AuthorKong, Hoi L.

In 1992, President F.W. de Klerk's National Party in South Africa held a referendum (the "1992 referendum") on the following question: "Do you support continuation of the reform process which the State President began on February 2, 1990 and which is aimed at a new constitution through negotiation?" (1) This mandate referendum sought to "better prepare the public for a peace settlement, incorporate early feedback into the negotiations, strengthen their credibility across ethnic antagonists and safeguard the peace process from subsequent and often unavoidable reversals in public opinion." (2) Of the eligible, white population, 68.7% voted "yes" and set the stage for the dismantling of a profoundly discriminatory system of laws and for the construction of a new constitutional order. (3)

As others have noted, the 1992 referendum has received little academic attention (4) and, to my knowledge, it has not been examined from the perspective of constitutional theory. This article aims to fill this gap in the literature and to use the referendum as a case study that will contribute to three related debates in constitutional theory.

First, the 1992 referendum was part of an extended act of constitutional self-definition and an analysis of it can assist us in understanding the nature of constituent power. (5) Theorists have long debated whether the legitimacy of "the people" who exercise constituent power can be called into question. (6) In Part II, I will answer "yes." In arriving at this conclusion, I will argue against authors who claim that "the people" pre-exist the exercise of constitution making, that the making of a constitution happens at a discrete moment and results in a clean break with the past, and that this exercise of the constituent power is not subject to normative evaluation. In that Part, I will use the 1992 referendum to illustrate the theoretical claims that the making of a constitution and the constituting of a polity are mutually constitutive processes that occur over time, and that constitution making can be evaluated against norms that are relevant to the affected polity. The fact that the 1992 referendum was limited to the white population may cause one to hesitate to characterize it as a genuine moment of constitutional self-definition. According to de Klerk, the only constituency divided to reform was the white population (7) and in light of this response, one might justify the limitation by saying that it was unnecessary to open the referendum to other segments of society. In Part II, I will advance a different kind of argument. I will claim that the limitation in the 1992 referendum evidenced the extent to which constitution making in South Africa did not represent a clean break with the past. I will further argue that, limited though it was, the referendum set in motion a wider process of constitutional reform that involved extensive and, indeed, unprecedented consultations with the general population. It was therefore an essential part of a genuinely inclusive process of constitutional self-definition.

The second debate that the 1992 referendum engages is about how to characterize moments of constitutional change. Scholars disagree about whether exercises of constituent power lie outside of constitutional constraints. (8) I will argue that constitutional moments can be evaluated against deliberative democratic standards of reasoned debate and I will claim that the 1992 referendum and the processes that flowed from it can be understood to have satisfied those standards. I will therefore reject the claim, sometimes made in the constitutional theory literature, that exercises of constitution making are unconstrained by constitutional standards.

The third debate that the 1992 referendum touches upon concerns the appropriate place of direct and representative modes of decision making in the making of a constitution. Critics argue that referendums ask the general populace to make decisions that are beyond their competence, that referendums are subject to manipulation, and that they are reductive. (9) I will argue that these pitfalls can be avoided through careful institutional choices and I will claim that the 1992 referendum is an example of this kind of institutional design.

Parts II, III, and IV address each of these debates and advance my position on each of them. Part I sets out the aspects of the 1992 referendum that are relevant to the present discussion.


Katherine Collin has noted that referendums in contexts of national conflict have a mixed record of success. (10) She has developed a typology of these "peacemaking referendums" that categorizes them along two axes. The first axis distinguishes between referendums that are "integrated in a peace process" and those that "stand alone". The second axis divides referendums between those that pose "process ballot questions" and those that pose "outcome ballot questions", and she categorizes the 1992 referendum within the category of "process ballot questions" as a "process mandating" case. (11) Furthermore, she identifies determinants of peacemaking referendum success, including a referendum's timing and "[t]he alignment of consensus between voters, political elites, and international actors" in favour of a "yes" vote. (12) In what follows, I will organize my discussion of the 1992 referendum around these determinants.

Consider first the timing of the 1992 referendum. After de Klerk's National Party won the 1989 election, he promised dramatic change and set in motion a series of initiatives that included the release of Nelson Mandela (13) and the initiation of negotiations with the African National Congress (ANC) at the Convention for a Democratic South Africa (CODESA). (14) As a way of addressing the white population's fears about the promised changes during the 1989 election campaign, the National Party promised that population (and in particular, traditional National Party voting blocs) a referendum that would enable them to have "the final say" over constitutional negotiations. (15) De Klerk's decision to announce the referendum came on the heels of a series of by-elections that the National Party lost to the Conservative Party and which "challenged his authority as chief negotiator for the white community." (16)

At this precarious moment and throughout the referendum campaign, de Klerk's National Party drew upon a consensus that had emerged among the general public, political elites, and the international community. In 1991, almost 54% of the black population agreed completely or to some extent with the statement "the current policies of Pres. De Klerk will bring lasting peace to South Africa". (17) And in a national survey held in July 1990, 74.5% of all respondents and 78.4% of white respondents said "yes" to the question: "Do you support current talks between the ANC and the National Party?" (18) Elite opinion also supported negotiations. The ANC issued a statement in support of the referendum and during the campaign, Mandela actively corrected misinformation issued by the Conservative Party. (19) A group of business leaders "launched a 'Private Sector Referendum Fund' in order to finance an aggressive advertising campaign" in favour of the "yes" side. (20) Executives in large firms (including Toyota), prominent economists, and vice-chancellors of Stellenbosch University and Cape Town University warned against the dangers of a "no" vote. (21) And leaders of the international community, including in the United States, Europe, (22) and Australia, (23) highlighted the increased investment and lifted embargoes that would accompany a "yes" vote and warned of the negative repercussions, including sanctions, that would accompany a "no" vote. In the end, the overwhelming vote in favour of the "yes" side underlined the limited support for right wing policies in the white population and in the public at large (24) and strengthened de Klerk's hand in the constitutional negotiations. (25)


With this outline of the 1992 referendum in view, I turn now to consider its implications for debates within constitutional theory and I begin with a long-standing debate about the nature of a constitutional "people" and exercises of constituent power. In political and legal theory, the relationship between popular and constitutional legitimacy has generated an enormous body of writing. One issue that has posed particularly acute problems is whether the question of the legitimacy of the people who are subject to constitutional institutions is one that can be practically or coherently posed. (26) According to Nasstrom, some theorists argue that because "unanimous consent is impossible to mobilize--no society is founded on voluntary agreement--they think it more prudent to leave the claims of legitimacy aside and presuppose the existence of a circumscribed people." (27) For this group of theorists, the question of constitutional legitimacy is incoherent because any act of constitution making by, for instance a vote, presupposes the legitimacy of the group engaging in that very act. (28) In Habermas' words a "constitutional assembly cannot itself vouch for the legitimacy of the rules according to which it was constituted." (29) As a consequence, these theorists accept the existence of a given polity that is the subject of a constitutional order to be a fact of historical contingency. As a consequence of this acceptance, these theorists are precluded from normatively assessing and critiquing the boundaries of political communities. (30)

A related problem has been identified by Daniel Lee in his study of the concept of sovereignty in early modern thought. (31) He notes that the concept of constitutionalism presupposes that public power is limited by law, whereas the idea of sovereignty is "in theory, unlimited and absolute, superior to all...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT