〈Semi-Presidential Systems: Dual Executive And Mixed Authority Patterns〉Matthew Søberg Shugart

1. Introduction
1. Duverger’s (1980) original and influential definition
1. A president who is popularly elected
2. The president has considerable constitutional authority
3. There exists also a prime minister and cabinet, subject to the confidence of the assembly majority.
2. A ‘pure’ parliamentary democracy
1. Executive authority, consisting of a prime minister and cabinet, arises out of the legislative assembly
2. The executive is at all times subject to potential dismissal via a vote of ‘no confidence’ by a majority of the legislative assembly.
3. Presidential democracy
1. The executive is headed by a popularly elected president who serves as the ‘chief executive
2. The terms of the chief executive and the legislative assembly are fixed, and not subject to mutual confidence
3. The president names and directs the cabinet and has some constitutionally granted lawmaking authority
2. Hybrids: Mirrors and Mixes
1. Pure types of executive–legislative structure and their mirror hybrids
   
2. semi-presidential system-dual executive
   One portion of this dual executive — the president — has both origin and survival separated from the assembly, while the other portion — the prime minister (and cabinet) — has its survival fused with the assembly majority. As we shall see below, semi-presidential systems vary in the origin of the prime minister, but usually involve considerable separation from the assembly in this stage.The key point is that semi-presidential systems mix elements of the two pure types for each portion of a dual executive structure.
3. institutional approach and behavioral approach
   An institutional approach defines the authority patterns of the executive and assembly and how they are constitutionally related to one another.A behavioral approach, on the other hand, focuses on extra-constitutional factors such as the party system and leadership dynamics
3. Patterns of Executive and Assembly Authority Relations: Hierarchies and Transactions
1. The basic theoretical underpinning of presidentialism has its origins in The Federalist Papers
   Madison wrote in Federalist 51, the design of government ‘consists in giving to those who administer each department (i.e. branch) the necessary constitutional means and personal motives to resist encroachments of the others’
2. a neo-Madisonian
   is interested in the organization of government in terms of the hierarchical and transactional authority patterns between institutions.In a hierarchy, one institution is subordinated to another.Hierarchy is thus about vertical relationships, in that one actor is superior to another.Transactional relationships, on the other hand, are among co-equals
   
   Two institutions or actors in a transactional relationship each have independent sources of authority, and must cooperate to accomplish some task, thereby implying a horizontal juxtaposition of co-equals.The neo-Madisonian perspective, as I use it here, specifies the formal hierarchical and transactional juxtaposition of authority between constitutionally defined actors. 
4. Authority patterns in presidential and parliamentary government
   in either pure type, behavioral practice originating from the constellation of political parties may temper the institutional authority patterns
1. Hierarchical and transactional relations in the ‘pure’ forms of executive–legislative structure.
   
5. Authority patterns in semi-presidential government
1. Hierarchical and transactional relationships in the two main forms of semi-presidential systems.
   
2. subdivision
   Some presidents that appear quite powerful on paper are actually observed to exercise few powers (e.g. Austria), while others seemingly have limited formal powers, yet can be dominant political players (e.g. France). Under the rubric of semi-presidentialism, there is much variation, both institutionally and behaviorally.This is why Shugart and Carey (1992) proposed a further subdivision of the concept into premier-presidential and president-parliamentary subtypes: Under premier-presidentialism, the prime minister and cabinet are exclusively accountable to the assembly majority, while under presidentparliamentarism, the prime minister and cabinet are dually accountable to the president and the assembly majority.This distinction has not always been appreciated in the literature, and has been criticized on various terms by Sartori (1994a) and Siaroff (2003)
3. premier-presidential system
   The fact that a president in such a system cannot guarantee that his or her preferred cabinet can remain in place is both what separates these systems from pure presidential systems and is a feature that restricts the president’s real choice of prime-ministerial candidate to someone he expects to be able to command parliamentary support (or at least acquiescence).Once appointed then, a cabinet that enjoys parliamentary confidence is not subordinated to the president but to parliament, and thus the relationship between president and cabinet is strictly speaking transactional
1. In behavioral practice, the cabinet may be subordinated to the president if:
1. the president and the assembly majority come from the same side of an ideological divide
2. the president is the de facto head of his or her party
2. What an institutional definition allows us to recognize is that it is precisely the relationship of the president to the constellation of parties in the assembly that is critical for determining the extent of presidential dominance, given that, institutionally, a premier-presidential system subordinates the cabinet to the assembly
4. president-parliamentary system
    the president selects the cabinet and also retains the possibility of dismissal
   
