstances.

It may be useful to note how some countries have done this. In South Africa [1996] the process for making the final constitution was set out in the interim constitution. This was itself the product of negotiations among the main political parties, and it seems to have worked well for that country. The parties had great legitimacy in the eyes of the people. But in some countries, the parties may be an element of the dispute that generated the desire for a new constitution. In Kenya the outlines for a process were drawn up at a national conference that involved not only political parties but civil society. In Nepal the parties tried to keep hold of the design process. In these countries certain groups within society, especially ethnic groups, kept forcing their agendas upon the parties. Because the parties were not inclined to take these groups seriously, the groups resorted to violence or disruption. Then the government would reluctantly listen and perhaps make some agreement about the process—usually about the system of representation in the constituent assembly. This was a very flawed, and protracted, design process.

In many countries the occasion for reform has been a sense of exclusion on the part of some sections of society. If the constitution is thought of as a remedy for this problem, it naturallymakes sense for the constitution-making process to be inclusive. This is what many groups in Nepal were pressing for. There is now a good deal of writing about the need for peace negotiations to be inclusive. Similar arguments apply to negotiations over the constitution- making process—which may be part of the peace process as well. United Nations Security Council Resolution 1325 calls for women’s equal and full participation in peace processes. In many cases they have been excluded, even if they have been engaged in armed struggle. Many countries have other communities that tend to be excluded; in South Asia this is true of the Dalits. And just as article 25 of the International Covenant on Civil and Political Rights is now accepted as including the right to participate in constitution-making, so it should include, if possible, the right to be involved or at least consulted in designing that process.

Various factors may hamper the careful design of the process. Though people tend to understand that peace negotiations need careful planning—the chair will be carefully selected, as will the location, and even the seating will be considered—the same care may not be given to the design of the constitution-making process. Perhaps lawyers will take over and insist that it is a technical matter best left to them. Perhaps there will be a tendency to follow the same plan as the previous time—even though the constitution made by that process was flawed.

It may be helpful to reach a preliminary agreement on some basic principles that the constitution-making process must follow, and even on some of the elements of the new constitution—a topic to which we return later in this section.

The contents pages of this book—especially for part 3, on institutions and processes—should be a good indicator of the agenda that process designers should bear in mind. Broadly, that agenda includes:

  • Who is to decide—including who is to be able to have input into the discussion, even if not to make the final decisions?
  • Funding—how much will it cost, where is the money to come from, and who will be accountable?
  • Timing—is there to be a timetable, and if so is it to be rigid or open to change? Is it to be tight or to allow a lot of time?
  • Adoption—how is the new constitution to be passed into law—by the body that discusses and decides, by the president who usually signs laws, or by the approval of the people through a referendum? Are there to be any other prerequisites?
  • Technical quality—how is the technical quality of the document to be assured?
  • Openness—how will the public be involved, what parts of the official proceedings are to be open to the public, and what will be the role of the media?

Who designs?

Decisions on many of these issues may turn out to be highly political, and require a good deal of hard bargaining. To take the case of Fiji: the makeup of the commission that drew up the draft constitution was fiercely debated by the government and the main political parties. How many members should the commission have (especially to be considered representative)? Should all the members be nationals or should some be foreigners? Should the chair be a local or a foreigner and, if the latter, from what country? Again in Fiji, the terms of reference for constitutional commissions have been much debated. In 1987 the original terms of reference included “proposing to the Governor-General amendments which will guarantee indigenous Fijian political interests and in so doing bear in mind the best interests of other people in Fiji;” when the opposition forces (largely “other people”) objected, the italicized words became “with full regard to the interests of other people in Fiji.” And in 1995 the government wanted to minimize changes to the 1990 constitution, and wanted the terms of reference to reflect the priority of ethnic Fijian interests, while the opposition wanted a full review and fairness for all communities.

Design may be at least in part carried out by negotiation in a peace process. In some countries it has involved public consultation or some body with popular legitimacy. In Ecuador in the mid-1990s, unusually, one element in the design was referred to a nonbinding popular referendum: whether the constituent assembly delegates were to be directly elected in the usual way for parliaments, or elected by social movements.


