Introduction

Semi-presidential countries have burgeoned recently in the third wave of democratization. In Europe, semi-presidentialism is now the most prevalent system of government (Amorim Neto and Strøm, 2006, 623). In the global context, there are now more than 55 countries operating semi-presidentialism (Elgie, 2007, 9). In terms of its popularity alone, semi-presidentialism merits the attention of scholars of comparative constitutional engineering. In this article, semi-presidentialism is defined as the situation where a country has a popularly elected fix-term president and a prime minister and cabinet who are responsible to the legislature (Elgie, 1999). However, the category of semi-presidentialism contains a widely diversified set of countries. On the one hand, there are semi-presidential countries with figurehead presidents and strong prime ministers, such as Austria and Ireland. On the other hand, there are semi-presidential countries with strong presidents, such as Russia and Ukraine. There are vast differences between the countries that are currently referred to as semi-presidential.

If we wish to gauge the institutional effect of semi-presidentialism on the prospects for democratic consolidation,¹ we need to analyze the effect of various sub-types of semi-presidentialism rather than the effect of semi-presidentialism as a whole. Various scholars use different variables to identify semi-presidential sub-types. This article proposes that semi-presidentialism can be classified into five sub-types: (1) president-dominant semi-presidentialism, (2) premier-dominant semi-presidentialism, (3) parliament-dominant semi-presidentialism, (4) cohabitation semi-presidentialism, and (5) balance-of-powers semi-presidentialism. How this framework differs from the previous frameworks is that it emphasizes the constitutional powers of parliament in tandem with two more standard variables, namely, the powers of the president and the party situation in parliament. On the basis of this framework, it suggests that when two opposing branches (president and parliament) achieve parity, they are more likely to contend for the power to pass legislation. This means that balance-of-powers semi-presidentialism is much more likely to generate political conflicts and political deadlock than any other sub-type. The severity of these conflicts can even lead to the point of democratic breakdown. Political deadlock here refers to the situation when the president and parliament, with both having nearly equal powers, contend for the authority of legislation.² The rest of this paper proceeds as follows: first, the definition of semi-presidentialism is reviewed, then the different sub-types of semi-presidentialism are constructed, and the occurrence of political deadlock under balance of powers semi-presidentialism is examined.

The Definition of Semi-presidentialism

Semi-presidential countries appear to have two completely different scenarios in terms of political performance. For example, a study from France indicated that semi-presidentialism is much better than the previous parliamentary system, and has consolidated the democratic transitions in both Portugal and Poland (Pasquino, 1997, 136, 2007). Following the same line, another study has noted that some semi-presidential countries in Central and Eastern Europe³ have not become less democratic in the 5 years since they adopted their constitutions (Metcalf, 2000). Alternatively, some scholars argue that semi-presidentialism is more risk prone than the modern parliamentarism that has evolved in European countries, with the exception of France, since the end of World War II (Stepan and Suleiman, 1995). They point out the ease with which a populist, anti-party candidate, may emerge and the difficulty in removing an incompetent president serving for a fixed term. In short, the debate regarding the effect of semi-presidentialism is still bifurcated. However, this debate assumes semi-presidentialism is a homogenous category of countries. In fact, there are vast differences in the countries that are categorized as semi-presidential. Methodologically, following Elgie (2007, 10), we argue that it is necessary to discuss the relative merits of different sub-types of semi-presidentialism and take these sub-types as our independent variables. In doing so, we believe that it can clarify how the nuances of different sub-types affect political performance.

The definition of semi-presidentialism⁴ is fiercely disputed in the field of political research. Maurice Duverger (1980) was the first political scientist to systematically discuss the concept. Duverger established three criteria for semi-presidentialism: the president is elected by universal suffrage; the president possesses quite considerable powers; and there also exists a prime minister and cabinet, subject to parliamentary confidence, who perform executive functions (ibid., 166). Some scholars, such as Sartori (1997), argue that the criterion of universal suffrage is ambiguous as to whether the president is elected in a direct or an indirect manner. Sartori adopts the principle that the president is elected by popular vote — either directly or indirectly.

Bahro et al. (1998) revised Duverger's third criterion — the president must face a prime minister possessing executive and governmental powers and remain in office only if parliament does not express its opposition. In addition, some scholars emphasize the dual authority structure (president and prime minister) of semi-presidentialism (Blondel, 1984; Bogdanor, 1987; Pasquino, 1997). Among all the definitions of semi-presidentialism, Robert Elgie (1999) proposes a concise definition of semi-presidentialism in order to avoid the confusion of measuring the president's powers in practice. According to Elgie's definition, semi-presidentialism is where a popularly elected fixed-term president exists alongside a prime minister and cabinet who are responsible to the legislature. This definition is based on the constitutional situation and is more comprehensive than other's definitions by identifying more than 55 countries as semi-presidential (Elgie, 2007, 9). Semi-presidentialism as a category contains a varied set of countries. It can refer to countries with strong presidents or countries with weak presidents. In order to clarify similarities and differences among semi-presidential countries to facilitate comparisons, it is better to examine the effects of different sub-types of semi-presidentialism.

The Sub-types of Semi-presidentialism

Various scholars have already tried to identify different sub-types of semi-presidentialism. The previous analytical frameworks for semi-presidentialism mainly focus on presidential powers (Duverger, 1980; Shugart and Carey, 1992), as well as presidential and prime ministerial powers⁵ (Elgie, 2005), and the party situation in parliament (Skach, 2005b). Table 1 outlines existing typologies and proposes a new typology.⁶ Using constitutional sub-types as a means of classifying semi-presidentialism can have some advantages. For example, Skach (2007, 93) argues that 'Constitutional sub-types help us to move away from the metalevel of constitutional analysis and better understand when well-intentioned constitutional designs are likely to be turned from virtue to vice, distorted through political battles across ideological lines.' In this paper, we label semi-presidentialism with adjectives to refer to five sub-types. All of the sub-types denote different phases of semi-presidentialism depending on the interaction between the constitutional situation and political circumstances. It is possible to place any semi-presidential regime into one of the five sub-types. For example, from 1991 to 1993 Russia experienced balance-of-powers semi-presidentialism; however, after 2000, it has experienced president-dominant semi-presidentialism due to Russia's constitutional changes and the election results in parliament.

