ROBERT ELGIE
Maurice Duverger first introduced the idea of semi-presidentialism as a regime type separate from presidentialism and parliamentarism in 1970 (Duverger 1970). He was the first person to complete a full-scale comparative work on semi-presidentialism, Echec au roi, in 1978 (Duverger 1978). He also published the first English-language article on the topic in 1980 (Duverger 1980). Even though Duverger's substantive addition to the literature on semi-presidentialism effectively ended with the publication of a book that he edited in 1986 (Duverger 1986),1 by this time the concept of semi-presidentialism was being widely debated and the comparative analysis of semi-presidential countries had already become a focus of international academic attention. The third wave of democratisation in the early 1990s further increased the interest in semi-presidentialism as the number of semi-presidential countries increased quickly and as the effects of regime types were systematically studied (for example, Linz 1994). The consolidation of many of these democracies, particularly in Central and Eastern Europe, has not diminished the interest in semi-presidentialism. Indeed, only recently one leading scholar wrote that semi-presidentialism ‘apparently is a regime type whose time has come’ (Shugart 2005: 344).
In France, the introduction of the direct election of the president, following the October 1962 constitutional amendment, established the Fifth Republic as a semi-presidential regime. This terminology is still not accepted within the French community of constitutional lawyers. However, within the international political science community France is almost unanimously acknowledged as having a semi-presidential system. More than that, France is often considered to be the archetypal example of such a system. For instance, while the list of semi-presidential countries can still vary from one writer to the next, France is always included in such a list. When the debate about the institutional effect of regime types was debated in the early 1990s, the French experience was often central to the discussion about the pros and cons of semi-presidentialism. In other words, within the political science community France remains a standard reference point for the study of semipresidentialism.
This paper examines France's claim to be the archetypal example of semi-presidentialism. Now that there are up to 60 semi-presidential countries in the world, 19 of which are in the greater European area including Russia, to what extent is France the typical example of semi-presidentialism, or even the typical example of European semi-presidentialism? In addressing this question, we reflect on Duverger's contribution to the study of semi-presidentialism. Focusing only on the experience of semi-presidentialism in Europe, we argue that semi-presidentialism is best understood as comprising an heterogeneous set of countries and that even in a European context France can at best be seen as an example of only one form of semi-presidentialism rather than as an example of semi-presidentialism as a whole. We suggest that Duverger's enduring contribution to the study of semi-presidentialism lies in his very identification of the concept and in his implicit insight that there are different types of semi-presidentialism.
Duverger's Idea of Semi-presidentialism and the Number of Semi-presidential Countries
The main contribution of Duverger to the study of semi-presidentialism lies in his formulation of semi-presidentialism as a separate regime type. Prior to his work, the parliamentary/presidential dichotomy was the long-established analytical focus of governmental systems. For example, in the early 1940s the well-known debate between Don Price (1943) and Harold Laski (1944) focused solely on the parliamentary/presidential dichotomy. Moreover, when Duverger first identified the concept of semi-presidentialism there were very few semi-presidential regimes in existence apart from France, and those that did exist were in small countries usually overlooked by political scientists at that time – Austria, Finland, Iceland and Ireland. Therefore, it is remarkable that it took Duverger less than a decade after the French shift to semi-presidentialism to conceptualise the idea coherently and to start to analyse the politics of semi-presidential countries comparatively.
While Duverger's conceptualisation of the idea was seminal, his definition of semi-presidentialism has proved to be problematic. By 1980, he had arrived at his standard definition:
[A] political regime is considered as semi-presidential if the constitution which established it combines three elements: (1) the president of the republic is elected by universal suffrage; (2) he possesses quite considerable powers; (3) he has opposite him, however, a prime minister and ministers who possess executive and governmental power and can stay in office only if the parliament does not show its opposition to them. (Duverger 1980: 166)
Elsewhere, I have demonstrated at length that the problem with this definition is the issue of what should count as ‘quite considerable’ presidential powers (Elgie 1999a; 2007). Different people make different judgement calls. As a result, the list of semi-presidential regimes varies from one person to next. This introduces the problem of selection bias when the effects of semi-presidentialism are studied. As a result, there is a tendency now to define semi-presidentialism simply as the situation where there is both a popularly elected fixed-term president and a prime minister and cabinet responsible to the legislature (see, for example, Shugart 2005; 2006). This greatly reduces the level of disagreement about the list of semi-presidential countries.
