Constitutional Choices Turned into Simple In-Period Choices: A Power Relations Reading of the Chadian and Cameroonian Constitutions

Abstract

In this study we apply a methodology proposed by Imbeau and Jacob to test the veil of ignorance hypothesis in the Chadian and Cameroonian cases.  In a first step, we present the context that led to the making of the Chadian and Cameroonian constitutions and in a second step, we analyze, in the light of that context, the uncertainty measures derived from a content analysis of those constitutions. Overall, our evidences are mixed: our uncertainty measures are sometimes found to be rather counterintuitive. We suggest that in cases where there are presumably legal writing styles and traditions, content analysis-based measures of uncertainty can be noisy. 

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In the constitutional political economy literature, an important element of distinction between constitutional choices and regular politics choices is that the former involve a great deal of uncertainty and are said to be made behind a “veil”: the veil of ignorance. However, the hypothesis that constitutional choices are made behind the veil of ignorance has mostly received a theoretical treatment.

The Veil of Ignorance Project (VOIP) proposed by Imbeau (Imbeau and Jacob 2015, 2011; Imbeau 2009) is mainly an empirical endeavor. It seeks to test the presence of a “veil” at the time of the making of constitutions and suggests, in order to do so, that we read constitutional texts in terms of power relations. That a constitutional text deals more with political power than economic or preceptorial power (H1), that it is most concerned with social power than instrumental power (H2), and that it tends to restrict power more than to ascribe power (H3) can be viewed as indications of the presence of a “veil” during the drafting/adoption of that constitution. Heuristically, the VOIP follows a logic known as abduction.¹ This abduction logic is strengthened as we make our inferences in light of the historical context of the drafting and adoption of a constitution.

The purpose of this case study is to implement the VOIP framework using the Chadian and Cameroonian constitutions to tell whether or not they were adopted behind a veil of ignorance. Beyond the methodological interest, one can derive a practical interest from this undertaking. Indeed, one can learn valuable lessons on how to design settings suitable to the making of democratic constitutions. This case study has the following structure: First, we present the major events that led to the current Chadian and Cameroonian constitutions; second, we present a power relation-based content analysis of those constitutions; and third, we discuss our findings.

1 Constitution Making in Context

The purpose of this section is to shed light on the context of different constitutional phases that Chad and Cameroon went through up to the adoption of their current constitutions. In that regard, the two countries have a quite similar trajectory except that Chad’s political history was marked by more violence and instability. Both countries were under French tutelage before 1960: Chad as a colony and a part of Cameroon as a mandate.² Following the Second World War, France initiated a gradual process of decolonization of its colonies that would ultimately result in their independences in 1960. At their independences, Cameroon and Chad adopted constitutions that resembled very much that of the French Fifth Republic: Obviously, the drafters and adopters of those first constitutions wished to see their new countries become and evolve as democracies.

At the time of their independences, there were legitimate concerns about the viability of the new nations: Many feared that the young countries would implode because of their ethnic and religious heterogeneity. As a consequence, the first executives of Chad and French Cameroon, respectively, Francois Tombalbaye and Ahmadou Ahidjo, prioritized among other things the reinforcement and the consolidation of national unity. In fact, they would later use that objective as a pretext to arrogate more constitutional powers and turn their nascent democracies into autocratic presidentialist regimes.

In Chad, it did not take too long before “centrifuge forces” blew off “national unity.” Northerners felt underrepresented in Tombalbaye’s government³ and criticized his autocratic leanings: In 1962, he promulgated a new constitution instituting a unique-party system and strengthening his powers as president. They reacted by forming in June 1966 a coalition of rebellions (Front de Libération Nationale du Tchad) against him, and Chad entered a cycle of civil wars and coups that lasted until recently.⁴ The coalition overthrew and killed Tombalbaye in 1975, suspended the constitution, and instituted a military regime led by General Felix Malloum. In 1982, General Hissen Habré a dissident of the coalition seized power after a coup and promulgated an interim constitution. Habré then became the president of a divided country with several regions under the control of warlords; he resorted to violence and brutality to restore his authority over the country.⁵ His dictatorial regime lasted for eight years till the Idriss Deby’s coup in 1990.