Table 1 - Classifications of semi-presidentialism.

Full table

There are three reasons why the classification proposed here better captures political dynamics in semi-presidentialism. First, it includes a greater number of sub-types, such as a parliament-dominant system and a balance-of-powers system, thus capturing a more comprehensive set of situations than in other work. Second, it is based on the interaction of all the possible dominant actors — the president, the premier, and parliament — thereby illustrating how varied semi-presidentialism is. Third, it highlights the constitutional powers of parliament and the executive–legislative relations (president vs parliament). These are crucial for grasping the operation of semi-presidentialism. In short, analytic differentiation has been increased in order to capture the diverse forms of semi-presidentialism that have emerged.

The president's constitutional power is an important indicator for analyzing the operation of semi-presidentialism. However, it is not necessary for there to be a linear relationship between a president's formal powers and a president's real political authority. For example, a president with weak formal powers can still have strong political authority if he/she can command a legislative majority in the parliament. By the same token, a president with strong formal powers can end up possessing less political authority when he/she only wins a minority of support in the parliament. In other words, only emphasizing a president's formal constitutional powers is not sufficient to grasp the whole picture of political practice in semi-presidentialism. As Skach (2005a, 355) argues 'A president's constitutional powers alone do not explain conflict within the semi-presidential model, for such powers remain constant over time, and yet the president's de facto power vis-à-vis the legislature, and thus the degree of institutional conflict, varies.' The relationship between the president's political authority (de facto power) and the president's formal power depends on the political circumstances in parliament. When the president also controls a majority of the parliament, it is highly possible for the president to intervene in legislation, since the president can ask the premier with the same political affiliation to initiate or pass bills in an effort to enforce the president's preferred policies. When the president only has a minority support in parliament, it is less possible for him to intervene in legislation unless he has some unilateral powers such as decree power to do so. This article argues that the constitutional powers of the president, the constitutional powers of the parliament, and the party situation in parliament can affect which branch holds the power to pass legislation. Figure 1 shows the classification of semi-presidential sub-types, based on which branch of government or political leader holds the power to pass legislation.

Figure 1.

The classification of semi-presidential sub-types.

Full figure and legend (28K)

The power to pass legislation denotes which branch or leader has the final say on legislation. The rules to decide the power to pass legislation are both the constitutional powers of the president and parliament and the institutional actor's partisan strength in parliament. Some studies have underlined the need to pay attention to which branch dominates the legislative process (Bahro, 1999). For example, holding the power to pass legislation is more important than controlling executive posts, since it can decide the content and direction of new policies. Generally speaking, parliament is the highest legislative branch in every democratic country, but, in practice, it is not necessarily the institution that holds de facto legislative power. Within semi-presidentialism, legislative power may be held by the president or parliament, or the president and parliament may be in competition with each other. In this context, five scenarios can be envisaged. First, when the president holds considerable powers such as decree power and veto power, controlling a legislative majority in the parliament, the president commands the power to pass legislation. Second, when the president does not have substantial powers and acts as a symbolic leader, then the premier controlling a legislative majority holds the power to pass legislation. Third, when the president does not have considerable powers such as decree power, when the cabinet is a minority government, and when the opposition controls a legislative majority against the president and government, parliament holds the power to pass legislation. Fourth, when the president holds reserved powers and the premier controlling a legislative majority comes from an opposing party or coalition, the president and premier share the power to pass legislation. Fifth, when two opposing branches (president and parliament) are equally empowered, they would contend for the power to pass legislation.

The first three sub-types (president-dominant system, premier-dominant system, and parliament-dominant) have clear authority structures. This means they are characterized by less political deadlock between executive and legislative branches than the remaining two sub-types. By contrast, cohabitation semi-presidentialism is marked by conflict between the president and the prime minister. The last sub-type (balance-of-powers semi-presidentialism) is a particularly risk-prone situation. This sub-type is likely to generate more political conflicts and/or deadlock than any other, thereby contributing to the breakdown of a democratic regime. In the rest of this article, we examine each sub-type of semi-presidentialism and illustrate why balance-of-powers semi-presidentialism is particularly problematic.

Sub-types of Semi-presidentialism

President-dominant semi-presidentialism

In semi-presidentialism, when the president simultaneously commands the government and a legislative majority in parliament, he/she holds the power to pass legislation. The president can dominate the initiation of policies by virtue of his/her party's legislative majority in parliament. Most of the time, prime ministers in semi-presidential countries are nominated by the president. If both of them are from the same party or coalition, the president can capitalize on the party structure and/or coalition agreement to enforce his/her authority. In addition, since presidents in semi-presidential countries are directly elected by popular vote, they have a higher legitimacy than the prime minister and at times the prime minister simply executes the policy of the president. The regime of President Charles de Gaulle in France is a sub-type representing president-dominant semi-presidentialism. The reasons why President de Gaulle was dominant were due to constitutional stipulations, constitutional change, and the political dynamics in parliament.

In France, Article 8 of the 1958 Constitution stipulates that the President of the Republic shall appoint the Prime Minister and he shall terminate the appointment of the Prime Minister when the latter tenders the resignation of the Government. Article 9 says that the President of the Republic shall preside over the Council of Ministers. Those two articles empower the French president to be influential in initiating the government's policies. After 1962, the fact that the French president was elected by popular vote has also granted the president more political authority. The universal suffrage reform of 1962 offers the president a principled argument to defend his de facto power to dismiss the prime minister despite constitutional language to the contrary (Rohr, 1994, 130). For example, according to Article 49 (2), the National Assembly may raise an issue of the Government's responsibility by passing a motion of censure. However, in practice, the president holds a great deal of authority on deciding the tenure of premiership, and the prime minister is relegated to the sideline. In terms of political dynamics, the UNR (Union for the New Republic) party led by President de Gaulle and his allies controlled a legislative majority in the parliament. Since then, the legitimacy of direct election and the control of a legislative majority in the parliament have increased the president's political authority. The president can also engage the powers granted by the constitution. These enable the president to be the chief executive and to wield the power to pass legislation.⁷ The latter French presidents Georges Pompidou and François Mitterrand have followed the role of President de Gaulle to become a super-president (Keeler and Schain, 1997, 95–97).

