Restoring Constitutional Order after Coups

Following successive military coups, Mali, Chad and Burkina Faso have been undergoing complex political transitions. Central to these post-coup transitions is a commitment to engage in a constitutional reform process to restore constitutional and democratic order. Mali enacted a new constitution on 22 July 2023. Chad promulgated a new constitution on 29 December 2023 and adopted a package of amendments on 8 October 2025. In Burkina Faso, transitional authorities have expressed their intention to draft a new constitution but have so far made only limited amendments to the existing 1991 Constitution. This report also discusses the 2017 draft Constitution, as stakeholders in Burkina Faso may use it as a reference or starting point for deliberations and negotiations.

Given the severe security crises and the structural economic and environmental challenges affecting these three countries, a key challenge for these constitution-making processes is to design constitutions that enable effective executive action to respond to the multidimensional crises while providing sufficient safeguards to ensure the regime does not descend into authoritarianism.

This report conducts a comprehensive comparative analysis of the 2023 Constitution of Mali, the 2023 Constitution of Chad and the 2017 draft Constitution of Burkina Faso. It analyses the similarities and differences among these constitutional texts and evaluates progress and potential setbacks to democratic governance within them. The main findings can be summarized as follows.

1. Individual assessment of the three constitutions

Mali’s 2023 Constitution can be described as an ‘emergency’ constitution. It establishes a hyper-presidential system and a centralized state structure that prioritizes presidential leadership and rapid, short-term decision making over inclusive deliberation, political compromise, and checks and balances. Although it ostensibly seeks to create a framework more capable than the 1992 Constitution of responding to the country’s severe security and economic crisis, this new political system concentrates tremendous, unchecked powers in the presidency. This new dispensation may increase the stakes of presidential elections, intensify winner-takes-all politics and pave the way for authoritarian drift. The Constitution also fails to reflect the commitments to extensive and asymmetric decentralization made in the 2015 Algiers Peace Agreement, which may lead to further tensions and destabilization.

The 2023 Constitution replaces the semi-presidential system introduced in the 1992 Constitution with one that is essentially unique and hyper-presidential, in which all executive power is vested in the directly elected president. The offices of the prime minister and cabinet are retained, but they are appointed and dismissed at will by the president, with no parliamentary authority to remove them. National policy orientations are defined by the president; the prime minister and cabinet operate essentially as the president’s implementing agents. Despite this shift, the text retains several constitutional devices of so-called rationalized parliamentarism found in francophone semi-presidential systems, which were originally designed to enhance government stability and enable the executive to govern without a stable majority in the legislature. Notably, the Constitution retains limits on the scope of the legislative domain and grants the executive autonomous regulatory power over policy areas not explicitly assigned to the legislature. It also preserves powerful tools that enable the executive to dominate lawmaking, bypassing parliament through direct legislative referendums and broad ordinance-making powers, effectively controlling parliamentary proceedings, including through a dominant role in proposing bills and amendments, control over the legislative agenda and procedures such as blocked voting. Crucially, the Constitution also retains the president’s discretionary power to dissolve the legislature and exercise broad emergency powers. Retaining such devices in a system where the executive is not accountable to the legislature strengthens the powers of the president significantly and distorts the logic of a presidential system. Overall, the combination of broad decree and referendum powers, considerable influence over the lawmaking process, and broad and unchecked emergency powers enables the president to govern and decide unilaterally on a wide array of policy matters. This expansion of presidential prerogatives lacks commensurate oversight and balance.

The lack of oversight of the president is partly reflected in the involvement of the president (as guarantor of the independence of the judiciary and chair of the Supreme Council of the Judiciary) in the administration of the judiciary, including in judicial appointments. While the Constitution re-establishes a strong constitutional court, with diverse membership and enhanced guarantees of independence, the court may lack constitutional authority to constrain a formally powerful president. In fact, the court may be more likely to constrain the legislature to the benefit of the president. Legislative oversight mechanisms are also weak. The impeachment procedure establishes exceptionally high thresholds, which, in practice, limit its usefulness as an accountability mechanism to a substantial degree. The limited oversight of the presidency also stems from the president’s broad appointment powers in the civil and military administrations, exercised without the involvement or scrutiny of the legislature, and from the lack of effective guarantees of independence for regulatory and oversight institutions. The armed and security forces—which are under the hierarchical authority of the president—are also assigned an expanded role, as they may now participate in the ‘economic, social, and cultural development and environmental protection of the country’ and are also responsible for ‘ensuring the enforcement of the law’ (article 89). In this context, the only possible checks on the presidency are presidential elections and the constitutional limit of two five-year presidential terms, which is unamendable, but these mechanisms will not be helpful between elections.

The bill of rights includes most recognized rights, but they are weakly protected, as most are subject to regulation by law. The Constitution introduces ex post judicial review of legislation through incidental referral. It allows any litigant, during proceedings before an ordinary court, to request that the court refer a contested legislative provision to the Constitutional Court for a ruling on its constitutionality. This is a positive development for the protection of constitutional rights.

Regarding the form of the state, the Constitution maintains a unitary decentralized state structure but provides limited guarantees and detail regarding decentralization arrangements. All the substance of these arrangements is left to ordinary law. Importantly, the commitments to extensive and asymmetric decentralization made in the Agreement for Peace and Reconciliation resulting from the Algiers Process concluded in 2015 between the Government of Mali and the Coordination of Azawad Movements are not reflected in the Constitution. Similarly, the commitment to increasing the representation of communities from northern Mali in decision making at the national level is weakly operationalized. Despite the establishment of a senate composed of three-quarters of members representing decentralized territorial authorities, the influence of decentralized entities on legislation is likely to remain limited, as the National Assembly, upon request of the government, may make the final decision on all bills in case of disagreement between the two chambers. Moreover, while the Constitution does not specify the appointment process, one-quarter of senators will be appointed, and the process is likely to involve the president of the republic.

Chad’s 2023 Constitution addresses the overconcentration of power in the presidency that characterized the 2018 and earlier constitutions. It establishes a premier-presidential semi-presidential system—uncommon in francophone Africa—which introduces more checks and balances within the executive and provides a basis for a relatively decentralized system of governance. In principle, these features could enhance political inclusion and deliberation, reduce winner-takes-all politics and, more broadly, support the emergence of a new political dispensation. However, some important concerns remain, and these have been further compounded by the constitutional amendments adopted by Chad’s Government in October 2025, which profoundly altered the balance of power in favour of the presidency.

The 2023 Constitution replaces the presidential system under the 2018 Constitution, in which a directly elected president held all executive power, with a semi-presidential system in which executive power is shared between a prime minister and government, accountable to the legislature, and a directly elected president. Crucially, although the president appoints the prime minister, only the legislature can dismiss the prime minister and cabinet (premier-presidential system). The prime minister holds significant powers, including the authority to select cabinet members, and jointly exercises with the president the power to appoint officials to key civilian and military institutions. Moreover, national policy orientations are defined by the government rather than by the president, as was the case under the 2018 Constitution. This shift to a premier-presidential system formally reduces the dominance of the president compared with the 2018 and earlier constitutions.

The original version of the Constitution envisaged a maximum of two consecutive five-year terms for the president, without limiting the total number of terms a person could serve (in contrast to the 2018 Constitution, which limited the presidency to two six-year terms). However, the term limit provisions were not among the unamendable provisions. The October 2025 amendments extended presidential terms to seven years and abolished the maximum of two consecutive terms, thereby allowing the incumbent president to be re-elected for an indefinite number of terms. These amendments also lifted the prohibition on presidential involvement in political parties, thus allowing incumbent President Mahamat Idriss Déby Itno to head the ruling party founded and previously led by his late father, who ruled the country as president for over three decades. Taken together, the October 2025 amendments recentralize authority in the presidency to a significant extent, weaken formal constraints on executive power and reduce institutional guarantees for political alternation.

At the national level, the legislature consists of two chambers (as under the original 1996 Constitution). The newly established Senate is intended to represent decentralized units of government, but the president of the republic appoints one-third of the senators, giving the president considerable influence over legislative matters. Regarding the judiciary, the Supreme Judicial Council, which is responsible for selecting judges and managing their careers, is now chaired by the president of the Supreme Court, and not by the president of the republic. While this change is commendable, the effective independence of both the Judicial Council and the judiciary will depend on organic laws. Moreover, the Constitution leaves crucial aspects of the mandate, composition and functioning of the Constitutional Court to legislation.

The bill of rights includes most of the rights recognized in international and comparative practice, but these rights are weakly protected, as most are subject to regulation by law.

Regarding the form of the state, the Constitution largely maintains the unitary decentralized state structure introduced in the 2018 Constitution. While this framework provides the basis for significant decentralization of powers, the substance of the arrangements is left to organic law, without specific timelines for adopting the necessary laws. Existing legislation does not provide for high levels of decentralization; this arrangement may persist despite the constitutional changes.

On paper, the constitutional amendment procedure appears to be sufficiently stringent to prevent self-serving or partisan changes—requiring a two-thirds legislative majority and approval in a referendum for most amendments, and a three-fifths legislative majority for so-called technical revisions. In practice, however, the country’s history of winner-takes-all politics, which regularly gives the president’s party a supermajority in parliament, combined with the Constitution’s overall minimalism, which defers key matters to legislation, may allow transient political majorities to unilaterally define and redefine, through amendments and legislation, a core set of institutions, procedures and individual rights that are critical for the functioning of constitutional democracy. The constitutional amendments adopted on 8 October 2025 as a technical revision illustrate this risk.

Burkina Faso’s 2017 draft Constitution aims to address the excessive concentration of power in the president. The proposed changes go further than the 2015 amendments to the 1991 Constitution adopted during the political transition following the fall of President Blaise Compaoré. In comparison, the 2017 draft Constitution seeks to create a better balance between the executive and the legislature, enhance the independence of the judiciary and strengthen guarantees for political alternation. However, the president of the republic retains important powers that may enable dominance over the executive branch.

The draft maintains a president-parliamentary form of semi-presidentialism in which executive power is shared between a prime minister and government, which is accountable to the legislature, and a directly elected president. However, the president retains the power to dismiss the prime minister and cabinet at will, a feature that strengthens presidential authority within the executive substantially. In president-parliamentary systems, this institutional arrangement has often been associated with reduced executive power sharing and weaker constraints on presidential authority. Currently, such systems of government are mainly found in authoritarian countries (e.g. Azerbaijan, Belarus, Russia, as well as some francophone African countries). Moreover, the draft lacks clarity with regard to the division of responsibilities between the president and the prime minister. The president would define the policy orientations of the state, and the government would implement them. In situations of cohabitation, where the president lacks a supportive legislative majority, the government would define and implement public policy except in key domains reserved to the president. However, the draft does not specify these reserved domains. Similarly, the prime minister must countersign presidential acts, except acts that fall within the exclusive powers of the president (article 79), but the draft does not specify which acts fall within this category. As in other semi-presidential regimes, the executive has available several tools for influencing the lawmaking process and marshalling the unicameral legislature. The draft contains robust provisions on presidential term limits, permitting a maximum of two five-year terms, with both the number of terms and the length of the terms made unamendable.

The draft potentially enhances the independence of the judiciary by granting the Supreme Judicial Council decision-making authority over the appointment of judges and prosecutors. However, the effectiveness of this improvement will depend on the organic laws defining the procedure for appointing Council members. The draft also strengthens the independence of the Constitutional Court by reducing the role of the president of the republic in determining its composition: the president would appoint only two out of nine members. The draft also broadens access to the Court by granting it the power to initiate proceedings on its own initiative (self-referral power) and by permitting individuals to directly petition the Court for ex ante judicial review of bills before their promulgation (alongside the existing ex post review of enacted bills in the context of court proceedings). The Court’s authority to review bills on its own initiative or at the request of individuals risks potentially disrupting the legislative process, overburdening the Court and exposing it to criticism for judicial activism. The draft also constitutionalizes several oversight and regulatory institutions but does not provide sufficient guarantees for their independence, as their composition and responsibilities are to be determined by organic statute.

The draft Constitution expands the list of constitutional rights. It also references international human rights instruments related to the rights of women and youth and includes provisions guaranteeing their right to participate in the conduct of public affairs. It also guarantees certain social and economic rights that are only recognized—not guaranteed—under the current Constitution. However, fundamental rights and freedoms are weakly protected, as most are subject to regulation by law.

Regarding the form of the state, the draft maintains a unitary structure, with decentralization arrangements to be defined by law.

2. Comparative analysis of the three constitutions

Compared with anglophone and lusophone constitutions in Africa, these three constitutional texts provide little detail and leave several important elements to be defined in legislation. The texts defer to organic or ordinary legislation the definition of important procedures and elements that are critical for the functioning of a constitutional democracy, including the criteria and procedures for limiting human rights; the system for electing the legislature, as well as key aspects of presidential elections; the procedures for appointing and removing constitutional court judges; the mandate and composition of important independent state institutions such as the judicial service commission, the electoral management body or the human rights commission; the status and prerogatives of the opposition; and the responsibilities of decentralized territorial entities. Although the three constitutions vary in their level of detail, this approach reflects the minimalism that characterizes francophone African constitutions. In contexts where a single party often dominates the legislature—as has mostly been the case in these three countries—this arrangement could allow transient majorities to unilaterally shape and reshape institutions, procedures and individual rights that are critical for the functioning of constitutional democracy. Although constitutions should not intend to be exhaustive, safeguarding multiparty competitive democracy requires regulating a core set of institutions, procedures and individual rights in the constitution itself and protecting them from being changed unilaterally by an incumbent governing majority. In contexts of post-military-coup transitions, however, where the constitution-making process may lack inclusivity and may be dominated by the military and transitional authorities (as in Chad and Mali), the level of detail in the constitution poses a dilemma. On the one hand, regulating all core components of multiparty democracy in the constitution may provide an opportunity for transitional authorities, with questionable legitimacy, to entrench rules in their favour and consolidate their power, making it difficult for future elected majorities to modify these provisions and limit their incentive for doing so, as they would themselves likely benefit from such arrangements. On the other hand, deferring key institutional issues and procedures to legislation risks majoritarian dominance, allowing newly elected majorities to change these rules to their advantage without the need for cross-party negotiations or the support of opposition parties. This tension highlights the importance of inclusive constitution-making processes and of constitutional texts that provide sufficient detail and guarantees to protect the core aspects of the democratic system.

These three texts concentrate significant powers in a directly elected president, although with important variations. This concentration of authority, including broad emergency powers, in the presidency may facilitate more centralized and rapid executive decision making in response to the security and economic crises affecting these countries and the region. However, concentrating power in the presidency may also increase the stakes of presidential elections and, in the absence of commensurate oversight mechanisms, risks enabling the president to govern unilaterally on a wide range of policy matters. In Mali, all executive powers are exercised by a directly elected president who can unilaterally appoint a large number of key civilian and military officials and possesses considerable tools for influencing the lawmaking process and the administration of the judiciary. In Burkina Faso, executive authority is shared between a directly elected president and a prime minister and cabinet responsible to the legislature. Nevertheless, the president may still dominate the executive branch, as they define the general policy orientation of the nation and can dismiss the prime minister at will. In Chad, executive authority is also shared between a directly elected president and a prime minister and cabinet responsible to parliament. Powers within the executive branch are divided and arguably better balanced because the president cannot dismiss the prime minister at will, and the general policy orientation of the state is defined by the Council of Ministers rather than solely by the president. Nevertheless, given Chad’s history of winner-takes-all politics—which regularly gives the president’s party a supermajority in parliament—combined with the abolition of presidential term limits in October 2025, this constitutional arrangement may still be insufficient to prevent presidential dominance.

These texts retain many similarities with the Constitution of France, some of which may indicate insufficient adaptation to the local context. Notably, they replicate several devices and powerful government instruments of so-called rationalized parliamentarism, including substantive limitations on the legislative domain, which leave the executive with authority to regulate policy matters that do not explicitly fall within the legislative domain (Burkina Faso, Chad and Mali); tacit voting procedures that enable the government to adopt legislation without a vote in parliament (Burkina Faso and Chad); ordinance procedures that enable the government to enact executive acts on policy issues that are normally within the legislative domain (Burkina Faso, Chad and Mali); exceptional powers of the president, which permit the president to unilaterally take any measures deemed necessary in emergency situations (Burkina Faso, Chad and Mali); and presidential referendum powers, which enable the president to bypass the legislature and submit a text directly to the people for approval (Burkina Faso, Chad and Mali).

These government instruments were introduced in France’s 1958 Constitution to address the dominance of the legislature and the resulting government instability that characterized the Third and Fourth French Republics. While some of these constitutional devices—provided that adequate safeguards are in place—may be relevant for the context in Burkina Faso, Chad and Mali in addressing severe security and economic crises, other instruments and features may indicate insufficient adaptation to the local context and a reliance on certain imported constitutional arrangements. Notably, presidents in these countries have historically enjoyed broad support in the legislature. Accordingly, replicating such instruments in contexts where the legislature and fourth-branch institutions have been relatively weak may formalize and further strengthen the dominance of the executive branch within the political system.

Although these countries are heterogeneous in terms of ethnicity, geography and history, their constitutional texts establish unitary state structures with weak guarantees for decentralization and no asymmetric arrangements. While all three constitutions recognize the existence of decentralized entities, the responsibilities and financing of such entities are deferred to legislation. The legislature therefore unilaterally determines and can freely modify the scope of decentralization or decide to recentralize certain responsibilities at will. In other words, significant demands (Chad) and commitments made in peace agreements (Mali) to extensive decentralization are not reflected in the constitutional texts. Existing laws on decentralization have not resulted in significant levels of autonomy for decentralized entities. In addition, existing laws grant deconcentrated authorities (i.e. agents of the central government in each decentralized entity) significant substantive power over decentralized authorities through their oversight and advisory roles, further limiting the autonomy of decentralized entities.

While these constitutional texts do not preclude the development of highly decentralized states, the actual degree of decentralization will depend on whether existing laws on decentralization will be amended, the substance of these amendments and their implementation. Given that these constitutional texts do not establish timelines or implementation processes and bodies, progress will depend on the will of the political majority. Furthermore, none of these texts provides for the possibility of asymmetric arrangements (although they do not specifically preclude them), whereby certain parts of the country could be granted more autonomy on specific policy issues than others. These constitutional arrangements may therefore be contested and could contribute to the resurgence of conflict between central governments and certain non-state armed groups in Chad and Mali, where demands for extensive (asymmetric) decentralized arrangements persist. Burkina Faso may also face increasing territorially based political claims that could affect constitutional debates.

Between August 2020 and August 2023, eight successful military coups d’état took place in Central and West Africa—in Mali (August 2020 and May 2021), Chad (April 2021), Guinea (September 2021), Burkina Faso (January and September 2022), Niger (July 2023) and Gabon (August 2023)—a situation unprecedented in the past two decades.

After seizing power, all the coup leaders announced a period of political transition during which constitutional and institutional reforms would be enacted to guide the return to constitutional order and transfer power to a democratically elected government. In each of these countries, the coup-makers sought to legitimize their position during the post-coup transition by enacting interim legal frameworks of some sort (Zulueta-Fülscher and Noël 2022). In Burkina Faso (after the January 2022 coup), Chad, Gabon and Mali (after the 2020 coup), transitional charters complementing the existing constitutions were unveiled; in Guinea, a transitional charter was enacted to replace the existing Constitution; and in Niger, the extant Constitution was suspended, and a series of decrees established two interim institutions. Besides establishing interim governing institutions, these temporary legal frameworks included commitments to, among other things, conducting constitutional reforms—either explicitly through a commitment to adopting a new constitution (Chad, Gabon and Guinea) or more implicitly through a duty to engage in institutional reforms with a view to rebuilding the state (Burkina Faso and Mali)—and to holding elections to end the transition. In all cases, constitutional and institutional reforms were planned before elections.

Among these six countries, Chad and Mali adopted new constitutions in 2023; Gabon, in 2024; and Guinea, in 2025, while Burkina Faso is conducting technical discussions on constitutional and institutional reform.

1. In Mali, following a series of national consultations that recommended drafting a new constitution, the president of the transition appointed two consecutive constitutional commissions by presidential decree, in June 2022 and January 2023, respectively (Republic of Mali 2022, 2023a). The first commission was mandated to produce a preliminary draft of the new constitution within two months. Then, a finalization commission was charged with finalizing the draft within two weeks. The commission submitted the final draft of the new constitution on 27 February 2023 (Republic of Mali 2023b), and its contents were validated by the president of the transition (Maiga 2023). According to the official results, the Constitution was adopted in a referendum on 18 June 2023, with a 96.8 per cent majority in favour and 38.2 per cent turnout (Cour Constitutionnelle de la République du Mali 2023), and promulgated on 22 July 2023 (Republic of Mali 2023c).
2. In Chad, transitional authorities organized a national dialogue in September 2022 that recommended drafting a new constitution based on the 1996 Constitution (Republic of Chad Dialogue National Inclusif et Souverain 2022). During the first half of 2023, a 15-member constitutional committee appointed by the executive prepared a preliminary draft constitution. The transitional government subsequently reviewed and approved the draft on 1 June 2023. The transitional legislature then endorsed it on 27 June 2023 (RFI 2023). The Constitution was approved in a referendum on 17 December 2023 and promulgated the following day (Republic of Chad, Présidence de la République du Tchad 2023). On 3 October 2025 the legislature adopted a series of constitutional amendments (Republic of Chad 2025).
3. In Burkina Faso, the Transitional Charter tasks transitional authorities with ‘[initiating] political, administrative and institutional reforms’ (article 2). Both the president of the transition and the prime minister have repeatedly expressed their commitment to drafting a new constitution to replace the existing 1991 Constitution (Le Monde 2023; Republic of Burkina Faso 2023; Dabiré 2024). The transitional authorities amended the 1991 Constitution in January 2024 (Burkina Faso 2024b) but have not yet initiated a comprehensive constitution-making process. When negotiating the basis for returning to constitutional order, stakeholders in Burkina Faso may not have to start from scratch (Noël 2022, 2023). If there is sufficient agreement, they could use the 2017 draft Constitution as a reference or basis for negotiating the country’s future constitutional text. The draft was developed by the multistakeholder 92-member constitutional commission from April 2016 to December 2017, following the transition period sparked by President Blaise Compaoré’s popular overthrow in October 2014. The commission developed and adopted the 2017 draft by consensus, but the draft was not submitted to a referendum for adoption. Although the circumstances have changed since the development of the draft, notably the worsening of the security crisis, the draft can still provide a useful framework to build upon.

This report conducts a comprehensive comparison of the content of the 2023 Constitution of Mali, the 2023 Constitution of Chad and the 2017 draft Constitution of Burkina Faso. The aims of this comparative analysis are as follows:

• to evaluate progress and potential setbacks for democratic governance within these constitutional texts—which would form the basis for the restoration of constitutional order—compared with the constitutions in force before the coups;
• to shed light on subtle yet significant differences among these texts, focusing particularly on four constitutional design issues pertinent to francophone jurisdictions in Central and West Africa: (a) the level of detail in the constitution (which determines the scope of freedom for elected majorities to unilaterally set key institutional rules of the democratic game); (b) the design of the political system and the powers of the president; (c) authority over and the mandate of the armed and security forces; and (d) arrangements for decentralization;
• to provide, where relevant, examples of comparative good practices, especially from anglophone and lusophone countries that have contributed to securing democratic progress and stability since the early 1990s; and
• to identify, where relevant, similarities between these constitutional texts and France’s Constitution, and to underscore provisions that might indicate a lack of adaptation to context or a default reliance on imported constitutional arrangements.

Mali, Burkina Faso and Chad are facing severe security crises and structural economic and environmental challenges. In Mali, the security crisis expanded in 2012 when the National Movement for the Liberation of Azawad aligned with extremist armed groups in the fight against the government to take control of the north of the country. A peace agreement was signed between the Government of Mali and the Coordination of Azawad Movements in 2015, but the extremist groups, which are pursuing broader claims with no boundaries, have become more prominent and have continued to expand their operations, notably in the northern and central regions of Mali. Growing insecurity contributed to the coup and has been central to the military government’s rhetoric and policies. In January 2024 the transitional government of Mali abrogated the peace accord.

Since 2015 Burkina Faso has been facing an escalating security crisis due to attacks by extremist armed groups targeting civilians and government infrastructure in several regions. It is estimated that, as of January 2026, over half of Burkina Faso’s territory lay outside effective government control. In these areas, various militant Islamist groups are fighting for control of territory. The military coups of January and September 2022 aggravated the security crisis. Since the coups, the number of people killed by Islamist militants has nearly tripled compared with the 18 months preceding the January 2022 coup (Africa Center for Strategic Studies 2023). The expanding reach of extremist activities around the capital, Ouagadougou, puts the country in a very fragile situation.

Chad has experienced persistent security threats from the Islamic State’s West Africa Province (formerly Boko Haram) in the Lake Chad Basin border areas. The country has also endured conflict between the central government and various politico-military groups demanding a greater share of power and more regional autonomy. On 8 August 2022 the transitional government and 43 politico-military groups signed a peace agreement in Doha, Qatar. The Doha Peace Accord provided for, among other measures, a ceasefire; a disarmament, demobilization and reintegration programme; and the participation of signatory armed groups in the national dialogue held from August to October 2022. However, 18 politico-military groups, including some of the main armed factions, did not sign the Peace Accord.

The security crises have exacerbated the economic and development challenges faced by these countries. Burkina Faso, Chad and Mali rank among the 10 countries with the lowest Human Development Index scores (UNDP 2022) and among the countries with the highest poverty rates, with over 40 per cent of their populations living below the national poverty line (World Bank n.d.a, n.d.b, 2022).

These contextual challenges impact the constitution-making processes. A key challenge that constitution-makers must address is the need to design constitutions that enable legitimate and effective government action to respond to security and economic crises while providing sufficient constraints to prevent the regime from descending into authoritarianism. The comparative analysis provided in this document focuses on the different approaches taken by the three constitutional texts to securing this delicate balance.

