The word ‘government’ has long been used to describe the way collective decisions are taken and implemented in a particular community by specially designated people.1 Be they a single individual, a small group, or the entire community, the decision-makers are called ‘rulers’, while the members of the community are ‘the governed’. People are governed in the sense that their conduct is directed (or influenced) by rules, institutions, and sanctions. The considerable power of the state is exercised through this process in modern democracies.
We are ‘governed’ by many people and institutions, at different levels. ‘Systems of government’ are the different arrangements according to which these institutions interact and co-operate in making and enforcing law. They are influenced by such things as how democratic a system is; political parties; the electoral system; the distribution of wealth; the degree of decentralisation of power in the state; and more. Constitutional law has a rather narrow approach to government, taking for granted that what matters is the decision-making process by state institutions, especially legislatures, executives, and courts. Since at least the late eighteenth century, the state has governed in ways unknown in past political forms. In particular, it legislates abundantly, and it relies on a specialized bureaucracy.
This chapter provides an introductory description of the main systems of government in democratic countries that seek to promote the rule of law.2
There is much more to government than law but law permeates the modern state. In modern societies, government is readily equated with the process of law-making (with many different kinds of laws, including statutes, regulations, and international treaties), and a developed administrative structure that implements laws and makes sure that ‘private’ entities act in accordance with them. As the application of laws is not always straightforward—laws may conflict, be ambiguous, or be wrongly applied—institutions, usually the courts, are needed to oversee their application.
The emphasis on law-making, implementation, and enforcement has led lawyers to tend to equate the state with law-making and implementation. The perceived benefit is that issues of power and politics can be set aside, if not altogether banished, from the business of describing the state in legal terms. On this view, ‘government’ is the business of political scientists, dealing with ‘positive’ facts, while ‘positive law’ is the proper study of lawyers.
The success of this scientific agenda is staggering, and so are its shortcomings. Politics is pervasive. It is about collective decision-making and how it takes place: preferences and demands of individuals and groups; the pressure they put on decision-makers; appointment and dismissal of officials at different levels; policy-making and the bargaining it entails; and a great deal more. Constitutions provide the framework within which this political activity is organized through expressing values, setting out decision-making procedures, and creating the institutions that give shape to political action and that seek to constrain its inherent propensity to conflict.
Constitutions do not always ‘work’, of course. While France’s record of recovering from World War II in the 1945–1958 era was impressive in economic and social terms, its 1946 constitution (the ‘Fourth Republic’) is generally seen as having significantly impeded its development, generating political instability and preventing the finding of a peaceful solution in the decolonisation conflicts, notably in Algeria. France, it was perceived at the time, was no longer ‘governed’. In the summer of 2011, the US showed a remarkable example of constitutional inadequacy with a Democratic President and a Republican majority in one of the Houses of Congress nearly unable to reach a deal on the issue of debt, the public deficit, and reining in public spending.
Moreover, constitutional arrangements that ‘work’ in one context may fail or operate very differently in another. The parliamentary system, which has endured in the United Kingdom, survived only a few years after independence in most of the UK’s former African colonies. The reasons for this failure are complex and diverse but in part it can be attributed to the very different political context of those young African states. In other cases, small variations in the political environment or rules mean that systems that are formally similar, with similar institutions and rules, work quite differently. Again, parliamentary systems provide examples: compare the relationship of the executive to Parliament in the UK—where vigorous Question Time sessions require the Prime Minister to defend government policy and enhance accountability and open government—with that in South Africa where Question Time is organised to benefit the majority party and members of the executive answer questions reluctantly. Consider also the way in which Germany has managed to govern with stable coalitions for decades while Belgium has repeatedly struggled to form governments.
These examples show how constitutional processes and institutions influence politics and vice versa. Ignoring the relationship between them can lead to misunderstandings of both law and politics.