In any participatory process, certain tasks must be performed, but different constitution-making processes have ordered some of these tasks differently. While the logical order of some events is obvious, there is room for different opinions about the sequence in which others should be carried out. The most important area of difference is about the stage at which a draft constitution should be prepared. The issue appears in two forms:

  • Should public consultation take place without there being any draft, or should the public be asked for its opinion on the basis of a draft or at least concrete proposals?
  • Should a constituent assembly begin its discussion on the basis of a draft?

The underlying issue is to what extent the voice of the people—whether through direct public consultation or through a constituent assembly—should be sought on the basis of concrete proposals, or more in the abstract.

Where would a draft come from if it were to come into being before the voice of the people was sought? There are three major possible answers: from one or more political parties, from a commission or committee, or from a single expert.

The major arguments in favor of “draft later” are that the existence of a draft is likely to inhibit free discussion, and that the draft tends to shape the views the public has of the assembly. Fears will be rather different if the alternative is having a draft put forward by political parties than if it is one put forward by an independent commission or committee, or if the alternative is a single technical expert’s draft. Are discussions going to be inhibited by the views of one political party, or by a political compromise reached in a committee, or even by the conservatism of an individual lawyer?

On the other hand, there is a fear that if public consultation takes place without any structure being given to the discussion by the existence of concrete proposals the people will flounder, not understanding the nature of a constitution and how they can contribute to its formation. Some will worry that the people will expect too much of the constitution. This argument itself perhaps reflects particular views of the appropriate scope of a constitution, and may be linked to the question of whether a constitution should contain economic, social, and cultural rights.

In Brazil the idea of the constituent assembly starting its work on the basis of a draft was firmly rejected because it would be a “dangerous instrument of control over the assembly.” But in Timor-Leste the dominant party in the assembly, Fretilin, was able to shape much of the discussion by ensuring that it was primarily based on its own proposed draft constitution.

In Fiji, Kenya, and Uganda, public consultation preceded the drawing up of a first draft. In South Africa the interim constitution was the outcome of interparty negotiation. Public consultation took place on the drafting of the final constitution, which drew heavily on the interim constitution. That interim document therefore to some extent served as a draft for discussion.

Box 2. The Kenyan process [2010] and the Bougainville process [2004]

The Kenyan process did not specify how long the first stage (preparing a draft) should take, but it did specify rigid limits for later stages. Consequently the Committee of Experts was supposed to respond to public comments and amend its draft in a twenty- one-day period that began on December 17 and continued through Christmas and New Year’s Day, when members and staff had already planned to be on holiday.

In Bougainville, major amendments to the Papua New Guinea constitution intended to give effect to the provisions of the Bougainville Peace Agreement 2001 were approved by the Papua New Guinea parliament in March 2002. The amendments did not come into operation, however, until Bougainville militias that had been involved in a complex secessionist conflict were certified by a United Nations observer mission as having completed agreed steps in the disposal of weapons. Only at that point (July 2003) could Bougainville take major steps provided for in the amended Papua New Guinea constitution toward adoption of a subnational constitution establishing the Autonomous Bougainville Government.

2.1.5 Timetables

A constitution-making process that is designed (as opposed to one of incremental constitutional change) will usually include a timetable and sequence of events. This is desirable for various reasons, including the need to seize a “constitutional moment” (see part 2.1.2), to control the costs of the process, and to ensure that narrow interests do not either rush the process or drag it on for too long. Getting the timetable wrong may mean that:

  • the document produced lacks legitimacy in the eyes of the public;
  • the document is of poor quality; or
  • no document at all is produced—because enthusiasm fades and people learn to live with what they have got, or an interim constitution turns into (or is amended to become) a permanent one.