The analysis is organized into thematic sections, each focusing on a specific constitutional design issue: (a) the structure and level of detail of the three constitutional texts (Chapter 1); (b) fundamental rights and freedoms (Chapter 2); (c) the design of the political system (Chapters 3 and 4); (d) presidential term limits (Chapter 5); (e) emergency powers (Chapter 6); (f) authority over and the mandate of the armed forces and security sector agencies (Chapter 7); (g) guarantees for judicial independence (Chapter 8); (h) the judicial review system (Chapter 9); (i) fourth-branch institutions (Chapter 10); (j) the status and rights of the parliamentary opposition (Chapter 11); (k) the form of the state and arrangements for decentralization (Chapter 12); and (l) the constitutional amendment procedure (Chapter 13).

The constitutions of Chad and Mali and Burkina Faso’s draft Constitution specify the main governing institutions of the state and their respective mandates. However, they leave important elements and procedures crucial for the functioning of constitutional democracy to be defined through legislation. To illustrate, these texts defer to legislative regulation issues such as the criteria and procedures for limiting human rights (Burkina Faso and Mali); the substantive conditions and effects of declaring states of emergency (Burkina Faso, Chad and Mali); the mandate and composition of important independent state institutions such as the electoral management body, the judicial service commission or the human rights commission (Burkina Faso, Chad and Mali); the electoral system for the legislature (Burkina Faso, Chad and Mali); the status and prerogatives of the opposition (Burkina Faso, Chad and Mali); and the responsibilities (Burkina Faso, Chad and Mali), structure and financing of decentralized territorial entities (Burkina Faso and Mali).

Although these three texts vary in their level of detail, this approach is often adopted in francophone constitutions. Compared with most anglophone and lusophone constitutions on the continent, these three constitutional texts provide little detail and leave important elements to be defined in legislation.

In contexts where a single party or coalition often dominates the legislature—as has mostly been the case in the countries examined—this approach allows transient political majorities to unilaterally shape and reshape institutions and procedures critical to the functioning of constitutional democracy, including the composition of independent state organs responsible for overseeing the executive. This approach may also lead to instability and opportunistic changes in the relevant institutional frameworks, undermining the consolidation of institutions and practices. Although constitutions need not be exhaustive, it is generally recommended that critical procedures and elements of constitutional democracy be regulated by the constitution. Drawing upon Dixon and Landau’s concept of the ‘minimum core of a democratic constitution’ (Dixon and Landau 2016), Bisarya and Rogers (2023) argue that safeguarding competitive multiparty democracy requires regulating a core set of institutions, procedures and individual rights in the constitution and protecting them from being unilaterally changed by an incumbent governing majority. Constitutionalization provides a higher level of legal protection and is more likely than the ordinary legislative process to necessitate agreement among a broad range of political forces for changes, thus reducing the risk of unilateral and opportunistic changes by governing majorities.

However, in contexts of post-military-coup transitions, where the constitution-making process may lack inclusivity and be dominated by military and transitional authorities with questionable legitimacy (as in Chad and Mali), the level of detail in the constitution poses a dilemma. On the one hand, regulating all core components of multiparty democracy in the constitution might provide an opportunity for transitional authorities to entrench rules in their favour, consolidate power and make it difficult for future elected majorities to modify these provisions. It may also reduce their incentive to do so, as they would themselves likely benefit from such arrangements. On the other hand, deferring key institutional issues and procedures to legislation risks majoritarian dominance, allowing newly elected majorities to change these rules to their advantage without the need for cross-party negotiations or the support of opposition parties. Proponents of constitutional democracy should therefore promote both inclusive constitution-making processes and the establishment of sufficient guarantees in the constitution to prevent the unilateral imposition or alteration of the rules for a democratic and accountable government framework.

Most constitutions define the fundamental rights and freedoms of both citizens and non-citizens living in the country. At a minimum, these should include the civil and political rights necessary for an open and democratic society (e.g. the right to life; the right to equality and non-discrimination; the freedoms of thought, expression, association and assembly; the right to participate in the conduct of public affairs; the right to a fair trial). Many constitutions go beyond this minimum to include social, economic, cultural and environmental rights, as well as rights specific to minority or disadvantaged groups.

However, including rights in a constitutional text is not sufficient. To ensure the full realization of these rights, the constitution must also provide effective enforcement mechanisms and safeguards. In particular, to prevent abusive restrictions by state authorities, the constitution must clearly specify the circumstances and criteria under which certain rights may be restricted. Constitutions do so through what are commonly known as limitation clauses.

All three constitutional texts considered in this analysis enshrine the main civil and political rights, as well as several social, economic and environmental rights. However, a reading of these provisions reveals three important shortcomings, varying in degree across the texts.

First, none of these texts contain non-derogable rights. These are rights that cannot be suspended or restricted under any circumstances, including in times of emergency or armed conflict. The International Covenant on Civil and Political Rights lists six non-derogable rights: (a) the right to life; (b) the right not to be subjected to inhuman or degrading treatment or acts of torture; (c) the prohibition of slavery; (d) the principles of the legality of offences and penalties and the non-retroactivity of criminal law; (e) freedom from imprisonment for the inability to fulfil a contractual obligation; and (f) the right to recognition before the law. While these rights are enshrined in the three constitutional texts considered, they are not explicitly designated as non-derogable. Consequently, state authorities could potentially impose restrictions on these rights in emergency situations.

Second, these three constitutional texts either do not include limitation clauses or include far-reaching ones. Mali’s 2023 Constitution does not include a limitation clause. Most rights are to be exercised ‘as determined by law’, which means that the legislature regulates the exercise and implementation of the rights and freedoms guaranteed in the Constitution (Title I, Chapter I). Such deferral to the legislature without specifying the procedural requirements and criteria according to which rights can be restricted increases the risk of excessive and arbitrary limitations. This arrangement effectively means that the legislative majority could restrict the content of these rights as it pleases, which undermines the very purpose of guaranteeing them in the Constitution.

Chad’s Constitution and Burkina Faso’s draft Constitution also provide that most rights are to be exercised ‘as determined by law’, but they attach a weak limitation clause to certain civil and political rights. Chad’s Constitution provides that the freedoms of opinion, expression, communication, conscience, religion, press, association, assembly, movement and demonstration ‘may be limited only by respect for the freedoms and rights of others and by the need to maintain public order and morality’ (article 28). This means that state authorities may restrict specified rights only to preserve the rights of others or to maintain public order or morality. Burkina Faso’s draft Constitution provides that the freedoms of thought, expression, conscience and religion ‘are guaranteed, subject to respect of the law, public order and morality’ (article 11). The formulation ‘subject to respect of the law’ means that the legislature can still restrict the content of these rights as it pleases.

From a comparative perspective, limitation clauses in constitutions generally include three core requirements to ensure that the limitation of rights is fair and does not undermine the rights altogether (e.g. Germany (1949), article 19; South Africa (1996), article 36; 2014 Constitution of Tunisia, article 49). The following requirements apply to limitation clauses:

1. They must have a legal basis. Any limitation of a right must be clearly set out in law. The public authority that enacts limitations on a specific right must therefore make clear the legal basis on which it limits that right. This also means that such law must be precise, certain and foreseeable. Laws need to be formulated with sufficient precision to enable individuals to regulate their conduct accordingly.
2. They must pursue legitimate aims. Any restriction of a right must serve a legitimate purpose. Some constitutions explicitly specify such purposes, but doing so does not necessarily protect rights because the terms constitutions often use, such as ‘public order’ and ‘morality’, are open to broad interpretation. Some constitutions clarify the idea of legitimate purpose by requiring limitations to be justified in a free and democratic society (e.g. South Africa (1996), article 36; Kenya (2010), article 24). Public authorities must demonstrate that any measures restricting rights are necessary in a democratic system. International human rights instruments can serve as a reference for identifying the legitimate aims that may justify restrictions on certain rights.
3. They must be proportionate to the purpose of the limitation. Measures taken by state authorities should be the least restrictive means of achieving a legitimate aim.

Third, there are important discrepancies among these three constitutional texts with regard to women’s rights. Some of the texts contain either no provisions, or only weak ones, to foster women’s political participation, and they lack sufficient guarantees to protect the specific rights of women and girls. Mali’s 2023 Constitution prohibits discrimination based on sex (article 1) but does not contain mechanisms or commitments to women’s representation in public institutions and decision-making bodies. By contrast, Chad’s 2023 Constitution (articles 34 and 273) and Burkina Faso’s 2017 draft Constitution (article 4) place a duty on the state to ensure women’s representation in public institutions but do not establish specific mechanisms—such as reserved seats or quotas—to foster women’s representation in such institutions. Specific measures are deferred to legislation.

Moreover, among these three constitutional texts, only Chad’s 2023 Constitution explicitly bans the practice of female genital mutilation (article 20). Mali’s 2023 Constitution simply guarantees the right to physical integrity and prohibits cruel, inhuman and degrading treatment (articles 2 and 4). Burkina Faso’s 2017 draft Constitution goes a step further by also imposing a duty on the state to take all necessary measures to eliminate all forms of violence against women and girls (article 4). Comparatively, several countries on the continent incorporate explicit prohibitions of female genital mutilation within their constitutional frameworks to strengthen the protection of women and girls against this practice (e.g. Côte d’Ivoire (2016), article 5; Ghana (1992), Chapter 5, article 26; Kenya (2010), article 53.1.d; Somalia (2012), article 15.4).

Similarly, Mali’s 2023 Constitution defines marriage as the union between a man and a woman (article 9) without specifically requiring that it be between consenting adults. This omission fails to address the issue of combating forced marriages, especially those involving young girls. Burkina Faso’s 2017 draft Constitution partially addresses this issue by providing that marriage is based on free consent between a man and a woman (article 36). While it adds the requirement of mutual consent, this provision does not specify that both spouses must be adults. Chad’s 2023 Constitution does not define marriage, but it explicitly prohibits child marriage (article 20).

The system of government (régime politique) is one of the most consequential constitutional choices a country needs to make. From a comparative perspective, three broad categories of systems of government exist:

1. Parliamentary systems. In such systems, executive powers are vested in a prime minister and government selected by, accountable to and removable by the legislature.
2. Presidential systems. In presidential systems, executive powers are exercised by a directly elected president who serves a guaranteed fixed term in office, who is not politically accountable to the legislature and who cannot dissolve the legislature.
3. Semi-presidential systems. Executive powers in semi-presidential systems are divided between a directly elected president serving a fixed term and a prime minister and government accountable to and removable by the legislature. There are two subtypes of semi-presidential systems: (a) premier-presidential systems, in which only the legislature can dismiss the government (régime semi-présidentiel moniste); and (b) president-parliamentary systems, where both the legislature and the president can dismiss the government at will (régime semi-présidentiel dualiste). It should be noted that, while there are both democratic and undemocratic regimes with premier-presidential systems, there has not been an established democracy with a president-parliamentary system. This system of government is often found in authoritarian countries (e.g. Azerbaijan, Belarus and Russia) and is common in francophone African countries.

All things being equal, if designed well, semi-presidential systems may help reduce the concentration and personalization of power in the presidency, while also enabling mechanisms to ensure the representativeness and legitimacy of the political system through a form of power sharing. Nevertheless, poorly designed semi-presidential systems may exacerbate the concentration of power in the presidency (notably in president-parliamentary systems) without enhancing effective representativeness and increase the chances of political paralysis. Alongside the design of the system, therefore, it is important to build, nurture and consolidate a democratic and political culture oriented towards deliberation, moderation and compromise.

Chad’s 2023 Constitution foresees a return to a semi-presidential system (akin to the 1996 Constitution), departing from the presidential system established in the 2018 Constitution. More specifically, the Constitution outlines a premier-presidential system, in which only the legislature can dismiss the prime minister and the government.

Burkina Faso’s 2017 draft Constitution maintains the semi-presidential system laid out in the 1991 Constitution. It proposes a president-parliamentary system, in which both the legislature and the president can dismiss the government at will.

In contrast, Mali’s 2023 Constitution abolishes the semi-presidential system of the 1992 Constitution and establishes a presidential system in which the president’s prerogatives are greatly amplified. In view of its unique features, the political regime laid out in Mali’s 2023 Constitution is analysed separately (see Chapter 4).

This chapter compares key design features of the semi-presidential systems in Chad’s Constitution and Burkina Faso’s draft Constitution: (a) the government formation and dismissal procedures; (b) the chairpersonship of the government; (c) the division of executive powers between the president and the prime minister; (d) the appointment of government officials in the civil service and bureaucracy; (e) presidential dissolution of the legislature; (f) removal and impeachment of the president; and (g) the balance of power between the executive and the legislative branches.

3.1. Government formation and dismissal

The procedure for appointing and removing the government affects both the balance of power between the executive and the legislature and the relationship between the president and the prime minister within the executive branch. If the president can unilaterally appoint a prime minister and a cabinet without agreeing or cooperating with the legislature, and if the president can dismiss the government at will, they can direct the government’s programme and dominate the executive branch. Executive power cannot be shared if the president determines the formation and survival of the government. A system that does not give the president a significant role in forming the government and that does not allow the president to remove the prime minister can ensure a balance of power, but it may also result in paralysis. Accordingly, in all semi-presidential systems, incentives for cooperation between the president and the legislature must be built into the process of government formation to ensure a workable power-sharing relationship within the executive branch and between the executive and legislature.

3.1.1. Appointment of the prime minister

There are three main design options for appointing a prime minister in semi-presidential systems (Center for Constitutional Transitions and International IDEA 2014): (a) the president has exclusive authority to select a prime minister (option 1); (b) the legislature has the power to appoint the prime minister, and the president nominally confirms the prime minister (option 2); and (c) the president and the legislature jointly appoint the prime minister (option 3).

Both Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution opt for option 1, with some variations.

Under Chad’s 2023 Constitution, the president has exclusive and unilateral authority to appoint the prime minister (articles 85, 97 and 100). The legislature plays no initial role in either selecting the prime minister or confirming the president’s choice. Moreover, the 2023 Constitution does not stipulate requirements—such as age, citizenship or qualifications—for the position of prime minister. As a result, the president can, in principle, appoint anyone—whether or not a member of parliament (MP)—as prime minister. This unilateral and discretionary power of the president of the republic is, however, balanced by the requirement for the newly appointed government to be subject to a parliamentary vote of investiture within 21 days (presumably from the formation of the government—article 101). This vote of investiture is held after the appointment of the prime minister and the formation of the government. During this vote, confidence in the government as a whole—not only in the prime minister—is at stake (see 3.1.2: Appointment of the rest of the cabinet). Therefore, while the president is not formally obliged to negotiate with or consult the legislature when appointing the prime minister, the requirement for a vote of investiture may incentivize the president to select a prime minister capable of securing the support of the National Assembly. However, if the National Assembly does not grant confidence to the newly formed government during the vote of investiture, and therefore forces the resignation of the cabinet and the president’s prime ministerial appointee, the president has the power to dissolve the National Assembly (except within one year of elections following a prior dissolution) and call for snap elections (articles 89 and 97; see 3.5: Presidential dissolution of the legislature). Under this arrangement, the balance of incentives may encourage the legislature to accept the president’s choice of prime minister rather than encouraging the president to defer to the legislature’s preference. Nevertheless, the choice would depend on electoral calculations—that is, a president who is less confident about the expected outcomes of new elections may instead be forced to accept the wishes of the legislature.

Under Burkina Faso’s 2017 draft Constitution, the president of the republic also has exclusive authority to appoint the prime minister (article 64). However, the text places some constraints on the president’s choice. First, the president must consult the parliamentary majority before appointing the prime minister. While the president is not bound by the recommendation that the parliamentary majority may formulate during that consultation, the requirement to consult the parliamentary majority fosters dialogue. Second, the president must appoint the prime minister from within the parliamentary majority. At first glance, this requirement seems to mean that the appointee must be an elected MP. However, despite the existence of this rule under the 1991 Constitution, former prime ministers were not always appointed from among sitting MPs. Importantly, the text does not define what constitutes a parliamentary majority and may thus leave some discretion to the president in case no party secures an absolute majority in the legislature. Moreover, although article 64 implies that the president alone appoints the prime minister, the 2017 draft Constitution does not specify which presidential acts are not subject to the prime minister’s countersignature (article 79). Clarifying this point would require that the presidential acts that do not require the prime minister’s countersignature be enumerated and that the appointment of the prime minister be included in this list (similar to article 57 of the 1991 Constitution). If the outgoing prime minister were required to approve the appointment of their successor, this would be a unique (and potentially unworkable) arrangement.

Similar to the presidential authority described in Chad’s 2023 Constitution, this unilateral power of the president is balanced by the requirement for the newly appointed government to be subject to a parliamentary vote of investiture within 30 days of the appointment of the prime minister (article 86). If the National Assembly does not grant confidence to the newly formed government through an absolute majority vote during the vote of investiture, and therefore forces the resignation of the cabinet and the president’s prime ministerial appointee, the president has the power to dissolve the National Assembly (except within one year of elections prompted by an earlier dissolution) and call snap elections (article 68; see 3.5: Presidential dissolution of the legislature). Under this arrangement, the president must consult the legislature, but the framework may favour the president’s choice. As in Chad, the balance of power will depend on the expectations about elections after a dissolution.

When compared with France’s Constitution of 1958 (which is also an example of option 1), the procedure for appointing the prime minister foreseen in Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution provides more incentives for cooperation between the president and the legislature. In France, the president unilaterally appoints the prime minister without consultation with the legislature (articles 8 and 19). The prime minister and their government are not subject to a mandatory parliamentary vote of investiture (article 49, para. 1). The only mechanism by which the legislature can balance the president’s prerogative in appointing the prime minister is through a motion of censure, which, if successful, results in the resignation of the prime minister and the government. If the National Assembly passes a motion of censure against the government, the president has the power to dissolve the Assembly (except within one year of elections following an earlier dissolution) and call snap elections (articles 12 and 19). Under this arrangement, the president may play a dominant role in appointing the prime minister, subject to electoral calculations. Despite the president’s apparent broad discretion in appointing the prime minister, past French presidents have refrained from appointing prime ministers who were not supported by the majority in the legislature. In periods of cohabitation (i.e. when the president is from a different political party than the prime minister and the majority in the legislature), the president typically appoints the leader of the legislative majority as prime minister. Opting for another individual would be viewed as a provocation to the parliamentary majority. Moreover, the president may not risk the blow of having their appointee immediately censured by the National Assembly. In addition, even though a parliamentary vote of investiture is not mandatory, many prime ministers since 1973 have sought a vote of confidence from the National Assembly for their newly formed government. This practice aims to underscore that the government derives its power from the confidence (i.e. support) of the National Assembly and not solely from the president.

Other countries also grant the president exclusive authority to appoint the prime minister (option 1) but have established additional measures to further foster cooperation between the president and the legislature and encourage the president to take the legislature’s preferences into consideration. Cabo Verde’s Constitution (1980) is a notable example, as it combines three such incentives. First, the president of the republic ‘appoints the prime minister after consultation with the parties represented in the National Assembly, and taking into account the election results, the existence or lack of a majority party, and the possibility of coalitions and alliances’ (articles 206.1 and 147.1.i). Second, the government must secure a vote of investiture from the National Assembly within two weeks of the appointment of the prime minister (article 209). And third, the president’s power to dissolve the National Assembly is restricted to a considerable extent, preventing dissolution if the legislature has not granted the newly formed government a vote of investiture (articles 155 and 156). In contrast to Burkina Faso, Chad and France, the Cabo Verde president’s decision to dissolve the National Assembly is not discretionary and cannot occur within the 12 months following legislative elections (see 3.5: Presidential dissolution of the legislature). Such an arrangement obliges the president to consult the legislature and incentivizes the president to consider the legislature’s preferences.

To mitigate the risk of the president playing a dominant role in appointing the prime minister, two additional safeguards may be considered in Chad’s Constitution and Burkina Faso’s draft Constitution: (a) requiring the president to choose prime ministerial candidates in a certain order or to consult with the parties in the legislature and take into account the election results before appointing the prime minister (in the case of Chad); and (b) providing that only the National Assembly can dismiss the prime minister and the government, while prohibiting the president from dismissing the prime minister and the government at will (in the case of Burkina Faso, essentially shifting to a premier-presidential system). Prohibiting discretionary presidential dissolution of the National Assembly for one year following legislative elections could also help limit the risk of the president playing a dominant role in appointing the prime minister. At the same time, such an arrangement may pose challenges to swift action, especially in countries facing significant security and economic crises, and would require building and nurturing a (new) political culture of deliberation and compromise.

3.1.2. Appointment of the rest of the cabinet

The procedure for appointing members of the cabinet other than the prime minister affects the balance of power within the executive between the president of the republic and the prime minister.

Designing a cabinet appointment procedure in which the prime minister chooses the composition of the cabinet, while the president has only a limited formal influence on the process, can counterbalance presidential dominance but can also lead to political confrontation.

Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution provide for a cabinet appointment procedure in which the prime minister selects cabinet members, and the president formally appoints them.

Under Chad’s 2023 Constitution, the president appoints members of the government upon the recommendation of the prime minister (articles 85 and 97). Essentially, this means that the prime minister has the authority to choose the cabinet members, who need not be MPs. It appears that the president’s role is intended to be purely formal—that is, the president may not refuse to appoint the prime minister’s nominees to the cabinet. The decree appointing cabinet members must be signed by both the president and the prime minister (article 97). The prime minister must then present the government’s programme to the National Assembly for a vote of confidence by a simple majority vote within 21 days of the formation of the government (article 101). This vote of investiture finalizes and confirms the appointment of the government as a whole. Until this vote of investiture has been carried out, the government’s appointment may be considered provisional. If a vote of investiture is refused within the specified time, the government as a whole (i.e. the prime minister and the cabinet members) must resign, and the government formation process starts again by default (articles 85 and 152).

Burkina Faso’s 2017 draft Constitution provides for a similar arrangement (articles 64, 78 and 86). However, the text does not explicitly specify whether the presidential decree appointing cabinet members, who may not be MPs (article 89), requires the prime minister’s countersignature (article 79). The prime minister must also present their political programme and obtain the approval of an absolute majority of MPs within 30 days of their appointment (article 86). This approval constitutes a vote of investiture for the government as a whole.

In principle, the arrangement in the two constitutions empowers the prime minister vis-à-vis the president while minimizing the risk of government deadlock. The prime minister chooses the cabinet members, and the president formally appoints them. A prime minister who has the power to compose their own cabinet can exercise closer control over that cabinet and ensure that it can act as a bulwark against presidential power. Furthermore, this approach may incentivize the emergence of an effective and unified government. Nonetheless, the prime minister’s authority to select their own government is not absolute. In reality, the prime minister remains an agent of the legislative majority and must compose a cabinet capable of securing a vote of investiture from the legislature and retaining its confidence. Moreover, in cases where the president may remove the prime minister at will (as foreseen in Burkina Faso’s 2017 draft Constitution), this arrangement may induce the prime minister to consider the president’s preferences.

In practice, the functioning of this appointment procedure may fluctuate depending on the prevailing political circumstances. In situations where the president enjoys a supportive legislative majority, the president is more likely to intervene and exercise decisive influence over the composition of the cabinet. In situations of cohabitation, decision-making authority remains in the hands of the prime minister. Under such circumstances, the prime minister composes the cabinet, and the president formally appoints cabinet members nominated by the prime minister. However, the president may refuse to appoint the prime minister’s nominees for ministerial positions in domains where the president holds specific constitutional responsibilities. In such instances, the prime minister may need to consider the president’s preferences and select individuals with whom the president would be willing to collaborate. The president cannot impose specific individuals but may constrain the prime minister’s choice. Such a situation occurred in France in 1986 during the first period of cohabitation when President François Mitterrand refused to appoint two individuals nominated by Prime Minister Jacques Chirac for the ministries of defence and foreign affairs. President Mitterrand thought that he would not be able to work effectively with them because of their past stances on certain political issues. Consequently, Prime Minister Chirac chose two alternative candidates.

In very rare cases, semi-presidential constitutions grant the president the authority to appoint certain ministers unilaterally in areas where the president holds specific responsibilities. These designated ministries are sometimes referred to as presidential ministries. In Ukraine, for example, from 2006 to 2010 the president appointed the ministers of defence and foreign affairs, while other cabinet members were appointed by the prime minister. Allowing the president to appoint their preferred, politically aligned candidates to these key ministries may facilitate rapid and authoritative responses in times of crisis. However, in cases of post-military-coup transitions, such as those in Burkina Faso, Chad and Mali, where military leaders may run in presidential elections, such an arrangement would raise serious concerns. It could enable the president to further strengthen direct control over defence and security forces and insulate decisions in certain policy areas from the rest of the government.

3.1.3. Government dismissal

When examining government dismissal procedures in semi-presidential systems, several questions must be considered, including the following:

1. Who can dismiss the government? In this regard, two subtypes of semi-presidential regimes exist. In premier-presidential systems, only the legislature can dismiss the government (but the president can also indirectly dismiss the government by dissolving the parliament). In president-parliamentary systems, both the legislature and the president have the power to dismiss the government.
2. What are the thresholds and other procedural requirements for dismissing the government? Most semi-presidential constitutions place various procedural limitations on the legislature’s power of dismissal. These limitations are often intended to prevent the misuse of the dismissal procedure and to enhance the stability of the government.
3. Can the government call a vote of confidence in the legislature?
4. Can government members be dismissed individually, and if so, who has the authority to dismiss them?