Contrary to his predecessor, Deby promised democracy to Chadians in his take-over speech. In fact, the geopolitical context at the time somewhat pressured him to do so. Indeed, France’s president Mitterrand at La Baule introduced the principle of conditional aid toward France former colonies in exchange of political and institutional reforms. Several Western donors later adopted the same policy. In addition, the collapse of the Berlin Wall in 1990 ignited among African people strong desires for democracy and liberties. In the early 1990s, almost everywhere on the continent, people were rioting and marching against the authoritarian and dictatorial regimes that had oppressed them almost since the independences.

In Chad, some leaders of the civil society and political parties’ leaders suggested the organization of a National Conference, a sort of political forum to discuss a thorough reformation of the state. That idea received strong support from the people, and in 1991, Deby promised to convene the Sovereign National Conference to draft a new constitution. He appointed at the end of that year a first committee of eighty members with the task of setting the conference agenda, proposing its rules and procedures, selecting the participants, and specifying the logistics (Buijtenhuijs 1993). The public opinion was, however, unsatisfied with the composition of that committee which it thought was partisan: The committee chair was one of Deby’s army general, and many members were in some way close to him (Buijtenhuijs 1993). To ease up the tension, Deby appointed a second preparatory committee called the Tripartite Committee (members came from the three main groups of the polity): Five members representing the public authorities (from which the chair was chosen), five others the political parties (including Deby’s party and the opposition), and two the civil society. The Tripartite job was to discuss, “second-read” and eventually amend the first preparatory committee’s report. The Sovereign National Conference scheduled to take place in 1992 finally started in January 15, 1993, and sat until April 12, 1993. There was a great confusion about how many people exactly attended the conference and even potential historical sources diverge on that issue (May and Massey 2000). Officials at the time said 830; Buijtenhuijs⁶ on the first day of the conference for the election of the presidium found 846 voter cards of which only 822 could be accounted for. According to Buijtenhuijs’ accounts (Buijtenhuijs 1993: 46–56), there were seven categories of delegates as follows:

• 116 public officials (ministers, prefects, mayors, etc.),

• 256 representatives of political parties (eight for each registered party and one for political parties in the process of being registered),

• 31 delegates from human rights and youths organizations,

• 99 delegates from professional corporations, unions, cultural, and humanitarians organizations,

• 168 delegates representing the people of Chad: sultans, chiefs representing cantons, farmers, and representatives from the diaspora,

• 8 delegates from religious groups: two Catholic bishops, two Protestant pastors, and four Muslim leaders,

• 144 Chadian notables, a heteroclite groups made up of former ministers, ambassadors, academicians, state corporations executives, former high-ranking military officers, etc.

During the debates, three themes received paramount attention from the delegates: the decisions/voting rules of the conference, its procedures, and its competencies (Buijtenhuijs 1993: 67). As far as the decision rules were concerned, the Tripartite Committee suggested the general rule of consensus and the voting by showing of hand. Delegates opponent to Deby’s viewed that voting method as an intimidation technique and wanted a secret ballot. Deby’s partisans argued in return that the voting method proposed was just a way to avoid electoral frauds. That controversy lasted for 10 days. Finally, it was agreed that only voting for official positions would be secret. The procedures were very much discussed for obvious practical reasons, but also because many delegates did not want the conference to be chaired by Deby’s partisans. In the end, a southerner close to Tombalbaye won the chairmanship. Another topic intensely debated was the competencies of the conference, for there were no juridical texts defining and specifying its mandate. According to delegates hostile to Deby’s regime, the conference had the power to overthrow him and to elect a new head of state that would lead the country till the adoption of a new constitution. Obviously, Deby’s partisans did not share that view. It was finally agreed that Deby would remain the president of Chad during the transition; nevertheless, the conference would elect a prime minister and appoint a transitory government.