A road map may work in various ways. It may:

  • have only a final date by which the new constitution must be adopted, giving no other indications of time periods or order of events (this is unusual, because if tasks are specified some intended sequence will usually be stated or implied, even if in general terms);
  • specify tasks in some detail and the order in which they are to be carried out, but without any time periods being fixed at all, or with only an ending date being specified;
  • specify tasks in very general terms, without a clear indication of when they are to be done (for example, a requirement might be “to consult the public” without an indication of whether this is to occur before any other work is done, only when a draft is prepared, or both);
  • spell out the entire sequence of events with precise time periods attached;
  • involve a mixture of these approaches; or
  • schedule events by reference not to time but to the occurrence of other events, such as other elements of agreement in a peace process (as in Bougainville, Papua New Guinea—see appendix A.9).

Box 3. Damaging deadlines

Some processes dominated by the United Nations or foreign governments have been seriously affected by externally imposed deadlines (e.g., Afghanistan [2004] and Iraq [2005]). This outside pressure not only caused resentment, but prevented local processes from being worked out fully. In Iraq, important issues were left unresolved, and remain unresolved. (See part 4.2, appendix A.1, and appendix A.6.)

How much detail should there be? It is impossible to anticipate exactly how long any stage of a process will take. Though it is possible to plan meetings for civic education and the collection of public views, the amount of public enthusiasm may affect how much time it will take to analyze the views. (See part 2.2.4.) More time or more resources may be needed. Each stage should ideally be planned to take some time. Otherwise there will be the risk of other stages being shortened, or extensions being sought.

Complex processes involving civic education and public consultation, the preparation of drafts, and discussions held by a constituent assembly or similar large-scale debating body typically take a few years.

Factors suggesting some urgency may include:

  • the risk of a return to conflict;
  • the risk of a coup (which is why the Philippines completed the process of drafting its 1987 constitution in six months);
  • the desire to complete a process before an election (either because it is clear that a new government will resist change or because before an election it may be difficult for participants to calculate how particular provisions will work to their advantage); and
  • foreign pressures (financial or other).

But the process must not be rushed. It is important to allow time to educate the public about what a constitution is, to educate the political actors, for people to formulate and submit views, for those views to be processed, and for consensus to be developed on difficult issues. In deciding what is “enough” time, the following might be relevant:

  • how much knowledge the people already have, including how much civic education has occurred;
  • terrain and communications—how people are to learn about the process and how they are to submit their views;
  • allowing time for views to be processed (because otherwise they may simply be put in a cupboard and forgotten);
  • whether it will be necessary to hold an election for a constituent assembly;
  • whether participants in the process (commission or constituent assembly members, for example) are engaged full-time or part-time; and
  • finances (which may determine whether a commission works full-time, or whether extra people can be hired to process submissions, for example).

Designers must take account of existing laws or constitutions prescribing some elements in a time frame (for example, that there must be a certain lapse of time between stages of deliberation in parliament).

Who should set the timetable?

Some people want a constitution quickly, some people want to extend the process, and some people may not want a constitution at all. Who sets the timetable may have serious consequences for success or failure. Factors involved include the following:

  • Politicians as a group may wish to enhance their electability by producing a constitution, regardless of its quality.
  • Individual politicians may believe that a new constitution will give them a chance to return to office (as some presidents have argued: if there is a new constitution, the old limit on the number of terms one can serve does not apply!).
  • Participants in a process who come from outside politics may find their role pleasurable, even financially rewarding, and wish it to last as long as possible.
  • International actors often want to have a clear exit date, and may be prepared to sacrifice quality for speed. (See part 4.2.)

Rigid or flexible?

It is not uncommon to allow for extensions of time. This flexibility may be given to the constitution-making body itself; if the time is without restriction, however, this may make nonsense of any timetable. But perceptions of the need for speed may change according to political circumstances. In other cases the control over extensions is given to the same body that imposed the timetable originally: parties to a peace agreement, parliament, foreign interests, and the like.

In Nepal the interim constitution provided the possibility of a six-month extension, but only if a state of emergency caused delay. This extension would require only a resolution of the constituent assembly itself (largely controlled by the parties), passed by a majority of those members present and voting.