Chad’s 2023 Constitution foresees the establishment of a premier-presidential form of semi-presidential system, in which only the legislature has the authority to dismiss the government (articles 85, 101, 151 and 152). More precisely, the National Assembly can remove the government by adopting a motion of censure (articles 151, paras. 2 and 3; and 152) or by refusing to grant confidence in a vote on the government’s political programme (articles 101; 151, para. 1; and 152). The National Assembly may introduce a motion of censure at any time. Such a motion must be proposed by at least one-tenth of the members of the National Assembly (article 151, para. 2). It is not clear whether this threshold, while allowing legislative control of the executive, is sufficiently high to prevent irresponsible obstruction by a few isolated members. Once a motion is introduced, the Constitution requires two days’ notice before the vote on the motion of censure is held (article 151, para. 3). Such procedural limitations are common in semi-presidential systems. They are intended to ensure that the government is not removed through a sudden vote that is scheduled to take MPs by surprise. The adoption of a motion of censure requires the support of an absolute majority of the members of the National Assembly (article 151, para. 3). If the motion is adopted, the prime minister must submit the government’s resignation to the president of the republic (articles 85, para. 2; and 152). If the motion is not adopted, the government retains its mandate, and the parliamentarians who submitted the motion cannot introduce another motion during the same legislative session. Other parliamentarians, including those belonging to the same party as the proponents of the unsuccessful motion, can submit a motion of censure during the same legislative session.

The National Assembly can also remove the government by rejecting a motion of confidence submitted by the prime minister. The prime minister can decide to commit the responsibility of the government through a vote on a government programme or a declaration of general policy (article 151, para. 1). The prime minister can decide to resort to this procedure unilaterally at any time, after deliberation in the Council of Ministers, although consent is not required. If the National Assembly approves the government programme or declaration through a simple majority vote, the government retains its mandate. If the National Assembly rejects the government’s programme by a simple majority vote, the prime minister must submit the government’s resignation to the president of the republic (articles 85, para. 2; and 152). Government-initiated votes of confidence are relatively common in semi-presidential systems. They enable the government to demonstrate or consolidate parliamentary support and strengthen its authority.

The president of the republic does not have formal power to dismiss the government. In case of cohabitation, the president will have to cooperate and compromise with the prime minister and the cabinet on issues over which they share authority. In situations where the president enjoys the support of a legislative majority (cases of majority alignment¹), however, the president may have de facto authority to force the prime minister to submit the government’s resignation. If the president is elected simultaneously with or shortly before the National Assembly, it may affect the outcome of legislative elections, in which case the government may become de facto accountable to the president. In situations of majority alignment, the dynamics can shift, rendering the regime functionally president-parliamentary or turning it into a reversed premier-presidential system in which the cabinet becomes primarily accountable to the president.

Finally, under Chad’s 2023 Constitution, neither the legislature nor the president of the republic can dismiss ministers individually. Only the prime minister holds actual authority to dismiss individual cabinet members with whom they are dissatisfied (articles 85, para. 2; and 100, para. 2). Under this arrangement, the president formally dismisses individual cabinet members on the proposal of the prime minister (articles 85, para. 2; and 100, para. 2). The fact that the prime minister can dismiss cabinet members individually, and that the president does not have this formal power, strengthens the authority of the prime minister over the cabinet, limits the influence of the president over the direction and priorities of the government, and thus fosters power sharing within the executive branch between the president and the prime minister. For this reason, several other premier-presidential systems empower only the prime minister to dismiss cabinet members individually (e.g. Cabo Verde (1980), article 147.2.d; France (1958), article 8; 2014 Constitution of Tunisia, article 92, para. 2).

By contrast, Burkina Faso’s 2017 draft Constitution foresees the establishment of a president-parliamentary form of semi-presidentialism, in which both the legislature and the president have formal authority to dismiss the government (articles 64 and 81). Under this arrangement, the president can dismiss the government at will (articles 64 and 81). The presidential dismissal power incentivizes the prime minister and the government to abide by the president’s preferences, thereby providing the president with leverage to dominate the executive branch. In addition, in situations of cohabitation, the respective dismissal powers of the president and the legislature might lead to recurrent appointments and dismissals of the cabinet, generating government instability and a potential power vacuum that could create an opportunity for the president to seize and concentrate power or increase the risk of a coup.

Burkina Faso’s National Assembly may remove the government by adopting a motion of censure (articles 118 and 122) or by refusing to grant confidence in a vote on the government’s political programme (articles 119 and 122). These procedures are similar to the ones foreseen in Chad, but with three notable differences. First, a motion of censure must be proposed by at least 25 per cent of the members of the National Assembly (article 118, para. 2), as opposed to 10 per cent in Chad. Comparatively, this threshold of legislative support for introducing a motion of censure appears relatively high and could make it difficult for opposition groups to propose such a motion. Second, if the National Assembly rejects the motion of censure, the parliamentarians who proposed it cannot introduce another motion for one year (article 188, para. 3), as opposed to Chad, where the restriction applies only to the duration of the same legislative session. In other words, Burkina Faso’s draft Constitution places stricter limitations on motions of censure than the Constitution of Chad does. Third, both motions of censure and government-initiated votes of confidence require an absolute majority vote of all members of the National Assembly to be approved (articles 118 and 119). By contrast, under Chad’s 2023 Constitution, a motion of censure requires an absolute majority vote of the National Assembly, whereas a government-initiated vote of confidence requires a simple majority vote of the National Assembly. Finally, similar to the situation in Chad, only the prime minister can dismiss individual cabinet members at will (article 78, para. 2). However, the prime minister would have limited authority over ministers in practice, as the president retains the authority to dismiss the cabinet at will.

A few semi-presidential constitutions limit the legislature’s power to dismiss the government by requiring the legislature to approve a new prime minister before dismissing the current government. In parliamentary systems, this procedure is known as a constructive vote of no confidence. In Poland, for example, the lower house of parliament may dismiss the government only by initiating and passing a vote of no confidence and approving a new prime minister (Poland (1997), article 158). While this procedure makes it more difficult for the legislature to dismiss the government, and therefore increases the chances of stasis, whereby an unpopular government may struggle to pass its policies in parliament, it has two main benefits. First, it eliminates the power vacuum that can arise between the dismissal of the incumbent government and the formation of a new one, thus mitigating the risk of a power grab by the president. Second, it prevents the removal of a government by a temporary alliance of opposing factions that would be unable to form a coalition government, which would otherwise trigger a dissolution and new elections (except within one year following snap elections after an earlier dissolution).

3.2. Chairpersonship of cabinet meetings

The chairpersonship of cabinet meetings is not merely a ceremonial function. The chair is responsible for convening cabinet meetings and approving their agenda. Arrangements for the chairpersonship of cabinet meetings in semi-presidential systems must be examined in light of the government formation and dismissal procedures. In premier-presidential systems, in which the president has no power to dismiss the prime minister and cabinet, and where policy orientations are defined by the government, granting the chairpersonship of cabinet meetings to the president can encourage discussion and negotiation between the president and the prime minister. It can foster presidential commitment to policy decisions. By contrast, in president-parliamentary systems, where the president can dismiss the prime minister and cabinet at will, and where general policy orientations are defined by the president, granting the chairpersonship of cabinet meetings to the president can result in the president dominating the policy process and reducing the prime minister and cabinet to mere implementing agents of the president.

Under Chad’s 2023 Constitution, the president chairs the Council of Ministers (article 86). The president may authorize the prime minister to convene and chair the Council of Ministers on a specific agenda (article 105, para. 2). The prime minister chairs the Council of Cabinet (article 105), which involves meetings of the government without the participation of the president of the republic. This body exists in several francophone jurisdictions (e.g. Cameroon, France—under the Third and Fourth Republics—and Mali) but is usually not provided for in the constitution and serves only a coordinating role, lacking decision-making power. Given that Chad’s 2023 Constitution foresees a premier-presidential system, where the president cannot dismiss the prime minister and cabinet at will and where the government defines policy orientations, this arrangement may foster discussion and negotiation within the dual executive. In situations of cohabitation, such an arrangement may enable the president to express reservations about policy proposals during cabinet meetings. The president could also delay discussion of a policy proposal that they disagree with by postponing its inclusion on the agenda until a later cabinet meeting. However, excessive use of the power to set the agenda may lead to a serious conflict with the government, as the government would not be able to determine general policy orientations as provided for in article 98 of the Constitution.

Burkina Faso’s 2017 draft Constitution provides for similar arrangements but would have different consequences. The president of the republic chairs the Council of Ministers (article 53). The prime minister replaces the president as chair of the Council in case the latter is unavailable (article 78, para. 3). Given that the draft establishes a president-parliamentary system, where the president can dismiss the prime minister and cabinet at will (article 64) and where the general policy orientations of the state are defined by the president (article 54), this arrangement would further strengthen presidential dominance over the cabinet.

3.3. Division of executive power between the president and the prime minister

Under Burkina Faso’s 1991 Constitution, the president defines the policy orientations of the state, and the prime minister and the government implement them (articles 36 and 61). In addition, the president exercises explicit control over national defence policy. The president is responsible for defining such policy, serves as the commander-in-chief of the armed forces, chairs the National Security Council and unilaterally appoints the chief of staff of the armed forces (articles 52 and 63). The prime minister implements the national defence policy defined by the president (article 63). Combined with the president’s power to dismiss the prime minister at will and to chair cabinet meetings, this arrangement ensures presidential domination over the executive branch and the policymaking process.

During negotiations over the 2017 draft Constitution, members of the constitutional commission agreed that the president should continue to define the policy orientations of the state, as the president is directly elected by the people based on their political programme. However, the division of responsibilities between the president and the prime minister in cases of cohabitation became contentious within the commission. The 2017 preliminary draft Constitution provided that the president would define the policy orientations of the state (2017 preliminary draft Constitution of Burkina Faso, article 56) and that the government, headed by a prime minister, would implement them (article 79). If the prime minister were to come from a parliamentary majority other than that supporting the president, both the president and the prime minister would define policy orientations through consensus. In the absence of consensus, the government would determine and conduct these policy orientations (2017 preliminary draft Constitution of Burkina Faso, article 56).

The constitutional commission modified this proposed arrangement in the final version of the 2017 draft Constitution after several members of the commission and President Roch Marc Christian Kaboré raised concerns that it could lead to a political crisis. According to the final draft, in cases of cohabitation the government would define and implement the policy orientations of the state except for ‘royal domains reserved to the president’ (2017 draft Constitution of Burkina Faso, article 54). This arrangement is problematic because the final draft does not specify these reserved domains, which are to be defined by organic law (2017 draft Constitution of Burkina Faso, article 54). Similarly, the final draft provides that the prime minister must countersign acts of the president with the exception of acts that fall under the exclusive powers of the president (2017 draft Constitution of Burkina Faso, article 79), without specifying which acts fall under this category. While these two provisions aim to limit the president’s dominance over executive decision making, they may have limited practical impact—regardless of their degree of clarity—because the president still retains the power to dismiss the prime minister and cabinet unilaterally without cause, even during periods of cohabitation.

The remaining constitutional provisions regulating the division of powers between the president and the prime minister are similar to the arrangements found in the 1991 Constitution. Notably, the president is responsible for defining national defence policy, serves as the commander-in-chief of the armed forces, chairs the National Security Council and unilaterally appoints the chief of staff of the armed forces (2017 draft Constitution of Burkina Faso, articles 65, 70 and 77). In situations of majority concordance, the prime minister implements the policy orientations defined by the president, including those relating to national defence (article 77).

While the arrangement foreseen in the draft Constitution grants the president exclusive competence over certain policy domains, particularly national defence, to address security crises, it would establish a principal–agent relationship between the president and the prime minister and cabinet across all policy areas. Combined with the president’s power to dismiss the prime minister at will, this arrangement could perpetuate presidential dominance over the executive branch and the policymaking process. Removing the power of the president to dismiss the prime minister at will and clearly defining the president’s reserved policy domains would be required to establish a more balanced intra-executive relationship while enabling the president to retain control over some key policy areas, such as national defence.

Under Chad’s 2023 Constitution, executive powers are shared between the president of the republic and the government (article 65). The president has a limited number of powers that they may exercise unilaterally (article 97), including appointing the prime minister (articles 85 and 100), sending messages to the legislature (article 96), making referrals to the Constitutional Council (article 179), appointing individuals to certain judicial positions and members of fourth-branch institutions (article 97), exercising the right of pardon (article 95) and issuing so-called simple decrees (article 97). The term ‘simple decrees’ is ambiguous, and it is unclear whether it pertains to the explicitly listed presidential actions that do not require the prime minister’s countersignature or whether it constitutes an additional category of presidential decisions exempt from countersignature. Should the latter interpretation prevail, statutes could potentially expand the unilateral regulatory powers of the president of the republic to a significant extent.

The president also exercises certain powers unilaterally but must solicit the non-binding opinion of other political actors beforehand. These powers include the dissolution of the National Assembly (article 89; see 3.5: Presidential dissolution of the legislature) and the use of presidential emergency powers (article 93; see Chapter 6: Emergency powers). Furthermore, certain decisions must first be proposed by other political actors besides the president, who then retains the discretion to make the final decision, as in the case of submitting a bill to a referendum (article 88; see 3.7.1: Tools of the executive to influence the lawmaking process). All other presidential acts must be countersigned (i.e. approved) by the prime minister and, where appropriate, the relevant ministers (article 97). The prime minister exercises regulatory authority over policy matters defined by the Council of Ministers (article 106). It is unclear whether the definition of the scope of the prime minister’s regulatory powers necessitates the president’s countersignature. If it is required, it would limit the autonomy of the prime minister and government vis-à-vis the president significantly.

Chad’s 2023 Constitution grants the government as a whole the authority to define national policy (article 98). More precisely, policy orientations are defined by the government and approved in the Council of Ministers (article 98). As the chair of the Council of Ministers, the president may engage in discussions regarding the government’s policy orientations and share their point of view or disagreement but, in principle, cannot impose their own priorities or veto policy priorities defined by the government. For example, the government can submit bills to the legislature without the signature (i.e. approval) of the president (article 142). The president acts as an arbiter and ensures the efficient functioning of the government, while the prime minister directs and coordinates the government’s actions (articles 66 and 102). However, the effective division of executive powers between the president and the prime minister will depend on the political context (i.e. whether or not one party or coalition controls both the presidency and the prime minister’s office), the scope of the president’s powers to issue simple decrees (article 97) and the interpretation of article 90 of the 2023 Constitution.

Article 90 provides that the ‘president of the republic signs the ordinances and decrees enacted in the council of ministers’ (article 90). While this wording implies that the president must countersign (i.e. cannot veto) these acts, the provision may be interpreted differently in practice, as demonstrated in France. Article 13 of France’s Constitution (1958) contains the same wording. During the period of cohabitation between President Mitterrand and Prime Minister Chirac, the former refused to sign three ordinances in 1986, arguing that article 13 granted the president the possibility, but not the obligation, to sign them.

During periods of cohabitation in France, presidents have occasionally refused to include decrees of nomination on the Council of Ministers’ agenda, effectively vetoing certain appointments proposed by the government. However, they have never exercised veto power over decrees of general application that either implement statutes or regulate policy areas falling within the regulatory domain. The interpretation and implementation of this provision in Chad will greatly influence the relationship and division of executive powers between the president and the prime minister.

When one party controls both the presidency and the prime minister’s office, this arrangement may generate a hierarchical system of governance in which policy priorities are set by the president and implemented by the prime minister. During such periods, the president could also expand their regulatory authority by requesting that decrees that do not require deliberation be added to the council of ministers’ agenda. If such decrees are added onto the council of ministers’ agenda, they must be countersigned by the president. These decrees may then be modified only through the same procedure—that is, after deliberation in the council of ministers and upon signature by both the president and the prime minister. This procedure ensures that, in case of a future period of cohabitation, the prime minister would not be able to modify these decrees unilaterally.

During periods of cohabitation, this division of responsibilities would result in a policymaking process in which the president may influence but cannot veto the prime minister’s national policy direction. The president could delay discussion of a policy proposal by postponing its inclusion on the agenda until a later cabinet meeting or could veto certain decrees and ordinances prepared by the government by refusing to countersign them. However, excessive use of veto and agenda-setting powers could lead to a conflict with the government, as the government would not be able to determine general policy orientations as provided for in article 98.

3.4. Appointment of government officials in the civil service and bureaucracy

The distribution of powers to appoint and dismiss high-level government officials in the civil service has an impact on the balance of power between the president and the prime minister. Dominance over these bureaucratic appointments by either the president or the prime minister may allow either office to capture the bureaucracy. This risk can be mitigated by establishing an appointment procedure that clearly enumerates the positions to be appointed by the president, specifies residual appointments to be made by the prime minister, requires countersignature for all high-level bureaucratic appointments made by either the president or the prime minister and provides an oversight role for the legislature concerning the appointment of certain key bureaucratic posts.

Under Chad’s 2023 Constitution, the president appoints individuals to high-level civilian and military positions. The constitutional text does not enumerate these positions. Instead, they are to be determined by an organic statute (requiring an absolute majority vote of all MPs). The Constitution includes two more constraints on presidential appointments to bureaucratic positions. First, these appointments are made in the Council of Ministers and must therefore be countersigned by the prime minister. This requirement forces the president to negotiate with the prime minister, particularly during periods of cohabitation. Second, these appointments must be made based on ‘the principles of equality, equity, competency and according to the territorial configuration’ of the country (article 90). If enforced by the courts, these principles may help reduce the risk of co-optation or appointments based solely on loyalty to the president. They may also foster an administration that reflects the country’s different territorial entities.

The Constitution is silent on residual appointment powers—that is, appointments to bureaucratic posts other than those designated for presidential appointment. This gap may lead to tensions between the president and the prime minister: the president may claim authority to make such appointments unilaterally through the so-called simple decrees referred to in article 97, while the prime minister may also claim such authority based on the fact that they have the civil service at their disposal (article 102) and that the government ensures the functioning of public services (article 104). This issue may need to be clarified through either a constitutional amendment or an organic law.

Under Burkina Faso’s draft Constitution, the power to make appointments to key civilian administrative positions is divided between the president and the prime minister. On the one hand, the president appoints high-level civilian and military positions that are deemed ‘of strategic character for the nation’ (article 65). As the constitutional text does not enumerate these positions, they are to be determined by ordinary statute (requiring a simple majority vote in parliament). Compared with Chad’s Constitution, Burkina Faso’s 2017 draft Constitution imposes fewer constraints on presidential appointments. It simply provides that some of these appointments require prior consultation with the National Assembly. The positions subject to such consultation are also to be defined by ordinary legislation. Importantly, the draft Constitution does not specify whether these presidential appointments require the prime minister’s countersignature (articles 65 and 79), creating the potential for the president to unilaterally appoint individuals to important bureaucratic roles.

On the other hand, the 2017 draft Constitution grants the prime minister responsibility for making appointments to civilian posts other than those designated for presidential appointment, following deliberation in the Council of Ministers, as determined by legislation (article 85). This provision lacks clarity for two reasons. First, the constitutional text does not specify whether decrees deliberated in the Council of Ministers need to be signed by the president. Second, it is unclear whether the prime minister can appoint all civilian posts other than those specifically designated for presidential appointment, or whether the prime minister’s authority is limited to civilian posts explicitly enumerated in legislation. While the latter arrangement may reduce the risk of deadlock, it carries the risk of allowing unilateral appointments by both the president and the prime minister rather than fostering a deliberative decision-making process.

From a comparative perspective, France’s Constitution provides for a similar arrangement but with greater clarity regarding countersignature requirements and includes a parliamentary oversight mechanism for the appointment of some key posts. In France, the Constitution grants both the president and the prime minister the authority to appoint individuals to bureaucratic positions (articles 13 and 21). Appointments to the highest bureaucratic positions are made by the president in the Council of Ministers or by presidential decree, depending on the post. Positions appointed by the president in the Council of Ministers include ambassadors, prefects, directors of administrative departments, academy rectors and management positions in public companies. Positions appointed by presidential decree include those that are most often determined by competitive examinations or rules governing the careers of civil servants (e.g. judges, university professors). In both cases, these presidential appointments are subject to the prime minister’s countersignature (articles 13 and 19).

The appointments made by the president that are subject to the prime minister’s countersignature are explicitly enumerated in the Constitution (article 13), as well as in organic statutes and decrees. Around 70,000 posts are filled through this procedure (Ardant and Matthieu 2020). Importantly, the constitutional amendments enacted in 2008 introduced a parliamentary oversight mechanism for some of these presidential appointments, in particular for positions relating to the guarantee of rights and freedoms or impacting economic and social life (article 13, para. 5). Organic statutes enumerate around 55 posts, primarily in major public companies and regulatory and fourth-branch institutions, that are subject to this parliamentary review. The relevant standing committees of each chamber of parliament hold public hearings with the nominees and may veto an appointment through a three-fifths majority of the votes cast by the two committees combined. Although attaining this majority is difficult, this procedure enables the legislature to express its opinion on important appointments and raises public awareness when appointments appear to prioritize loyalty over competence.

Residual appointment powers to bureaucratic positions in France (i.e. positions other than those designated for presidential appointment) fall under the responsibility of the prime minister as head of the administration (article 21). In practice, the prime minister delegates a significant portion of these appointments to individual ministers, allowing them to appoint civil servants within their respective ministerial departments. These appointments do not require presidential countersignature.

3.5. Presidential dissolution of the legislature

In semi-presidential systems, the prime minister and government govern with the confidence (i.e. political support) of the majority of the lower house of the legislature. When the lower house is fragmented and struggles to enact necessary legislation or to provide stable support to the government, the presidential power to dissolve it can offer a way out. This presidential prerogative is intended to serve as a deadlock-breaking mechanism. It enables the president to dissolve a deadlocked lower house before its term expires and call for new elections. Dissolution of the lower house results in the dismissal of the prime minister and cabinet. Given its far-reaching implications, restrictions on this presidential prerogative may be necessary. The Constitution may provide substantive, procedural and temporal restrictions to prevent potential abuse of the presidential dissolution power. Specifically, provisions regulating presidential dissolution should be crafted to prevent the president from using this procedure to override a stable legislature dominated by an opposition party or coalition and thus avoid cohabitation.

While both Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution grant the president the power to dissolve the lower house, the former provides more restrictions on the use of this presidential power than the latter does.

These two texts establish similar procedural and temporal restrictions on the president’s power to dissolve the lower house of the legislature but diverge regarding substantive grounds. In Burkina Faso, the president’s decision to dissolve the National Assembly is discretionary, as the text does not foresee substantive grounds for triggering this mechanism (article 68). In contrast, Chad’s 2023 Constitution provides that the president may decide to dissolve the National Assembly ‘when the regular functioning of the public authorities is threatened by persistent crises between the executive and legislative powers, or if the National Assembly dismisses the Government twice within a year’ (article 89). Although the first part of this provision leaves room for interpretation, the president’s decision to trigger dissolution is not entirely discretionary, as it requires justification based on a conflict between the government and the legislature or governmental instability. Both texts establish similar procedural conditions. Before dissolving the National Assembly, the president must consult the prime minister and the speaker of the unicameral parliament in Burkina Faso or the speaker of each of the two chambers of the bicameral parliament in Chad. While these consultations are mandatory, the opinions of the authorities consulted are not binding on the president. Thus, the president’s authority to dissolve the lower house of parliament remains a unilateral power.

Additionally, both texts impose three comparable temporal restrictions. The lower house of the legislature cannot be dissolved during the exercise of emergency powers by the president (Chad (2023), article 94; 2017 draft Constitution of Burkina Faso, article 73), during an interim period of the presidency (Chad (2023), article 83; 2017 draft Constitution of Burkina Faso, article 62) or if a dissolution has already occurred within one year following snap elections after an earlier dissolution (Chad (2023), article 89; 2017 draft Constitution of Burkina Faso, article 68). By limiting the president’s ability to dissolve the lower house of parliament during a political crisis or emergency situation, such constitutional constraints can help prevent the president from manufacturing and exploiting crises to dissolve the legislature and consolidate power. However, Burkina Faso’s draft Constitution does not prohibit presidential dissolution during impeachment proceedings against the president (article 160). This omission could allow the president to dissolve the legislature in order to evade impeachment proceedings, thereby undermining presidential accountability to the legislature. According to Chad’s Constitution, the president of the republic is suspended from office in the event of indictment by a two-thirds majority vote of MPs and would therefore not have the authority to dissolve the National Assembly during impeachment proceedings (article 194).

Both texts also raise concerns regarding the effects of dissolution. Chad’s Constitution provides that new legislative elections must be held within 90 to 180 days following dissolution (article 89). While the Constitution should allow sufficient time for organizing new elections and campaigning, new elections should be held within a reasonable timeframe following dissolution to limit the risk of a power vacuum and prevent the president from ruling by decree and consolidating power, as dissolution would entail the immediate disbandment of the legislature.

The timeframe for holding snap elections varies from country to country, but a six-month window may be considered excessively long. Under Burkina Faso’s draft Constitution, new elections must be conducted within 90 days following dissolution (article 68). However, unlike Chad’s Constitution, which guarantees that the newly elected National Assembly convenes as of right on the 15th working day following the elections (article 89), Burkina Faso’s draft Constitution does not specify that the newly elected legislature can convene as of right following elections. It simply stipulates that it can convene during ordinary sessions or through an extraordinary session, which requires a request from the prime minister or a majority of MPs. To ensure that the newly elected legislature can exercise its mandate and powers within a reasonable timeframe following snap elections, the constitution should explicitly provide that the newly elected legislature convenes as of right within a specified and reasonable period following elections.

