Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=fwep20
West European Politics
ISSN: 0140-2382 (Print) 1743-9655 (Online) Journal homepage: http://www.tandfonline.com/loi/fwep20
Duverger, Semi-presidentialism and the Supposed
French Archetype
Robert Elgie
To cite this article: Robert Elgie (2009) Duverger, Semi-presidentialism and the Supposed French
Archetype, West European Politics, 32:2, 248-267, DOI: 10.1080/01402380802670453
To link to this article: https://doi.org/10.1080/01402380802670453
Published online: 10 Mar 2009.
Submit your article to this journal
Article views: 2403
View related articles
Citing articles: 15 View citing articles
Duverger, Semi-presidentialism and
the Supposed French Archetype
ROBERT ELGIE
The concept of semi-presidentialism was first operationalised by Maurice Duverger.
There are now 17 countries with semi-presidential constitutions in Europe. Within this
set of countries France is usually considered to be the archetypal example of semipresidentialism.
This article maps the main institutional and political features of
European semi-presidentialism on the basis of Duverger’s original three-fold schema.
The most striking feature is the diversity of practice within this set of countries. This
means that semi-presidentialism should not be operationalised as a discrete explanatory
variable. However, there are ways of systematically capturing the variation within semipresidentialism
to allow cross-national comparisons. This diversity also means that
France should not be considered as the archetypal semi-presidential country. At best,
France is an archetypal example of a particular type of semi-presidentialism. Overall,
Duverger’s main contribution to the study of semi-presidentialism was the original
identification of the concept and his implicit insight that there are different types of
semi-presidentialism. In the future, the study of semi-presidentialism would benefit from
the development of theory-driven comparative work that avoids a reliance on France as
the supposed semi-presidential archetype.
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
Correspondence Address: robert.elgie@dcu.ie
West European Politics,
Vol. 32, No. 2, 248–267, March 2009
ISSN 0140-2382 Print/1743-9655 Online ª 2009 Taylor & Francis
DOI: 10.1080/01402380802670453
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 semi-
presidentialism.
This paper examines France’s claim to be the archetypal example of semipresidentialism.
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 semipresidentialism.
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 semipresidentialism
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
Duverger, Semi-presidentialism and the Supposed French Archetype 249
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 semipresidentialism
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
250 R. Elgie
rest of this article, we compare France with 14 other European semipresidential
systems. We leave aside consideration of Belarus, because
democracy collapsed almost immediately after the introduction of semipresidentialism;
Russia, because only part of the country is in the European
geographical area; and both Montenegro and Serbia, because semipresidential
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 semipresidentialism
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 semipresidential
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.
Duverger, Semi-presidentialism and the Supposed French Archetype 251
Since Duverger, the measurement of presidential powers has become a
cottage industry. Here, we report the scores for the 15 European semipresidential
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 recent measure by Siaroff (2003) and is
based on a very different set of indicators. This measure is also contestable,5
but, again, it is replicable and it has started to be used in comparative
analyses (see Tables 1 and 2).
The measures reinforce the idea that European semi-presidentialism is
heterogenous. On the basis of the Shugart and Carey criteria, the president’s
constitutional powers range from a score of 0 in Ireland to 13 in Ukraine
prior to the 2006 reforms. In theory, the maximum score possible is 40.
However, in their comparative work the maximum score that Shugart and
Carey (1992: 155) recorded was 15 in Chile. Most presidential systems they
measured had a score of around 11–12. So, the calculations here confirm
that European semi-presidential systems range from the situation where
practice is purely parliamentary, as in Ireland, to at least one case where
practice is consistent with politics in a pure presidential regime. The modal
score is 2, suggesting that in Europe the dominant form of semipresidentialism
is parliamentary-like. However, only three of the 20 cases
recorded this score. So, again, the inherent heterogeneity of semipresidentialism
is noticeable. The average score is 5.7, suggesting some sort
of limited but not insignificant role for the president.
On the basis of the Siaroff criteria, the same heterogeneity is present. The
measures range from the lowest possible score of 1 – Austria, Iceland and
Slovenia6
– to a high of 7 of a possible 9 – France and Ukraine. The average
score is 4.3, again suggesting a limited but not insignificant presidency,
whereas the modal score is 6. Again, the spread of scores suggests that
European semi-presidentialism ranges from parliamentary-like systems to
presidential-like systems, given that 7 is the modal score for the presidential
regimes that Siaroff measures (Siaroff 2003: 296–7).