Premier-dominant semi-presidentialism

Although some presidents in semi-presidential countries hold certain constitutional powers, in practice, their role is more like a national symbol, or figurehead. The major role and task of the president are to sign and promulgate the laws passed by parliament. A premier with a legislative majority controls the power to pass legislation. Some scholars argue that the constitution of such countries is semi-presidential, but in practice it works like a parliamentary system (Duverger, 1980; Elgie, 2007). Similar to a parliamentary system, the premier with a legislative majority in this type of semi-presidentialism calls the shots in the process of legislation.

The Austrian case is a typical case of premier-dominant semi-presidentialism. According to Article 60-1 of the Austrian constitution, the Federal President is elected by the nation on the basis of equal, direct, secret, and personal suffrage. This stipulation follows one of the definitions of semi-presidentialism — that the president is elected by universal suffrage. In addition, the president in Austria has considerable political powers such as the power of nomination and termination of the premiership, and the power of dissolution of the parliament. First, the Federal Chancellor and on his recommendation the other members of the Federal Government are appointed by the Federal President; no recommendation is requisite to the dismissal of the Federal Chancellor or the whole Federal Government; and the dismissal of individual members of the Federal Government ensues on the recommendation of the Federal Chancellor (Article 70-1). In this instance, the title of Federal Chancellor is considered the premier or prime minister. Second, the Federal President can dissolve the House of Representatives, but he may avail himself of this prerogative only once for the same reason. If we only consider the formal constitutional articles, the Austrian president holds considerable political powers. But if we ponder some other constitutional articles and constitutional customs, we can conclude that the premier is dominant in the operation of semi-presidentialism.

In Austria the president takes the role of chief of state. Article 70-2 says that only persons eligible for the House of Representatives can be appointed Federal Chancellor, Vice-Chancellor, or Federal Minister; members of the Federal Government need not belong to the House of Representatives. In other words, the Chancellor and Vice-Chancellor have to be a legislator in the parliament before they become the premier or the vice premier. This stipulation sets a limitation on the nomination powers of the president. The president must passively accept the political dynamic in the parliament and nominate a legislator from a party or a coalition with a legislative majority to become the chancellor. Moreover, a further constitutional stipulation constrains the operation of presidential powers. According to Article 67 of the Austrian constitution, unless otherwise provided for by the Constitution, all official acts of the Federal President shall be based on recommendation by the Federal Government or the authorized Federal Minister. This means that except for the powers reserved for the president, all official acts of the president have to acquire consent from the premier. This article has resulted in the Austrian president becoming less involved in the operation of government and has, in essence, diminished his political authority.

In past constitutional development, the former presidents in Austria have also played a symbolic role as chief of state (Muller, 1999, 23–26). By controlling a legislative majority in the parliament, the premier is responsible for the power to pass legislation. Austrian semi-presidentialism resembles a parliamentary system, but the difference is that the president in Austria has dissolution power. Since that power has seldom been used, it is possible to view the president as powerless and incorrectly define or classify the whole government as a parliamentary system. In fact, the constitution grants the president such a power that he/she could enforce when facing political contingencies. Hence, the Austrian case should be treated as a premier-dominant semi-presidential system, not as a parliamentary system.

Parliament-dominant semi-presidentialism

In a system where the president does not have considerable powers such as decree power, and when the cabinet is a minority government, or when the opposition controls a legislative majority against the president and government, parliament holds the power to pass legislation.⁸ For parliament, controlling the power to pass legislation not only means that it can amend or veto the proposals from the government, it also indicates that the parliament can decide the direction and content of policies. In this case, even when the president has executive authority to nominate the premier or form the cabinet, his/her political authority is still decreased by the fact that the parliament is dominant. The president cannot control the legislative agenda. As some scholars argue, by controlling the majority the opposition can agree on policy packages that are more preferred by its members than government proposals (Laver and Schofield, 1998). Under a minority government, the power to pass legislation shifts to parliament and it is therefore forced to implement the laws or policies passed by the parliament. The minority government can initiate the policies or bills, but the outcome can be a far cry from the original proposals. This sub-type of parliament-dominant semi-presidentialism has seldom been explored and discussed. Portugal is an excellent example to illustrate this particular sub-type.

Portugal

The Portuguese political system is a semi-presidential system (Martins, 2006). According to the Portuguese constitution, the president shall be elected by universal, direct, and secret suffrage (Article 121). The president has considerable powers such as nomination of the premier, veto power, and dissolution of the parliament. In regard to the nomination of the premier, the constitution stipulates that the president shall appoint the prime minister after consulting with the parties that hold seats in Assembly of the Republic and in the light of the electoral results (Article 187). There is no requirement of investiture, which means that the president can nominate a premier without the approval of parliament in accordance with Portuguese law (Bergman, 1993). This gives the president some autonomy in selecting the premier. With regard to veto power, the constitution stipulates that within 20 days of the receipt of any decree from the Assembly of the Republic for enactment as a law, or the publication of a Constitutional Court ruling that does not declare any of the decree's provisions unconstitutional, the President of the Republic shall either enact the decree or exercise the right of veto (Article 136-1). However, if the Assembly of the Republic confirms its original vote by an absolute majority of all the Members in full exercise of their office, the President of the Republic shall enact the decree within 8 days of receiving it (Article 136-2). The first article grants the president the power to veto any legislation from the Assembly; however, the second article diminishes that political power by allowing any veto to be overridden by a simple majority vote of the Assembly. In addition, the president in Portugal does not have decree power to unilaterally enact policies. However, the Portuguese president can dissolve the parliament after first consulting both the parties with seats in the Assembly and the Council of State. In addition, the Assembly of the Republic cannot be dissolved during the first 6 months following its election or during the last 6 months of the President of the Republic's term of office. Also, other laws stipulate that the assembly cannot be dissolved during a state of siege or a state of emergency (see Article 132 and Article 172-1). In short, the president in Portugal has considerable powers that can make him/her a strong president. However, in practice, the president's political authority can be restrained by political dynamics in parliament.