3.6. Impeachment and removal of the president

In semi-presidential systems, the president is directly elected for a fixed term and is not politically accountable to the legislature. The legislature cannot dismiss the president because of disagreement with their policy positions. However, as the president exercises certain prerogatives as part of the dual executive, most semi-presidential constitutions provide mechanisms for removing the president from office in cases where they exceed their mandate or otherwise take actions detrimental to the integrity of the presidential office or that threaten the country’s interests. In some jurisdictions, the president may be dismissed through an impeachment procedure which involves a vote in the legislature followed by court proceedings (e.g. Cabo Verde, Madagascar and Senegal). Impeachment typically requires justification and a finding that the president has committed a serious act or omission detrimental to the presidential office or the country. In such cases, the president often becomes ineligible to run in future presidential elections. In some other jurisdictions, the president can be dismissed through a political removal procedure that requires only a supermajority vote in the legislature (e.g. France, Lithuania and Namibia). Since removal procedures are more political in nature, a president removed through this mechanism may remain eligible to run in future presidential elections, provided that they have not met the constitutional term limit. This approach enables voters to ultimately determine the fate of the removed president (see Abebe 2022).

Under Chad’s 2023 Constitution, impeachment of the president is both a political and a judicial procedure that involves a vote in the legislature and a ruling by the High Court of Justice. The president can be removed from office for high treason (articles 193 and 194), which is ‘any act that undermines the republican form, the unity and secularity of the State, and the sovereignty, independence and integrity of the national territory’ (article 194, para. 1). The text also provides a list of specific acts that qualify as high treason, including ‘serious and flagrant violations of human rights, embezzlement of public funds, corruption, bribery, drug trafficking and the introduction of toxic or dangerous waste for transit, deposit or storage on national territory’ (article 194, para. 2). While high treason seems to constitute narrow grounds for impeachment, the definition of high treason and the list of acts that qualify as such are broad enough to encompass conduct detrimental to the country and the presidential office, such as gross violations of human rights, corruption or fraud.

The initiation of impeachment proceedings requires a two-thirds majority vote of all members of the bicameral parliament in a joint sitting (article 195). The High Court of Justice then tries the president on the charges to determine whether the president has committed high treason. The High Court of Justice is a hybrid body responsible for trying the president and high government officials for cases of high treason. It consists of 15 members—4 members of the lower house of parliament, 4 members of the upper house of parliament, 4 members of the Supreme Court and 3 members of the Constitutional Council (article 192). They are elected by their peers within their respective institutions. The Constitution does not specify a threshold for the ruling of the High Court of Justice. Rules governing proceedings before the Court are deferred to an organic statute (article 198). The president is removed from office if found guilty by the High Court of Justice. The text does not specify whether a removed president can run in later presidential elections.

Under Burkina Faso’s draft Constitution, the removal of the president is handled entirely within the legislature. The president may be removed from office ‘in case of high treason or a breach of duties patently incompatible with the exercise of their office’ (article 160). The initiation of removal proceedings requires the support of two-thirds of the members of the unicameral legislature. The president may then be removed by a three-quarters majority vote of all members of the legislature. The draft does not specify what constitutes high treason or a breach of duty patently incompatible with the exercise of the presidential function. Under this arrangement, the determination of the grounds for removal is left to the discretion of MPs: the alleged conduct may involve criminal acts or other present or past behaviour whose disclosure is clearly detrimental to the office.

The required thresholds may be considered relatively high. While the two-thirds majority vote necessary to initiate the removal procedure serves as a safeguard against politically motivated attempts at removal, the three-quarters majority required for actual removal may be seen as excessively high, potentially undermining the effectiveness of the mechanism as a means of holding the president accountable. Importantly, the draft provides that the decision to remove the president may be appealed to the Constitutional Court (article 160, para. 4). However, the provision does not specify who can lodge such an appeal or the scope of the Court’s review. The draft does not specify whether the Court’s authority extends solely to procedural matters, ensuring the fairness of the process, or whether it may also encompass an assessment of whether the alleged facts indeed qualify as high treason or a breach of duties incompatible with the presidential office. If the Court were permitted to assess the factual elements, there would be a risk of politicizing the Court and removing actual decision-making power from the legislature.

The decision to remove the president results in a vacant presidency, acknowledged by the Constitutional Court (article 160). A new presidential election must be held within 60 to 90 days (article 62). The removed president may run in this election unless the removal occurred during their second term (article 57). Thus, voters have the last word.

3.7. Balance of power between the executive and the legislative branches in the legislative process

Chad’s Constitution and Burkina Faso’s draft Constitution retain several elements from France’s 1958 Constitution, particularly in their provisions regulating the relationship between the executive and the legislative branches in the legislative process.

These two constitutional texts contain several tools to promote government stability which originated in France’s 1958 Constitution as part of so-called rationalized parliamentarism. Title V of France’s 1958 Constitution contains multiple constitutional devices that were designed to enable the executive to govern without a stable majority in the lower house of the legislature. These tools were designed to respond to specific contextual needs—namely, to address government instability and the dominance of the legislature in the political system that characterized the preceding Third and Fourth French Republics. Among the tools developed for this purpose that are retained in Chad’s Constitution and Burkina Faso’s draft Constitution are the following: the limitation of the scope of the legislative domain and the granting of autonomous regulatory power to the executive (3.7.1.1 below); the government’s power to submit bills to the legislature (3.7.1.2); influence over the legislature’s agenda (3.7.1.3); amendment procedures for bills (3.7.1.4); blocked voting (3.7.1.5); the commitment of the government’s responsibility on a text (or tacit voting) (3.7.1.6); the ordinance procedure (3.7.1.7); and legislative referendums (3.7.1.8).

Furthermore, Chad’s Constitution and Burkina Faso’s draft Constitution include provisions constraining the legislature’s oversight role that are similar to some of those found in France’s 1958 Constitution (3.7.2).

3.7.1. Tools of the executive to influence the lawmaking process

3.7.1.1. Legislative domain and autonomous regulatory power

Chad’s Constitution and Burkina Faso’s draft Constitution limit the scope of the legislative domain to certain policy areas. They provide a list of such areas that must be regulated through legislation (domaine de la loi) (Chad (2023), article 132; 2017 draft Constitution of Burkina Faso, article 137). Other policy areas fall under the regulatory authority of the executive (domaine réglementaire) (Chad (2023), article 133; 2017 draft Constitution of Burkina Faso, article 138).

A similar distinction between legislative and regulatory domains can be found in previous constitutions of Chad and Burkina Faso, as well as in most francophone constitutions on the continent. This arrangement originates in the French Constitution, where article 34 lists policy areas that fall under legislative authority, and article 37 provides that all policy matters not explicitly assigned to the legislative domain fall under regulatory authority.

This arrangement carries two major implications. First, the legislature can legislate only on policy matters explicitly assigned by the constitution. Both texts thus enumerate similar broad lists of policy areas within the legislative domain, encompassing critical aspects of democratic governance, the enforcement of fundamental rights, and matters of importance for society and future generations. One notable difference, however, is that under Chad’s Constitution, electoral systems and rules must be regulated by legislation, whereas this requirement is absent in Burkina Faso’s draft Constitution. This omission creates the possibility that the executive could define and modify the electoral rules, including for legislative elections. Defining the main electoral rules in constitutional or statutory frameworks could address this issue. Similar to the Constitution of France, both the Constitution of Chad and the draft Constitution of Burkina Faso distinguish between policy areas where legislation defines the rules and those where legislation simply establishes fundamental principles. In theory, the former refers to policy areas that are deemed most important in a democratic society and subject to detailed legislative regulation, while the latter refers to policy areas that the legislature regulates in a broad manner and that require derived regulatory acts from the executive for implementation. In practice, however, strict adherence to this distinction may be difficult, as the legislature may not be capable of regulating each and every detail of a policy matter, and complementary regulatory acts may be necessary for comprehensive implementation.

Second, this arrangement grants the executive two types of regulatory power. The executive holds ordinary (or derived) regulatory power—that is, the power to adopt regulatory acts that implement or provide further details about a piece of legislation. Additionally, the executive has autonomous regulatory power—that is, the power to regulate through regulatory acts policy areas that do not fall within the scope of the legislative domain.

One important difference between Chad’s Constitution and Burkina Faso’s draft Constitution concerns the possibility of expanding the legislative domain. In Chad, article 132 mandates the legislature to refine and extend the list of matters falling within the legislative domain through organic laws, thereby providing a potential mechanism to constrain unilateral presidential lawmaking power. No such mechanism exists in Burkina Faso’s draft Constitution, where any formal extension of the legislative domain would require a constitutional amendment.

Both constitutional texts establish mechanisms to enforce the delineation between the legislative domain and the autonomous regulatory domain. In Chad, the Constitution does not explicitly prohibit parliament from legislating on matters falling within the autonomous regulatory domain (Chad (2023), articles 133 and 145). Instead, it grants the executive the authority to determine whether parliament can legislate on such matters. If the executive does not want parliament to legislate on such matters, the president can invoke inadmissibility to prevent passage of the bill (Chad (2023), article 145), or the government may amend the legislation ex post through decree following a decision of the Constitutional Council declaring that the legislation falls within the autonomous regulatory domain (Chad (2023), article 133).

In Burkina Faso, the draft Constitution provides that bills (and amendments) sponsored by MPs that do not fall within the legislative domain are prohibited and must be declared inadmissible by the speaker of the National Assembly (article 127). However, the government may submit bills that fall within its autonomous regulatory domain.

Although this arrangement may seem to restrict the legislature’s lawmaking power significantly, this may not necessarily be the case in practice. First, the list of policy issues that fall within the legislative domain is relatively broad, limiting the scope of the government’s autonomous regulatory power. Second, the government may exercise its autonomous regulatory power sparingly, preferring to regulate policy issues that do not fall within the legislative domain through legislation. The government may adopt this approach to grant such rules a higher degree of legal protection, broaden their acceptance and ownership, and demonstrate respect for the legislature’s lawmaking mandate. In France, for example, despite lacking an absolute majority in the 16th legislature of the National Assembly (28 June 2022 to 9 June 2024), the government enacted an average of one autonomous decree per month, similar to figures during the 15th legislature, when the government held a stable majority (Quinart 2024). Most of these autonomous decrees did not regulate policy areas but instead addressed the functioning of the national administration or deconcentrated services. While the constitutional texts of Chad and Burkina Faso grant autonomous regulatory powers to the government, it remains to be seen whether governments will fully exercise this authority in practice.

3.7.1.2. Initiation of bills

Pursuant to the constitutional texts of Chad and Burkina Faso, both MPs and the government have the authority to propose bills (Chad (2023), article 142; 2017 draft Constitution of Burkina Faso, article 133). This arrangement is common in semi-presidential systems. It enables MPs to fulfil their mandate to enact legislation and allows the government to submit bills necessary for implementing its policy agenda.

In Chad, all preliminary government-sponsored bills (projets de loi) as well as MP-sponsored bills (propositions de loi) must be examined by the Supreme Court for a non-binding opinion. This requirement somewhat mirrors the procedure outlined in France’s Constitution, where government-sponsored bills are examined by the Council of State before discussion and approval in the Council of Ministers (France (1958), article 39). The Council of State issues a non-binding legal opinion assessing whether the provisions of the bill fall within the legislative domain or the regulatory domain and evaluating its coherence with existing laws and the normative system, but it does not comment on the bill’s political purpose. In Chad, concerns may arise regarding the structure of the Supreme Court. While in France these legal opinions are produced by the advisory section of the Council of State (rather than its judicial section), in Chad it is unclear how the Supreme Court will be structured and which section would issue these ex ante non-binding opinions. It is important to ensure that those providing such an opinion on a bill are not later responsible for adjudicating cases involving the implementation of that law.

In both Chad and Burkina Faso, preliminary government-sponsored bills are deliberated in the Council of Ministers (Chad (2023), article 142; 2017 draft Constitution of Burkina Faso, article 135). The decision to transform a preliminary bill into a formal bill and to submit it to the legislature takes the form of a decree signed by the prime minister and the relevant ministers.

Neither constitutional text explicitly requires the attachment of additional documents when bills are submitted to the legislature, such as a justification or an impact assessment, with a view to improving the quality of the bill, fostering informed debates in the legislature, and facilitating monitoring and evaluation during the implementation phase.

The two constitutional texts also outline three types of limitations on MPs’ right to initiate bills. First, MPs may propose bills only on topics that fall within the legislative domain. If a bill falls outside this domain, the president of the republic (in the case of Chad) or the president of the National Assembly (in the case of Burkina Faso) may declare it inadmissible (Chad (2023), article 145; 2017 draft Constitution of Burkina Faso, article 127). In cases of disagreement between the executive and the legislature, the question of admissibility may be adjudicated by the constitutional council or the constitutional court.

Second, MPs are barred from introducing bills (or amendments to bills) that would result in a decrease in public revenues or an increase in public spending. This restriction, which is also common in parliamentary regimes, protects the state budget (prepared by the government to implement its policy programme) from being questioned by MPs who might be tempted to propose tax cuts or spending increases to the electorate.

Third, the government may amend the content of MP-sponsored bills before they are submitted to a parliamentary committee for initial discussion and examination. According to Burkina Faso’s draft Constitution, the government may propose amendments to MP-sponsored bills within two months before they are examined by the legislature (article 134). In Chad, MP-sponsored bills are submitted to the Council of Ministers for deliberation before being sent to the legislature (article 142). Moreover, the legislature must consider and debate the version of the MP-sponsored bill that was submitted by the government (article 146). Essentially, this procedure empowers the government to modify the content of MP-sponsored bills before they reach the legislature. Additionally, it remains unclear whether the government could decide not to submit an MP-sponsored bill to the legislature. These restrictions favour the government in the lawmaking process to a large extent and may constrain the legislature’s lawmaking mandate. In comparative practice, while the government may have the right to propose amendments to MP-sponsored bills during the first examination of the bill by the relevant parliamentary committee and during the first reading, it typically does not have the power to decide whether such bills are sent to the legislature or to impose changes to the text before they are considered in the legislature.

3.7.1.3. Parliamentary agenda setting

Constitutional provisions regulating the establishment of the legislature’s agenda influence the relationship and balance of power between the executive and the legislature. On the one hand, the government should have mechanisms to shape the legislative agenda in order to ensure the consideration and adoption of bills necessary for implementing its policy programme. On the other hand, the legislature should have sufficient independence to legislate (including drafting and proposing bills itself) and effectively oversee and scrutinize the government’s actions and performance.

Under Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution, the legislative chambers determine the parliamentary agenda (Chad (2023), article 150; 2017 draft Constitution of Burkina Faso, article 123).

In Chad, the bureau (conférence des présidents) of each legislative chamber determines the agenda of that chamber. The composition of the bureau is determined in the rules of procedure of the respective legislative chamber, but a member of the government attends its meetings by right. While this arrangement allows the government to convey its legislative priorities by attending bureau meetings, it does not grant the government the formal right to place bills or topics on the parliamentary agenda. Instead, the government must rely on its affiliated MPs within the bureau to secure placement on the agenda.

Under Burkina Faso’s draft Constitution, the National Assembly determines its agenda after consulting the government. Furthermore, the prime minister can include an item on the daily agenda of the National Assembly as a matter of priority. While this arrangement may appear commendable at first sight—granting the legislature authority over its agenda while enabling the government to introduce priorities—it may raise concerns. First, ambiguity surrounds the meaning of ‘an item’. It is unclear whether it refers to a demand to consider a single bill or to hold a debate on a single topic, or whether it could encompass multiple bills or topics. Second, the prime minister’s authority to add a priority item to the parliament’s daily agenda is not limited to a specific number of days during a parliamentary session. The prime minister could potentially exercise this prerogative every day of the parliamentary session. Such an arrangement could lead to government monopolization of the parliamentary agenda. To safeguard against executive dominance and ensure greater balance in the legislative agenda-setting process, further limitations could be placed on the prime minister’s prerogative to add priorities to the parliamentary agenda. For instance, this authority could be restricted to a specified number of days or weeks per parliamentary session, or priority status on the daily agenda could be restricted to one bill or one topic for debate.

3.7.1.4. Right to submit amendments to bills

The right to submit amendments to bills constitutes a fundamental component of the lawmaking process. In semi-presidential systems, this right is usually granted to both MPs and the government. It enables them to suggest changes and influence the text, especially when the bill originates from the other branch. As most bills are prepared by the government, MPs’ right to propose amendments is crucial. It enables their meaningful participation in the legislative process, allowing them to influence the content of legislation rather than simply accepting or rejecting proposed government bills as a whole. Consequently, MPs’ right to submit amendments is a key mechanism for fostering effective parliamentary debate and elevating the legislature beyond being a mere rubber stamp for the government’s agenda.

However, the parliamentary opposition may also use this right to impede the legislative process. Therefore, constitutions and parliamentary rules of procedure often impose limits on MPs’ ability to submit amendments during the legislative process. Similarly, while the government usually has the right to propose amendments to bills during this process, this right should be carefully defined. Overly broad powers risk enabling the government to exercise ultimate decision-making power on the content of legislation, thus weakening the legislature’s lawmaking mandate. In this context, the right to propose amendments, both for MPs and for the government, must be carefully crafted to ensure that the lawmaking process remains a deliberative exercise through which the government can negotiate the adoption of legislation necessary to implement its programme, rather than merely securing the rubber-stamping of its agenda.

Chad’s Constitution and Burkina Faso’s draft Constitution both grant MPs and the government the authority to submit amendments to bills (Chad (2023), article 148; 2017 draft Constitution of Burkina Faso, article 125).

Both texts establish limits on MPs’ right to submit amendments. First, MPs are prohibited from proposing amendments (or bills) that do not fall within the legislative domain or that would result in a decrease in public revenues or an increase in public spending (Chad (2023), article 144; 2017 draft Constitution of Burkina Faso, article 126). Second, under Chad’s Constitution, the government may reject amendments submitted by MPs during plenary debate if the amendments were not previously submitted to the parliamentary committee that first examined the bill (article 148). Such a restriction—which is also found in France’s Constitution—aims to reduce the risk of late-stage amendments.

Both texts also briefly address the government’s right to propose amendments, allowing the government to propose amendments to MP-sponsored bills before their first examination in the legislature (see 3.7.1.2: Initiation of bills). However, they do not precisely specify the timeframe or the stages during which the government can submit amendments. These matters may be regulated by the parliamentary rules of procedure.

3.7.1.5. Blocked voting

Another device found in these two constitutional texts designed to facilitate governance in the absence of a stable legislative majority is blocked voting (Chad (2023), article 148; 2017 draft Constitution of Burkina Faso, article 120). This tool enables the government to force the legislature to vote not article by article—the normal procedure—but in a single vote on the text as a whole, or on parts of it, while retaining only amendments proposed or accepted by the government. Essentially, it enables the government to put to a vote only the precise text it prefers or accepts. It leaves MPs with a binary choice—accept the government’s preferred text or reject the text entirely. Blocked voting can serve the government in multiple ways. The government can use it to maintain the content of the proposed text against substantial alterations by the legislative majority or to counter amendments from the opposition that could gain support from certain members of the legislative majority.

Under Chad’s Constitution, the blocked voting procedure is available to the government for any ordinary bill or parts thereof at any time (article 148). By contrast, Burkina Faso’s draft Constitution restricts this tool to a significant extent, allowing the government to use it only once per ordinary session of the legislature (article 120).

In addition to limiting the number of times the government may resort to blocked voting, another way to constrain its use would be to require a higher approval threshold—such as an absolute majority vote—for adopting a bill through this procedure. Such a requirement could help balance expediency and deliberation: the government could still advance its preferred version of the bill in the legislature, but its adoption would require broader legislative consensus.

3.7.1.6. Tacit voting

In several francophone jurisdictions operating under a semi-presidential system, the prime minister can commit the government’s responsibility on the adoption of a bill (e.g. France (1958), article 49; 2010 Constitution of Niger, article 107; Senegal (2001), article 86; see also Romania (1991), article 114; Moldova (1994), article 106.a). When this procedure is triggered, discussion of the bill stops in the legislature, and the text is deemed adopted unless the legislature passes a motion of no confidence against the government. In essence, it enables the government to pass a bill without a vote in parliament. This powerful tool in the hands of the government is balanced by the possibility for the legislature to introduce and adopt a motion of no confidence within a specified timeframe, in which case the bill fails, and the government is dismissed (OSCE/ODIHR 2025). However, such a scenario poses risks for MPs, as a government dismissal may be followed by presidential dissolution of the legislature and snap elections to resolve the ensuing conflict.

Initially introduced in France’s 1958 Constitution, this mechanism was designed to enable a government with a relative or undisciplined majority in parliament to pass key legislative texts, allowing the government to implement its priorities. It aims to strengthen government stability but also encroaches on the legislature’s role. Due to its significant impact on the balance of power between the government and the legislature, this procedure must be carefully crafted.

Chad’s 2023 Constitution and Burkina Faso’s 2017 draft Constitution provide for such mechanisms under similar arrangements. Both texts enable the prime minister, following deliberation in the council of ministers, to commit the responsibility of the government on the adoption of a text (Constitution of Chad (2023), article 151, paras. 5 and 6; 2017 draft Constitution of Burkina Faso, article 121). These constitutional texts provide very weak procedural requirements and no substantive restrictions on the use of this tool. The prime minister can unilaterally decide to resort to this procedure. The only procedural requirement is that this decision be deliberated in the council of ministers before it is announced in the national assembly. Deliberation in the council of ministers does not necessarily mean approval by the council. The prime minister simply has to mention their decision during the meeting of the council of ministers and have it recorded in the minutes for the requirement to be met, regardless of the reactions of the other participants, including the president of the republic. Critically, there are no restrictions on the subject matter, the type of act or the number of texts on which this procedure may be triggered. As currently designed, the tacit voting procedure in both Chad’s Constitution and Burkina Faso’s draft Constitution appears highly favourable to the government and could enable it to encroach upon the legislature’s deliberative and lawmaking authority to a significant degree.

As soon as this procedure is triggered, debate on the text in the national assembly is suspended for 24 hours. The text is then automatically considered adopted—without a vote in the legislature—unless a motion of censure is submitted within those 24 hours and subsequently adopted by an absolute majority vote of the members of the national assembly. If a motion of censure is submitted, the debate in the national assembly focuses on the question of dismissing the government rather than the substance of the text. If the motion of censure is rejected, the text is deemed adopted, and the government remains in power. If the motion of censure is adopted, the bill fails, and the government is dismissed. This procedure forces MPs into a challenging position—either accept the adoption of a text they may fundamentally oppose without a vote or trigger an institutional crisis that could lead to the dissolution of parliament and snap elections.

Given the far-reaching impact of this procedure, several countries have introduced constitutional restrictions. In Senegal, this procedure may be used only for budget bills and one other bill per parliamentary session (which lasts about one year) (article 86, para. 4). Similarly, in France, this procedure may be used only for finance bills, social security finance bills and one other bill per session (article 49, para. 3). Such restrictions effectively limit the number of texts for which the procedure may be triggered (OSCE/ODIHR 2025). Additional safeguards may include prohibiting the use of the tacit voting procedure on certain policy matters, excluding the possibility of combining it with other tools of so-called rationalized parliamentarism and requiring that it be activated only after a thorough parliamentary debate on the bill.

Niger’s 2010 Constitution takes a different approach, providing that a text may be considered adopted only if the legislature approves a motion of confidence by an absolute majority vote (articles 107 and 108). The logic is reversed. The government must obtain a vote of confidence for the text to be considered adopted, whereas under the arrangement described above the legislative majority must pass a motion of censure to reject the text. If the legislature withholds its confidence, the bill fails and the government is dismissed. This approach places the legislature in a stronger position, as adoption of the bill—and survival of the government—requires a vote in the National Assembly. This arrangement seeks to balance expediency, deliberation in parliament and government accountability: the government may still advance its preferred version of the bill in the legislature, but its adoption requires a broader legislative consensus, and the government assumes a greater risk of dismissal. A combined approach, limiting the type and number of bills on which this procedure may be used and requiring a vote of confidence for the adoption of the bill, could help mitigate potential government overreach into the legislature’s lawmaking authority while maintaining government stability (OSCE/ODIHR 2025).

3.7.1.7. Ordinance procedure

The two constitutional texts examined in this section also enable, under certain conditions, the government to enact regulatory measures on policy matters that would normally be regulated by parliament through legislation (Chad (2023), article 137; 2017 draft Constitution of Burkina Faso, article 140). This is called the ordinance procedure. This procedure, which originated in France’s 1958 Constitution, is designed to enable the adoption of rules that are urgently needed but that would normally have to go through the lengthy ordinary lawmaking process. It also aims to ensure that necessary decisions can still be made even when the parliament is divided and unable to pass legislation. While originally conceived as an extraordinary procedure, the ordinance procedure, if not properly regulated, can be used extensively by governments to regulate policy matters that normally fall within the legislative domain and thus significantly encroach upon the parliament’s role as a deliberative lawmaking body—as witnessed in the past 15 years in France.² To prevent the transformation of this extraordinary procedure into an ordinary practice, constitutional provisions regulating the ordinance procedure should be carefully crafted. Such provisions are crucial for safeguarding the parliament’s role in making and adopting laws.

Chad’s Constitution and Burkina Faso’s draft Constitution specify several steps and conditions under which the government may act through ordinances (Chad (2023), article 137; 2017 draft Constitution of Burkina Faso, article 140).

First, the government must request the authorization of the legislature to enact an ordinance by submitting an enabling bill. As both constitutions provide that the government can submit such requests with a view to implementing its political programme, the government must indicate in the enabling bill how the measures it intends to take relate to that programme. Chad’s Constitution additionally requires that the enabling bill specify the legislative matters on which the government intends to act through ordinance and to justify the request. Such requirements prevent overly broad or vague language that could enable blanket or frequent resort to the ordinance procedure. Chad’s Constitution also prohibits policy matters falling within the remit of organic laws from being subject to delegation (article 139). The enabling bill must also set a timeframe for the duration of the power to enact ordinances granted to the government. In this regard, Chad’s Constitution simply provides that the delegation can be granted only for a limited period of time, whereas Burkina Faso’s draft Constitution limits it to three months. The enabling bill must also include a timeframe for the ratification of ordinances by parliament, which may go beyond the duration of the delegation itself.

Second, the legislature then considers the enabling bill and can amend it in line with the ordinary lawmaking procedure before either approving or rejecting the delegation.