Factors likely to result in completion on time

A realistic timetable is more likely to be adhered to; in addition, the following factors may be important:

  • effective chairing and management: the chair of any body, or, more broadly, the management structure, must develop a work ethic for participants, and generally promote the understanding that timing matters;
  • political commitment, so the actors feel they will “lose face” by seeking an extension; • making sure the public understands—and, ideally, supports—the timetable; • mechanisms for resolving particularly difficult issues (see part 2.5.2); • not making involvement in the process too lucrative;
  • trying to ensure that financial support is sufficient; and
  • generally managing the process effectively, including instituting measures to avoid corruption (see part 2.3).

What happens if a deadline simply expires?

This will depend, from a legal perspective, on the document that sets forth the deadline.

  • If the result is that the whole process dies, the existing constitution may remain. This may be only an interim document and not suitable for long-term application, or it may be a full, if unsatisfactory, constitution.
  • The document setting forth the deadline may require an election and then a revival of the process (as in Iraq).

How can a timetable be enforced?

Timetables imposed from outside, whether they purport to be legally binding or not, may be “enforceable” for the same reason that they were imposed in the first place: hard-nosed political or economic reality. If security will be withdrawn, if foreign aid may be released, pressures to finish may be real. States emerging from conflict are often vulnerable and divided.

Compromises are sometimes possible. In Timor-Leste the original timetable was completely unrealistic and had to be extended. But it remained unsatisfactory, and the constitution might well have been improved if additional time had been available.

In Iraq a constitution was produced according to a timetable largely dictated by the United States. But even the United States could not prevent the constituent assembly from amending the constitution five minutes before the expiration of the deadline under the interim constitution. A long extension was not practicable, and the result was an incomplete constitution, with certain issues not being properly resolved. Even before the constitution was adopted, a promise was made to certain sections of the community that it would be reviewed and finalized immediately after the elections.

If a constituent assembly is also the legislature, it may not be possible to prevent it from amending the constitutional or legislature framework (as in Nepal)—though this may have a political cost.

Some examples of timetables

The constitutional convention for the United States took nearly four months; ratification by the states took a further forty months. The Indian constituent assembly sat from 1946 to 1949 (though it was severely affected by Partition). The Eritrean process took thirty-eight months from the proclamation of the constitutional assembly to ratification of the constitution. The South African process took five years from the beginning of multiparty negotiations to the adoption of the final constitution.

The Ugandan commission took from 1989 to 1993 to prepare a draft constitution, and the final constitution was adopted in 1995. There were various reasons for the length of time the process took, including the time needed to carry out extensive public consultation as required by the law establishing the commission, and lack of resources.

The Timor-Leste constituent assembly [2002] took five weeks to prepare its rules of procedure, and the Bolivian constituent assembly [2009] seven months. The Bangladesh assembly [1972] took only two days, because it used the existing parliamentary rules.

Public consultation early in the process took about four months in Kenya [2005], and nearly three months in Fiji [1997]; in Rwanda [2003], civic education and public consultation combined took about six months.

The drafting of a constitutional document to be submitted for public consultation took two months in Rwanda [2003] and about one month in Kenya [2005]—but considerable preliminary work had been done in the latter case, and perhaps in the former also.

Public consultation on a draft constitution or concrete proposals has taken from one week in Timor-Leste [2002] to about four months in Eritrea [1997]. The Timor-Leste period was recognized as too short, but that was the result of pressure to complete the process.

Debate on a draft constitution in a constituent assembly took two weeks in Afghanistan [2004], five months in Timor-Leste [2002], and eleven months in India [1950] and Kenya [2005]. The Afghan process was largely a rubber-stamp operation; the Timor-Leste assembly oversaw almost the entire process and there was no separate commission. In Afghanistan [2004] and Kenya [2005], a separate commission prepared the constitution. In India the constituent assembly was also the parliament, and in Kenya all members of parliament were members of the constituent assembly, which could not sit when the parliament was in session.

Referendum campaigns for and against a complete draft have taken less than four weeks in Albania, one month in Venezuela, five weeks in Spain, and three months in Kenya.

These figures will be of limited use to planners, and are intended just to show the range of times and some of the factors that may affect them.