The word ‘governance’ has recently been revived to capture an understanding of government that is broader than formal rules, and to cover ‘a variety of recent changes in governmental practice arising from globalizing social, economic and technological developments’ (Loughlin).
To a large extent, modern governance is a universal phenomenon and states with different systems (be they parliamentary, presidential, or whatever else) are faced with similar problems (environment, economic policy-making, multiculturalism, ‘law and order’ concerns), under similar constraints (scarcity of state resources, rising demands for a better protection of individual rights), and have found solutions of a broadly similar nature, often through processes of cross-fertilisation. Moreover, ‘governance’ is now often of a transnational nature: regional courts may have authority over a number of states (such as the European Court of Justice and the East African Court of Justice); and institutions such as the UN World Trade Organization, the Basel Banking Committee mechanism, North American Free Trade Agreement (NAFTA), the Southern African Development Community (SADC), and, at a more integrated level, the European Union are examples of transnational governance. A country’s international and regional economic, security, and other obligations constrain its constitutional choices and its law-making and implementation just as international human rights obligations do—a point dramatically illustrated by the replacement of Italy and Greece’s governments under extreme pressure from the EU by ‘governments of experts’ in the 2011 financial crisis.
1. A poorly charted territory called ‘comparative government’ lies at the crossroads of political science and law. It consists in an attempt to understand scientifically the way in which countries are governed through constitutions and institutionalised politics. The vocabulary of comparative government has a reassuring quality: we talk of ‘systems’, ‘structures’, ‘régimes’. This lends a veneer of rationality to our approach but it can be quickly displaced by disenchantment. As a result, the very word ‘system’, which appears in the title of this chapter, has aroused increasing scepticism from the ranks of political scientists, some of whom have tried to replace it with ‘situations’, ‘fields’, or even ‘games’ to describe patterns of action in which actors and even rules are constantly changing.
The theory of comparative government has struggled to reduce the immense variety of existing systems of government to a small number of identifiable categories but it has proved extremely difficult to categorise particular régimes. Even more disconcertingly, it produces mere categories, rather than fully fledged explanations of the way in which each régime actually works. Nonetheless, although the explanatory power of the intellectual categories is limited, they have proved rather robust and have not been superseded. Despite its shortcomings, the standard categorisation of comparative government has the advantage of being standard: people speak more or less the same language.
One concluding caveat about the categories of governmental systems concerns their scope. They fit Western liberal democracies fairly well. But, first, the standard categorisation described here simply does not apply to many important non-Western political régimes, China being the main example. Secondly, the widespread use of Western constitutional categories or principles in non-Western political cultures does not mean that these government systems can always be understood in Western terms. The parliamentary system of, say, India, generates a kind of party politics and institutional configurations remarkably similar to a Western ‘Westminster-style’ régime. On the other hand, although Thailand has a parliamentary government and the Republic of the Philippines an American-style presidential system, the culture, politics, and values in these countries differ vastly from those of their Western constitutional role models, and the functioning of their systems has to be described quite differently.
2. The standard categorisation distinguishes between two main forms of government: parliamentary and presidential. We follow suit, while acknowledging that other, less easily identifiable forms, such as the Swiss system or the institutions of the European Union, need to be identified.
The categorization ‘parliamentary/presidential’ is structured around the concept of separation of powers: it aims to describe the ways in which separation of powers is positively transcribed into a set of working institutions. Separation of powers should not be misunderstood as a single, straightforward, constitutional blueprint. It is first and foremost a principle of political morality, expressed as a core requirement to secure political liberty. As such, its only demand is that all the powers of the state should not be devolved to a single institution. All Western democracies achieve this. The particular way in which different countries arrange the institutions of the state to distribute competences is what differentiates the several ‘systems’ of government. There is only a limited number of models available and they are mostly the products of history and, in particular, of struggles to contain power. The mother of all government systems, the British parliamentary government, has arisen from practice— and an impressive amount of background political thinking (from Locke to, say, Walter Bagehot)—rather than from deliberate, voluntary constitution-making. The ‘prototype’ presidential system, the US system, was developed in a more formal process of constitution-making but it too was very much a product of its political and historical context as was, most famously, the French system in which executive power is shared between a directly elected head of state and a Prime Minister who emerges from Parliament.