For instance, from 1985 to 1987, Portugal experienced a period of minority government, that was led by the Social Democrat Party and which only captured 110 of the total 250 seats, thus giving the opposition coalition a legislative majority (Leston-Bandeira, 1999). During this period, the parliament was a policy-making legislature. It could not only modify or reject government measures but could also substitute those policies or even formulate entirely separate policies altogether (Leston-Bandeira, 1999, 49). The substantial evidence of this scenario is that the percentage of government's bills approved at the final vote during the period of minority government (1985–1987) was 34.1%, which is much lower than the percentage of government's bills (94.9%) approved at the final vote during the period of a unified government from 1987 to 1991 (Leston-Bandeira, 2001, 146). The minority government experience in Portugal shows that the president can nominate the premier or dominate in the formation of the executive branch, but the final authority of legislation rests in the hands of parliament. Even when the president can use veto power against the parliament, parliament can override the veto by a majority vote. Clearly, the opportunity for veto success is very low if the president's party does not have control of the parliament. Political dynamics in parliament not only decreased the president's political authority in legislation but also constrained the operation of presidential powers of veto.

Cohabitation Semi-presidentialism

Cohabitation describes the situation that the president and the premier share executive power but come from opposing parties and have different sources of legitimacy. The legitimacy of the president is from direct election, but the legitimacy of the premier is from the control of the majority of seats in parliament. In this sub-type of semi-presidentialism, the premier is in charge of most of the legislation, but in some reserved domains the president assumes certain responsibilities. The cohabitation experience originated from the emblematic semi-presidential country, France. Article 20 of the French constitution states that government decides and directs the policy of the nation and that it is responsible to parliament, and Article 21 stipulates that the constitution entrusts the prime minister with national defense and the execution of laws. In other words, the premier is at the helm of government, but it does not mean that the premier can be preponderant in every domain during the period of cohabitation. The president in France has reserved powers pertaining to defense and diplomatic affairs. In those domains, the premier has to ascertain the president's opinion before the enforcement of related policies. The president also holds some constitutional weapons that balance his/her relationship with parliament. The president can dissolve the parliament or propose the passed bills by parliament to the constitutional court to be reviewed, or send the passed bills back to the parliament for further deliberations (Helms, 2004, 36–37). In addition, Article 9 of the French constitution states that the president in France shall preside over the Council of Ministers. The president can affect the formation of legislation by chairing the council meeting. In cohabitation, the president in semi-presidential countries is not like the president in parliamentary countries. The president under cohabitation is not a symbolic national chief, but is entrusted with some considerable powers to affect decision-making and the conduct of policies. In brief, during cohabitation, the premier dominates most of the legislative process while the president holds the final decision-making authority in some reserved domains.

From 1986 to 1988, France experienced its first period of cohabitation. The leftist president, François Mitterrand, from the Socialist Party, nominated a rightist premier, Jacques Chirac, controlling a legislative majority (Poulard, 1990). With that majority, Chirac was the most powerful prime minister of the Fifth Republic (Pierce, 1991, 278–279). However, the president under cohabitation did not turn out to be a powerless president. President Mitterrand insisted on the primacy of the president in foreign affairs, exploited his capacity to command the attention of the public, frequently registered his disapproval of the government's conduct, relied on certain veto powers specifically conferred on the president by the constitution, presented himself as the champion of parliament in the face of the overbearing government, sought to delay enactment of several of the government's measures, but never sought to create a situation from which a united parliamentary could not extricate itself (Pierce, 1991, 279). Despite cohabitation, the president can claim and enforce some considerable powers granted in the constitution. Among those powers, the veto power of the president is a powerful weapon against the parliament. The president can ask the parliament to reconsider a law or certain part of its articles (Article 10). Certainly, the veto power of the president cannot succeed every time, especially when there is an opposition with a legislative majority in the parliament, but the president can exercise this prerogative to delay the legislation in parliament and force a negotiation.

In short, the experience of cohabitation in France did not result in a pure parliamentary system since the president could still enforce some constitutional powers. On the other hand, the difference between the semi-president and the US-style president is evident: he/she is not the chief executive and can only have effect on the executive body through the government if he/she has a majority in parliament; if the president under semi-presidentialism does not have such a majority or even if he is opposed by parliament as in cohabitation, he can do neither (Bahro, 1999, 7). The possible conflicts in this sub-type stem mostly from the competition between the president and premier for the sharing of legislative power.

Balance-of-powers semi-presidentialism

Balance of powers between president and parliament denotes that the president and parliament hold considerable powers but confront each other. This sub-type occurs where no party or coalition controls a legislative majority. In this situation, political dynamics can quickly shift and the government has to establish different ad hoc alliances when attempting to pass any legislation (Skach, 2005b). For some bills, when a legislative majority is established, the parliament holds the final authority of legislation. For other bills, it is hard to form a legislative majority producing legislative delay, which makes a president with unilateral powers (such as decree powers) and emergency powers more likely to bypass parliament and unilaterally execute his preferred policies. Moreover, the president and parliament compete for the final authority of legislation, forming a dual authority of legislation. The features of this sub-type are that the president and parliament each hold significant powers and compete with each other for legislation, and it usually ends in political deadlock in which no branch can gain an advantage. We use the Weimar Republic and the Russian First Republic (1991–1993) to illustrate this sub-type of semi-presidentialism.⁹

The Weimar Republic

Constitutional practice in the Weimar Republic was a classic example of the balance-of-powers semi-presidentialism. The president appointed the personnel of cabinet, could dissolve parliament, enjoyed decree power, and had the power to call referenda. The constitution stated that the president could appoint and dismiss the Federal Chancellor and, on the latter's recommendation, Federal Ministers (Article 53). This meant that the president chose the cabinet at will. As to the dissolution of power, the President could dissolve the Reichstag (parliament), but only once for any one cause (Article 25). Technically, this article set a limitation on the use of dissolution, but an arbitrary president could use different causes to dissolve the parliament. The constitution also stipulated that if public order and security were seriously disturbed or endangered within the Federation, the President could take all necessary steps to restore order, which included the use of armed forces (Article 48-3). This emergency power was an important unilateral power for the president, especially when he wanted to implement his preferred proposals. The president in the Weimar Republic could invoke the power of referenda against parliament. The constitution said that a law passed by the Reichstag could, before its promulgation, be submitted to popular vote if the President decided to do so (Article 73). In short, the Weimar constitution created a strong president, endowing the president with considerable powers to enforce the nation's tasks.