The transitory government, called the High Council for the Transition (Conseil supérieur de la transition), was to implement the resolutions adopted by the conference and to appoint a committee of specialists to draft a constitution. In 1996, the draft was submitted to the Chadian people for approval. The proposed constitution, as far as civil liberties and rights were concerned, followed very much the orientations set by the conference; however, the opinion was not satisfied with its distribution of power—especially the powers attributed to the president—(May and Massey 2000). Despite that, many opposition leaders campaigned for the “Yes.” They reasoned that the victory of the “No” would prolong an already tiresome transition period; it was a better strategy, in their eyes, to save their strengths for the upcoming presidential battle which they thought would be easily won because of Deby’s unpopularity (Buijtenhuijs 1996). That argument received the support of the voters, and the constitution was adopted (with 71.19 % of the electorate in its favor). However, contrary to the prediction of many opposition leaders, Deby won the presidential election that followed. Later in 2005, the 1996 Constitution underwent major amendments; among other things, the presidential term limit was lifted up, the senate was suppressed, and another constitutional body was instituted (the Economic, Social and Cultural Council).

The current Cameroonian constitution was adopted in a context similar in many extents to that of Chad. As mentioned earlier, Cameroon’s first constitution was an adaptation of French 1958 constitution nonetheless with some noticeable presidentialist deformations (Bouopda 2008). Ahidjo⁷ pushed, at the time, for a strong executive arguing that the parliamentary system experimented in Cameroon before the independence was a source of instability and political deadlocks. In 1961, the reunification of French Cameroon and a part⁸ of British Cameroon into a federation prompted for a new constitution. That new constitution saw the increase of presidential powers. In 1963, with a dose of ruse and “force,” President Ahidjo united all the political parties of the country under the banner of Union Nationale Camerounaise establishing de facto a unique-party system.

Afterward, Ahidjo would use the rhetoric of fight against the “rebellion” to embellish his authoritarian drifts. In the early days of the independence, he launched a crusade against Union des Populations du Cameroun (UPC), the independentist party known for its strong opposition to the former colonial master. Although the leaders of UPC vowed to keep fighting Ahidjo’s regime which they viewed as the continuation of the French government, UPC was no longer very much of a threat. The French army, just prior the independence, had already seriously weakened UPC by killing most of its important figures and many of its followers.⁹ For the following years, almost until early 1970s, Ahidjo managed to maintain in the country an artificial atmosphere of “civil war” which made it easy for him to take “exceptional” measures and to strongly repress any form of opposition. Ahidjo quickly acquired the reputation of an autocrat; he was feared and even revered¹⁰ throughout the polity.

On May 20, 1972, the country adopted, through a referendum, a new constitution establishing a unitary state. To justify that change, Ahidjo’s government put forth the necessity to end inefficiencies and wastes caused by the federalism in place since 1961. In fact, the passage from a federal republic to a unitary one was the ultimate phase of his power concentration agenda: “The unification of the country” ended the “inconvenient” diarchal executive¹¹ of the former federal republic and weakened the powers of the Anglophone state. The 1972 Constitution was amended later in 1975 and 1979; when Ahidjo stepped down in November 1982, it became obvious that those amendments aimed at preparing his succession.¹² Ahidjo’s prime minister, Paul Biya, then became the country second president. In the early years of his presidency, Biya would struggle to impose his authority over Ahidjo’s regime barons and over Ahidjo himself¹³; in 1984, a coup was attempted against him. The same year the 1972 Constitution was again amended: The office of prime minister was suppressed, and more discretion was given to the president to choose a successor in case of incapacity.