Separation of powers is most closely associated with the US where it dominates descriptions of the political arrangements and particularly the sharing of power between Congress, the President, and the Courts derived from the 1787 Constitution. While generally accepted, this nomenclature has obvious flaws: ‘powers’ in the sense of functions (legislative, executive, judicial) are not entirely separated in the several institutions of government. Executives routinely make laws (often called regulations), fulfil quasi-judicial functions in tribunals, and may have the power to veto laws; legislatures may appoint members of courts or the executive; courts, and particularly courts with the power to review laws for constitutionality, may make law.
The inadequacy of the description ‘separation of powers’ is even more evident in parliamentary régimes that follow, more or less closely, the British (Westminster) model. There, members of the executive are often members of Parliament and, until 2005, the House of Lords was at the same time a part of the legislature and the highest court in the land. Nonetheless, even in parliamentary systems, in practice functions are separated to some extent. For example, the bureaucracy is separate from the legislature, and courts are not subject to executive or parliamentary control.
1. ‘Parliamentary government’ is the most widespread form of democratic government. This quantitative success may be due to the fact that the label ‘parliamentary’ hardly denotes a uniform structure of government. At best, it identifies a basic criterion that many systems satisfy. The criterion can be stated thus: a collective executive is accountable to an elected legislative chamber. This accountability is expressed through expressions of ‘confidence’ by the chamber. Should this confidence be withdrawn (by what is usually called a ‘vote of no confidence’), a (written or unwritten) constitutional rule prescribes that the executive (i.e. Prime Minister and major cabinet ministers) must resign. Nothing else is needed to join the club of parliamentary governments, not even some other common features such as a separate head of state (sometimes a monarch) and head of government, or the executive’s power to dissolve Parliament. This apparently straightforward criterion expresses a more subtle underlying constitutional reality. Parliamentary systems are normally found in contexts where the best solution to legitimising executives has proved to be a functional link between the executive and the elected chamber(s) of Parliament. Votes of no confidence have always been rare.
In practice in a parliamentary system, after each parliamentary election the largest party in the legislature, or largest coalition of parties, forms a government (the executive). Usually, the leader of this group becomes head of the executive (and is usually called Prime Minister) and chooses a cabinet with which to govern collectively. Thus, the executive generally commands the support of Parliament and ‘confidence’ is mostly a positive political relationship. It means that a parliamentary majority expresses its approval of the cabinet’s political intentions (the ‘programme’ of art 49 of the French Constitution of 1958, or the ‘mandate’ of British democratic theory) and is willing to support its legislative implementation. Co-operation between Parliament and the executive maintains the executive in office. Conversely, rejecting the smallest piece of legislation that has been declared to be a confidence issue (a single item in an Appropriation Act, an unimportant measure that has nothing to do with the hot political issues of the day) can convey the message that the cabinet has been defeated and that its political existence is in jeopardy.
The requirement in parliamentary systems that the executive must retain the confidence of Parliament does not mean that Parliament is politically dominant in the state. Rather, in most cases, the executive is in charge of ‘developing and implementing national policy’ (section 85(2)(b) South African Constitution of 1996) or ‘determining the nation’s policies’ (art 20, French Constitution of 1958). Despite many claims to the contrary (including those expressed in some constitutions), the function of the deliberative Houses of Parliament is not to devise policies or make ultimate political decisions but rather (to borrow from a medieval phrase still to be found in every British Act of Parliament) to ‘advise and consent’ on policies and statutes devised by better equipped executives. Executives make decisions that are scrutinised and ultimately approved in legally authoritative form by democratic, deliberative chambers of Parliament. The executive can act in this way only because it enjoys the ‘confidence’ of Parliament. In many cases, this has involved the decline of the power of heads of state. Being left outside the magic circle of legitimacy, the monarch or the elected President is unable to wield effective power or engage in political action. He/she retains formal powers of appointment and dismissal, and perhaps some leeway on the appointment and dismissal of the Prime Minister or the decision to dissolve Parliament. But generally in modern ‘monistic’ parliamentary government, the head of state is deprived of an active part in political decision-making. As the 1975 dismissal of the Whitlam government by the Australian Governor-General demonstrates, political decision-making by an unelected head of state in a parliamentary system is deeply problematic.