However, the Weimar constitution also entrusted the parliament with powers such as censure power and the power to abolish the president's decrees, making rival claims as an equal to the president. The Reich chancellor and the Reich ministers, in order to exercise their mandates, were required to have the confidence of Reichstag; any one of them had to resign if the Reichstag voted by explicit decision to withdraw its confidence (Article 54). The parliament could vote no confidence in the premier or any individual cabinet member. Also, the president was required to inform the parliament immediately about all measures undertaken and the measures could have been suspended immediately if parliament warranted (Article 48-3). This meant that parliament had a proactive power to override the president's executive measures.

In practice, the features of the Weimar Republic meant that 'it was a cabinet without a parliamentary majority, which it tried to secure by dissolution of the Reichstag and new election; where the elections did not provide a majority (as in fact they did not during the Presidential period) the cabinet would govern by the use of the President's emergency powers, continuing in office as long as the Chancellor had the President's confidence' (Needler, 1959, 692). The President of the Weimar Republic, Paul Von Hindenburg, quoted Article 48 of the constitution repeatedly in order to unilaterally govern the country. In 1931, parliament only passed 34 items of law while the president enacted 44 emergency decrees. Even in 1932, parliament only passed five items of law as the president's emergency decrees reached a new high of 66 items (Lepsius, 1978, 49 and see Table 4). By enactment and enforcement of emergency decrees, the Weimar president held the final authority of legislation. Moreover, if parliament suspended an emergency decree, the president could have invoked the power of dissolution of parliament to hold a new legislative election. In fact, in July 1930, an emergency decree on the budget was not tolerated by Parliament and the president resorted to his right to dissolve the Parliament (Lepsius, 1978, 48). The power to dissolve the parliament became a potent weapon for the Weimar president when confronting an intransigent parliament. When the opposition united together to pass a bill or to suspend a president's emergency decree (as in the aforementioned case), the parliament grasped the final authority of legislation. In a nutshell, the constitutional practice in the Weimar Republic was a vicious circle. When the parliament delayed the passing of a law, the president would choose to enact emergency measures in lieu of parliament bills, and then the parliament would cease the president's emergency measures, and finally the president would retaliate via the dissolution power. As Juan Linz (1994, 54) argues, 'The situation ends up being similar to the worst of the true presidential systems with an ineffective and rebellious congress.' The Weimar Republic case demonstrates that when two branches both hold momentous powers to check each other and jockey for the final authority of legislation, serious political conflicts and deadlock occur, even to the point of breaking down the democratic regime. The Weimar Republic was a type of balance-of-powers semi-presidentialism. We suggest that this sub-type is most likely to generate a political deadlock. In order to reach a generalization, we will illustrate another case of the type of balance-of-powers semi-presidentialism — the Russian case.

The Russian First Republic (1991–1993)

In Russia, the constitution of the First Republic stated that the president nominated the premier (Article 104-10). The constitution said that the president may issue decrees on subjects within his/her competence, as long as they did not contradict existing law or the Constitution (Article 121-8). The president held an important unilateral power. Decree powers mean that the executive branch has the authority to establish law in lieu of action by the assembly (Carey and Shugart, 1998). In addition, the president could return bills to the assembly for reconsideration (Article 121-5). The constitution entrusted the president with numerous, considerable powers. However, the powers of parliament were not inferior to the powers of the president, even surpassing it. The Congress was the supreme organ of state power (Article 185). Parliament also held investiture power, the power to abolish executive measures, and the power to override a veto. The president could nominate the prime minister, but it required the approval of a majority of deputies in the Congress (Article 104:10). When the president did not control a legislative majority, the Congress dominated the decision of choosing the prime minister. Existing law or the Constitution trumped any presidential decree in the case of conflict (Article 121:8). That means the parliament could pass laws to replace or rescind executive decrees. According to Article 121:5, the president's veto could be overridden by a simple majority vote in the Congress. Weighing relative powers between the president and parliament in the Russian First Republic case, we can classify it as a case of a strong president vis-à-vis a strong parliament. Although the president of the Russian Republic was less powerful than the president of the Weimar Republic (the Russian president in the First Republic did not have the power to dissolve the parliament), the Russian president possessed unilateral powers that allowed him/her to become strong.

The president of the First Republic, Boris Yeltsin, extensively used constitutional decree power and delegated decree power granted by the parliament to carry out political and economic reforms during the democratic transition (Parish, 1998). President Yeltsin capitalized on his presidential mandate and popular support to legitimatize rule by decree and ignored the voice of opposition in the parliament (Schleiter, 2003). Although Yeltsin won the powers he asked for, his victory did not reflect a strong partisan majority supporting Yeltsin; no such majority existed (Parish, 1998, 73). Hence, during the reforming process, the parliament locked horns with the strong president, even passing related laws and resolutions to undermine the president's reforms (Kubicek, 1994). The executive–legislative relations reached a serious deadlock. For instance, in 1993, President Yeltsin issued a decree to accelerate the process of voucher privatization, the parliament voted to suspend that decree and asked the Constitutional Court to consider its constitutionality, and Yeltsin responded with a second decree, which reiterated the provisions of the first one — and even further accelerated the timetables for privatization (Parish, 1998, 76–77). The parliament countered with another suspension and Yeltsin again enacted a third decree on the same issue (Gualtieri, 1993, 12–13). In brief, the interaction between the executive and legislative branches became a war of laws (Huskey, 1992, 91). The president and parliament contended for the power to pass legislation, making political conflicts between two branches ascend to regime crises or even a breakdown of the democratic regime. The First Russian Republic case proves that when the constitution entrusts the president and parliament with almost equal powers, it is easy to bring about political conflicts and deadlock over the power to pass legislation.