Cameroon did not escape the wind of democratization that blew on sub-Saharan Africa in the early 1990s. Indeed, in 1990, protests for multipartyism broke out throughout the country, and despite the government yielding to that demand, the political climate during the next two years would remain tense: civil disobedience, general strike, etc. Leaders of newly created political parties, just like in Chad, asked for a Sovereign National Conference. Biya’s government refused and instead convened the Tripartite Conference that sat from October 30, 1991, to November 17 of the same year. Contrary to what happened in Chad (and in many francophone countries where the Sovereign National Conference was experimented), Biya’s government arranged to make the Tripartite Conference a simple consultative body. Furthermore, he limited the orders of business of that conference to two items: the drafting of an electoral code (elections bill) and the modalities of an equitable media access to all political parties even though the opposition also wanted the theme of constitutional reform to be included (Fodouop 2010: 280). In the view of several observers at the time, delegates close to Biya represented more than half of the total of delegates attending (Fodouop 2010: 278–279), a non-surprising outcome since Biya’s Prime minister,¹⁴ who also chaired the Tripartite Conference, unilaterally selected the participants. In the Tripartite Conference final declaration, delegates asked for the creation of a technical committee to reflect on a new and modern constitution. Afterward, the president appointed an experts committee (known as the Technical committee) to do just that; however, the upcoming legislative and presidential elections would push the issue of constitutional reform in the background.

Indeed, amidst an important economic and political crisis, legislative and presidential elections were held in March and October 1992, respectively. Biya’s party won 88 of the 180 seats in the National Assembly and had to coalesce with Mouvement démocratique pour la Défense de la République (MDR) in order to have a majority necessary to run the government. The presidential election was as disputed as the legislative election: Biya defeated John Fru Ndi by a thin margin: 39.9 % for Biya versus 35.5 % for Fru Ndi. It was after those elections that the constitutional reform debate was brought back on the table. In 1994, Biya appointed a consultative committee of 57 members (representing political parties, religious leaders, tribal chiefs, and experts) to advise him on constitutional reform proposals submitted years earlier by the technical committee. Instead of calling a referendum for the adoption of a new constitution, Biya chose to submit a bill to amend the 1972 Constitution. The scope of the amendment was so large that many Cameroonian public law experts¹⁵ simply refer to the “Law No. 06 of 18 January 1996 to amend the Constitution of 2 June 1972” as the Constitution of 1996. It is reported that during the twenty-day debate of the bill in the National Assembly, the topics most discussed concerned:

• the voting system in presidential election: the opposition favored the two-round system, while the majority supported the one-round system used for the 1992 election;

• the majority system in presidential election: the opposition advocated the absolute majority system;

• the designation method of senators and regional executives (Owona 2012: 82).

The National Assembly finally adopted the Law No. 06 of 18 January 1996 to amend the Constitution of 2 June 1972 with 160 votes for, only 2 votes against, and 8 abstentions. That law would be amended as well in 2008: The amendment mostly concerned the office of president and the successoral mechanism in case of vacancy.

After this look into the background of Cameroonian and Chadian constitutions, we now turn to their analysis in terms of power relations as stipulated by the VOIP framework.

2 Chad and Cameroon Constitutions in Terms of Power Relations

The constitutional corpus used in our analysis for Cameroon is made up of the Law No. 06 of 18 January 1996 to amend the Constitution of 2 June 1972, also known as the Constitution of 1996: The 2008 amendment was not included. For Chad, we used only the initial 1996 Constitution (the subsequent amendment of 2004 was therefore excluded). We used the official French version of the aforementioned texts. Reading Chadian and Cameroonian constitutions in terms of power relations involves first unitizing and second coding.

2.1 Description of Constitutions in Terms of Power Relations

Unitizing here is the process of identifying power relations in the constitutional text under analysis. A power relation in that process is understood as a discursive structure that links an agent with a capacity (Imbeau and Jacob 2012). In the Chadian and Cameroonian constitutions, we found, respectively, 541 and 415 units of analysis or power relations with an average Krippendorf statistics¹⁶ of 0.92, well above the commonly accepted threshold of 0.80. In relative terms, however, we found the Chadian constitution less dense than the Cameroonian one: 49.6 power relations per thousand words versus and 53.1. Of the 16 constitutions analyzed in the VOIP, the two African constitutions are above the mean density of 46.7 (Table 1).