2. In ‘continental’ Europe, the British constitutional model has long attracted praise for its conduciveness to political stability. Starting with the Belgian Constitution of 1831, written constitutions have tried to reproduce the British model’s main features. This increasingly took the form of an attempt to formulate the unwritten ‘conventions’ of the British Constitution into entrenched constitutional clauses and the development of new constitutional rules to respond to differing circumstances. This process was brought to another level by the German and Austrian Constitutions of 1919–1920 and the process known as the ‘rationalization’ of parliamentary government. Specific rules and detailed procedures purported to bring more clarity and greater constraining power to the core elements of the parliamentary model. At least three aspects were dealt with.
3. There are many other variations on the traditional Westminster-style parliamentary regime. In the UK and other Commonwealth countries with parliamentary systems, ministers are usually appointed from Parliament, or, if they are drawn from outside Parliament, an arrangement is made to secure them parliamentary seats as soon as possible. But, in a number of European parliamentary systems, for instance, ministers need not be members of Parliament (e.g. Denmark, Germany, Finland, Italy) or may not be (e.g. Luxembourg, Norway, the Netherlands) (Andeweg & Nijzink, 160). In South Africa, the National Assembly elects the President, who is both head of state and head of the executive, from amongst its members but he or she must leave Parliament once elected. None of these variations affects the basic feature of parliamentary systems: the accountability of a collective executive to the legislature.
The most profound variation on the traditional parliamentary model is found in the French system, which has been followed in Francophone African countries and forms the basis of many of the new systems in Eastern Europe. While often ranked among ‘presidential’ forms of government (and variously referred to as hybrid, premier-presidential, or semi-presidential), these systems are often specific instances of parliamentary government in which the head of state is elected by universal suffrage and granted important constitutional prerogatives. At the same time, the cabinet is accountable to Parliament and the criterion of parliamentary government is therefore satisfied. In France, where the President is elected by universal suffrage, there can be two different configurations. If the parliamentary majority is the same as that which has elected the President, it becomes ‘the President’s majority’. It is elected on his/her own political agenda and supports the cabinets that the President has appointed and can freely dismiss. This is why France’s system is sometimes called ‘semi-parliamentary’ or ‘semi-presidential’. Yet there can be a second political configuration in which the President and the parliamentary majority do not belong to the same political side. In these (rarer) phases of ‘cohabitation’, the President’s formal constitutional powers are intact, but his/her political authority is diminished. The French system then functions as a classic ‘monistic’ parliamentary government, in which the Prime Minister sets the political agenda and is accountable only to the majority in Parliament.
The tendency to call these régimes ‘presidential’ stems from the fact that the actual seat of power appears to be the head of state. He or she appoints and dismisses ministers according to his/her own political will and does not have to bow to the will of Parliament. This can lead to a form of political authoritarianism (Russia) or to a régime that works very differently from the Westminster model (France’s Fifth Republic, Finland, Austria). Yet calling these régimes presidential creates a misleading analogy with the American model. Rather, they are parliamentary regimes that co-exist with a strong, and sometimes overbearing, President elected by universal suffrage.
The American system is the classic presidential system. In the United States, the label ‘presidential’ should not be construed as implying that the constitution puts the President, the constitution’s executive authority, to the forefront at the expense of Congress (the legislative branch). When Woodrow Wilson, then a professor of political science, wrote an authoritative account of the régime in 1885, he called it ‘congressional government’ in order to emphasize (and deplore) Congressional prominence. This prominence has later waned, but the US model of government still relies on a balanced relationship between both branches.