Balance of powers and political deadlock

Formal presidential powers do not exist in a political vacuum and are influenced by other factors such as parliamentary powers and the political dynamics in the parliament. Most notably, constitutionally strong presidents are sometimes politically weak and constitutionally weak presidents are sometimes politically strong (Elgie, 1999, 14). In other words, presidential formal powers do not accord with presidential political authority. Theoretically, their relationship has three scenarios: (1) when the president has strong formal powers and also controls a legislative majority, he holds the highest political authority; (2) when the president has weak formal powers but controls a legislative majority, the president still holds considerable political authority; and (3) when the president has strong formal powers but does not control a legislative majority, his/her political authority is severely weakened.

For instance, when the president wants to use his/her veto power, she/he has to consider parliamentary powers and the political dynamics in the parliament that are prevalent in the parliament. When it needs a two-thirds of vote in the parliament to override a presidential veto and the president commands at least one-third votes in the parliament, the president can exercise his/her veto power and with a high rate of success. The Russian president, Boris Yeltsin, invoked his veto powers 219 times from 1994 to 1999. According to the Second Republic constitution in Russia, the parliament needed a two-thirds of majority to override a presidential veto. It was not easy for both the Duma and the Federation Council to muster the necessary two-thirds majority to overturn a presidential veto (Chandler, 2001, 506). This allows the president to exercise his veto power with fewer limitations. However, when the Duma and the Federation Council only needed a simple majority of votes in order to override a presidential veto and the president controlled less than half of the members in the parliament, it restrained the president from using his veto power.

In semi-presidential countries, parliamentary powers can be divided into two categories: proactive powers and reactive powers. Proactive powers include censure power and power to initiate bills. Reactive powers are investiture power and the override power. The reactive override power is considerable when parliament only needs a simple majority to override the president's veto. When parliament needs a two-thirds of majority to override the president's veto, its reactive power is restrained. Overall, when the parliament is entrusted with investiture, censure, and override powers requiring a simple majority vote, the parliament is a powerful legislature. In addition to formal powers, the political dynamics in parliament also affects its political authority. When the opposition commands a legislative majority against the president, parliament has its autonomy and substantial authority. By contrast, when the legislative majority is controlled by the president, the parliament's autonomy and substantial authority are decreased.

The Interaction of Executive and Legislative Branches in Semi-presidentialism

There are four scenarios that reveal the dynamic between the president and parliament. First, when the president has decree power, veto power, the power to nominate a premier, and the power to dissolve the parliament, he/she can use those powers to confront an opposing parliament. By contrast, when the parliament can override vetoes or rescind decrees or propose a censure motion with a simple majority, it has strong powers to counteract the president or the government. The pendulum of power can swing back and forth between the president and parliament. This is a so-called strong president vis-à-vis strong parliament mode. Second, when the president has many powers and the parliament by contrast has few powers, it is the mode of strong president vis-à-vis weak parliament. Third, when the president has few powers, but the parliament has many powers, it is the mode of weak president vis-à-vis strong parliament. Fourth, when the president and parliament both have few powers, it is the mode of weak president vis-à-vis weak parliament.

Political conflicts and deadlock occur easily when the country has a mode of strong president vis-à-vis strong parliament. The scenario of strong president vis-à-vis strong parliament is typically as follows. First, the parliament passes a bill preferred by the opposition. Second, the president vetoes the parliament's bill. Third, the opposition in parliament overrides the president's veto by a simple majority. Fourth, the president issues a decree to substitute the bill in dispute. At this point, political conflicts loom large between the president and parliament. Fifth, the opposing parliament rescinds the decree by a simple majority or passes another law to surpass it. Political deadlock culminates in this situation. In brief, the balance of powers interchangeably shifts in favor of the president or in favor of parliament (see Figure 2).

Figure 2.

The mode of strong president vis-à-vis strong parliament.

Full figure and legend (11K)

When one branch surpasses another branch, such as the mode of strong president vis-à-vis weak parliament or the mode of weak president vis-à-vis strong parliament, political conflicts between the executive and legislative branches can still occur but do not ripen into political deadlock. Overall, the Weimar Republic and the Russian First Republic belong to the mode of strong president vis-à-vis strong parliament and this situation is more likely to generate political conflicts and deadlock, even to the point of breaking down the democratic regimes. Portugal is in the mode of parliament-dominant semi-presidentialism and the president's powers and political authority are restrained by the parliament. During the non-cohabitation period, the president in France dominates and during the cohabitation period, the premier guides legislation in most domains, but the president prevails over some reserved domains. The Austrian premier dominates the legislation and executive enforcement. These different cases show that semi-presidentialism embodies different sub-types and the distribution of power between the president and parliament makes a difference as to how the democratic regimes function.

Conclusion

The Weimar Republic was the first country embodying the mode of semi-presidentialism in the history of constitutional engineering. The democratic regime broke down and an authoritarian regime replaced it. The Russian First Republic ended with a confrontation between an imperial president and an intransigent parliament. These two cases are often cited by constitutional scholars to explain the perils of semi-presidentialism. However, some countries have implemented semi-presidentialism during the process of democratic transition and have successfully consolidated. What leads to the success or failure of semi-presidentialism has become a focal point for comparative constitutional engineering. Past research has focused on presidential powers, but has ignored parliamentary powers or the interaction of presidential and parliamentary powers. In semi-presidential countries, presidential powers play an important role, but they have to be considered alongside parliamentary powers and in the context of the interaction between the two.