The reformulated version of Duverger's definition means that we can identify semi-presidential countries simply by reading their constitution. In a European context, there are currently up to 19 semi-presidential countries, depending on where the boundaries of Europe are set2 (see Figure 1).3 In the
FIGURE 1
EUROPEAN COUNTRIES WITH SEMI-PRESIDENTIAL CONSTITUTIONS, 2008 | Country | Year | Country | Year |
| Austria | 1945 | Montenegro | 2007 |
| Belarus | 1994 | Poland | 1990 |
| Bulgaria | 1991 | Portugal | 1976 |
| Croatia | 1990 | Romania | 1990 |
| Finland | 1919 | Russia | 1993 |
| France | 1962 | Serbia | 2007 |
| Iceland | 1944 | Slovakia | 1999 |
| Ireland | 1937 | Slovenia | 1991 |
| Lithuania | 1992 | Ukraine | 1991 |
| Macedonia | 1991 | | |
rest of this article, we compare France with 14 other European semi-presidential systems. We leave aside consideration of Belarus, because democracy collapsed almost immediately after the introduction of semi-presidentialism; Russia, because only part of the country is in the European geographical area; and both Montenegro and Serbia, because semi-presidential systems were only formalised there following the formal dissolution of the union of the two countries in 2007. In the next section, we use Duverger's standard schema for analysing the politics of semi-presidentialism to engage in an institutional mapping exercise of the 15 European countries under consideration. This mapping exercise will help to determine whether or not France can be considered the archetypal semi-presidential country.
Explaining Presidential Influence: Duverger's Three Variables
Duverger's main focus was the operation of semi-presidentialism in France. For example, in Echec au roi (Duverger 1978) over half of the book is taken up with an analysis of the French situation. However, he applied his analysis of French politics comparatively. In particular, he was concerned with the presidentialisation of political systems in semi-presidential countries and with the variation in presidential influence over time. So, we can take Duverger's dependent variable to be the systemic influence of the president. A problem, though, is that he does not provide a measure of such influence. Instead, he identifies a number of ‘modalities’, including a hegemonic presidency and a limited presidency (see, for example, Duverger 1996: 518–87). Unfortunately, these modalities are poorly specified and a contemporary researcher would be expected to identify them much more rigorously.
To explain the different modalities of presidential influence, Duverger identified three key variables: the president's constitutional powers; the founding context of the regime; and the president's relationship with the parliamentary majority. In the rest of this section, we identify these variables and compare the situation in France with the situation in the 14 other European countries under consideration to see whether France is part of a modal group or an outlier.
The President's Constitutional Powers
Duverger was one of the first academics to try to measure the president's constitutional powers. In Echec au roi, he identified 14 constitutional powers and then identified the situation for each of the six West European countries with which he was concerned (Duverger 1978: 22–3). He did not come up with a score as such, but he did offer a ranking of presidents using his indicators (see Figure 2). On the basis of Duverger's rankings, France is not an archetypal semi-presidential country. The French president is constitutionally one of the weakest in Western Europe.
FIGURE 2
DUVERGER'S 1978 RANKING OF THE PRESIDENT'S CONSTITUTIONAL POWERS IN WEST EUROPEAN SEMI-PRESIDENTIAL SYSTEMS (IN DESCENDING ORDER)
Finland
Iceland
Weimar Germany
Portugal
Austria
France
Ireland
Source: Duverger (1978: 22–3).
Since Duverger, the measurement of presidential powers has become a cottage industry. Here, we report the scores for the 15 European semi-presidential countries on the basis of two such measures. The first is the long-standing measure identified by Shugart and Carey (1992). This way of measuring presidential power has been criticised (Metcalf 2000), but it has been widely used in the academic literature and it has the advantage of being replicable.4 The second is a more rece...