The core features of the presidential model, US style, can be expressed in the language of separation of powers. The constitution creates three ‘powers’ or three branches of government: the legislative function is devolved to Congress; the executive power to the President (and that institution alone); and the judicial power is ‘vested’ in the Supreme Court, and in ‘inferior Courts’ (s 1, art III). This vesting of powers, which takes place in the first three Articles of the Constitution and dominates understanding of it, is not absolute. Functions overlap. In part, this is because it is impossible to define functions absolutely—the interpretation of law necessarily involves some law-making; executing legislative programmes necessarily involves some law-making and adjudication of interests. In part, however, functions are shared to prevent any branch of government from monopolising power: it is, famously, a system of ‘checks and balances’. For example, the grant of legislative power to Congress is mitigated by the presidential power to veto laws, and the executive power of the President is tempered by the involvement of the Senate in certain appointments to the executive and in approving international agreements. Moreover, even though the requirements of justice are such that courts are expected to enjoy greater independence than the two other branches of government, there are some limits on the ‘separation’ of the judicial branch. Most notably, judges are chosen by the other two branches.
The three branches of government are deprived of the means of institutional pressure that characterise parliamentary government. The separately elected President and Congress have independent legitimacy. Accordingly, the President cannot dissolve either House of Congress and, in turn, neither the House of Representatives nor the Senate can force the President to resign by way of a vote of no confidence. The President can be removed from office only through a deliberately complex process of ‘impeachment’ by Congress in which it is established that he or she has (or has not) been ‘guilty’ of ‘Treason, Bribery, or other high Crimes and Misdemeanors’. Certain aspects of this procedure replicate some features of criminal justice and, unlike the Queen of England, who enjoys full constitutional immunity, the US President is both subject to criminal liability and can be removed from office through the impeachment procedure. As far as the President is concerned, impeachment is better approached as a very specific instance of political accountability. The President cannot be removed merely because a majority (or even a super-majority) in Congress disapprove of his or her policies. Political disagreements between Congress and the President must be resolved by negotiation or, at worst, at the next scheduled election. Unlike votes of no confidence in a parliamentary government, impeachment does not sanction poor (but legal) political choices (‘maladministration’). Rather, it aims at protecting the constitution and the system of government against egregious behaviour that is likely to imperil the whole constitutional edifice and is perceived as a ‘breach of trust’.
In the past, about 30 countries around the world have followed the United States and adopted ‘presidential democracy’, many of them in Latin America. However, many deviate from the US model in significant ways. In many South American and African countries, the powers of the President are substantially greater than those of the US President and checks and balances weaker. For example, many Presidents have substantial law-making powers. When such power is granted by a law passed by the legislature, the legislative check on executive power remains in place, at least formally, as the legislature can revoke the grant of power (see Chapter 10). However, presidential constitutions may grant the President the power to make laws (often referred to as decrees) and sometimes these powers are broad. Emergency powers also often give Presidents the power to suspend rights. Sometimes legislative oversight of this is required (Namibia art 26), but not universally (see Chapter 7). Similarly, few presidential systems follow the US example and require legislative approval for cabinet appointments. Often too, the judiciary is weakened because the President controls the appointment of judges. And, some Presidents have the power to dissolve the legislature either under circumscribed conditions or more generally.
These and other variations in the design of presidential systems may affect the way in which the doctrine of separation of powers operates. The terms ‘hyper-presidentialism’ and ‘imperial presidency’ have been coined to describe cases in which few democratic checks operate, and many scholars argue that these are not examples of a classic presidential system at all. These examples, together with the poor record of presidential regimes in South America, have provoked a lively debate about whether presidential or parliamentary systems are better.