This research shows that the perils of semi-presidentialism mostly emerge from a balance of a strong president vis-à-vis a strong parliament. When both the president and parliament in semi-presidentialism hold strong powers, and when opposing parties control the presidency and parliament, the situation is more likely to generate political conflict and deadlock. When one branch prevails over the other branch, political conflict still occurs but it does not necessarily evolve into a political deadlock or spiral into a regime crisis. When the president in semi-presidentialism faces a confrontational parliament, he/she may choose not to confront the parliament or resort to institutional brinkmanship. It all depends on how much power the president holds. Unless the president holds considerable unilateral powers, a president with weak powers may choose to negotiate or make concessions with the parliament.

All in all, semi-presidentialism is a variegated constitutional type, since it has many sub-types and different sub-types can generate different degrees of political conflicts and/or deadlock. The advice we should give to the founders of a young democracy is this: 'Do not grant the president and parliament almost equal powers.' If you do, your newfound democracy will be in jeopardy.

Notes

¹ Regarding the dependent variable, the past research on semi-presidentialism stresses political stability or political instability (see Wu, 2000; Roper, 2002; Skach, 2005b). In addition, regime survivial and regime performance are the most common dependent variables in the research of comparative constitutional engineering. Related dicussions can be seen in Elgie (2004).

² The definition of political deadlock here is different from Cheibub's definition. He argues that political deadlock occurs when parliament passes bills preferred by the opposition, the president vetoes these bills, and the opposition does not have enough votes to override the presidential veto (Cheibub, 2002, 289).

³ These semi-presidential countries are Croatia, Macedonia, Bulgaria, and Poland (see Metcalf, 2000, 679).

⁴ Skach (2005b) proposes semi-presidentialism is not a mixed system nor does it combine the features of presidentialism and parliamentarism and indeed is the third type of constitutional system. In this vein, we second Skach's idea. In addition, Shugart and Carey (1992, 23) substitute the premier-presidential and president-parliamentary system for semi-presidentialism. They argue that the term 'semi-presidential' implies a regime type that is located midway along some continuum running from presidential to parliamentary. Sartori disagrees with their argument, asserting that the word 'semi' does not assume any continuum because it precedes continuum-mania by well over 2,000 years (1997, 137). He also cautions that we should not read the label too literally. However, M. Shugart in a recent paper forgos his previous argument, agrees to use the term of semi-presidentialism, and treats the premier-presidential regime and president-parliamentary regime as the sub-types of semi-presidentialism. Related discussions can be seen in Shugart (2005).

⁵ Premier and prime minister will be used interchangeably in this paper.

⁶ Precisely, the classification here has to be called taxonomy, since it is based on empirical entities, not logical types. However, the term taxonomy is more generally used in the biological sciences, while typology is used in the social sciences (Bailey, 1994, 6). Accordingly, we still use typology here to refer to the classification.

⁷ Suleiman even argues that the French president is the most powerful executive in the Western world or an imperial president, who has more powers than the American president. Related dicussions can bee seen in Suleiman (1980, 103–105) and Suleiman (1994).

⁸ This situation only applies to cases in which there is no investiture. Studies of minority governments in Europe have found that the reason why some countries have a high frequency of minority governments is that they usually adopt the negative form of investiture, that is, no vote of investiture is required and a government remains in power until the opposition passes a vote of confidence or the government resigns (Bergman, 1993). One of the semi-presidential countries, Taiwan, has the same design as the formation of government, which does not need the investiture from the parliament. From 2000 to the present, Taiwan has experienced a period of minority governments. The president, Chen Shui-bian, chose to form minority governments and they had to face an opposition controlling a legislative majority in the parliament. Taiwan's case is similar to Portugal's case being discussed here. However, we only analyze Portugal as a case of parliament-dominant semi-presidentialism due to the limitation of the paper's length.

⁹ Skach uses the sub-type of divided minority government to also explain the Weimar Republic (2005b). Here, this paper emphasizes parliamentary powers and balance of powers between the president and parliament, which is different from Skach's argument.