Third, once the delegation is approved by parliament, the government can draft one or more ordinances and adopt them in the council of ministers during the period of delegation. Before their adoption by the council of ministers, ordinances are submitted to the Supreme Court in the case of Chad and to the Council of State in the case of Burkina Faso for a non-binding opinion.

Fourth, the president of the republic signs the ordinances.

Fifth, the ordinances are published in the official gazette and enter into force immediately as regulatory measures (as opposed to legislation).

Sixth, the government must submit ratification bills to parliament before the expiration of the ratification date specified in the enabling law. If the government does not submit the ratification bills before the deadline, the ordinances are considered null and void.

Seventh, if the government submits a ratification bill on time, the parliament has two choices—adopt or reject the ratification bill. Parliament can ratify the ordinance as adopted by the government or amend or repeal certain provisions before ratifying it. Burkina Faso’s draft Constitution provides that the ratification by the legislature must be explicit, meaning that the legislature must vote on and adopt the ratification bill. This requirement strengthens the role of the legislature. If the ratification bill is adopted, the ordinance(s) become law. If it is rejected, the ordinances are considered null and void.

While the two constitutional texts offer some constraints on the government’s ordinance powers, the scope of this power could be further narrowed—for example, through the following measures (OSCE/ODIHR 2025):

1. Limiting policy matters that may be subject to delegation. The constitution could provide that certain subjects may not be delegated, such as laws relating to elections, the judiciary, citizenship, fundamental rights and organic laws. A similar approach would be to explicitly enumerate policy matters that may be delegated. Given the context of the countries concerned, these matters could include, for example, laws relating to the security sector and the economy.
2. Requiring special voting rules for the legislature to adopt an enabling bill. Such rules could include the requirement for an absolute majority vote or a qualified majority vote.
3. Ensuring that the delegation is specific and limited. For instance, the enabling bill could be required to specify the purpose and scope of the measures to be taken by ordinance, and parliament could be allowed to set out guiding principles in the enabling law to inform the content of the government’s ordinance enacted under that delegation.
4. Providing a maximum duration for the legislative delegation and specifying that the enabling law automatically lapses upon the dismissal of the government, dissolution of the parliament or the expiration of its term.
5. Prohibiting the government from sub-delegating its delegated lawmaking power to another executive body, regulatory institution or decentralized institution unless explicitly authorized by the legislature in the enabling law.
6. Providing that ordinances automatically lapse at the end of the delegation period unless ratified by parliament, or rendering ordinances null and void if the ratification bill is not considered by parliament—rather than simply submitted by the government—within the specified timeframe.

3.7.1.8. Referendums

Chad’s Constitution and Burkina Faso’s draft Constitution both grant the president of the republic the power to submit a legislative bill to a referendum. While a legislative referendum can serve as a tool for directly engaging the people in policy decisions, it also carries the risk of enabling the president to effectively bypass the legislature. Consequently, the president’s power to hold a legislative referendum should be carefully crafted.

The Constitution of Chad includes two important constraints on the president’s power to call a legislative referendum (article 88). First, the president cannot initiate a legislative referendum unilaterally. The president can decide to initiate such a referendum only upon receiving a proposal from either the government or the legislature. The government can propose holding a legislative referendum to the president only during ordinary sessions of the legislature. This requirement enables MPs to discuss and share their views on this proposal and, if they disagree with it, to initiate a motion of no confidence against the government. While the legislature can also propose a legislative referendum to the president, the requirement that such referendums can be held only on government-sponsored bills (projet de loi) decreases the likelihood that proposals for a referendum will originate from the legislature. Upon receiving a proposal from the government or the legislature, the president can accept or refuse to hold a referendum at their own discretion. This decision does not require the countersignature of the prime minister.

Second, article 88 greatly restricts the scope of subject matters that can be submitted to a legislative referendum. Legislative referendums can be held only on government-sponsored bills relating to ‘the organization of the public powers’, the approval of a union agreement or the ratification of international treaties that would affect the functioning of the Constitution (for example, a treaty transferring certain policy responsibilities to a regional organization). As a result, legislative referendums cannot be held on topics that concern the people directly, such as major societal issues. While these procedural and substantive restrictions limit the risk of abuse of legislative referendums by the president, Chad’s Constitution does not explicitly exclude constitutional bills from the president’s referendum power. In the absence of such an explicit prohibition, the president could potentially push through a popular constitutional amendment without political consensus among the various political forces represented in parliament. Such a measure would undermine the spirit of the provisions governing constitutional amendment, which require legislative deliberation and approval.

Burkina Faso’s draft Constitution explicitly prohibits the president from using the legislative referendum procedure to submit a constitutional amendment bill to a referendum (article 67). Nevertheless, the draft Constitution falls short of providing sufficient safeguards to prevent the president from using the legislative referendum to bypass a legislature opposed to them. The president can, after non-binding deliberation in the Council of Ministers and a non-binding opinion from the speaker of the National Assembly, submit to a referendum any bill sponsored by the government that they consider to require direct consultation of the people. Under this arrangement, the president can bypass the legislature and unilaterally submit any legislative bill to a referendum at their own discretion. The only balancing mechanism available to the legislature is that, since a law adopted by referendum does not take precedence over legislation adopted by the legislature, parliament could later modify the law adopted by referendum. However, amending a law directly endorsed by the people may incur significant political costs for the legislature.

3.7.2. Legislative oversight of the executive

Chad’s Constitution and Burkina Faso’s draft Constitution both explicitly affirm the oversight functions of the legislature. Both texts provide that, besides adopting laws, the legislature ‘oversees the government’s actions’ (Chad (2023), article 130; 2017 draft Constitution of Burkina Faso, article 95). To operationalize this power, both texts grant the legislature several prerogatives commonly found in semi-presidential and parliamentary systems.

First, these texts establish mechanisms to ensure that MPs are informed of the government’s decisions and actions. The government presents its political programme and can make a declaration on general policy orientation in the legislature (Chad (2023), articles 101 and 151; 2017 draft Constitution of Burkina Faso, articles 86 and 119). Members of the government have the right to access the legislature and engage in debates, both in plenary sessions and in committees (Chad (2023), article 138; 2017 draft Constitution of Burkina Faso, articles 128 and 129). In Chad, the Constitution places an obligation on the government to provide the legislature with all requested information and explanations regarding its management and activities (article 154). MPs can submit both written and oral questions as well as questions on current issues to the government (Chad (2023), article 154; 2017 draft Constitution of Burkina Faso, article 130).

Second, these texts establish mechanisms for scrutinizing and holding the executive accountable. In Chad, the legislature can issue resolutions (article 130) to express its stance on various matters or to critique governmental priorities or positions without adopting legislation. Such resolutions are not legally binding. Both constitutional texts also enable the legislature to establish commissions of inquiry to investigate specific matters such as allegations of corruption or to scrutinize the management of public services or public enterprises (Chad (2023), article 154; 2017 draft Constitution of Burkina Faso, article 113).

In both countries, the legislature’s rules of procedure regulate oversight mechanisms. Although parliamentary rules of procedure go beyond the scope of this analysis, it is crucial that these mechanisms remain accessible to the legislative opposition or minority.

Both constitutions also enable MPs to initiate motions of censure or votes of no confidence against the government and to dismiss it for political reasons (see 3.1.3: Government dismissal). They also enable the legislature to initiate impeachment proceedings against the president of the republic in cases of high treason (see 3.6: Impeachment and removal of the president).

To perform the oversight function and utilize these oversight tools effectively, dedicated time should be allocated during parliamentary sessions. The provision of such time could be ensured by providing in the constitution that a specific portion of the parliamentary session be reserved for oversight and post-legislative scrutiny, and by granting the legislature, including the legislative minority, the authority to set the agenda for a portion of the parliamentary session. The provisions in the two constitutional texts appear relatively weak in this regard. In Burkina Faso, the draft Constitution provides that at least one day per week during the parliamentary session should be allocated to MPs’ questions and government responses (article 130). In Chad, the Constitution does not allocate any portion of the parliamentary session to legislative oversight. While it empowers the legislature to determine its own agenda, it does not grant a role to the legislative minority in setting the agenda (see 3.7.1.3. Parliamentary agenda setting and Chapter 11: Status and rights of the parliamentary opposition). Neither text explicitly reserves time for post-legislative scrutiny.

Mali’s 2023 Constitution provides for a transition from the semi-presidential system established by the 1992 Constitution to a presidential system. One of the rationales behind this transition was to strengthen the power of the executive in order to address the enduring security crises and economic challenges facing the country. As examined in more detail below, however, the Constitution expands the president’s prerogatives to such an extent and provides so few and weak checks and constraints on presidential actions that this new political system can in fact be characterized as hyper-presidential.

In presidential systems, the legislature and the president are both directly elected for fixed terms. The president serves as both head of state and head of government. As such, the president exercises executive power and appoints, dismisses and directs cabinet members at will without the involvement of the legislature. The president and the cabinet are not politically accountable to the legislature. This means that the legislature cannot dismiss the president or cabinet members via a vote of no confidence. The president can only be removed for cause—that is, on limited grounds—through an impeachment procedure. In return, the legislature also serves for a fixed term and cannot be dissolved by the president. In such systems, there are several factors that help balance power between the president and the legislature. First, as there is no relationship of confidence between the two branches, the president must negotiate with the legislature to pass legislation necessary to pursue their policy agenda. In addition, several constitutional devices are often established to serve as checks and balances between the two branches, such as the president’s power to veto legislation or the requirement for certain appointments made by the president to be confirmed by the legislature. Independent oversight and regulatory institutions also play an important role by providing politically neutral oversight of the president and the administration and by regulating certain areas of government activity (such as elections, judicial administration, the media, etc.).

In some cases, however, presidential systems have resulted in an excessive concentration of power in the presidency, which has contributed to authoritarian outcomes. This situation is often influenced by constitutional design features, including the following:

• the power of the president to govern unilaterally through broad decree, referendum and emergency powers;
• the lack of oversight of the executive, resulting from the president’s broad powers to make appointments in the administration, independent institutions and courts, combined with a lack of legislative oversight mechanisms;
• presidential oversight of security sector agencies; and
• weak provisions on presidential term limits.

Therefore, when examining the design of a presidential system, these elements should be carefully assessed. In the case of Mali’s 2023 Constitution, a detailed examination of these elements is provided in this and subsequent chapters (see Chapters 5–10).

Mali’s Constitution establishes a presidential system but retains certain features that are usually found in semi-presidential systems, particularly in the semi-presidential system of France. These features are likely to operate differently in a presidential system.

The president of Mali, who is directly elected for a five-year term, is both head of state and head of the executive (articles 43 and 45). The president defines the national policy orientation (article 44), while the prime minister and the government implement it (article 76). The president unilaterally appoints the prime minister and the government and holds them accountable (articles 57 and 58). While in most presidential systems the president unilaterally appoints and directs the cabinet, which advises the president and implements the policy programme, it is relatively rare in comparative practice to have a prime minister in a presidential system. Indeed, such a position may appear to be unnecessary because the president holds all executive power—thus, executive power is not divided between the president and the prime minister and the government—and because there is no relationship of confidence between the legislature and the government. In the region, Côte d’Ivoire has a similar arrangement, with a presidential system that retains the office of prime minister. In presidential systems, the prime minister can serve as a ‘fuse’ for the president in the event of a crisis or criticism of the president’s policy. Such an arrangement may reduce presidential accountability, as the prime minister can be blamed and dismissed for policy failures despite the president’s sole authority to determine the policy agenda and exercise all executive power.

While the Constitution states that executive powers are exercised by the prime minister (article 77), the president retains ultimate decision-making authority in executive matters. Indeed, the president defines the national policy orientation (article 44), chairs the Council of Ministers (article 58), and signs ordinances and decrees adopted in the Council of Ministers (article 66), while the prime minister and government cannot take executive decisions that the president opposes, as they are directly accountable to the president (article 57). In addition, although the Constitution establishes a presidential system, it retains the limitation of the scope of the legislative domain to certain policy areas (article 115), a feature commonly found in francophone semi-presidential systems. The Constitution lists policy areas that must be regulated through legislation (the legislative domain), while other policy areas fall within the regulatory authority of the executive and may be regulated through regulatory acts (the regulatory domain) (article 116). Coupled with the president’s dominance of the executive branch described above, this arrangement grants the president two types of regulatory power. The president de facto exercises ordinary (or derived) regulatory power—that is, the power to adopt regulatory acts that implement or provide further detail to legislation. Additionally, the president exercises autonomous regulatory power—that is, the power to regulate policy areas that do not fall within the legislative domain through regulatory acts. While in semi-presidential systems these executive powers are divided between the government and the president, in Mali’s presidential regime they are concentrated in the presidency.

The Constitution maintains the requirement that most regulatory acts must be signed by both the president and the prime minister, as well as the relevant ministers (articles 66, 71 and 77). In a semi-presidential system, the countersignature requirement is designed to ensure that certain decisions are taken with the agreement of the two main holders of executive power (i.e. the president and the prime minister). Thus, during periods of cohabitation, the president and the prime minister must negotiate and reach a compromise in order to take decisions subject to countersignature. Conversely, in a presidential system, the countersignature requirement enables members of the government to engage in discussions with the president. However, as they are politically accountable to the president, ultimate decision-making authority rests with the president. Thus, in a presidential system, countersignature serves primarily to affirm the president’s position as head of the executive rather than to share and balance power within the executive.

In legislative matters, the powers of the president are also significantly strengthened. The president can submit bills to parliament (article 119) and can require that parliament, through the government, vote on a bill in a single vote on only those amendments proposed or accepted by the government (article 122). In addition, the government can request authorization from parliament to regulate policy matters falling within the legislative domain through ordinances (article 121). This provision establishes limited procedural and substantive requirements for the ordinance procedure, such that the authorization must be limited in time and that the measures taken by the government must be necessary to implement its action plan (article 121). The executive also exercises strong influence over the parliamentary agenda, since the chambers of parliament must give priority to bills sponsored by the president and to bills sponsored by MPs in the order determined by the government (article 119). The president also presents their political priorities to parliamentarians by delivering an annual address to parliament (article 61) and by communicating by message with each chamber of parliament at any time (article 62). Importantly, the president is granted broad referendum powers. The president can unilaterally submit to a referendum any question of national interest or any bill relating to the organization of state institutions after receiving the non-binding opinion of the Constitutional Court (article 60). Constitutional amendment bills are not explicitly excluded from the president’s power of referendum. This provision thus enables the president to pass popular legislation (and potentially constitutional amendments) without the consent of the legislative majority, effectively bypassing the legislature.

This combination of broad decree and referendum powers, significant influence over the lawmaking process, and extensive and unchecked emergency powers (article 70; see Chapter 6: Emergency powers) would enable the president to govern and decide unilaterally on a wide array of policy matters.

Additionally, the president also has the power to dissolve the National Assembly (article 69). As the Constitution does not foresee substantive grounds for dissolution, the president’s decision to dissolve the National Assembly is discretionary. In terms of procedural requirements, article 69 simply requires that, before dissolving the National Assembly, the president must consult the speakers of the two chambers of parliament and the chairperson of the Constitutional Council. While these consultations are mandatory, the opinions of the authorities consulted are not legally binding on the president. Thus, the president’s authority to dissolve the lower house of parliament remains a unilateral power. The text also imposes certain temporal restrictions. The president is prohibited from dissolving the National Assembly during the first 12 months of the legislative term, when an impeachment procedure against the president is under way, if a dissolution has already occurred within a year following snap elections (article 69) or during the exercise of emergency powers by the president (article 70). While these temporal limitations may reduce the risk of abuse of dissolution, granting the president the power to dissolve the legislature distorts the logic of a presidential system and raises significant concerns with regard to its impact on the balance of power. In presidential systems, both the president and the legislature are directly elected for fixed terms. The president cannot dissolve the legislature, and the legislature cannot remove the president without cause. To advance their policy agenda, the president must negotiate with the legislature. Granting the president the power to dissolve the National Assembly shifts the balance of power to a large degree in favour of the president. The president can leverage the threat of dissolution to marshal the legislature and force it to enact legislation necessary to implement the president’s agenda without negotiation.

The Constitution also grants the president significant discretionary powers of appointment within the civilian and military administration, without the legislature’s involvement or scrutiny (article 67).

What’s more, although the president is now the head of the executive branch and thus a political actor, the president is also the guarantor of the independence of the judiciary (article 134) and, as such, chairs the Supreme Council of the Judiciary (article 64). Irrespective of the type of system of government, this arrangement raises concerns, as it enables the president to be involved in various aspects of judicial administration, such as judicial appointments, promotions and disciplinary proceedings against judges. This arrangement does not correspond to internationally recognized standards and good practices regarding judicial independence on the continent.

In addition, the mechanisms for legislative oversight of the executive appear limited. The legislature oversees government actions through written or oral questions to government members and through the establishment of commissions of inquiry. The National Assembly can no longer dismiss the government through a vote of no confidence, as the government is no longer accountable to the legislature. The legislature can remove the president from office in the event of high treason, through a particularly stringent impeachment procedure requiring multiple stages, where the procedure can be stopped at each stage, and a final three-quarters majority vote of the members of both houses of parliament in a joint sitting (article 73). The procedural conditions governing the impeachment procedure appear so restrictive that the mechanism is nearly impossible to implement in practice.

In summary, the president holds regulatory powers, exerts significant influence over the legislative process (despite lack of political accountability to parliament), is involved in the management of the judiciary and has discretionary authority to appoint much of the civilian and military administration. The only possible checks on this formally powerful presidency are presidential elections and provisions on term limits, which are not helpful between elections.

In this context, Mali’s 2023 Constitution can be described as an ‘emergency’ constitution (Abebe 2024). It establishes a hyper-presidential system that prioritizes presidential leadership and rapid, short-term decision making over inclusive deliberation, political compromise and checks and balances. Although it ostensibly seeks to create a framework more capable than the 1992 Constitution of responding to Mali’s severe security and economic crises, this new political regime concentrates tremendous, unchecked powers in the presidency. This new dispensation may increase the stakes of presidential elections, intensify winner-takes-all politics and pave the way for authoritarian drift.

Presidential term limits—constitutional limitations on the ability of a president or former president to seek re-election to the office of president—are a common feature in modern democratic constitutions worldwide. Over 80 per cent of presidential and semi-presidential constitutions in force today include provisions on presidential term limits (International IDEA 2022). Although the 2001 Economic Community of West African States (ECOWAS) Protocol on Democracy and Good Governance does not explicitly require such limits, all ECOWAS member states except The Gambia set limits on presidential re-election in their constitutions—either a maximum of two consecutive terms with no restriction on the total number of terms (Cabo Verde, Guinea Bissau and Senegal) or a maximum number of two terms in total (Benin (1990), Burkina Faso (1991 as amended), Côte d’Ivoire (2016), Ghana (1992), Guinea (2010), Liberia (1986), Mali (2023), Niger (2010), Sierra Leone (1991) and Togo (1992 and 2024)). Operating alongside a set of constitutional checks and balances, provisions on presidential term limits can serve as an important mechanism for ensuring regular alternation of power and helping prevent autocratic rule. In presidential and semi-presidential systems, where the president tends to exercise significant powers and responsibilities, the regular change of office holders through term limits can help prevent excessive and prolonged concentration of power within the presidency. When well designed, term limits may strengthen the guarantees of political alternation at the highest level of executive power and help ensure more orderly, peaceful transfers of power.

Nevertheless, experiences in the region and beyond demonstrate that constitutional term limits can be circumvented through various means, such as constitutional amendment or replacement, and thus do not constitute an absolute guarantee of political alternation (e.g. Cameroon (2008), Côte d’Ivoire (2016), Togo (2002 and 2024), Central African Republic (2023)). However, according to Ginsburg, Melton and Elkins (2011), well-designed constitutional presidential term limits are still valuable because they establish clear rules and increase the political cost associated with attempts to remain in office beyond the established limits. Robust presidential term limits make it more difficult for incumbents to remain in power once their term limit is reached.

When examining the robustness of provisions on presidential term limits, there are generally five key issues to consider: (a) the length of the presidential term; (b) the number of terms an individual may serve as president; (c) whether or not the term limit constitutes an absolute maximum or applies only to consecutive terms; (d) whether a new term limit provision applies only prospectively or also retroactively (i.e. whether it counts the incumbent’s current term as well as terms served previously by others); and (e) whether or not the constitutional term limit provisions can be amended and, if so, how.

The three constitutional texts considered in this analysis all set limits on presidential re-election, albeit with nuanced differences.

In Mali, the 2023 Constitution retains the limit of two five-year presidential terms (article 45). The Constitution remains silent on the issue of the retroactive or non-retroactive application of the term limit provision. In this context, it is unclear whether the current term of the incumbent president of the transition (during which the 2023 Constitution was adopted) would count towards the maximum two terms permitted under the Constitution. Experiences both within the region and beyond (e.g. in Burkina Faso in 2005, in Senegal in 2012, in Burundi in 2015 and in Côte d’Ivoire in 2020) have shown that failing to address this potentially contentious issue specifically could require a future judicial resolution, which carries a risk of politicization and manipulation of the Constitutional Court, given its exclusive jurisdiction in matters of constitutional interpretation. Furthermore, even though the 2020 Transitional Charter explicitly prohibits the president of the transition from running in the presidential election at the end of the transition, the Constitution does not provide such a prohibition in its section on transitional provisions.

One notable difference compared with Mali’s 1992 Constitution is that the 2023 Constitution stipulates that the provision establishing the two-term limit cannot be amended (article 185). The unamendability represents a significant advancement in safeguarding political alternation compared with the 1992 Constitution. However, for this constitutional safeguard to be truly effective, it could have been strengthened by stipulating the procedure for adopting a new constitution and specifying that even a new constitution cannot change the term limit provision. Such additional constitutional safeguards could have helped prevent future incumbents nearing the end of their term from resetting their tenure by adopting an ostensibly new constitution with a new non-retroactive term limit provision or removing term limits altogether (as has occurred in some countries).

Turning to Chad, the 2023 Constitution adjusted the term limit provisions outlined in the 2018 Constitution. While the 2018 Constitution allowed for a maximum of two six-year terms in total (article 66), the original version of the 2023 Constitution envisioned a maximum of two consecutive five-year terms but with no restriction on the total number of terms (article 67). This arrangement could potentially facilitate a situation where a president nearing the end of their term supports the election of a loyal successor, assumes the position of prime minister and then runs again in a subsequent presidential election. In this scenario, term limits might not effectively prevent prolonged concentration of power in the hands of a single individual. In addition, the 2023 Constitution is silent on the issue of the retroactive or non-retroactive application of the term limit provision. Thus, it is likely that the current term of the incumbent president of the transition will not be counted towards the maximum two-term limit stipulated in the Constitution. Importantly, neither the length of the presidential term nor the limit of two consecutive terms is specifically protected against constitutional amendment. The relevant provisions can be amended through a two-thirds majority vote in each chamber of the bicameral parliament followed by approval by a simple majority in a referendum, or potentially through a three-fifths majority vote of both chambers of parliament in a joint session (articles 279–81; see 13.1: Constitutional amendment procedures). Indeed, on 3 October 2025 the legislature adopted—by a three-fifths majority vote in a joint sitting—a package of constitutional amendments that extended the presidential term to seven years and abolished the limit of two consecutive terms (2023 Constitution of Chad, article 67, as amended through 2025; Republic of Chad 2025), thereby permitting the incumbent president to be re-elected indefinitely. These amendments also lifted the prohibition on presidential involvement in political parties, thus enabling the incumbent President, Mahamat Idriss Déby Itno, to head the ruling party founded and previously led by his late father, who ruled the country as president for over three decades. Taken together, the October 2025 amendments recentralize authority in the presidency to a significant extent, weaken formal constraints on executive power and reduce institutional guarantees for political alternation (see also Leubnoudji Tah 2025).

In Burkina Faso, the 2017 draft Constitution replicates the 1991 Constitution (as amended in 2015) with regard to the provisions on presidential term limits. It allows for a maximum of two five-year terms in total (article 57). Neither the total number of terms nor the length of the presidential term may be amended (article 191). Unlike the 2023 constitutions of Chad and Mali, which do not specify the retroactive or non-retroactive application of new term limits, Burkina Faso’s 2017 draft Constitution explicitly provides that the term of the then-incumbent president would count towards the maximum of two terms under the proposed Constitution. Retaining this provision in the current context would signify that the president of the transition would be eligible to be re-elected only once under the proposed Constitution (and not during the first presidential election, which is prohibited by article 4 of the Transitional Charter). However, the draft Constitution does not explicitly regulate the procedure for adopting a new constitution and does not prevent a new constitution from altering the term limit provision.

Any country may experience public emergencies arising from war, invasion, armed uprisings, terrorist attacks, natural disasters, pandemics or other types of crises or catastrophes. Democratic states must respond to emergencies within the framework of legally guaranteed rights and institutional checks and balances associated with the democratic constitutional order, which can pose their own unique challenges. Most of the world’s democratic constitutions therefore include emergency provisions that allow the authorities, in times of urgent necessity, to take actions required to safeguard national security, maintain law and order, protect the lives and property of citizens, keep essential public services working and in general restore normalcy. If they are well designed and properly applied, emergency provisions are a self-defence mechanism for democracy—a way of ensuring democratic resilience by providing the powers needed to respond effectively to serious crises within the framework of the democratic constitution. However, many governments have also misused emergency powers—for example, by needlessly prolonging or renewing states of emergency or by using emergency powers to bypass normal channels of democratic accountability, harass dissidents, rig elections, restrict the press and ultimately impose a dictatorial regime. Great care therefore needs to be taken in designing the constitutional provisions governing emergency powers (Bulmer 2018).