The standard analysis has to make some room for long-standing exceptions to the parliamentary/presidential divide. An obvious example is Switzerland. In Switzerland, the seven-member Federal Council, the federal executive, is elected by the Federal Assembly, a joint assembly of the two houses of the Federal Parliament. Annually, the Assembly must also elect a Federal President from the members of the Council, to serve for one year only. Like executives in parliamentary systems, the Council operates as a collective (it is subject to the principle of collegiality: art 177). But, although the Assembly has ‘supervisory control’ over the Federal Council (art 169), the Council is not subject to the confidence of the Assembly— there is no procedure for the dismissal of the government by a vote of no confidence. Thus the Swiss system has been described as ‘non-parliamentary’ (because the executive is not dependent on legislative confidence) and ‘non-presidential’ (because there is no head of state directly elected by the people) (Linder & Steffan, 298).3
The Swiss system also provides an excellent example of how poorly formal descriptions of systems of government explain the processes of government. For more than 40 years, the Federal Council has been composed of a coalition of four political parties, holding seats in the ratio of 2:2:2:1, which reflects their representation in the elected legislature. Thus, a hallmark of the Swiss system is its consensual nature. As Linder and Steffan comment, ‘political decisions are not found by majority decisions but through negotiations and compromise among the important political forces’ (Linder & Steffan, 291–92).
Parliamentary government and presidential government are faced with major challenges. Most of these challenges are common to all systems of government: instability; insufficient accountability; and ‘crisis in governance’ because, in the words of Davina Cooper, ‘welfare provisions can no longer be sustained, political authority and legitimacy are in decline, [and] the state as we know it is in jeopardy’ (Cooper). Yet experience has shown that each main model is prone to specific pathologies.
Parliamentary government is often perceived as being beset by political instability and powerlessness: Belgium has not had a cabinet enjoying a majority in Parliament from June 2010 to December 2011; during the Fourth Republic (1946–1958), France had 26 ministries and more than a year of interim administrations. There have been 61 administrations in Italy since 1945. Other parliamentary systems seem to have escaped this fate: since 1949, Germany has succeeded in achieving the aim of having one administration per legislative term.
The presidential system, on the other hand, is associated with problems of immobilisation. In the United States, there are ever-growing signs that political polarisation and failure to reform the Constitution have brought about a state of recurring tension between President and Congress resulting in ‘endless backbiting, mutual recrimination, partisan deadlock … the House will harass the executive and the President will engage in unilateral action whenever he can get away with it’ (Ackerman).
The case has also been made that presidentialism has sometimes been detrimental to the countries to which it has been exported. In Latin America, presidentialism has coincided with a record level of regime instability, and a propensity to give way to military dictatorships. Juan Linz has been the outspoken proponent of the view that this is due to some inherent features of presidentialism as a form of government. In the ‘Linzian’ account, presidentialism is based on ‘mutual independence’ between the executive and the legislature. This generates irresolvable conflicts between a President elected by the electorate and who is also the head of government on one side, and on the other side an elected assembly that cannot either be dissolved by the President or force him or her to resign. Interdepartmental conflicts are more likely than co-operation (Linz).
Other scholars are less prone to hold the institutional framework of presidential government responsible for the failure of South American presidential régimes. As Cheibub demonstrates, there is no evidence that presidential systems are less able to deal with deadlock than parliamentary ones. Instead, he argues, ‘the reason for the instability of presidential democracies (…) lies in the fact that presidential institutions tend to exist in countries that also more likely to suffer from dictatorships led by the military’ (Cheibub, 3). In this context, he concludes, democratic failure is due to the ‘military–presidential nexus’, namely that ‘democracies following military dictatorships are more likely to become a dictatorship and that presidential democracies are more likely to follow military dictatorships’ (Cheibub, 141).
