The program for the constitution-making process may have various legal bases, and many processes are based on a combination of documents and arrangements, some of them laws and some administrative or informal considerations.
Constitution-making processes may be designed in an atmosphere of some euphoria—especially if peace has only recently broken out. But the process has to work even when the euphoria has passed, when tensions reassert themselves. The implication is not only that the process should be well designed; the legal form the blueprint takes may become important. Few constitutions provide in detail for the process. It is important for designers of a process to study carefully the requirements that do exist, to make sure that the new document is validly adopted.It may be necessary to pass a law to set out the process. An example is the Constitution of Kenya Review Act. Sometimes an executive order is used. This may be done under a general constitutional power or under a statute. In some countries a constitutional review has been carried out by a commission under an act that provides for the government (or just the president) to set up a commission of inquiry. In a truly revolutionary situation there may be no existing constitution. In such situations the design may have to be included in agreements between the parties. There may even be no existing way to make laws.
What are the implications of the form of the blueprint? Perhaps the most important is whether the process, once having started, can be stopped. An executive order can be “unmade.” A law can be repealed. An agreement between parties may have no legal force in a technical sense (for example, a court might well refuse to hear a case arguing that one side to such an intensely political agreement had reneged on it). In Kenya there was constant debate about whether the process ought to be “entrenched” in the existing constitution—because of fears that vested interests would interfere with it. Those fears turned out to be well founded. In Kenya even the courts have been used to slow down or stop a process.
There may be other significance attached to the form of the blueprint. Certain forms may have greater acceptance or legitimacy in the eyes of the previously disputing groups in society. The executive order of one person may lack the sense of national support that some people would wish to see.
There are many questions about the blueprint that require careful judgment. How detailed should it be? Should it be able to be changed? Should it be only a guideline, or should it be legally binding on the bodies that are to make the constitution?
There are aspects of this matter of legal form that lawyers may attach particular importance to, particularly the matter of legal continuity: should the new regime, whether interim or permanent, be legal in terms of the previous regime? In South Africa the regime that existed until the 1990s had no legitimacy in the eyes of most of the citizens, and had been the subject of international sanctions. One could say that it was illegal. Yet a decision was made to move step by step to a new constitution, always using the procedures of existing law. So the 1993 interim constitution was passed pursuant to the procedures set forth in the apartheid constitution. It was felt that this would give a solid legal foundation to the 1996 permanent constitution.
On the other hand, some countries have deliberately made a legal break with the past. They may insist that the foundation of the new constitution is not to be laid in the laws of the past (including perhaps the laws of a colonial power) but rather to come from the will of the people.
Sometimes, it must be said, legal continuity is something of a fig leaf. Where there has been a coup, and a military regime has come to power, clearly illegally, it sometimes insists that parts of the constitution remain in force. Yet parts that the regime does not like, or cannot operate because there is no parliament, for example, have to be abrogated or suspended. The regime may be using this approach to give itself some (rather spurious) legitimacy. The significance of a constitution is that it binds people and authorities even when they do not like it. If you say that you are bound by only the bits of a constitution that suit you, there is in effect no constitution.
Some countries have no choice about the nature of the blueprint: it is prepared by the United Nations or other international actors, and takes the form of a Security Council resolution or international agreement. But it is rare for the international framework to give much detail about the process, though the timetable that may also be imposed may have inevitable effects on certain aspects of the process, such as the amount of public consultation.
If the idea of the constitutional moment has any significance, it is probable that the really key moment is when the process is being designed, rather than when the design is being applied. This may be the moment of greatest optimism; it may be the moment when concessions will be most readily made. This gives particular importance and responsibility to the designers. Yet they may not realize this, or they may focus only on certain aspects. In Nepal it was noticeable that everyone concentrated on the makeup of the constituent assembly—who was going to get how many members. Little attention was given to the issues of how the constituent assembly would work, or the participation of the public in the deliberations.
Peace agreements—which sometimes are wholly internal to a country and sometimes involve the international community, including perhaps the United Nations and key interested international actors—have provided much of the framework for constitution-making in Afghanistan  and Nepal [ongoing process].
Box 4. “Revolutionary legality” and “necessity”
After a military takeover, or some other crisis that makes ordinary processes of governing and constitution-making impossible (or so say those in power), courts will sometimes have to decide on the legality of a regime. They will usually find justification for the regime under which they are serving. They may do so on the basis of arguments of “necessity,” “revolutionary legality,” or both. The former expression was used in the Eastern European socialist republics—their legality was based on their revolution, and on the supposed will of the people. The phrase was also used by a famous Austrian legal theorist, Hans Kelsen, and has been used in a number of legal cases in nonsocialist contexts (Kelsen 2002). The key argument is usually framed in terms of coup leaders having formed a government that is effectively in control and accepted. (This is sometimes also known as the “successful coup” doctrine.) In 2000, Fiji’s Court of Appeals rejected such an argument because the military was unable to show that the people really did accept the situation.