FIGURE 2
DUVERGER’S 1978 RANKING OF THE PRESIDENT’S CONSTITUTIONAL POWERS
IN WEST EUROPEAN SEMI-PRESIDENTIAL SYSTEMS (IN DESCENDING ORDER)
Source: Duverger (1978: 22–3).
252 R. Elgie
According to the Shugart and Carey criteria, France scores 5. This figure
suggests that the French president’s constitutional powers are slightly less
than the European semi-presidential average, but that they are not out of line.
Thus, the 1958 constitution would appear to establish a moderately powerful
presidency. By contrast, according to the Siaroff criteria, France scores 7,
which is considerably higher than the average score and which is consistent
with the modal score for presidential regimes. Given Duverger ranks France
as having one of the weakest presidencies in Western Europe, taken together
these rankings and scores probably tell us more about the perils of measuring
presidential powers than the extent of the powers themselves. Nonetheless, we
can conclude that only the Shugart and Carey criteria indicate that France
has anything like an archetypal semi-presidential system in terms of the
president’s constitutional powers. However, even this statement ignores
the great variation across the set of semi-presidential countries.
Overall, on the basis of constitutional powers alone, it may be
more accurate to conclude that there is no such thing as an archetypal
semi-presidential regime and that at best France is only an archetypal
TABLE 1
SHUGART AND CAREY’S MEASURES OF PRESIDENTIAL POWERS IN EUROPEAN
SEMI-PRESIDENTIAL SYSTEMS
PKV PTV DC EXL BUD REF CF CD CEN DIS TOT
Austria 0 0 0 0 0 0 1 0 0 3 4
Bulgaria 1 0 1 0 0 0 0 0 0 0 2
Croatia 1990–99 0 0 1 0 0 2 1 2 2 1 9
Croatia 2000- 0 0 1 0 0 2 1 0 2 1 7
Finland 2000- 0 0 1 0 0 0 0 0 0 0 1
France 0 1 0 0 0 0 1 0 0 3 5
Iceland 0 0 1 0 0 2 4 0 0 4 11
Ireland 0 0 0 0 0 0 0 0 0 0 0
Lithuania 1 0 1 0 0 0 1 0 2 1 6
Macedonia 1 0 0 0 0 0 1 0 0 0 2
Poland 1992–96 2 0 0 0 0 2 1 0 2 1 8
Poland 1997– 1 0 0 0 0 2 1 0 1 1 6
Portugal 1976–82 1.5 0 0 0 0 0 1 4 0 4 10.5
Portugal 1983– 1.5 0 0 0 0 0 1 2 0 3 7.5
Romania 0 0 1 0 0 4 1 0 0 1 7
Slovakia 1999–2000 0 0 0 0 0 0 1 0 0 1 2
Slovakia 2001– 1 0 0 0 0 0 1 0 0 1 3
Slovenia 0 0 1 0 0 0 1 0 1 1 4
Ukraine 1992–2005 2 0 4 0 0 2 1 4 0 0 13
Ukraine 2006– 2 0 1 0 0 2 0 0 0 1 6
Average 5.7
Mode 2
Notes: PKV ¼ Package veto; PTV ¼ Pocket veto; DC ¼ Decree powers; EXL ¼ Exclusive
introduction of legislation; BUD ¼ Budgetary powers; REF ¼ Referendum proposal; CF ¼ Cabinet
formation; CD ¼ Cabinet dismissal; CEN ¼ Cabinet censure; DIS ¼ Dissolution of
Assembly.
Sources: See Shugart and Carey (1992: 150) for scoring schema. Scores for Central and Eastern
Europe taken from Elgie and Moestrup (2008a). Scores for Western Europe calculated by
author for this article.
Duverger, Semi-presidentialism and the Supposed French Archetype 253
example of a particular form of semi-presidentialism, whether presidentiallike
on the basis of the Siaroff criteria or a mixed premier-presidential form
on the basis of the Shugart and Carey scores.