   Nonetheless, these systems are semi-presidential because the assembly majority may dismiss the cabinet even if the president would prefer to retain it.Thus the president and assembly must engage in transactions, as shown in the figure, but unlike in pure presidential systems these transactions are not only over policy-making.They are ongoing transactions over the composition and direction of the cabinet, brought on by the dual accountability that defines the president-parliamentary subtype
6. Semi-Presidentialism in Practice
   Nonetheless, I hope to show that there is a clear divide between these two subtypes, and that this division has important behavioral and performance implications once we situate the formal features into the party-system constellations in which they operate.Before considering the role of parties in the behavior of premier-presidential and president-parliamentary systems, it would be useful to consider which actual cases fall into the respective subtypes.
7. Variations among semi-presidential constitutions
   Two questions are especially critical to the differentiation of subtypes of semi-presidentialism: Whether the president has discretion to dismiss a prime minister and cabinet,and whether the assembly is restricted in voting no confidence in a cabinet.As can be seen immediately, the systems classified as premier-presidential all have an indication of ‘no’ for both of these questions. All the president-parliamentary systems have ‘yes’ to the question of dismissal by the president, though they vary on the presence of restrictions on assembly-initiated dismissal (i.e. no confidence).
8. Premier-presidential systems
1. initiative in naming a premier
    if the president does not have initiative in naming a premier, the regime is barely distinct from parliamentarism. However, even when the president has this initiative it may be almost meaningless if there is a partisan or coalitional majority that organizes the assembly.
2. dissolve the assembly
1.  the president may do so on his or her own initiative (with restrictions on frequency)
2. dissolution power can be exercised only in response to assembly behavior, that is, a vote of no confidence or failure to invest a new government.
   Presidents can be expected to exercise this authority only when they expect the voters to take their side in the disagreement and provide a fresh mandate to an assembly more sympathetic to the president than the one dissolved
3. veto power
1. Even if the president has no discretion in the forming of cabinets or the right to dissolve parliament, his or her constitutional authority can be regarded as ‘quite considerable’ in Duverger’s sense if cabinet legislation approved in parliament can be blocked by the people’s elected agent
2. Even if the veto override requires only an absolute majority (50% +1 of the total membership), the veto may result in a transactional situation that would not occur in a parliamentary system: the inability of a government to pass a law with a plurality while some parties or members either abstain or are absent from the vote
9. President-parliamentary systems
1. restrict the assembly’s right to dismiss a cabinet
   (Georgia, Russia, Taiwan, and Ukraine (1996–2005))
1. Russia
   In Russia, for example, a no-confidence vote does not have immediate effect.If it is repeated within 3 months, then the president still has a choice: accept the resignation of the cabinet and nominate a new prime minister, or retain the cabinet but call a new election (which need not be for four months).As Morgan-Jones and Schleiter (2004) observe, the Russian configuration of powers induces bargaining tied to electoral expectations, and thereby ‘reinforces the authority of the most recently elected branch of government
2. allows the president to dissolve at any time
   Namibia allows the president to dissolve at any time but places a novel negative incentive on his exercise of the right: He must stand for a new election at the same time as the new assembly elections
10. Parties, government, and opposition in the performance of semi-presidential systems
    It is clear from the foregoing overview that semi-presidentialism is not actually a coherent category to the same degree as parliamentarism or presidentialism; that is why it is useful to consider how its further subdivisions, premierpresidential and president-parliamentary, function differently.As with any regime type, interactions among political parties, as well as formal constitutional provisions, matter for system performance
11. Conclusion
    In this conclusion, I hope to sketch a rough map of where we should be going in an effort to understand this regime type.
1. comparative presidential studies should extend beyond the Americas and include post-communist and African countries and a consideration of the potential impact of various forms of cabinet responsibility for various aspects of executive–legislative bargaining.Even better, cross-national studies that include countries of different institutional configurations should structure their analysis so as to allow tests of whether semi-presidentialism makes a difference

覺得這篇文章想要討論什麼?

用新的角度詮釋半總統制，以及細論其子類型

我覺得這篇文章有哪些重點?或是我的心得?

用階層與交換制度的概念，感覺值得深究