The other argument is necessity—“the tyrant’s plea,” as the poet Milton described it in a different context (Milton 1968). Courts that have accepted this argument have said that the implication is that the government must as soon as possible return the country to constitutional rule; necessity is not a basis for a long-term, fundamentally illegal, regime.
What to do if it is hard, or impossible, to follow the existing procedures
Some constitutions make it so difficult to change the constitution that it is virtually impossible. Some, as we have seen, provide no mechanism for replacing the constitution—only for making amendments in a piecemeal fashion. For others, the circumstances may dictate that, although amendments or even new documents are widely believed to be necessary, certain groups will stand in the way of using the existing rules, or certain institutions will no longer exist. Especially in societies emerging from conflict, it is sometimes not possible to observe all the requirements.
Here are some examples of situations in which constitutions have been hard to change, or to replace, at least when change was called for:
- Australia requires a referendum that must receive a positive vote from half the population nationally and from a majority in each of more than half the states (four when there are six states, as is the case now); in some circumstances a majority in every state must approve.
- In some countries doubts have been expressed whether the power provided in a constitution for the legislature to “alter” the constitution includes the power to make radical changes or to replace the constitution.
- In South Africa, the legislature in 1990, which would have had the power to amend, was dominated by the white community; black South Africans did not have the vote or any members.
- In Nepal, the 1990 constitution could be amended by votes of the two houses, and signed by the king; after the People’s Movement of 2006, the Senate was not reinstated (and even the legal status of the House of Representatives was doubtful), and no one wanted to involve the king.
Techniques that have been, or might be, used have included:
- using as many of the existing provisions as possible (Nepal);
- by negotiation, using the existing mechanisms even if they are against the interests of those who must take the necessary steps (South Africa);
- accepting that the constitution is hard to change and working within the constraints (Australia); and
- acting outside the constitution entirely by calling a national conference or a constitutional convention or assembly.
In countries emerging from conflict, some departure from the letter of the law may be necessary. Once a new constitution is adopted, it is unlikely that it will be attacked for lack of legal validity. Courts rarely hold that the regime and the constitution under which they are serving are illegal, though it happened in Nigeria (in 1966) and Fiji (in 2000 and 2009). Courts often accept arguments based on the principles of “necessity” or “revolutionary legality.” (See box 4.)
More detailed frameworks are usually provided in an interim constitution. (See part 2.1.9.) Often this is the primary purpose of such a constitution. Careful thought should, if possible, be given to how much detail is put into the constitution. Constitutions are hard to change—this is their strength, but sometimes their weakness as well. Constitution-making processes set up under an administrative procedure may be easily halted, and ordinary laws can be repealed or changed by a simple majority of parliament. Therefore to include the basic structure of the process of constitution-making in a document with constitutional force may be wise. But if all the details are included, problems may arise. The most common of these is: what happens if the deadline cannot be adhered to? If the document establishing the deadline does not allow for any extension, and an extension is essential, it may be necessary to change that document. In Nepal, the provisional legislature (the constituent assembly by another name) amended the interim constitution, and the president signed the amendment (although some people had argued that to allow such a change would defeat the whole purpose of a deadline—and even that it would be unconstitutional to use it, and the president ought to refuse to sign). But if the forces not wanting a new constitution had not been able to muster a two-thirds majority, the constitution-making process would presumably have ground to a halt.
Few countries will have an ordinary law on the books that provides for a fully participatory constitution-making process. So some countries establish their processes by means of a special law (perhaps setting out the membership of the constitution-making bodies, the processes for public consultation, and the timetable). For an example, see the case study of Uganda in appendix A.12. Although laws are easier to change if circumstances change (as they often do, especially in postconflict societies), it is not necessarily a simple matter to change a law. So the extent of detail to be included, especially concerning timing, should be considered carefully. But if there are no timetables or procedures with a legal basis, it is both harder for people to understand the process and easier for it to be subverted. Certain matters will be dealt with by existing laws, such as the auditing of the accounts of public bodies.
In American Samoa  the constitutional review process was set up by a series of executive orders issued by the governor of the territory. This was under the existing constitution, which gave powers to “issue executive regulations not in conflict with laws of the United States applicable to American Samoa, laws of American Samoa, or with this Constitution.” Such an order could presumably be as easily unmade as made, and so it would give even less security to the process than an ordinary piece of legislation.
Less formal sets of rules, such as codes of conduct (see appendix C) and administrative arrangements, will also be necessary to fill in the details of the process. Even giving orders to the police to facilitate public meetings, or not to enforce existing laws banning such meetings, may be important in some processes.
In short, the details of the process may be found in any or all of the following documents or decisions:
- peace agreements (which are usually not legally enforceable, especially through a country’s courts);
- constitutions—previous or interim;
- existing laws;
- specially made laws (which might have to be made by extraordinary means in some situations—by military decree, for example);
- regulations made under laws;
- codes of practice (which may not have the status of law);
- informal agreements among political parties, or made through national conferences; or
- administrative instructions.