The Founding Context
In his 1980 article Duverger stressed that the founding context of semipresidentialism
helps to explain the subsequent extent of presidential
influence over the political system. He underlines the importance of the
‘combination of tradition and circumstances’ (Duverger 1980: 180) that
affect the evolution of semi-presidential regimes and he states that these
factors are important not merely at the moment of constitution building but
also in the early years of the new regime. Duverger argues that the founding
context is a major reason why the extent of the president’s constitutional
powers is a poor predictor of the president’s actual influence over the system
as a whole.
TABLE 2
SIAROFF’S MEASURES OF PRESIDENTIAL POWERS IN EUROPEAN
SEMI-PRESIDENTIAL SYSTEMS
PE CE AP CM VT EDP FP GF DL Total
Austria 1 0 0 0 0 0 0 0 0 1
Bulgaria 1 0 1 0 1 0 0 0 0 3
Croatia 1990–2000 1 0 1 1 0 1 1 1 0 6
Croatia 2001– 1 0 1 1 0 1 0 0 0 4
Finland 1919–56 1 0 1 1 1 0 1 0 0 5
Finland 1957–94 1 0 1 1 1 0 1 1 0 6
Finland 1995–99 1 0 1 1 1 0 1 0 0 5
Finland 2000– 1 0 0 0 1 0 0 0 0 2
France 1 0 1 1 1 0 1 1 1 7
Iceland 1 0 0 0 0 0 0 0 0 1
Ireland 1 0 1 0 1 0 0 0 0 3
Lithuania 1 0 1 0 1 0 1 0 0 4
Macedonia 1 0 1 0 1 0 1 0 0 4
Poland 1992–96 1 0 1 1 1 0 1 1 0 6
Poland 1997– 1 0 1 0 1 0 0 0 0 3
Portugal 1976–82 1 0 1 0 1 0 1 1 1 6
Portugal 1983– 1 0 0 0 1 0 0 0 1 3
Romania 1 1 1 1 1 0 0 0 0 5
Slovakia 1 0 0 1 0 0 0 0 0 2
Slovenia 1 0 0 0 0 0 0 0 0 1
Ukraine 1992–96 1 0 1 1 1 0 1 1 0 6
Ukraine 1997– 1 0 1 1 1 1 1 1 0 7
Average 4.1
Mode 6
Notes: PE ¼ popularly elected; CE ¼ concurrent presidential and legislative elections;
AP ¼ discretionary appointments powers; CM ¼ chairs cabinet meetings; VT ¼ veto power;
EDP ¼ long-term emergency or decree powers; FP ¼ central role in foreign policy; GF ¼ central
role in government formation; DL ¼ ability to dissolve the legislature.
Source: Scores taken from Siaroff (2003: 299–300).
254 R. Elgie
Figure 3 provides a snapshot of the founding context of European semipresidential
systems. We can see that France is slightly unusual in terms of
the context in which semi-presidentialism was adopted. Most countries
chose semi-presidentialism as part of an overall constitutional package.
Most countries also chose semi-presidentialism at the point of democratisation
and/or statehood. However, with Poland, Romania and Ukraine,
FIGURE 3
THE FOUNDING CONTEXT OF EUROPEAN SEMI-PRESIDENTIALISM
Source: Country chapters in Elgie (1999b) and Elgie and Moestrup (2008b). For Portugal, Martins (2006).
Duverger, Semi-presidentialism and the Supposed French Archetype 255
France is part of a small group of countries that adopted semipresidentialism
as a constitutional amendment rather than as a full
constitutional package. Even so, Poland, Romania and Ukraine quickly
went on to adopt full constitutions that incorporated the semi-presidential
amendment. So, France is unusual within this set of countries. In addition,
with Ireland, France is part of an even smaller group that did not adopt
semi-presidentialism at the point of democratisation and/or statehood. So,
the context in which semi-presidentialism was adopted in France is
somewhat unique in a European context. France is the only country where
semi-presidentialism was adopted as a constitutional amendment when
democracy was already established.7
Again, this suggests that France should
not be viewed as an archetypal example of European semi-presidentialism.