Given the severe security crisis facing countries in the Sahel due to the cross-border operations of non-state armed groups, the emergency provisions contained in the three constitutional texts considered in this chapter are of particular importance. The challenge is to draft emergency provisions that grant the executive sufficient powers to respond effectively to security threats while preventing emergency powers from being misused for authoritarian purposes.

The emergency provisions contained in the Constitution of Mali (2023, article 70), the Constitution of Chad (2023, articles 93 and 94) and the 2017 draft Constitution of Burkina Faso (article 73) are very similar to those found in the Constitution of France (1958, article 16). In France, under very specific circumstances and when the proper functioning of the constitutional public authorities is interrupted, the president may take ‘all measures required by these circumstances’ (France (1958), article 16). Essentially, this provision temporarily concentrates all state power in the hands of the president. These so-called exceptional powers of the president were introduced in the 1958 Constitution at the request of Charles de Gaulle to prevent a recurrence of the paralysis witnessed in June 1940, when the government of the Fourth Republic was unable to take swift decisions in response to the invasion of the country by Nazi Germany.

The constitutions of Mali (2023) and Chad (2023) and the 2017 draft Constitution of Burkina Faso provide for similar presidential emergency powers. However, there are subtle yet significant differences among these texts regarding the substantive conditions and procedural requirements for triggering these powers. There are also differences in terms of temporal limitations and in the scope of authority granted to the president during such exceptional circumstances.

First, all three texts—using the same wording as that found in the French Constitution—provide that these emergency powers can be triggered only under specific circumstances. These powers can be invoked when there is ‘a serious and immediate threat’ to the institutions of the republic, the independence of the nation, the integrity of its territory or the fulfilment of international commitments, and when the proper functioning of the constitutional public authorities is interrupted. While the criterion of ‘interruption of the proper functioning of constitutional public bodies’ requires more than mere operational difficulties or a potential disruption, the seriousness of the threat provides a degree of discretion to the president. In Chad and Mali, these two conditions are cumulative—that is, there must be both a serious and immediate threat and an interruption in the functioning of the constitutional authorities. By contrast, in Burkina Faso’s draft Constitution, these conditions are not cumulative; in other words, emergency powers may be invoked either in the presence of a serious and immediate threat or when the constitutional authorities are no longer able to function properly. Consequently, presidential emergency powers could potentially be invoked in more situations in Burkina Faso than in Chad or Mali.

Under all three texts, the decision to invoke these emergency powers is taken unilaterally by the president. It constitutes an exclusive competence (pouvoir propre) of the president and does not require the countersignature of the prime minister. Although the president of the republic unilaterally assesses whether the circumstances justify recourse to emergency powers, they must consult various institutions and receive their non-binding opinion before deciding to use emergency powers. In Mali, the president must consult the prime minister, the speakers of both chambers of parliament and the Constitutional Court. In Chad, the president’s decision to invoke emergency powers is taken during a meeting of the Council of Ministers (without the need for the prime minister’s countersignature), following consultation with the speakers of both legislative chambers and the president of the Constitutional Council. In Burkina Faso, the decision is taken following deliberation (i.e. non-binding discussion) in the Council of Ministers and consultation with the speakers of both legislative chambers and the Constitutional Court. To enhance transparency and presidential accountability, a good practice would be to require that the non-binding opinion of the respective constitutional court or constitutional council be published. Although an opinion concluding that the conditions for invoking emergency powers are not met would not legally prevent the president from proceeding, making such an opinion public would increase the political cost of doing so. Once the president decides to invoke emergency powers, they must announce the decision publicly.

The scope of authority granted to the president during such exceptional circumstances varies among the three texts. Under Burkina Faso’s draft Constitution, the president can take ‘the measures required by the circumstances’ (2017 draft Constitution of Burkina Faso, article 73). The president could thus take not only regulatory but also legislative or judicial decisions. All state powers could thus be concentrated in the presidency, albeit subject to two restrictions: the president cannot dissolve the legislature (article 73) or amend the Constitution (article 192) while exercising emergency powers. These restrictions are critical, as they prevent the president from setting aside the legislature and imposing a permanent dictatorial regime through changes to the Constitution.

In Mali, the president can also take ‘the measures required by the circumstances’ (2023, article 70). As in Burkina Faso, all state powers could be concentrated in the president, who could take any regulatory, legislative or judicial decisions necessary. However, Mali’s Constitution imposes more limitations on the scope of these measures, which must not compromise the country’s national sovereignty or territorial integrity. The president may neither dissolve nor suspend any of the institutions of the republic. Importantly, the exceptional measures taken by the president must aim to ensure the continuity of the state and the restoration, as quickly as possible, of the regular functioning of state institutions. This means that the exceptional measures taken by the president must directly address the ongoing crisis and that the president’s emergency powers should cease as soon as state institutions are able to function properly and resolve the crisis by themselves.

In Chad, the president of the republic can also take ‘the measures required by the circumstances’ (article 93). However, these exceptional measures must be ‘designed to provide the constitutional public authorities, as swiftly as possible, with the means to carry out their duties’ (article 94). As in Mali, this requirement means that the exceptional measures taken by the president must directly address the ongoing crisis and that the president’s emergency powers should cease as soon as state institutions are again able to function properly and resolve the crisis. Furthermore, Chad’s Constitution provides that these exceptional measures cannot dissolve the National Assembly (article 94) or infringe the rights to life, to physical and moral integrity or to the jurisdictional guarantees of individuals (article 93). The latter restriction could be interpreted to imply that the president is precluded from making judicial decisions while exercising emergency powers. To prevent the president from taking exceptional measures that go beyond the scope of the crisis, these texts could require that the exceptional measures adopted by the president be justified and submitted to the respective constitutional court or constitutional council, which could assess whether the measures are directly linked to the crisis and aimed at restoring the proper functioning of state institutions.

The three constitutional texts also differ significantly with regard to temporal restrictions. Mali’s Constitution does not establish a time limit for the use of exceptional powers. It simply provides that exceptional measures taken by the president that fall within the legislative domain automatically lapse if they are not confirmed by parliament within 90 days of their enactment. This 90-day period not only appears excessive in comparative perspective, but it also applies only to the implementation of emergency measures falling within the legislative domain and not to the use of emergency powers themselves. Consequently, the president can continue to exercise exceptional powers until the threat is no longer considered immediate or serious and until state institutions are able to function properly again. However, only the president is competent to determine whether the crisis is over and whether state institutions can function properly. The Constitution does not establish a procedure for legislative approval or a mechanism for judicial oversight by the Constitutional Court to verify whether the conditions required for using exceptional powers are still met. This framework therefore grants the president unilateral discretion to determine the length of the state of exception. It appears to imply that neither the parliament nor the Constitutional Court can end this state of exception. The only temporal constraint on the president’s use of exceptional powers is that exceptional measures falling within the legislative domain lapse after 90 days if they are not confirmed by parliament. If this interpretation is correct, the provision grants extremely broad powers to the president, increasing the risk of abuse. The only safeguard against the prolonged misuse of exceptional powers would be the potential impeachment of the president by parliament. However, as discussed in Chapter 4, the procedural conditions for impeachment are excessively restrictive, rendering the procedure nearly impossible to complete.

Burkina Faso’s draft Constitution provides that the state of exception takes effect immediately and lapses automatically after 90 days if it is not confirmed by the legislature (article 73). While this provision effectively provides a time limit, 90 days appears excessive from a comparative perspective. The provision implies that parliament cannot end the state of exception before the expiry of the 90-day period (i.e. only the president can decide to stop the emergency within this period). If this interpretation is correct, the provision grants extreme powers to the president, increasing the risk of abuse.

Chad’s Constitution establishes a similar arrangement but with a much shorter timeframe for legislative confirmation. The state of exception enters into force following the president’s decision to invoke exceptional powers but lapses automatically after a period of 15 days if it is not confirmed by both chambers of the legislature (article 93). To further reduce the risk of the president’s prolonged misuse of exceptional powers, these texts could incorporate provisions requiring the legislature’s approval for the continuation of the state of exception, with progressively higher majority thresholds over time (e.g. an absolute majority after three weeks, followed by a two-thirds majority after two months, etc.). In addition, the texts could also specifically empower the respective constitutional court or constitutional council to assess, after a specified period, whether the circumstances justifying the exceptional powers are still in place. In France, for example, since the 2008 constitutional amendments, MPs can request that the Constitutional Council conduct a review after 30 days. After 60 days, the Council can conduct a review on its own initiative (France (1958), article 16).

7.1. The role of the armed and security forces

In democracies, there is typically a clear delineation between the role of the armed forces or military (to defend territorial integrity and protect against external threats), the role of the security forces or police (to maintain internal law and order) and the role of intelligence agencies (to gather information of national interest and assess internal and external security threats). This delineation is often defined through a constitutional and statutory framework that articulates and limits the mandates and missions of different security sector agencies, while also establishing a distinct institutional architecture for each. The 2001 ECOWAS Protocol on Democracy and Good Governance requires member states to ensure the apolitical and non-partisan nature of the security forces, to subject the security actors to civilian rule and to clearly delineate the role of the armed forces and the role of the security forces. It provides that the ‘role of the armed forces shall be to defend the independence and the territorial integrity of the State and its democratic institutions’ (article 19.1), while ‘the police and other security agencies shall be responsible for the maintenance of law and order and the protection of persons and their properties’ (article 19.2). Although the Protocol foresees the possibility that members of the armed forces may be called upon to participate in national development projects (article 19.5), it does not specifically foresee situations where the armed forces may contribute to the maintenance of internal law and order (such as in the case of emergencies that exceed the capacity of the police and other regular law enforcement bodies).

In this regard, Mali’s 2023 Constitution raises three notable issues. First, it defines the roles of the armed and security forces but does not distinguish their respective missions (article 89). As a result, the different security sector agencies seem to have the same general mandate. Second, the Constitution assigns the armed and security forces an expanded role that goes beyond their usual missions. In addition to protecting territorial integrity against external threats, maintaining public order, and protecting people and their property, the armed and security forces now participate in the ‘economic, social, and cultural development and environmental protection of the country’ (article 89). Furthermore, they are also responsible for ‘ensuring the enforcement of the law’ (article 89). This expanded role may not be consistent with the principle of civilian leadership and could create the potential for military intervention in response to allegations of civilian institutions breaking the law. Third, the Constitution defers the responsibility for determining the conditions under which the military may contribute to maintaining internal law and order to the legislature through ordinary statutes (article 91). Although Mali’s security situation may not easily accommodate a strict separation of functions between armed and security forces, the constitutional text could have specified conditions to better regulate the military’s role in maintaining internal law and order (see, for example, article 92 of the 2017 draft Constitution of Burkina Faso).

In Burkina Faso, the 1991 Constitution does not regulate the roles and missions of security sector agencies, including the military. Existing laws and decrees allow the military to intervene in the maintenance of public order at the request of the executive. By contrast, the 2017 draft Constitution defines the principal mission of the military—defending the country’s territorial integrity and national sovereignty (article 92). In addition, it acknowledges the contribution of the military in ensuring internal security but establishes a broad constitutional limitation on the use of military force against civilians. The draft provides that ‘the army contributes to the preservation of internal security’ and that ‘in no case [may] the army ... be authorized to use military force against all or part of the unarmed people’ (article 92).

In Chad, the 2018 Constitution contains detailed provisions on the structure and missions of the different security sector agencies (Title XII). However, some of these provisions lack clarity. National defence is ensured by the military, the gendarmerie and the National and Nomadic Guard, while internal public order is maintained by the police, the gendarmerie and the National and Nomadic Guard (article 188). Although article 188 assigns a role in national defence to some security forces (the gendarmerie and the National and Nomadic Guard), this role is not reflected in the respective missions of these security agencies (articles 192 and 195). Additionally, the mandates of the different security forces seem to partially overlap (articles 192, 193 and 195). By contrast, the missions of the military are well defined: it is tasked with defending the country’s territorial integrity, national unity, independence and security against external aggression or threats (articles 188 and 189). The military may also participate in economic and social development tasks as well as in humanitarian operations (article 190). However, it does not have an explicit role in the maintenance of internal public order (article 188).

Chad’s 2023 Constitution contains similarly detailed provisions on the structure and missions of the different security sector agencies (Title XVIII). Four defence and security sector agencies are listed—the military, the gendarmerie, the police, and the National and Nomadic Guard (article 242). These agencies are designated as apolitical (article 244) and subject to civilian authority (article 243). In an improvement on earlier constitutions, responsibility for national defence is now reserved for the military only (and no longer the gendarmerie and National and Nomadic Guard), and internal public order and security are the responsibility of the gendarmerie, police and National and Nomadic Guard (articles 246 and 247). This arrangement provides a clearer separation between the role of the military and that of the security forces. However, the missions of the various security forces continue to overlap. Another important change is that the 2023 Constitution removes the president’s previous power to expand the mandate of the armed and security forces by assigning them ‘any other tasks of public interest’ (article 94 of the 2018 Constitution).

7.2. General mobilization

From a comparative perspective, it is not uncommon for constitutions to assign a role to ordinary citizens in defending the territorial integrity of their country. Most often, this role takes the form of a general duty for citizens to defend the state (such a duty exists in Chad and in all ECOWAS member states except Guinea-Bissau and Liberia) and sometimes to take part in compulsory military training (Benin, Burkina Faso, Chad, Ghana, Niger and Togo).

In recognition of the security threats posed by non-state armed actors within their territories or by geopolitical developments, certain constitutional texts also incorporate explicit provisions for the declaration of partial or general mobilization. Mobilization entails the summoning of reserve component units of the defence forces—where applicable—and/or the enlistment of ordinary citizens to address the requirements of an emergency situation that poses a threat to national security. Given its significant implications, the procedure for declaring general or partial mobilization should be carefully defined in a way that balances the need for timely decision making and deliberation.

Mali’s 2023 Constitution places a duty on citizens to defend the homeland and provides that ‘all citizens from the age of 18 years old may be mobilized alongside the armed and security forces for the defence of the homeland’ (article 24). Regarding the procedure for declaring mobilization, the Constitution provides that ‘the President of the Republic orders general mobilization and determines the modalities of citizen participation in the defence of the homeland when the security situation so requires’ (article 63). Under this arrangement, the president may unilaterally order general mobilization without approval from or consultation with other state authorities, notably the Council of Ministers or the legislature.

In Chad, the 2018 Constitution empowered the president to order mobilization after deliberation in the Council of Ministers and in accordance with rules established by the legislature (articles 86, 100 and 127). The presidential act declaring mobilization had to be signed by the relevant ministers (article 100). However, as the ministers were accountable to (and could be dismissed by) the president, this countersignature was more a formality than a requirement for approval. If the president declared mobilization as part of the exceptional measures stipulated in article 96, the declaration had to be made by the president after deliberation in the Council of Ministers and following prior consultation with the speaker of the National Assembly and the president of the Supreme Court. In such cases, the countersignature of the relevant ministers was not required (articles 96 and 100).

Similarly, under Chad’s 2023 Constitution, the president may order mobilization after deliberation in the Council of Ministers and in accordance with rules established by the legislature (articles 97 and 127). The presidential act declaring mobilization should be signed by the prime minister and the relevant ministers (article 97). As the government is accountable to the lower house of the national legislature (and no longer formally accountable to the president), this countersignature equates to a requirement for approval. In other words, the decision to declare mobilization would not be a unilateral act on the part of the president but would require an agreement between the president and the government. If the president declares mobilization as part of the exceptional measures stipulated in article 93, the president makes the declaration after deliberation in the Council of Ministers and following consultation with the speakers of the two chambers of parliament and the president of the Constitutional Council (article 93). The countersignature of the prime minister is not required, as the use of exceptional powers under article 93 constitutes an exclusive competence of the president. However, the continued use of these exceptional powers after 15 days requires the approval of the legislature.

From a comparative perspective, countries that have explicit constitutional provisions on general mobilization usually grant the president the power to issue a mobilization order but establish certain constraints to prevent unilateral decision making. These constraints range from the limited requirement to consult the prime minister (Togo (1992), article 72), to mandatory approval by the legislature within a specified period following the declaration of mobilization (Ukraine (1996), article 85.31) to mandatory prior approval of parliament (Moldova (1994), articles 66.i and 87.2). In France, statutes provide that general mobilization is declared by a decree enacted in the Council of Ministers, thus requiring the signature (i.e. agreement) of the president, the prime minister and the relevant ministers (pursuant to article 13 of the Constitution). This arrangement reflects the constitutional distribution of responsibilities in matters related to national defence between, on the one hand, the president, who is the guarantor of national independence and territorial integrity (article 5) and commander-in-chief of the armed forces (article 15), and, on the other hand, the government, which is responsible for national defence (article 21) and has the armed forces at its disposal (article 20).

Burkina Faso’s draft Constitution retains a judicial structure similar to the one established under the 1991 Constitution and adds a separate Court of Audit responsible for overseeing public finances (Title X). The judiciary is thus structured with the Court of Cassation at the apex of the civil judicial order, the Council of State at the apex of the administrative judicial order, a specialized Constitutional Court as the only institution competent to assess the constitutionality of laws and a specialized Court of Audit.

Similarly, Mali’s Constitution retains a judicial structure similar to the one established under the 1992 Constitution, alongside a new Court of Audit (article 156). The judiciary thus comprises a civil and administrative judicial order with the Supreme Court at its apex, a separate Constitutional Court and a separate Court of Audit (Title V).

Chad’s Constitution introduces two main changes to the structure of the judiciary compared with the 2018 Constitution. While the 2018 text established a single order of jurisdiction with the Supreme Court at its apex, the 2023 Constitution re-establishes a separate Constitutional Council (Title VII) and establishes a Supreme Court of Audit (Title VIII).

As in their respective predecessor constitutions, the three constitutional texts affirm the independence of the judiciary as an institution (Mali (2023), article 129; Chad (2023), article 155; 2017 draft Constitution of Burkina Faso, article 147) and the independence of judges by guaranteeing their security of tenure (Mali (2023), article 132; Chad (2023), article 163; 2017 draft Constitution of Burkina Faso, article 149). However, despite a few positive changes, the three texts contain important shortcomings that may undermine the realization of judicial independence. The changes most likely to have an impact on the independence of the judiciary relate primarily to the composition of the judicial self-governing body and the procedures for judicial appointments.

8.1. Judicial self-governing body

All three texts establish a supreme council of the judiciary (CSM). Judicial councils are independent public institutions responsible for judicial administration. Their responsibilities typically encompass selecting, evaluating and promoting judges; developing codes of ethics and general norms governing judicial practice; conducting disciplinary proceedings against judges; and advising the government and the legislature on the budgetary needs of the justice system. Given their role in safeguarding judicial independence, judicial councils should themselves be independent and impartial. International standards provide that judicial councils should be composed of a majority of judges elected by their peers and representing all levels of the judiciary, and members representing other legal professions, and should have a gender-balanced composition. There should be no members of the executive or legislative branches on judicial councils, or, if present, they should not have a role in decision making (OSCE/ODIHR 2023).

In Burkina Faso, under the 1991 Constitution, the CSM was chaired by the president of the republic and vice-chaired by the minister of justice. The 2015 constitutional amendments removed both the president and the minister of justice from the CSM. The CSM is currently chaired by the first president of the Court of Cassation and vice-chaired by the first president of the Council of State (article 132). The composition of the CSM and details about its mandate are still specified in organic legislation (article 133).

The constitutional commission that prepared the 2017 draft Constitution included more details about the composition of the CSM. According to the draft, the new CSM would be composed of members with voting rights and members with consultative rights. Among members with voting rights, three-fifths would be magistrates, with the remaining two-fifths being non-magistrates. Non-magistrate members with voting rights would be selected respectively by the president of the republic, the speaker of the National Assembly, the Bar Association and human rights organizations. However, the draft does not specify how magistrate members would be selected or who would be responsible for selecting them, nor does it indicate the procedure for selecting non-magistrate members or the tenure of members of the CSM. These issues are deferred to organic laws. This omission represents a significant gap in the draft, which could impair the independence of the CSM through decisions of transient political majorities.

The constitutional commission decided to include non-magistrates in the CSM (while keeping a majority of magistrates) to reduce the risk (and the public perception) of corporatism, to strengthen the accountability of judges and ultimately to enhance public trust in the judiciary. The constitutional commission also maintained the exclusion of the president and the minister of justice from the CSM. Under the draft Constitution, the new CSM would be chaired by the first president of the Court of Cassation (article 158). As the head of state, the president of the republic would still be the guarantor of judicial independence (article 148). Accordingly, the president would chair a meeting every year with members of the CSM to discuss issues related to the functioning of the judiciary and could communicate with the CSM at any time (article 53) but would not hold a decision-making role in the CSM.

According to Chad’s Constitution, the president of the republic is still described as the guarantor of the independence of the magistracy (article 159) but is no longer chair of the CSM. The Constitution provides that the president of the Supreme Court chairs the CSM (article 159), rather than the president of the republic. However, no further details are provided about the composition of the CSM or about the procedure for selecting its members. These matters are to be determined by organic statute (article 164), which raises the same concerns as those mentioned above about Burkina Faso’s draft Constitution. In addition, although the president of the republic no longer chairs the CSM, they apparently retain full discretion in the appointment of the president of the Supreme Court and thus of the president of the CSM. The only constitutional constraints on this appointment are that the president of the Supreme Court must be drawn from the magistracy and that the president of the republic must consult the presidents of the two legislative chambers before making the appointment (article 167).

In Mali’s Constitution, shortcomings with regard to the independence of the judiciary are mainly due to the role of the president of the republic (who is now both head of state and head of the executive) in the functioning of the judicial system. While the presidency of the republic is now designed to be a political authority exercising all executive power (rather than a neutral arbiter), the president of the republic is still the guarantor of the independence of the judiciary (article 134) and still chairs the CSM (article 64). In addition, the Constitution does not specify the composition of the CSM, the procedure for appointing its members or every aspect of its powers. These elements are left to be determined by organic law, which raises similar concerns to those mentioned above about Chad’s Constitution and Burkina Faso’s draft Constitution.

8.2. Judicial appointments

Under Burkina Faso’s 1991 Constitution, judges (magistrats du siège) of the apex courts (the Court of Cassation, the Council of State and the Court of Accounts) were appointed by the president of the republic based on non-binding proposals from the CSM. Judges in lower courts were appointed by the president of the republic based on non-binding proposals from the Ministry of Justice, with the CSM providing its opinion on those proposals. Prosecutors (magistrats du parquet) were appointed by the president of the republic based on non-binding proposals from the Ministry of Justice (1991 Constitution of Burkina Faso, article 134). To depoliticize the appointment process and enhance the independence of judges from the executive, the 2015 constitutional amendments granted the CSM the authority to appoint all judges and prosecutors (1991 Constitution of Burkina Faso, as amended through 2015, article 134). The 2017 draft Constitution provides that the decisions of the CSM regarding the nominations of magistrates (i.e. judges and prosecutors) are binding on the authority that formally appoints them (article 157). In principle, this arrangement reflects good practice, as decision-making authority over judicial appointments is granted to the CSM. However, the effectiveness of this arrangement will depend on the composition of the CSM and the procedure for selecting its members, which are deferred to organic law.

In Chad, under the 2018 Constitution, the CSM proposed candidates for magistrate positions. Magistrates were then formally appointed by the president of the republic following the binding opinion of the CSM (articles 152 and 153). Under this arrangement, the president of the republic could exercise some influence over judicial appointments. The president of the republic—who also chaired the CSM—could choose candidates from among the nominees proposed by the CSM and then submit this list to the CSM for a binding opinion. If the CSM approved the list of candidates selected by the president from among its shortlist, the president could then appoint the selected candidates to magistrate positions. Under the 2023 Constitution, however, the president no longer plays any decision-making role in judicial appointments. Magistrates are nominated by the CSM and formally appointed by presidential decree (articles 160 and 161). The nominations made by the CSM are binding on the president of the republic. As in Burkina Faso, the effectiveness of this arrangement will depend on the composition of the CSM and the procedure for selecting its members, which are deferred to organic law. Notably, as mentioned above, the president of the republic appears to retain full discretion in the appointment of the president of the Supreme Court (Chad (2023), article 167). In addition, members of the Supreme Court serve renewable terms of seven years (Chad (2023), article 168). This arrangement may weaken judicial independence, as judges may be inclined to issue decisions that will increase their chances of reappointment.

In Mali, the 1992 Constitution simply provided that the president and vice-president of the Supreme Court were appointed by the president of the republic upon confirmation by the CSM (1992 Constitution of Mali, article 84). Other Supreme Court judges were appointed by decree in the Council of Ministers (1992 Constitution of Mali, article 47), which required the agreement of the president of the republic and the prime minister. Thus, it appears that the CSM did not play a role in the selection of other Supreme Court judges. The appointment procedure for other judicial positions was regulated by law. Organic law No. 03-033 of 7 October 2003, which regulates the CSM, provides that ‘judges of other jurisdictions are appointed by presidential decree upon proposal from the Ministry of Justice and following [the] binding opinion of the CSM’ (article 12). For prosecutors, the CSM issues a non-binding opinion (article 12). Under this arrangement, the executive selected judicial candidates, and the CSM could simply approve or reject these nominations. Under the 2023 Constitution, members of the Supreme Court are nominated by the CSM and formally appointed by the president of the republic. The nominations of the CSM are binding on the president. By contrast, members of the Court of Audit are appointed by the president of the republic upon confirmation by the CSM. The president of the republic has some decision-making power, as they proposes nominees who should then be approved by the CSM before the president formally appoints them (article 162). The 2023 Constitution does not regulate the appointment procedure for other judicial positions. This constitutes a significant gap, as this procedure would be defined by legislation, subject to the legislative majority. This arrangement risks weakening the independence of judges in first instance and appellate courts.

8.3. Court of audit

While oversight of public finances is an important function of legislatures and is often performed by legislative committees established for this purpose, it is also common to establish a specialized body with the technical expertise necessary to conduct a thorough audit of public finances and hold the government accountable for the use of public funds (Bulmer 2019).