Many of the political reforms that try to address the pathologies are not specific to the particular model of government. For instance, as a solution to issues of ‘good governance’, accountability and a crisis in ‘state delivery’, significant state competences have been devolved to independent agencies that are not subject to the authority of parliaments or executives and are not directly accountable to them (‘quangos’ in the UK, ‘autorités administratives indépendantes’ in France). In systems with newer constitutions, the inability of classic systems to control abuses of power has led to a proliferation of constitutionally entrenched ‘independent institutions’, which are sometimes characterised as additional branches of government: Central Banks; Electoral Commissions; Human Rights Commissions; Auditor Generals/Cours de Audit. (The 2009 Kenyan Constitution establishes 12 such independent institutions.) (See Chapter 10.)
Inside each type of government, critiques often focus on the shortcomings of the classic separation of powers model. Experts are often prone to think that the grass is greener on the other side of the separation of powers fence. In Europe, initial fascination with the Westminster model was replaced with attempts to repatriate some features of the US presidential system. In France, this used to mean increasing the President’s domination and moving away from parliamentarism. Yet the 2008 overhaul of the Constitution took the reverse approach: it aimed at creating a greater balance between the executive on one side, courts and parliament on the other.
On the other side of the fence, in the face of what was perceived as a failure of presidential government outside US borders, notably in South America, a plea has been entered in favour of ‘constrained parliamentarism’, a system that would resort to the core mechanisms of parliamentary government while constraining law-making through a greater involvement of the people and increased judicial review of higher ‘substantive political principles’ (Ackerman, 663 ff).
There is much debate about what system of government is ‘better’. The question of what system of government to adopt is perhaps the most difficult question for constitutional designers in a new state (like South Sudan) or in the many countries emerging from authoritarian rule or contemplating substantial constitutional change (such as Eastern European countries after 1989; South Africa in 1994; Pakistan in 2010; Tunisia in 2011; and Kenya between 2001 and 2010).
In this context, abstract discussions about the vices and virtues of government systems can be misleading. A better understanding of just how constitutional engineering can help fix political problems may depend on a closer look at national contexts and their interaction with the details of the design of each system. This is a lesson lawyers might draw from the active debate triggered in the field of political science by the work of Juan Linz. Linz argues that his analysis of presidential systems demonstrates their particular unsuitability for societies deeply divided by ethnic or other differences.
Current discussion of what system is best in ethnically divided societies is framed by a famous debate between Arend Lijphart and Donald Horowitz. Lijphart’s position is that ‘majority rule in plural societies spells majority dictatorship and civil strife’ (Lijphart 1985, 19)—in other words, neither a ‘pure’ parliamentary or presidential model is appropriate—and that in such circumstances the answer is ‘consociational democracy’, a system in which as many people as possible are involved in decision-making. For Lijphart, ‘the consensus model tries to share, disperse and limit power in a variety of ways’ (Lijphart 2008). A consociational system has many components—autonomy for different groups (for instance through federal arrangements), power sharing, and the careful distribution of civil service positions, among other things. In the context of systems of government, Lijphart prescribes ‘executive power sharing in broad multi-party coalitions’ and ‘executive–legislative balance of power’. Here, what is important is that all significant political groupings should have a say in decision-making. The system in Switzerland and Kenya’s ‘grand coalition’ government established in 2008 after ethnically based violence tore the country apart are examples of the type of arrangements Lijphart favours. The Kenyan system saw a President and Prime Minister from opposing sides share power with a cabinet composed of an equal number of ministers from each side. A similar system in South Africa under the 1993 ‘Interim’ Constitution provided for the appointment of Executive Deputy Presidents from any party with more than 20 percent of the seats in Parliament and required the President to consult them on most executive decisions.