What difference does the founding context make? Figure 3 suggests that
Duverger was surely right that the founding context can be crucial in
determining the subsequent operation of a system. For example, in Croatia
semi-presidentialism was adopted in the context of the break-up of the former
Yugoslavia and the subsequent conflict situation. There was a new majority
party that had a strong leader. The direct election of the president was a way
of legitimising the party’s rule and the leader’s position. Unsurprisingly,
therefore, Croatia developed a strong presidency at least in the early years of
independence. By contrast, in Slovenia the threat of war was very quickly
dispersed. The incumbent president was a popular figure and was considered
likely to win a direct election. However, the president’s party did not have a
majority and the other political forces were wary of giving the presidency too
many powers in the near certainty that their candidates would lose the
presidential election. In other words, in contrast to its geographical
neighbour, very different geopolitical and party political circumstances in
Slovenia encouraged the creation of a figurehead presidency.
The equivalent situation in France is well known. The 1958 constitution
established a parliamentary system with an indirectly elected president.
However, de Gaulle was an active president in the early years of the
presidency, especially in foreign and defence policy, but not exclusively so.
Major reforms, in education and the economy for example, would not have
been presented without the support of the president and the intervention of the
president’s advisers. He appointed prime ministers who were loyal supporters
of his vision of reform. In April 1961 he invoked Article 16 and assumed
emergency powers until 30 September. On 7 October 1962, three weeks before
the referendum that approved the constitutional amendment to directly elect
the president, de Gaulle dissolved the National Assembly following the
government’s defeat in a motion of confidence. Overall, the president was the
key political actor even prior to the 1962 reform. The 1962 reform merely
confirmed existing political practice and was deliberately designed to
institutionalise the presidency’s pre-eminence over the system in the future.
Duverger is surely right to suggest that the founding context can be
crucial for determining the subsequent operation of a system, but the extent
256 R. Elgie
to which this concept is useful in the comparative study of semipresidentialism
can be questioned. Duverger presents the founding context
as an explanatory variable. He argues that it can explain why some
presidents with strong constitutional powers are very weak and vice versa,
and he applies this argument to the French case very persuasively. However,
the founding context is a catch-all term that incorporates a set of other
variables – party political, external events and so forth. Therefore, the
founding context cannot serve as an explanatory variable in the simple way
that Duverger suggests. More than that, Duverger does not specify the
conditions under which the founding context makes a difference as opposed
to those when it does not and, when it makes a difference, he fails to specify
the conditions that lead to a strong presidency despite the constitutional
situation, as opposed to those that lead to a weak presidency similarly. In
short, while the founding context seems to make a difference to the
subsequent operation of semi-presidentialism, much more work needs to be
done to unpack the notion of the founding context so as to allow us to
predict both the circumstances under which the founding context has a
major effect on subsequent events and, in that case, the circumstances under
which it leads to either a strong or a weak presidency.
The President’s Relationship with the Parliamentary Majority
While Duverger emphasises the constitutional powers of president and the
founding context of semi-presidential regimes, his main explanatory variable
for the different modalities of presidential influence is the president’s
relationship with the parliamentary majority. In his political science
textbook, Duverger outlines his argument very clearly and in the context
of West European semi-presidentialism generally: ‘In practice presidential
power depends above all on the presence or absence of a parliamentary
majority and the president’s relationship with the forces that comprise it’
(Duverger 1996: 511).8
As regards the latter variable, Duverger identifies
three basic scenarios: the president as the leader of the majority; the
president opposed to the majority; and the president as a disciplined
member of the majority. The extent of the president’s influence over the
system as a whole, he says, will vary as a function of whether or not there is
a majority and, if there is, which of these three scenarios is present
(Duverger 1996: 514–17). According to Duverger (1996: 515), when there is
no majority, the president’s constitutional powers are weakened. When the
president leads a majority, the prime minister’s constitutional powers are
weakened and the president has ‘almost absolute control’ of the legislature
(Duverger 1996: 516). When the president is opposed to the majority
(under cohabitation), the prime minister acts as a British-style head of
government but the president is still able to use any constitutional powers
granted to the institution (Duverger 1996: 517). When the president is a
disciplined member of the majority then, contrary to the situation under
Duverger, Semi-presidentialism and the Supposed French Archetype 257
cohabitation, the president cannot act against the will of the prime minister
(Duverger 1996)
Table 3 maps the frequency of these relationships in terms of three
indicators: 1) cohabitation, namely where the president and prime minister
are from different parties and where no representatives of the president’s
party are in government; 2) the situation where the president and prime
minister are from different parties but where the president’s party is part of a
coalition government; and 3) whether the government has majority or
minority support in the lower house of parliament.