All three constitutional texts discussed in this chapter establish such an institution—namely, a court of audit. In Burkina Faso, this institution already exists under the 1991 Constitution. In Chad, a Court of Audit was established in 2015 following constitutional amendments to the 1996 Constitution but was removed with the enactment of the 2018 Constitution, which transferred responsibility for financial audits to the Supreme Court. In Mali, the Court of Audit is a new institution. The establishment of an independent court of audit is also required by subregional treaties (for Burkina Faso and Mali, the Treaty of West African Economic and Monetary Union, article 68; for Chad, the Economic Community of Central African States Directive No. 01/11-UEAC-190-CM-22 on Money Bills, article 72 (2011).

In Burkina Faso and Chad, the court of audit is established as a separate institution and is no longer part of the judicial system (2017 draft Constitution of Burkina Faso, Titles VII and X; Chad (2023), Title VIII). Burkina Faso’s draft Constitution affirms the independence of the Court of Audit but leaves the determination of its composition and functioning to organic legislation (article 173). Deferring the appointment procedure to an organic statute raises similar concerns to those mentioned above regarding the CSM. In Chad, the Constitution does not guarantee the independence of the Court of Audit, and the president of the republic effectively determines its membership (Chad (2023), article 186). In addition, members are appointed for a six-year term, renewable once (Chad (2023), article 187). These provisions may undermine the independence of the Court of Audit.

In Mali, the Court of Audit is established as the apex court of the financial judicial order and thus as a part of the judicial system (Mali (2023), articles 129 and 156). Members of the Court of Audit are appointed by the president of the republic upon confirmation by the CSM. Under this arrangement, the president of the republic has some decision-making power, as they propose candidates who must then be approved by the CSM before the president formally appoints them (article 162).

8.4. Military justice

The three constitutional texts adopt different approaches with regard to military justice. Mali’s Constitution does not refer to military justice. Therefore, military courts are established and regulated by ordinary statutes and decrees enacted prior to the promulgation of the Constitution.³

Burkina Faso’s draft Constitution provides that the military justice system will be regulated by organic law (article 154). By contrast, Chad’s Constitution provides for the establishment of a military justice system. While the Constitution provides for a single order of jurisdiction with the Supreme Court at its apex (Chad (2023), article 156), it also establishes a military justice system that appears to be separate (Chad (2023), Title X). Compared with the 2018 Constitution, the organization of military courts has been changed slightly to comprise three levels of courts, with the High Military Court exercising final jurisdiction over judgments issued by military courts of first instance and appellate military courts. Thus, there seems to be no mechanism for military courts’ decisions to be subject to judicial review by the regular apex court (i.e. the Supreme Court).

While a comprehensive examination of the military justice systems in these three countries goes beyond the scope of this paper, it is important to note that military courts remain problematic in some countries. Ideally, courts that hear matters relating to the military and military personnel should be an integral part of the judiciary, and members of the military should be entitled to the same guarantees of a fair trial as civilians. Allegations of human rights violations (even those committed by military personnel) should be heard by ordinary courts, and cases involving civilians should not be adjudicated by military courts.

Mali’s Constitution and Burkina Faso’s draft Constitution maintain a concentrated system of constitutional justice, whereby one specialized institution—a constitutional court—reviews the compliance of laws with the constitution. Chad also previously had a concentrated system of constitutional justice until the enactment of the 2018 Constitution, which abolished the Constitutional Council and transferred judicial review competences to the Supreme Court. However, Chad’s 2023 Constitution reinstated a concentrated system of constitutional justice by re-establishing the Constitutional Council. Despite the continuity of these concentrated systems, each of the three constitutional texts introduces some changes regarding the composition, mandate or accessibility of this specialized judicial review body, compared with the previous constitution they aim to replace (Burkina Faso) or have effectively replaced (Chad and Mali).

9.1. Composition of the constitutional court or council

The constitutional court or constitutional council foreseen in these three constitutional texts consists of nine members. However, there are variations among the three texts regarding the procedure for selecting members and their tenure.

Mali’s Constitution and Burkina Faso’s draft Constitution provide that the members of their respective constitutional court are selected by several different authorities. This approach aims to ensure a balanced court by including members who reflect the political landscape and different political views, with a view to ensuring the body’s inclusiveness and legitimacy.

Under Mali’s 1992 Constitution, the Constitutional Court consisted of nine members—three selected by the president of the republic, three selected by the president of the National Assembly and three selected by the Supreme Judicial Council (1992 Constitution of Mali, article 91). Members of the Constitutional Court served a seven-year term, renewable once. The 2023 Constitution strengthens guarantees of the Court’s independence by introducing a non-renewable seven-year term for its members and a more balanced distribution of appointing authorities. Under the new Constitution, the Court consists of nine members—two selected by the president of the republic; one, by the speaker of the National Assembly; one, by the speaker of the Senate; two, by the Supreme Judicial Council; two, by rectors of law faculties; and one, by the Bar Association (Mali (2023), article 145). Under this arrangement, the three branches of government select two-thirds of the members, while the other third is selected by law professionals (i.e. academia and the Bar Association). Although this arrangement introduces a more balanced and diverse distribution of appointments among the different branches of government, in situations of majority alignment, the president could still strongly influence the appointment of six of the nine members of the Constitutional Court—the two members appointed by the president, the member appointed by the president of the National Assembly, the member appointed by the president of the Senate and the two members appointed by the CSM since the president chairs the CSM and the Constitution does not specify the method for appointing the other members of the CSM.

Burkina Faso’s current Constitution and the 2017 draft Constitution also provide for a selection procedure in which several different authorities appoint members separately. Since the 2015 constitutional amendments, the Constitutional Council has consisted of nine members—three high-level judges appointed by the president of the republic on the proposal of the minister of justice, three members (including the president of the Constitutional Council) appointed by the president and three members appointed by the speaker of the legislature (article 153). The current arrangement may raise concerns because, in situations of majority alignment, the president may exert a strong influence on the selection of all members of the Council.

The 2017 draft Constitution diminishes the role of the president in the appointment process significantly. Indeed, the constitutional commission decided to retain the process of appointing members by different nominating authorities but ensured that no single authority would appoint a majority of members of the Court. The proposed Constitutional Court would have nine members—two of them appointed by the president of the republic; two, by the speaker of the National Assembly; two, by the Supreme Judicial Council; one lawyer nominated by the Bar Association; a professor of public law designated by their peers; and a representative of human rights and democracy organizations designated by their peers (article 166). Members would serve a non-renewable term of six years, and half of the members would be replaced every three years (article 167). Compared with Mali’s Constitution, the arrangement foreseen in Burkina Faso’s draft Constitution offers more safeguards for the independence of the Constitutional Court. Even in situations of majority alignment, the president of Burkina Faso could potentially have a strong influence on the appointment of only four of the nine members of the Constitutional Court (the two members appointed by the president and the two members appointed by the speaker of the National Assembly).

Chad’s Constitution also introduces changes to the procedure for appointing members of the Constitutional Council compared with the 1996 Constitution. Under the 1996 Constitution, the president of the republic and the presidents of the National Assembly and the Senate each appointed three members of the nine-member Constitutional Council (1996 Constitution of Chad, article 165). This arrangement gave the president significant influence over the appointment of all members of the Council because the president’s party held a majority in the legislature. The 2023 Constitution provides that the Constitutional Council consists of nine members appointed by the president of the republic but defers the selection procedure to organic legislation (Constitution of Chad, article 183). This arrangement may further weaken the Constitutional Council’s independence, as the selection procedure will be determined by an absolute majority vote of MPs. Members of the Constitutional Council are appointed for a non-renewable nine-year term.

A major gap in all three constitutional texts concerns the lack of clear provisions stipulating the grounds and process for removing judges from the constitutional court or council. Regardless of the appointment process, if judges are not protected from arbitrary or politically motivated threats of removal, the independence of the court or council may be compromised. Leaving these fundamental issues to organic laws may allow transient political majorities to interfere in the functioning of the court or council.

9.2. Mandate of the constitutional court or council

The three constitutional courts or councils have the power to review the constitutionality of organic and ordinary laws, as well as treaties. Only Burkina Faso’s draft Constitution explicitly provides that the Constitutional Court has the power to review constitutional amendment bills (on both procedural and substantive grounds), bills to be submitted to a referendum and ordinances (articles 164 and 188). This arrangement is commendable, as it clarifies the jurisdiction of the Constitutional Court in these matters and thus contributes to legal certainty. Conversely, in Chad and Mali, the constitutional silence with regard to judicial review of constitutional amendment bills and referendum bills creates uncertainty and therefore potential controversy. In particular, if the constitutional court or council refuses to review constitutional amendment bills on substantive grounds, it would be unclear how the unamendable provisions could be protected in practice.

All three constitutional texts provide for mandatory judicial review of organic laws and the rules of procedure of the legislative chambers before they enter into force. Burkina Faso’s draft Constitution additionally provides for mandatory judicial review of constitutional amendment bills, money bills and resolutions of the National Assembly (articles 165 and 188). Chad’s Constitution also provides for mandatory judicial review of bills concerning fundamental rights and freedoms (article 174).

The three constitutional courts or councils are also competent to adjudicate conflicts of competence between state institutions and electoral disputes (Mali (2023), articles 149, 150 and 151; Chad (2023), article 174; 2017 draft Constitution of Burkina Faso, articles 163, 164 and 165). In Mali and Burkina Faso, the Constitutional Court also announces the results of referendums and national elections, whereas in Chad the Constitutional Council announces the results of referendums only (Mali (2023), articles 149, 150 and 151; Chad (2023), article 174; 2017 draft Constitution of Burkina Faso, articles 163, 164 and 165).

9.3. Accessibility to the constitutional court or council

All three constitutional texts establish similar arrangements regarding access to the constitutional court or council for ex ante judicial review—that is, review after a bill has been enacted by the legislature but before its promulgation by the president. Under the three texts, the political authorities entitled to refer a bill to the constitutional court or council for review of its compliance before promulgation include the president of the republic, the prime minister, the speaker of the national assembly, the speaker of the senate (Mali and Chad), one-tenth of the members of the national assembly or one-tenth of the members of the senate (Mali and Chad) (Mali (2023), article 147; Chad (2023), article 179; 2017 draft Constitution of Burkina Faso, article 168). This arrangement enables the legislative minority to seek a ruling on the constitutionality of a bill to ensure that the majority abides by the terms of the constitution (see Chapter 11: Status and rights of the parliamentary opposition).

Fourth-branch institutions are not entitled to submit cases to the constitutional court for review. However, Burkina Faso’s draft Constitution differs from the constitutions of Mali and Chad in two important respects. First, it allows individual citizens to file a direct petition to the court for ex ante judicial review. Such an arrangement is rare in comparative practice and could risk disrupting the legislative process and overloading the court with ex ante petitions. This mechanism could therefore be reconsidered or, at a minimum, accompanied by strict criteria for individual standing. Second, Burkina Faso’s draft Constitution grants the Constitutional Court self-referral power, seemingly for both ex ante and ex post judicial review. Self-referral may be relevant in situations where the legislative minority (in ex ante review) and individuals (in ex post review) lack access to the Constitutional Court. However, given that Burkina Faso’s draft Constitution enables the legislative minority and individuals to initiate proceedings before the Constitutional Court, granting the Court self-referral power could raise the spectre of judicial activism and possibly expose the Court to criticism and attacks from political actors if the Court used this power too frequently.

All three constitutional texts also provide for ex post judicial review through incidental referral (Mali (2023), article 153; Chad (2023), article 180; 2017 draft Constitution of Burkina Faso, article 168). Incidental referral allows any litigant, during proceedings before an ordinary court, to argue that a legislative provision infringes on the rights and freedoms guaranteed by the constitution and to request that the ordinary court refer the matter to the constitutional court or council for a ruling on the constitutionality of the contested legislative provision.

In Mali, incidental referral is limited to cases where a legislative provision is alleged to infringe the litigant’s constitutional rights and freedoms. By contrast, in Chad and under Burkina Faso’s draft Constitution, incidental referral can be triggered by a litigant when any legislative provision relevant to the case is alleged to be incompatible with any provision of the constitution, not only those provisions concerning constitutional rights. In addition, in Mali, incidental referral can be triggered by a litigant regardless of citizenship. Under Chad’s Constitution and Burkina Faso’s draft Constitution, on the other hand, incidental referral may be triggered only by litigants who are citizens. This citizenship requirement may constitute discrimination and undermine the right to equality before the law.

While incidental referral already existed under Chad’s previous Constitution and continues to exist under Burkina Faso’s current Constitution, it constitutes an innovation in Mali. The introduction of incidental referral in Mali affords additional protection of constitutional rights and affirms the role of the Constitutional Court as guarantor of these rights and freedoms. Ex post judicial review makes it possible to identify unconstitutional provisions that become apparent during the implementation of a law and that might not have been detectable at first glance before its promulgation.

Fourth-branch institutions are public bodies that are politically neutral and independent from the three main branches of government and whose purpose is to ensure the integrity and improve the quality of democratic governance (Bulmer 2019). These institutions are referred to by various terms, such as fourth-branch institutions, independent regulatory and oversight institutions, state institutions supporting democracy (South Africa) or independent administrative authorities (particularly in francophone jurisdictions). Broadly, three types of fourth-branch institutions exist:

• institutions that safeguard the procedural fairness and integrity of the political system (e.g. electoral management bodies, judicial service commissions or superior councils of the magistrature, media commissions and anti-corruption commissions);
• institutions that scrutinize and report on government performance in particular policy areas and that sometimes formulate recommendations for decision-making authorities (e.g. national human rights institutions, gender and minority commissions, ombudspersons or mediators of the republic, or consultative bodies such as economic, social and environmental councils); and
• institutions that have an administrative or regulatory role within a specialized policy sphere (e.g. central banks, financial markets authorities, competition authorities).

If properly established, these institutions can help to improve the quality of democratic governance. They do so either by insulating certain types of state activity (such as the holding of elections or judicial appointments) from partisan politics or by providing a dedicated mechanism for publicly scrutinizing and reporting on other types of state activity (such as enforcing human rights or gender equality) (Bulmer 2019). Importantly, in political regimes where the directly elected president tends to exercise significant powers and responsibilities, fourth-branch institutions can provide additional checks and balances alongside relationships within and among the three traditional branches of government.

In older and stable democracies, fourth-branch institutions are often established by ordinary statute or are referred to in general terms in the constitution, with details about their design left to legislation. However, reliance on statutory provisions makes these institutions dependent on the goodwill of the legislative majority for their powers, resources and autonomy. Such arrangements can expose these institutions to partisan manipulation and hinder their ability to perform their duties independently. For this reason, several more recent constitutions include detailed provisions not only about the existence of fourth-branch institutions but also specifying their composition, methods of selection and removal, required qualifications, terms of office, mandate and funding. Such detailed constitutional provisions may be particularly helpful in situations where these institutions (a) are being set up anew by the constitution, (b) have been weak or have lacked independence in the past, or (c) operate in contexts where the legislature is relatively weak and likely to be dominated by the executive.

The three constitutional texts examined in this study provide for fourth-branch institutions, albeit with significant variations, especially in terms of the level of detail provided in the constitution. Notably, all three texts foresee the establishment of a consultative body aimed at providing a dedicated platform for representatives of socio-professional groups involved in economic and social life to share their experiences and perspectives on societal issues in order to contribute to certain policy decisions. Such consultative bodies can be found in several civil law jurisdictions.

Mali’s 2023 Constitution simply provides that fourth-branch institutions are to be established by ordinary statute (article 88). The number of such institutions as well as their respective mandates, composition and resources are left to regulation through ordinary legislation (which requires a simple majority vote in the legislature). Although this arrangement represents a positive improvement compared with Mali’s 1992 Constitution—which did not mention fourth-branch institutions—it defers to the legislative majority to design these institutions and therefore creates a risk of capture by that majority. The only fourth-branch institution that is constitutionalized is the Economic, Social, Environmental and Cultural Council (Title VI). This institution is a consultative body composed of representatives from various functional groups (unions, socio-professional groups, women’s and youth organizations, the diaspora, etc.) tasked with developing reports and formulating recommendations on economic, social, environmental and cultural policy issues for decision-making authorities. The Council plays a consultative role across relatively broad areas compared with similar institutions found in other francophone jurisdictions. It must submit an annual report that highlights concerns and expectations of the public and contains recommendations for the president, the prime minister and the presidents of the two chambers of parliament; it can also develop reports on its own initiative and submit them to these officials; and it can be consulted by the president of the republic on any government-sponsored bill, ordinance or decree (articles 165, 166 and 167). However, certain key design features that are critical for the independence of this institution—notably the process for selecting its members and determining their tenure and resources—are deferred to organic law (article 173).

Burkina Faso’s 2017 draft Constitution takes a slightly different approach. It refers to fourth-branch institutions in general terms but explicitly identifies several such institutions—namely, an anti-corruption commission, an electoral management body, an information and communication regulatory body, a mediation body and an ‘organ of decision-making support and anticipation of social changes’ (article 179) (which would presumably replace the economic and social council provided for under the 1991 Constitution). Their respective mandates, composition, tenure, functioning and resources are deferred to organic statutes (Title XI). Under this arrangement, the institutions have constitutional status, but their entire design is left to organic laws, which require an absolute majority vote in the legislature and are subject to mandatory ex ante review by the Constitutional Court (article 132).

Chad’s 2023 Constitution provides more detail on fourth-branch institutions, but some of these institutions lack sufficient constitutional guarantees for their independence. It establishes several institutions designated as independent—an electoral management body (Title XVII), a human rights commission (Title XII), a media and audiovisual regulatory authority (Title XIII) and a mediator of the republic (Title XVI). The Constitution specifies the overall mandate and functions of each of these bodies and regulates the composition of some of them. The mediator of the republic is appointed unilaterally by the president of the republic (articles 97 and 235). The media and audiovisual regulatory authority consists of nine members appointed unilaterally by the president of the republic by decree (without prime ministerial countersignature) for a three-year term, renewable once (articles 97 and 219). The combination of unilateral appointment by the president and a short renewable mandate does not bode well for the independence of these institutions.

The electoral management body is responsible for organizing and managing all electoral and referendum operations (article 238). The Constitution provides that it is independent and has no hierarchical links with any other state institutions (articles 239 and 240), but the key design features that are critical for its actual independence—notably the procedure for selecting its members and determining their tenure and the body’s resources—are deferred to organic statute (article 241). The national human rights commission consists of 11 members, including 9 members elected by their respective corporations and 2 appointed (presumably unilaterally by the president, as article 97 suggests) from among resource persons (article 212) for a four-year term that may be renewed once (article 213). The precise mandate, details of the appointment procedure and functioning of the commission are deferred to ordinary legislation (article 215).

The Constitution also describes several other institutions, including the Court of Audit (Title VIII) and the Economic, Social, Cultural and Environmental Council (Title XI). The Court of Audit, which oversees the implementation of the state budget (article 185), is not an independent body. It consists of 21 members appointed for a six-year term, renewable once (articles 186 and 187). The chairperson is appointed by decree by the president of the republic following the non-binding opinion of the speakers of the two legislative chambers (article 186). The Constitution does not specify the procedure for selecting the remaining 20 members. The president of the republic and the presidents of the two legislative chambers may consult the Economic, Social, Cultural and Environmental Council to provide opinions on issues of an economic, social, cultural or environmental nature (article 204). Article 204 seems to imply that the Council may also issue reports on its own initiative. It comprises 21 members, including one president appointed by the president of the republic following the non-binding opinion of the presidents of the two legislative chambers (articles 205 and 206). The procedure for selecting the other members, their tenure and the functioning of the Council are to be determined by organic statute (article 207).

In summary, in comparison with the constitutional texts of Mali and Burkina Faso, Chad’s 2023 Constitution provides stronger constitutional recognition of fourth-branch institutions but lacks effective constitutional guarantees for their independence.

To strengthen the independence and effectiveness of fourth-branch institutions, the three constitutional texts could explicitly list each fourth-branch institution and specify its composition, selection and removal procedures, required qualifications and terms of office, as well as its mandate, powers and funding. These provisions could be designed in a way that ensures that these institutions are independent from other branches of government and politically neutral, that they have a clear mandate and sufficient powers and resources to perform their functions, and that their members are professionally competent and broadly representative of the country’s diversity. For example, Ghana’s Constitution provides such detail for the national media commission (Ghana (1992), Chapter 12). Other comprehensive provisions can be found in the constitutions of Kenya and South Africa, which regulate most of these aspects of fourth-branch institutions (Kenya (2010), Chapters 12 and 15; South Africa (1996), Chapters 9 and 10).

A functioning and effective parliamentary opposition (i.e. individual MPs or parliamentary groups that do not support the government) is essential to democracy (Venice Commission 2010; Democracy Reporting International 2013; Bulmer 2021). Opposition parties oversee the actions and decisions of the governing majority and hold it accountable. They also formulate political projects and policy alternatives, channel different opinions in public debate and offer voters an alternative to the incumbent governing majority. As such, a functioning parliamentary opposition is a key component of a genuine pluralistic multiparty system.

To be effective, the opposition needs a robust legal framework. Opposition parties rely upon a wide range of constitutional protections, such as freedoms of association, assembly and expression, backed by an independent judiciary. These broad constitutional protections ensure that opponents of the governing majority continue to enjoy equal rights and are not criminalized, harassed or otherwise disadvantaged. Additionally, some democratic systems grant the parliamentary opposition specific roles, rights and means to enable it to operate effectively and perform its function. These opposition rights vary in scope but generally aim to empower the opposition to scrutinize the actions and decisions of the governing majority, to participate in the work of parliament and to influence the lawmaking process, as well as to participate in the appointment of certain senior officials and members of fourth-branch institutions. These rights can be implicit, granted to all parliamentarians, parliamentary groups or a qualified minority of MPs, or explicit, granted to the opposition as such. These specific parliamentary opposition rights can be established at different levels of the legal hierarchy and are often detailed in parliamentary rules of procedure. However, because sub-constitutional rules (such as ordinary statutes or parliamentary rules of procedure) can be more easily modified, some countries have decided to constitutionalize the status and rights of the opposition to provide stronger guarantees for such rights (Democracy Reporting International 2013; Bulmer 2021).

In particular, in political systems where the directly elected president tends to exercise significant power and is often supported by a majority in the legislature (as has largely been the case in the countries considered in this study), anchoring the status of the parliamentary opposition in the constitution and granting specific constitutionally guaranteed prerogatives to opposition groups may help strengthen the balance of power. During periods of majority alignment (i.e. when the president has a majority in parliament), the traditional separation and balance between the executive and legislative branches tend to erode, and the dynamic shifts towards the opposition and the governing majority. To counterbalance the concentration of power in the presidency and the recurring majority rule (le fait majoritaire), a good practice consists in granting specific constitutionally guaranteed powers and functions to the parliamentary opposition so that it can effectively perform its function of opposing the majority, scrutinizing its actions and formulating policy alternatives.

Similarly, in countries lacking a strong tradition of fair play and mutual respect between the government and the opposition, stipulating the status and main rights of the opposition in the constitutional text (as opposed to leaving them to ordinary statutes, parliamentary rules of procedure or unwritten conventions) provides a higher level of legal protection. While sub-constitutional rules may be easily circumvented or amended by the governing majority to weaken the opposition and prevent effective scrutiny, constitutional provisions are usually more difficult to change and typically require a qualified majority and therefore the support of multiple political forces, possibly including those in the opposition. The constitutional protection of opposition rights can limit the risk of the resurgence of a de facto one-party system. It can also institutionalize majority–opposition relations and provide assurances to all parties that they will have a role in the political system, even in the event of an electoral loss.

The ECOWAS Protocol on Democracy and Good Governance does not provide specific guidelines on the status or role of the opposition (ECOWAS 2001). However, the African Charter on Democracy, Elections and Governance (article 3.11) provides that states parties must work to enhance political pluralism, including by formally recognizing the role and rights of opposition political parties (African Union 2007).

The three constitutional texts examined in this analysis either lack or contain only minimal constitutional provisions guaranteeing explicit rights for the opposition as such, albeit with important variations.

Mali’s 2023 Constitution does not enshrine the notion of opposition. There is no constitutional obligation for the legislature to enact laws or other sub-constitutional rules defining the opposition and granting it specific roles and prerogatives. Under this arrangement, the establishment of specific rights for the opposition depends on the political will of the governing majority. The only indication of implicit opposition rights in the 2023 Constitution is that it allows a legislative minority to appeal to the Constitutional Court. One-tenth of the members of the National Assembly or the Senate can refer a bill to the Constitutional Court for a ruling on its compliance with the Constitution before its promulgation (article 147). This mechanism enables the legislative minority to seek a ruling on the constitutionality of a bill to ensure that the majority abides by the terms of the Constitution. Even if a constitutional challenge is not successful, it can draw public attention and shed light on issues of controversy.

Chad’s 2023 Constitution recognizes a general right to democratic opposition but does not define the opposition or assign it specific roles and powers. The status of the opposition is to be defined by ordinary legislation (article 32). Under this arrangement, the Constitution places an obligation on the legislature to define opposition rights through ordinary legislation. It is important to note, however, that the existing legislation on the status of the opposition was enacted by the executive through an ordinance procedure in 2018 rather than by the legislature. The Constitution implicitly provides two rights to the parliamentary opposition. It allows one-tenth of the members of the National Assembly to initiate a motion of censure against the government (article 151) and one-tenth of the members of either chamber of parliament to refer a bill to the Constitutional Council for a ruling on its constitutionality before its promulgation (article 180).