Horowitz thinks consociational systems of government proposed by Lijphart will not work when they are most needed, in part because they rely on greater idealism from the leaders of ethnic groups than is likely to exist. He proposes an alternative, dubbed ‘centripetal’ because, in his words, ‘its principal tool is not a regime of ethnic guarantees but the provision of incentives, usually elective incentives, that accord an advantage to ethnically based parties that are willing to appeal … to voters other than their own’ (Horowitz 2008; see also Horowitz 2002). The goal is to develop institutions that will encourage accommodation amongst groups before elections rather than to secure the representation of groups. Carefully designed electoral systems may contribute to achieving this goal. For Horowitz, the choice of system of government is less important as a tool in managing ethnic division but not irrelevant. In particular, he disagrees with both Lijphart and Linz in his preference for a—specially designed—presidential system. The model he points to is Nigeria where the President must secure at least 25 percent of the vote in at least two-thirds of all the states (Nigeria art 134), a requirement that is intended to require candidates to build multi-ethnic alliances and to lead to ‘pan-ethnic’ presidents.
A nation’s system of government is the backbone of its constitutional system—if the system of government is ineffective, free and fair elections are unlikely, rights will not be implemented or protected by courts, and national resources will not be distributed to the people. Yet we lack systematic knowledge of these systems’ consequences. In particular, relatively small changes in design (the scheduling of legislative and presidential elections, for example), in the political context (such as the number and stability of political parties) or in the broader social and economic context (the economic dependency of constituents on public representatives or the importance of consensus in a society perhaps) may have unpredictable consequences for the way in which a system actually operates. Moreover, the ‘choice’ of system, in those newer cases where it has been a deliberate choice, is frequently driven by history and the short-term interests of those making the choice and their understanding of the options available. Nonetheless, as we pay more attention to systems of government, their relationship to constitutionalism and the rule of law, and their design and implementation, we may come to understand better their strengths and weaknesses. The theory of governmental systems does not provide ready-made constitutional blueprints or panaceas. But it is a necessary appendage to the science of constitutions and (at times) a helpful guide for constitution drafters.
1 Sometimes terminological confusion arises because scholars of the British system, and more generally parliamentary systems, often use the word ‘government’ to refer only to the executive government. We use the term to encompass all ‘branches’ of government comprehensively.
2 This chapter does not discuss non-democratic systems in which power is concentrated in the hands of a few.
3 Linder & Steffan, 298 citing Riklin, A & Ochsner, A, ‘Parliament’, Handbuch Politisches System der Schweiz, eds Klöti U, Knoepfel P, Kriesi H, Linder W & Papadopoulos Y (Zurich: NZZ, 2004) p 210–11.
Ackerman, B (2000). ‘The New Separation of Powers,’ Harvard Law Review 113, no. 3 (1 January): 646.
Andeweg, RB & Nijzink, L (1995). ‘Beyond the Two-Body Image: Relations between Ministers and MPs’, in Herbert Doring (ed), Parliaments and Majority Rule in Western Europe, 152.
Cheibub, JA (2006). Presidentialism, Parliamentarism, and Democracy (Cambridge University Press).
Cooper, D (1998). Governing out of Order: Space, Law and the Politics of Belonging (Rivers Oram, London and New York University Press, New York, USA).
Horowitz, DL (2002). ‘Constitutional design: proposals versus processes’, in Reynolds, A (ed), The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy (Oxford: Oxford University Press).
—— (2008). Conciliatory Institutions and Constitutional Processes in Post-conflict States, 49 Wm & Mary L Rev 1213.
Linder, L & Steffan, S (2006). ‘Switzerland’ in Legislative, Executive and Judicial Governance in Federal Countries, Saunders, C & Le Roy, K (eds) (McGill-Queens University Press) p 298.
Lijphart, A (1985). Power-Sharing in South Africa (Univ California Press, Berkeley).
—— (2008). Thinking About Democracy: Power Sharing and Majority Rule in Theory and Practice, New edn (Routledge).
Linz, JJ (1990). ‘The Perils of Presidentialism’, Journal of Democracy 1, no. 1: 51 – 69.
Loughlin, M (2009). ‘In Defence of Staatslehre’, Der Staat (Berlin: Duncker und Humblot) 48 (1), 1 – 28.