References

1. Amorim Neto, O. and Strøm, K. (2006) 'Breaking the parliamentary chain of delegation: Presidents and non-partisan cabinet members in European democracies', British Journal of Political Science, 36(4): 619–643. | Article |
2. Bahro, H.H. (1999) 'Virtues and vices of semi-presidential government', Journal of Social Sciences and Philosophy, 11(1): 1–37.
3. Bahro, H.H., Bayerlein, B.H. and Veser, E. (1998) 'Duverger's concept: semi-presidential governments revisited', European Journal of Political Research, 34: 201–224.
4. Bailey, K.D. (1994) Typologies and Taxonomies: An Introduction to Classification Techniques, California: Sage Publications Ltd Press.
5. Bergman, T. (1993) 'Formation rules and minority government', European Journal of Political Research, 23: 55–66. | Article |
6. Blondel, J. (1984) 'Dual Leadership in the Contemporary World: A Step towards Executive and Regime Stability?', in D. Kavanagh and G. Peele (eds.) Comparative Government and Politics: Essays in Honour of S. E. Finer, Boulder: Westview Press, pp. 73–91.
7. Bogdanor, V. (1987) 'Semi-presidential Systems', in V. Bogdanor (ed.) The Blackwell Encyclopedia of Political Institutions, Oxford: Basil Blackwell Publishers, pp. 561–562.
8. Carey, J. and Shugart, M. (1998) 'Calling out the Tanks or Filling Out the Forms?', in J. Carey and M. Shugart (eds.) Executive Decree Authority, New York: Cambridge University Press, pp. 1–29.
9. Chandler, A. (2001) 'Presidential veto power in post-communist Russia, 1994–1998', Canadian Journal of Political Science, 34(3): 487–516.
10. Cheibub, J.A 2002 'Minority governments, deadlock situations, and the survival of presidential democracies', Comparative Political Studies, 35(3): 284–312. | Article |
11. Duverger, M. (1980) 'A new political system model: semi-presidential government', European Journal of Political Research, 8: 165–187. | Article |
12. Elgie, R. (1999) 'The Politics in Semi-presidentialism', in R. Elgie (ed.) Semi-presidentialism in Europe, New York: Oxford University Press, pp. 1–21.
13. Elgie, R. (2004) 'Semi-presidentialism: concepts, consequences and contesting explanations', Political Studies Review, 2: 314–330. | Article |
14. Elgie, R. (2005) 'Variations on a theme', Journal of Democracy, 16(3): 98–111. | Article |
15. Elgie, R. (2007) 'What is Semi-presidentialism and Where is it Found?', in R. Elgie and S. Moestrup (eds.) Semi-presidentialism Outside Europe, London: Routledge, pp. 1–13.
16. Gualtieri, D. (1993) 'Russia's new 'War of Laws'', RFE/RL. Research Report, 2(3), 10–15.
17. Helms, L. (2004) 'Five ways of institutionalizing political opposition: lessons from the advanced democracies', Government and Opposition, 391(Winter): 22–54. | Article |
18. Huskey, E. (1992) 'Executive–legislative Relations', in E. Husky (ed.) Executive Power and Soviet Politics: The Rise and Decline of the Soviet State, NY: M.E. Sharpe Inc Press, pp. 83–105.
19. Keeler, J.T.S. and Schain, M.A. (1997) 'Institutions, Political Poker and Regime Evolution in France', in K. Von Mettenheim (ed.) Presidential Institutions and Democratic Politics: Comparing Regional and National Contexts, Baltimore: The Johns Hopkins University Press, pp. 83–105.
20. Kubicek, P. (1994) 'Delegative democracy in Russia and Ukraine', Communist and Post-Communist Studies, 27(4): 423–441. | Article |
21. Laver, M. and Schofield, N. (1998) Multiparty Government: The Politics of Coalition in Europe, Ann Arbor: The University of Michigan Press.
22. Leston-Bandeira, C. (1999) 'The roles of the Portuguese parliament based on a case study: the discussion of the budget, 1983–95', The Journal of Legislative Studies, 5(2): 46–73.
23. Leston-Bandeira, C. (2001) 'The Portuguese parliament during the first two decades of democracy', West European Politics, 24(1): 137–156. | Article |
24. Lepsius, M.R. (1978) 'From Fragmented Party Democracy to Government by Emergency Decree and National Socialist Takeover: Germany', in J.J. Linz and A. Stepan (eds.) The Breakdown of Democratic Regimes, Baltimore: The Johns Hopkins University Press, pp. 34–79.
25. Linz, J. (1994) 'Presidential or Parliamentary Democracy: Does it Make a Difference?', in J. Linz and A. Valenzuela (eds.) The Failure of Presidential Democracy, Baltimore: The Johns Hopkins University press, pp. 3–87.
26. Martins, A. (2006) 'The Portuguese semi-presidential system: about law in the books and law in action', European Constitutional Law Review, 2: 81–100. | Article |
27. Metcalf, L.K. (2000) 'Measuring presidential powers', Comparative Political Studies, 33(5): 660–685. | Article |
28. Muller, W.C. (1999) 'Austria', in R. Elgie (ed.) Semi-presidentialism in Europe, Oxford: Oxford University Press, pp. 22–47.
29. Needler, M. (1959) 'The theory of the Weimar Republic', The Review of Politics, 21: 692–698.
30. Parish, S. (1998) 'Presidential Decree Authority in Russia', in J. Carey and M.S. Shugart (eds.) Executive Decree Authority, New York: Cambridge University Press, pp. 62–103.
31. Pasquino, G. (1997) 'Semi-presidentialism: a political model at work', European Journal of Political Research, 31: 128–137.
32. Pasquino, G. (2007) 'The Advantages and Disadvantages of Semi-presidentialism: A West European Perspective', in R. Elgie and S. Moestrup (eds.) Semi-presidentialism Outside Europe, London: Routledge, pp. 14–29.
33. Pierce, R. (1991) 'The executive divided against itself: cohabitation in France, 1986–1988', Governance, 4(3): 270–294. | Article |
34. Poulard, J.V. (1990) 'The French double executive and the experience of cohabitation', Political Science Quarterly, 105(2): 243–267. | Article |
35. Rohr, J.A. (1994) 'Executive power and republican principles at the founding of the fifth Republic', Governance, 7(2): 113–134. | Article |
36. Roper, S.D. (2002) 'Are all semi-presidential regimes the same? A comparison of premier-presidential regimes', Comparative Politics, 34(2): 253–272. | Article |
37. Sartori, G. (1997) Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and Outcomes, New York: New York University Press.
38. Schleiter, P. (2003) 'Mixed constitutions and political instability: Russia 1991–1993', Democratization, 10(1): 1–26.
39. Shugart, M.S. (2005) 'Semi-presidential systems: dual executive and mixed authority patterns', French Politics, 3(3): 323–351. | Article |
40. Shugart, M.S. and Carey, J.M. (1992) Presidents and Assemblies: Constitutional Design and Electoral Dynamics, Cambridge: Cambridge University Press.
41. Skach, C. (2005a) 'Constitutional origins of dictatorship and democracy', Constitutional Political Economy, 16: 347–368. | Article |
42. Skach, C. (2005b) Borrowing Constitutional Designs: Constitutional Law in Weimar Germany and the French Fifth Republic, New Jersey: Princeton University Press.
43. Skach, C. (2007) 'The "newest" separation of powers: semi-presidentialism', International Journal of Constitutional Law, 5(1): 93–121. | Article |
44. Suleiman, E.N. (1980) 'Presidential Government in France', in R. Rose and E.N. Suleiman (eds.) Presidents and Prime Ministers, Washington, DC: American Enterprise Institute Press, pp. 94–138.
45. Stepan, A. and Suleiman, E. (1995) 'The French Fifth Public: A Model for Import? Reflections on Poland and Brazil', in H. E. Chehabi and A. Stepan (eds.) Politics, Society and Democracy: Comparative Studies, Boulder, CO: Westview Press, pp. 394–414.
46. Suleiman, E.N. (1994) 'Presidentialism and Political Stability in France', in J. Linz and A. Valenzuela (eds.) The Failure of Presidential Democrcay, Baltimore: The Johns Hopkins University Press, pp. 137–162.
47. Wu, Y-S. (2000) 'The ROC's semi-presidentialism at work: unstable compromise, not cohabitation', Issues and Studies, 36(5): 1–40.