Table 1 Power relations in Chadian and Cameroonian constitution: summary statistics

The Cameroonian constitution has a preamble and 69 articles distributed in thirteen parts. Interestingly, we found 38 power relations in the preamble. Moreover, parts with most power relations are part 4 which treats the relationship between legislative and executive, part 3 which deals with legislative powers and part 7 on the Constitution Council: 87, 80, and 45 power relations, respectively. On the other hand, the sections with the least power relations are part 6 on treaties and international agreements, part 9 on The Economic and Social Council with both no power relations at all, and part 8 on the court of impeachment that has only 6 power relations. Table 5 in Appendix gives us in detail the number of power relations in each section of the Cameroonian constitution.

The 1996 Chadian constitution is made up of a preamble and 239 articles arranged in fifteen sections or parts. Like in the Cameroonian case, we found power relations in the preamble: 25. The sections with outstanding number of power relations are as follows: part 3 on executive power with 139 power relations, followed by the section on the relationship between the executive and legislative power which has 92 power relations and finally, part 2 on the fundamental rights and liberties and duties which has 60 power relations. Sections with the least power relations are part 12 on traditional chiefs, part 8 on the Court of impeachment, and part 14 on the amendment with, respectively, 4, 6 and 9 power relations.

In summary, unlike the Cameroonian constitution, the Chadian constitution has at least a power relation in every section. Like in the Cameroonian case, we found a relatively important number of power relations in the preamble of the Chadian constitution. Furthermore, we found, both in the Cameroonian and Chadian cases, that the relationship between the executive and legislative branches was an area filled with power relations; however, in the case of Chad, the section on the relationship between the legislative and the executive is behind the section on the executive branch in terms of number of power relations. This presentation of Chadian and Cameroonian constitutions already gives us some clues on the areas of uncertainty at the time of the drafting/adoption of those texts. We need to characterize the power relations extracted to make some inferences.

2.2 Characterization of Power Relations

Coding here is understood as the process of characterizing power relations according to the source, the type, and the direction of the power involved. The source of power is the resource that an agent can mobilize to “perform the action he has the capacity to perform or to make another perform [that] action” (Imbeau and Jacob 2012); that resource can be authority, wealth, or knowledge. Of the 415 power relations extracted in the Cameroonian constitution, we found that authority is the resource mobilized by agents 314 times, versus 1 and 38, respectively, for wealth and knowledge. In the Chadian constitution, the power resource used by an agent is authority in 401 power relations over 541, which represents a proportion close to that found in the Cameroonian case. Also in the Chadian constitution, wealth and knowledge are mobilized by an agent only 10 and 48 times, respectively. These results are consistent with the VOIP first hypothesis according to which power relations found in constitutions would be more political than economic or preceptorial. Tables 2 and 3 provide a characterization of power relations found in Chadian and Cameroonian constitutions according to the source of power.

Table 2 Source and type of power relations in the Cameroonian constitution

Table 3 Source and type of power relations in the Chadian constitution

Power relations in the VOIP framework can be of two types: social and instrumental. They are coded as social when the exercise of a capacity by the influencing agent can alter the choice set of another explicitly identified agent. They are coded as instrumental when the influencing agent is said to have the capacity to act on things or events, and they are coded as indeterminate when the text does not provide us with enough information to unequivocally characterize a power relation according to the two types just mentioned. In the Chadian constitution, we found 297 instrumental power relations and 244 social power relations. In the Cameroonian constitution, we found 211 social power relations versus 204 instrumental ones. Tables 2 and 3 show the total number of power according to the type of power involved in the Chadian and the Cameroonian constitutions.

In both cases, the difference between the number of social and instrumental powers is small. It is therefore with a dose of caution that we say VOIP second hypothesis according to which social power will dominate instrumental power is not verified in the Chadian case, but it is, loosely, in the Cameroonian case. Even when we cross the type and the source of power, we still cannot say decisively for each categories of source of power whether social power dominates instrumental power.