Burkina Faso’s 2017 draft Constitution is the only text examined in this analysis that constitutionalizes specific rights of the opposition as such. It establishes an obligation for the president of the republic to consult the opposition on issues of national interest (article 75) and allows the opposition to submit a proposal in the National Assembly twice a year to establish and chair a commission of inquiry (article 113). These innovations are a step forward compared with the 1991 Constitution, as they aim to grant the opposition a consultative role in identifying important policies (through consultations with the president) and to empower the opposition to scrutinize the executive and the majority in parliament (through commissions of inquiry). However, these two provisions have certain gaps. For example, the draft does not define what constitutes ‘issues of national interest’ and may therefore leave the president some discretion to determine when consultation with the opposition is required. Additionally, the status of opposition groups and the criteria for qualifying as such would be regulated by ordinary statute, therefore requiring a simple majority in the National Assembly. By contrast, the conditions and procedures for establishing a commission of inquiry would be regulated by the rules of procedure of the National Assembly. Those rules of procedure would have a status similar to that of organic statutes, requiring adoption by an absolute majority vote and being subject to mandatory ex ante judicial review. As in Chad, the status of the opposition and potentially additional prerogatives would be defined by ordinary statute (2023, article 137). In addition, the draft Constitution provides two rights to the legislative minority. It allows one-quarter of the members of the National Assembly to initiate a motion of censure against the government (article 118) and one-tenth of the members of the National Assembly to refer a bill or a statute to the Constitutional Council for a ruling on its constitutionality (articles 164 and 168).

By comparison, several other recent or recently amended presidential or semi-presidential constitutions provide stronger constitutional guarantees for the empowerment of the parliamentary opposition. These constitutionally guaranteed opposition rights can be classified in the three categories laid out in Table 11.1.

While the three countries analysed in this study are heterogeneous in terms of ethnicity, geography and history, their constitutional texts maintain a unitary state structure with limited guarantees for decentralization, albeit with variations (Mali (2023), article 30; Chad (2023), article 1; 2017 draft Constitution of Burkina Faso, article 42). Notably, none of these texts explicitly provide for the possibility of asymmetric decentralized arrangements, whereby certain parts of the territory could be granted more autonomy in certain policy issues than others. The texts also do not explicitly preclude such asymmetric arrangements, which could potentially be pursued through legislation.

Mali’s Constitution (2023) contains little detail regarding decentralization. It simply guarantees the existence of elected decentralized entities and the free administration of decentralized institutions, and it commits the central authorities to ensuring the ‘harmonious development’ of decentralized entities (articles 174, 176 and 178). The Constitution defers to ordinary legislation (and therefore to a simple majority) the task of determining the number of tiers, delineation, structure, composition, responsibilities, operation and financing of decentralized entities (articles 115, 176 and 177). Importantly, the Constitution does not explicitly allow decentralized entities to have their own administration (civil service) or to hire and manage their own personnel. Additionally, it does not regulate the oversight powers of deconcentrated authorities (representing the central government) over decentralized entities. This arrangement raises concerns, as the representatives appointed by the central government in each decentralized entity wield significant substantive authority through oversight and advisory roles. Such arrangements may limit the autonomy of decentralized entities. While this subject goes beyond the scope of the present analysis, legislation on decentralization could narrow the role of deconcentrated administrations vis-à-vis decentralized entities to oversight functions only, limiting the powers of deconcentrated officials to conducting audits, referring decisions of decentralized authorities to the courts (but without the power to void bills themselves) and limiting their role in local development planning.

The decentralization provisions in Mali’s Constitution are particularly noteworthy because they do not reflect the commitments made in the Agreement for Peace and Reconciliation resulting from the Algiers Process concluded in 2015 between the Government of Mali and the Coordination of Azawad Movements (Republic of Mali 2015). The Algiers Agreement included a commitment to establish decentralized territorial authorities with ‘extensive powers’ (article 5) and to transfer 30 per cent of state budgetary revenues to decentralized authorities (article 14) in order to enable northern populations to administer their own affairs (article 6). These commitments to extensive and asymmetric decentralization, made during the 2014 peace negotiations, are not reflected in the Constitution. While these arrangements could be introduced through legislation, they would remain vulnerable to subsequent amendments, thereby undermining the autonomy of the decentralized entities.

Similarly, the Algiers Peace Agreement included a commitment to increasing the representation of northern populations in decision making at the national level, notably through the establishment of a second chamber at the central level of the legislature. In this regard, the 2023 Constitution establishes a Senate composed of members representing decentralized territorial authorities, traditional authorities and Malians in the diaspora (article 97). Nevertheless, the influence of decentralized entities on legislation may be limited, as the National Assembly can take the final decision on all bills in case of disagreement between the two chambers (article 123).

Burkina Faso’s draft Constitution provides for similar decentralization arrangements (article 46). The only notable difference lies in the fact that the delineation, structure, operation and financing of decentralized institutions are to be determined by organic legislation—rather than ordinary legislation as in Mali—and thus require an absolute majority vote in the unicameral legislature and ex ante judicial review by the Constitutional Court (articles 46 and 132). The competences of decentralized entities are to be defined by ordinary legislation (article 137). A simple legislative majority can thus determine the scope of decentralization by transferring or withdrawing responsibilities from decentralized entities.

In Chad, the 2023 Constitution maintains a decentralized unitary state structure, despite demands made by several non-state armed groups during the peace negotiations in Doha and the ensuing inclusive national dialogue in 2022 to consider the establishment of a federal state. However, compared with Mali’s Constitution and Burkina Faso’s draft Constitution, Chad’s Constitution regulates more aspects of the decentralization system (Chad (2023), articles 1, 2 and 254–72). It guarantees the existence of elected decentralized entities and the free administration of decentralized institutions and specifies the number of tiers of decentralized entities, the structure of the sub-state deliberative and executive bodies and their relationship. The Constitution also specifies the funding sources of decentralized entities, provides that transfers of competences from the central government should be accompanied by adequate resources, and affirms that decentralized entities have their own administration and may recruit their own personnel. The Constitution defers to organic law (and therefore to an absolute majority vote in each of the two chambers of the central legislature) the task of specifying the number, delineation and responsibilities of decentralized entities (articles 139, 255 and 272).

At first glance, the arrangement in Chad appears to provide stronger protection for decentralization, as several important aspects are constitutionalized. Furthermore, the participation of decentralized entities through the Senate in the formulation of organic laws could, in principle, give them some influence over defining the scope of their responsibilities. However, the president’s unilateral power to appoint one-third of the senators gives the president considerable influence over legislative matters, potentially diminishing the weight of decentralized entities in the chamber. It should also be noted that the constitutional provisions regarding decentralization in the 2023 Constitution are almost identical to those contained in the 2018 Constitution. The 2023 Constitution promises—as the 2018 Constitution did—the establishment of what could be a highly decentralized state. However, despite the inclusion of similar provisions in the 2018 Constitution, decentralization in Chad remained limited. Whether the provisions of the 2023 Constitution will be implemented in a way that establishes a highly decentralized state depends on two main elements: (a) the content of the laws needed to implement them; and (b) whether those arrangements are implemented. The Constitution does not provide timelines or monitoring mechanisms for implementation (as was the case in Kenya, for instance, where the 2010 Constitution set out timelines for the implementation of its devolution arrangements and established an implementation commission).

13.1. Constitutional amendment procedures

Constitutions need to evolve over time to adjust provisions that have proven to be inadequate or unworkable; to respond to emerging understandings or needs or changing public demands; and to reflect evolving concepts of rights, values and institutions. Constitutional amendment provisions, which regulate the conditions and procedures for modifying the content of a constitutional text, have therefore become a near-universal feature of modern constitutions. Constitutional amendment procedures should be designed in a way that (a) allows a constitution to be adapted for the public good when necessary, when supported by a sufficient consensus and after careful consideration; but (b) prevents it from being changed by transient political majorities to extend or secure their tenure, marginalize the opposition or minorities, or pursue other self-interested, partisan or short-term motives (Böckenförde 2014).

When assessing constitutional amendment procedures, at least five key elements need to be examined: (a) who has the right to initiate a constitutional amendment procedure; (b) how constitutional amendment bills are adopted; (c) whether there are different thresholds for amending different parts of the constitution; (d) whether certain provisions are unamendable; and (e) whether restrictions apply to amendments in certain circumstances (e.g. during war or a state of emergency).

In Mali, the 2023 Constitution retains essentially the same constitutional amendment procedure as the one provided in the 1992 Constitution but takes into account the introduction of the Senate. The president of the republic and all MPs may initiate an amendment procedure. Amendment bills must be approved by each chamber of parliament separately by a two-thirds majority of their members and then approved by referendum (article 184). Although constitutions are more difficult to amend than legislation, this revision procedure appears restrictive from a comparative point of view, as it requires the holding of a referendum. Mali’s history underscores the rigidity of this amendment, as three attempts to amend the 1992 Constitution failed (in 2000, 2011 and 2019). The establishment of a new Senate, which will have to adopt amendment bills by a two-thirds majority, may make constitutional change even more difficult. This difficulty may, in turn, encourage the adoption of ostensibly new constitutions through referendums to bypass the amendment requirements, as has happened in other francophone African countries (Abebe 2024).

Malian constitution-makers could have drawn inspiration from comparative examples by providing for different amendment procedures depending on the provisions in question. Another alternative could have been to provide different routes for amendment. For example, the 2019 constitutional amendment bill proposed a new procedure that distinguished between constitutional amendment bills introduced by MPs (proposition de loi constitutionnelle) and those introduced by the president (projet de loi constitutionnelle). The former would first need to be approved by a simple majority vote in each of the two chambers of parliament and then by a referendum. The latter would first need to be approved by a simple majority vote in each of the two chambers of parliament, but then the president could decide between submitting it to a referendum or to the legislature in a joint sitting for adoption by a three-fifths majority vote. While referendums may be justified for the adoption of certain constitutional changes in order to ensure popular legitimacy of those changes, the arrangement proposed in the 2019 amendments could have avoided the systematic holding of a referendum, a costly and challenging exercise given the country’s security situation.

With regard to circumstantial restrictions on constitutional amendment, the 2023 Constitution (like the 1992 Constitution) simply prohibits constitutional amendments during periods when the territorial integrity of the state is infringed (article 185). This provision raises two main issues. First, given the security situation in Mali, and in particular the fact that non-state armed groups control part of the country, prohibiting amendments when territorial integrity is infringed could have the effect of rendering the Constitution unamendable until the state regains control over the entire country, which could take years. In fact, this provision is the result of constitutional borrowing from France. The prohibition, provided for in article 89 of France’s 1958 Constitution, was originally introduced in France’s 1946 Constitution to prevent future constitutional changes during periods of foreign occupation. On 10 July 1940, under Nazi occupation, the parliament of the Third Republic enacted a constitutional law that granted full powers to the government of Philippe Pétain and authorized it to adopt a new constitution. To prevent future constitutional reforms enacted under the pressure of occupation, this prohibition was introduced in the 1946 Constitution of the Fourth Republic and reiterated in the 1958 Constitution. Such a provision may not be well suited in the case of Mali, as some parts of the territory are controlled by non-state armed groups, and the conflict may continue for years.

Second, the 2023 Constitution does not prohibit constitutional amendments during a state of exception (i.e. a state of emergency, war or the exercise of exceptional presidential powers under article 70). Therefore, the president could potentially resort to exceptional crisis powers (article 70) to amend the Constitution unilaterally. Since states of exception result in a temporary hyper-concentration of power in the hands of the executive, a good practice consists in expressly prohibiting, or at least strictly limiting, constitutional amendments during such periods. Such measures would help prevent hasty amendments adopted in times of crisis and uncertainty and would explicitly prohibit the president from resorting to exceptional crisis powers to unilaterally amend the Constitution.

Furthermore, the 2023 Constitution does not expressly exclude constitutional amendment bills from the president’s referendum power (regulated by article 60), meaning that the president of the republic could potentially push through a popular constitutional reform without a political consensus between the country’s various political forces represented in parliament. From a comparative perspective, Burkina Faso’s 2017 draft Constitution expressly excludes constitutional amendment bills from the president’s referendum power (article 66).

Finally, the 2023 Constitution provides that the republican form of the state, secularism, the number (but not necessarily the duration) of presidential term limits and multipartyism are unamendable. The entrenchment of the number of presidential term limits in the 2023 Constitution is a new substantive limitation compared with the 1992 text (see Chapter 5: Presidential term limits).

Under Burkina Faso’s 1991 Constitution, constitutional amendment bills can be introduced by the president, a majority of MPs or at least 30,000 registered voters (article 161). Such bills must first be considered by the National Assembly (it is unclear whether this step entails a vote) and then be adopted either by a three-quarters majority vote of all members of the National Assembly or by a simple majority vote in a referendum (articles 163 and 164). The 2017 draft Constitution contains a similar amendment procedure but with some important changes. First, the threshold required for MPs to introduce an amendment bill is lowered from an absolute majority of MPs (in the 1991 text) to one-fifth of the MPs (article 186 in the 2017 draft). This proposed change would enable opposition groups to introduce constitutional amendment bills (article 186). Second, the 2017 draft increases the number of registered voters required to initiate constitutional amendments from 30,000 to 50,000. Third, the 2017 draft specifies that consideration by the National Assembly requires a three-fifths majority vote of all its members (article 187). This addition ensures that consideration by the National Assembly entails more than debate or consultation. As the president retains the power to propose a constitutional amendment bill in the 2017 draft Constitution, this addition would prohibit the president from bypassing the parliament and putting an amendment bill to a referendum without approval by a qualified majority in the National Assembly. Fourth, once considered (i.e. approved by a three-fifths majority vote of all members of the National Assembly), the amendment bill would need to be adopted by a four-fifths majority of all members of the National Assembly (as opposed to three-quarters under the 1991 Constitution) or through a referendum. This amendment procedure appears particularly rigid and cumbersome from a comparative perspective. Although the requirement of a four-fifths majority vote of all MPs would encourage negotiation and compromise across parties and help prevent partisan amendments, it could be difficult to achieve and could prevent the adoption of amendments that are otherwise needed. The alternative of a referendum, on the other hand, is costly and may be difficult to organize in all parts of the country given the security situation.

Constitution-makers could have drawn inspiration from comparative examples by providing different amendment procedures depending on the provisions concerned. Such an arrangement may help provide stability, certainty and strong guarantees for some parts of the constitution that need to be more rigid while allowing flexibility in other areas. For example, the draft could have included provisions that could be amended by a two-thirds majority vote of all members of the National Assembly, as well as provisions that could be amended by a three-quarters majority vote of all members of the National Assembly or by a two-thirds majority vote of all members of the National Assembly and a referendum.

The 2017 draft also proposes expanding the list of circumstantial restrictions on amendments. While the 1991 Constitution prohibits constitutional amendments during periods when territorial integrity is infringed, during states of siege and during states of emergency, the 2017 draft retains these restrictions and also adds a prohibition on amendments during the exercise of the president’s crisis powers (regulated by article 73) and during war (article 192). As in Mali, given the security situation in Burkina Faso, and in particular the fact that non-state armed groups control large portions of the country, prohibiting amendments when territorial integrity is infringed could risk rendering the Constitution unamendable until the state regains control of the entire country. However, in its decision of 12 January 2024, the Constitutional Council of Burkina Faso defined an attack on territorial integrity as control of the territory by a foreign power with the aim of removing the authority of the state. The Constitutional Council thus ruled that the current security crisis does not constitute an attack on the territorial integrity of the country, as it stems from the activities of non-state armed groups rather than a foreign power (Burkina Faso 2024a). The explicit prohibition of constitutional amendments during the exercise of the president’s crisis powers is a notable improvement. It may help prevent hasty amendments during times of crisis and uncertainty and explicitly prohibits the president from resorting to exceptional crisis powers to amend the Constitution unilaterally.

Importantly, and contrary to Mali’s 2023 Constitution, the 2017 draft explicitly excludes constitutional amendment bills from the president’s unilateral power of legislative referendum (article 67). Therefore, the president would not be able to push through a popular constitutional reform while bypassing parliament. This arrangement ensures that constitutional amendments can be adopted only through the formal amendment procedure provided for in article 184.

Regarding substantive limitations on constitutional amendments, the 1991 Constitution provides that the republican nature of the state, the multiparty system and territorial integrity, as well as the two-term limit on the presidency and the duration of presidential terms (since the 2015 amendments), are unamendable. The 2017 draft Constitution retains the same list of substantive restrictions on constitutional amendments and in addition entrenches secularism (article 192).

Notably, the 2017 draft introduces mandatory prior judicial review of constitutional amendment bills on both procedural and substantive grounds by the Constitutional Court (articles 164 and 188). This mandatory review would take place after the amendment bill has been considered by the National Assembly (i.e. after the approval of the bill by a three-fifths majority of its members) and before the submission of the bill to a referendum or to the plenary of the National Assembly for approval by a four-fifths majority vote.

Turning to Chad, the 2023 Constitution provides three separate procedures—one for the adoption of an entirely new constitution (articles 278 and 279) and two for amending the existing Constitution (articles 280–83). Amendments to the existing Constitution may be initiated by the president, following a decision of the Council of Ministers, or by MPs (article 280). This means that the president cannot unilaterally initiate an amendment procedure, and any presidential initiative requires approval of the government, headed by the prime minister. As the government is no longer formally accountable to the president under the 2023 Constitution, the government could, in case of cohabitation, block the president from initiating a constitutional amendment procedure. Under the first amendment procedure, constitutional amendment bills must first be adopted by a two-thirds majority vote of all members in each of the two chambers of parliament and subsequently approved in a referendum (articles 279, para. 2; and 280, para. 1). Under the second amendment procedure—limited to so-called technical revisions—a constitutional amendment can be adopted by a three-fifths majority vote of the legislature in a joint sitting (article 281, para. 2). While this arrangement foresees two possible amendment routes depending on their subject matter—and therefore provides more flexibility, as it does not always require a referendum—it suffers from significant ambiguities. Importantly, the Constitution does not define what would constitute a technical revision. Moreover, it is unclear whether the requirement of a two-thirds majority vote of all members in each of the two chambers of parliament provided in article 279, para. 2, also applies to technical revisions. It remains uncertain whether a technical revision must first secure a two-thirds majority vote in each chamber of parliament and then a three-fifths majority vote in a joint sitting, or whether only the latter vote is required. This lack of clarity is problematic and could cause controversy.

The 2023 Constitution specifies two circumstances in which amendment procedures are prohibited (article 283): (a) when the president uses their exceptional crisis powers (regulated by article 93); and (b) when the speaker of the Senate acts as interim president in the event of a vacancy of the presidency (regulated by article 82). However, constitutional amendment procedures during other types of states of exception (e.g. a state of emergency, war) are not explicitly prohibited.

Like Mali’s 2023 Constitution, Chad’s 2023 Constitution does not explicitly exclude constitutional amendment bills from the president’s referendum power (regulated by article 88). The absence of such a prohibition raises similar concerns to those mentioned above—that is, the president of the republic could potentially push through a popular constitutional reform without approval in parliament. However, unlike the situation in Mali, the president’s power to hold a referendum on a bill on certain matters is not unilateral. The president may decide to hold such a referendum only upon the proposal of the government or upon the joint proposal of the two chambers of parliament (article 88).

Chad’s 2023 Constitution retains the same substantive limitations on constitutional amendments as the 2018 and 1996 constitutions. It provides that territorial integrity, the independence and national unity, the republican form of the state, the principle of separation of powers, secularism, fundamental rights and freedoms of citizens and political pluralism are unamendable (article 282). Notably, the provisions limiting the president to two consecutive terms are not entrenched and were actually abolished by the constitutional amendments enacted on 8 October 2025 (see Chapter 5: Presidential term limits).

On paper, the constitutional amendment procedure appears to be sufficiently stringent to prevent self-serving or partisan changes—requiring a two-thirds legislative majority and endorsement in a referendum for most amendments, and a three-fifths legislative majority for so-called technical revisions. In practice, however, the country’s history of winner-takes-all politics, which regularly gives the president’s party a supermajority in parliament, may allow transient political majorities to unilaterally amend the Constitution (see also Leubnoudji Tah 2025).

13.2. Adoption of a new constitution

Beyond constitutional amendments, Mali’s Constitution and Burkina Faso’s draft Constitution do not outline a procedure for the adoption of new constitutions, nor do they clarify whether the adoption of new constitutions must follow the procedure provided for constitutional amendments. Experience from many francophone African countries shows that incumbent leaders have resorted to the process of adopting ostensibly new constitutions—invoking the constituent power theory and disregarding the extant amendment procedure—mainly to bypass unamendable presidential term limits. In light of this experience, constitutional safeguards could be strengthened if constitutions specifically regulated how entirely new constitutions may be enacted, in addition to specifying the process for constitutional amendment.

In this regard, Chad’s 2023 Constitution outlines a procedure for adopting a new constitution (articles 278 and 279). The adoption of a new constitution can be proposed by the president after consultation with parliament and must be approved by a two-thirds majority vote of all members of the legislature in a joint sitting, followed by approval in a referendum. Regulating within the constitution itself the procedure through which it may be replaced by an entirely new constitution can help enhance legal certainty and preserve constitutional continuity.

The post-coup constitutions of Mali and Chad and the 2017 draft Constitution of Burkina Faso all aim to restore constitutional order. However, each takes a different approach to balancing the need to enable legitimate and effective government action to respond to the security crisis and economic challenges the region faces with the need to provide sufficient safeguards to prevent a slide into authoritarian rule.

Mali has an ‘emergency’ constitution, which establishes a hyper-presidential regime and a centralized state structure that prioritizes presidential leadership and rapid, short-term decision making over inclusive deliberation, political compromise and checks and balances (see Abebe 2024). This new political regime concentrates tremendous, unchecked powers in the president and may intensify winner-takes-all politics and pave the way for authoritarian drift. The Constitution also fails to reflect the commitments to extensive and asymmetric decentralization made in the 2015 Algiers Peace Agreement, which may lead to further tensions and destabilization.

Chad has established a premier-presidential semi-presidential system—uncommon in francophone Africa—which introduces more checks and balances on and within the executive compared with previous constitutions and provides a basis for a relatively decentralized system of governance. In principle, these features could enhance inclusive deliberation, encourage political compromise and more broadly reduce the historical dominance of the presidency. However, the country’s history of winner-takes-all politics, which regularly gives the president’s party a supermajority in parliament, combined with the Constitution’s overall minimalism, which defers key matters to legislation, allow transient political majorities to unilaterally impose or alter the rules governing the framework for democratic and accountable government. The constitutional amendments enacted on 8 October 2025 as a technical revision—which abolished presidential terms limit, extended presidential terms to seven years and enabled the president to lead the ruling party—illustrate this risk.

Although elaborated in a different context, Burkina Faso’s 2017 draft Constitution may serve as a reference for future deliberations and negotiations in the country to restore constitutional order. Overall, the draft proposes a more balanced relationship between the executive and the legislature, strengthens the independence of the judiciary and enhances guarantees for political alternation. Nevertheless, it preserves a president-parliamentary form of semi-presidentialism in which the president of the republic retains important powers—including the authority to remove the prime minister and government at will—enabling presidential dominance over the executive branch.

Based on a comparison of the three constitutional texts, several concluding remarks can be made:

1. Compared with anglophone and lusophone constitutions in Africa, these three constitutional texts provide little detail and leave important elements to be defined through legislation. They defer to organic or ordinary legislation the definition of important procedures and elements that are critical for the functioning of a constitutional democracy. Although the three constitutions vary in their level of detail, this approach reflects the minimalism that characterizes many francophone African constitutions. In contexts where a single party often dominates the legislature—as has mostly been the case in these three countries—this could allow transient majorities to unilaterally decide issues that are critical for the functioning of constitutional democracy.
2. The three constitutional texts concentrate important powers in a directly elected president, although with important variations. This concentration of powers, including broad emergency powers, in the presidency appears primarily intended to create government systems capable of responding more effectively to the severe security crisis and structural economic challenges affecting these countries and the region. However, the concentration of such powers in the presidency may increase the stakes of presidential elections and, in the absence of commensurate oversight mechanisms, risks enabling the president to govern unilaterally on a wide range of policy matters.
3. These texts retain many similarities with France’s Constitution, some of which may indicate insufficient adaptation to the local context. Notably, they replicate several mechanisms and powerful government instruments of so-called rationalized parliamentarism. These instruments were introduced in France’s 1958 Constitution to address the dominance of the legislature and the ensuing governmental instability that characterized the Third and Fourth Republics. While some of these constitutional devices—provided that adequate safeguards are in place—may be relevant in Burkina Faso, Chad and Mali in addressing severe security and economic crises, other instruments and features may indicate a lack of adaptation to context and a reliance on certain imported constitutional arrangements. Notably, presidents in these countries have historically enjoyed broad support in the legislature. Accordingly, replicating such instruments in contexts where the legislature and fourth-branch institutions have been relatively weak may formalize and further strengthen the dominance of the executive branch in the political system. Similarly, retaining such instruments in systems where the executive is not accountable to the legislature (as in Mali)—rather than in a premier-presidential system—risks granting extremely broad powers to the president.
4. While these countries are heterogeneous in terms of ethnicity, geography and history, their constitutional texts establish unitary state structures with weak guarantees for decentralization and no asymmetric arrangements. While all three constitutions recognize the existence of decentralized entities, the responsibilities and financing of such entities are deferred to legislation. As a result, the scope of decentralization will depend on whether existing laws on decentralization are amended, the substance of these amendments and their implementation. Given that these constitutional texts do not establish timelines, implementation processes or monitoring bodies, the realization of decentralization will depend on the will of the political majority. Furthermore, none of these texts provides for the possibility of (but do not specifically preclude) asymmetric arrangements, whereby certain parts of the country could be granted more autonomy on certain policy issues than others. As a result, these constitutional arrangements may be contested, bringing about renewed conflict between central governments and certain non-state armed groups in Chad and Mali, where demands for extensive (asymmetric) decentralized arrangements persist. Burkina Faso may also face increasing territorially based political claims that could have an impact on future constitutional debates.