Power relations can also be characterized according to their direction which can be either positive when the text ascribes agents with a capacity or negative when it denies agents the capacity to do something. We found in the Chadian constitution 521 positive power relations and in the Cameroonian constitution a similar overwhelming number of positive power relations: 411 which represent about 99 % of all power relations extracted. Clearly, the VOIP third hypothesis that constitutions would tend to restrict rather than ascribe power is rejected. Table 4 presents power relations extracted in the Cameroonian and Chadian constitution according to the direction.

Table 4 Direction of power Cameroonian and Chadian constitutions

In summary, the results presented here confirm the VOIP first hypothesis in the Cameroonian and Chadian cases. The second hypothesis is rejected in the case of Chad, but it is not firmly accepted in the case of Cameroon. Finally, the third hypothesis is unequivocally rejected in both cases. What can be made of these results? Can we tell whether or not Cameroonian and Chadian Constitutions were adopted behind a veil?

3 Discussion

Only unambiguous consistency between the results of our three tests of the VOIP hypotheses would have allowed us to affirm that the Cameroonian and Chadians constitutions were written behind a veil. Looking at the origins of the two documents, we saw how, throughout time, the office of the president became the center of political and institutional life in Chad and Cameroon. The first presidents of these countries succeeded in obtaining several constitutional powers thanks to the repressive resources at their disposal. This gradual concentration of powers in the office of the president eventually made it the most sought after and coveted political office those countries. Even during the popular conferences held in Chad and Cameroon in the early 1990 supposedly to discuss the general reform of the state and to provide new economic and social orientations, the rules concerning the office of the president remained the actors’ main focus. It therefore comes as no surprise that constitutions adopted in those contexts dealt essentially with political power as we found. That political power is the primal concern of these constitutions indicates according to the VOIP, the presence of a veil at the time of their making.

In the case of Chad and Cameroon, using only public records, there is no easy way to determine the level of uncertainty faced by the actual drafters of the Chadian and Cameroonian constitutions. Nonetheless, because of their central role throughout the making process, we are not losing too much by looking at uncertainty from the standpoint of Deby and Biya who were, respectively, president of Chad and Cameroon at the time. In the light of the context surrounding the making of Chadian and Cameroonian constitutions, one can say that Deby and Biya faced some uncertainty in regard notably to their political powers in the early 1990s, but that uncertainty quickly faded out as the process advanced. Indeed, in the early 1990s, for the first time and even the only time since their independence, the presidents of Chad and Cameroon faced a sort of “social revolt”: openly confrontational and critical protests against their governance. However, both Biya and Deby, despite that threat to the stability of their regime, skillfully managed to mitigate the uncertainty they faced. In the midst of social contestations, they feigned yielding to the people’s demand, but they subtly made sure to remain in control. In the case of Chad, Deby ensured his control over the Sovereign National Conference during the participants’ selection process. For example, the prefects of his administration were in charge of selecting the 168 delegates to represent the Chadian people (tribal chiefs, sultans, etc.). Another thing that helped Deby stay in control was probably the lengthening of the constitutional making process. The process started in 1991 in the midst of social protests, but ended up only in 1996: This certainly contributed to wear the opposition out.

In the case of Cameroon, although Biya promulgated a set of laws allowing and extending some civil liberties¹⁷ to ease up the tensions, he never yielded to the demand of a Sovereign National Conference and proposed a mild form instead: the Tripartite Conference. Moreover, his government selected the participants and he himself fixed the agenda of that Tripartite Conference. Also, Biya choose to amend the 1972 Constitution instead of proposing a new Constitution. As we said earlier, the scope of the amendments and the orientations of the Law of 1996 were such that the option of a new constitution was probably more suitable. However, having faced strong social protests in 1990–1991 and having almost lost the 1992 presidential election, it was not in Biya’s interest to seek Cameroonian people approval for a new constitution which favored a strong presidency. Passing the amendment through the National Assembly was a better option: Although his party did not have the absolute majority, the main opposition party (SDF) was not represented in the Assembly during that term, for it had boycotted the 1992 legislative elections and the rest of the political parties were ideologically close to Biya’s CPDM. Finally, like in Chad, the making process was somehow unnecessarily lengthened with the same consequence: the weariness of the opposition. In 1996, when the amendment was passed in the Assembly, the country was back on pre-1990 routine tracks: The bill was passed almost without any opposition. All this showed that we could not reasonably expect Cameroon and Chad Constitutions to mark high on uncertainty measures developed by the VOIP. The measure on sources of power is contrary to that expectation; however, the results on the two other measures tend to support it.

Assuming that our reading of the historical context was correct (that there was little uncertainty toward the end of the making process), we expected more instrumental power relations than social ones in the Chadian and Cameroonian constitutions. This was the case for Chad but not for Cameroon, with a dose of caution though. These results raised a question not explicitly addressed in the VOIP. Since unitizing and coding were done by humans, the VOIP provided us with an overall measure of reliability: The Krippendorf alpha statistics mentioned above. The VOIP, however, failed to present results adjusted for a probability of error. This adjustment is very important when the call is really close like in the Cameroonian case where we had 211 social power relations versus 204 instrumental power relations. Some form of adjustment is needed to tell whether or not the difference between the numbers of social power relations found in a constitution is statistically different than that of instrumental power relations found in the same constitution. At first, the VOIP second hypothesis is confirmed in the case of Cameroon; however, if we were to take into account possibilities of wrong unitizing/coding, we are not sure to confidently make the same call. It is only with adjusted results that we could make strong interpretations of the uncertainty measured developed by the VOIP.

When looking at direction of power, the adjustment is not very much a big deal because the difference between the number of positive and negative power relations in both constitutions was important. That we find such a big number of positive power relations is in line with our expectations after reading the context of the making of the Chadian and Cameroonian constitutions: These numbers indeed suggest the absence of uncertainty at the time of the making. However, even here, we are unable to say it confidently. It might be the case that we found more positive power relations simply because writing in positive form is tradition in legal and juridical drafting. Indeed, Ard (2010) make the point on the existence of a legal writing tradition. He explained that in many legislative bodies like the US Senate or the Congress, there are explicit writing rules provided to drafters that definitely affect their legal writing style. On a US government Web site, one can read the following instruction on drafting legal documents:

6. Write positively: If you can accurately express an idea either positively or negatively, express it positively. DON’T SAY: The Governor may not appoint persons other than those qualified by the Personnel Management Agency. SAY: The Governor must appoint a person qualified by the Personnel Management Agency.¹⁸

These drafting instructions are not peculiar to US legislative bodies, they can be found on several legislative bodies’ Web sites around the world: in France, Australia and New Zealand, to name a few. It seems also that those who have studied law at some point in their curriculum have encountered a legal drafting manual since there are so many of them in the field. All this suggests that there might indeed be a tradition among lawyers to “write positively” when it comes to drafting legal texts. If it is the case, we could not interpret the big number of positive power relations found in Chadian and Cameroonian constitutions as indication of the absence of uncertainty at the time of the making of those texts.

4 Conclusion

The goal of this case study was to apply the VOIP framework to the Chadian and Cameroonian constitutions. Specifically, we wanted to know whether or not those constitutions were made under the settings described by the theoretical literature on constitutional choices: a setting marked by uncertainty. The VOIP first step consists in identifying the major events that led to and prevailed during the making of the texts under analysis. In the second step, we use that information to interpret three measures of uncertainty: the source, the type, and the direction of power. The source of power in a constitution written under uncertainty will be political, while the type will be social and the direction negative. Our interpretation of the context of the making of our two texts led us to believe that incumbent presidents faced little uncertainty at the time. We therefore expected the uncertainty measures to be consistent with that intuition. However, results on source of power indicated the presence of uncertainty, results on type of power indicated the presence of uncertainty in the case of Cameroon but not in the case of Chad, and results on direction of power in both cases indicated the absence of uncertainty at all. This inconsistency in our results prompted us two remarks: first, in VOIP, we need to present extracting and coding results adjusted for errors; second, lawyers’ legal writing style more than the presence/absence of uncertainty can be a better explanation of the direction of power in constitutions.

Appendix

Table 5 Power relations in the Cameroonian constitution per section