On the basis of these indicators there is a clear distinction between West
European and Central and East European semi-presidential countries. The
latter have experienced more minority government on average, though this
is hardly surprising given the chaotic party systems in that region in the
early 1990s. In turn, West European semi-presidential countries have
experienced more cohabitation and more periods when the president and
prime minister have been from different coalition parties. Indeed, with the
TABLE 3
FREQUENCY OF COHABITATION AND MINORITY GOVERNMENT IN EUROPEAN
SEMI-PRESIDENTIAL COUNTRIES (% OF YEARS AS SP)
Cohabitation – president and
PM from different parties
and no representative
of president’s party
in government
President and PM
from different parties
but president’s
party represented
in government
Minority
government
Austria 11.3 56.5 6.5
Bulgaria 40.0 0.0 0.0
Croatia 0.0 0.0 37.5
Finland 19.5 43.7 26.4
France 20.5 4.5 7.0
Iceland 22.6 19.4 6.3
Ireland 24.3 7.1 32.9
Lithuania 13.3 0.0 26.7
Macedonia 6.7 0.0 0.0
Poland 35.3 5.9 29.4
Portugal 54.8 9.7 32.3
Romania 0.0 12.5 62.5
Slovakia 37.5 62.5 0.0
Slovenia 13.3 0.0 6.7
Ukraine 0.0 0.0 75.0
WE average 22.8 26.7 18.3
CEE average 16.0 6.4 28.0
Average 21.0 21.4 22.2
Notes: All calculations are from the beginning of semi-presidentialism to the end of 2006, except
Iceland minority government from 1975 to 2004.
Sources: For cohabitation and coalition, dataset primarily constructed from www.
worldstatesmen.org/. For minority government, World Bank’s Database of Political Institutions
(DPI), entry MAJ. Available at http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/
EXTRESEARCH/0,,contentMDK:20699744*pagePK:64214825*piPK:64214943*theSitePK:
469382,00.html. Data for minority government prior to 1975: Austria and France from the
country chapters in Elgie (1999b); Finland from Strom (1990); Ireland from Mitchell (2001).
258 R. Elgie
notable exception of Slovakia, it is very rare for presidents and prime
ministers to have been from different coalition parties in Central and East
European semi-presidential countries. However, this situation is mainly due
to the tendency towards non-party presidents in this region. Given that such
presidents have often had clear party backgrounds in the past, the figures for
the incidence of presidents and prime ministers from different parties in
Central and Eastern Europe is probably underestimated. In addition to
certain regional differences, there are some trends across European semipresidentialism
as a whole. For example, three countries have never
experienced cohabitation, while four countries have experienced it more
than 33 per cent of the time. So, it would appear that some countries
are more ‘cohabitation-prone’ than others. Equally, seven countries have
experienced minority governments less than 10 per cent of the time, whereas
the other eight countries have all experienced them more than 25 per cent of
the time. Clearly this difference is a function of party politics, but it also
points to ways in which different types of semi-presidential governments
might be systematically identified. Overall, though, across the set of
countries the outstanding feature of these indicators is the diversity of
European semi-presidentialism.
For its part, France has experienced cohabitation almost exactly as often
as the European average overall. However, it has experienced the situation
where the president is from one party and the prime minister is from a
coalition party much more infrequently than average. The same is true of
minority government. In the French case, the presence or absence of a
majority and the president’s relationship with the majority helps to explain
the president’s influence over the system. The difference in the president’s
weak position under cohabitation compared with the president’s dominant
position when he has appointed a prime minister with the support of a loyal
presidential majority in the legislature is very clear. In addition, the relative
weakness of the presidency in the period 1988–93 is also at least partly
explained by the absence of an absolute majority of support in the
legislature for President Mitterrand. Figure 4 updates Duhamel’s (1995)
extrapolation of Duverger’s logic to identify the various relationships in
more detail still.
In this context, the significance of the introduction of the quinquennat is
apparent. The reduction of the president’s term of office to five years and the
reordering of the electoral calendar such that National Assembly elections
now take place in the period immediately following the presidential election
means that cohabitation is less likely than before. The successful candidate
at the presidential election will have won a majority of the vote at the second
ballot and there is a strong likelihood that the new Assembly will reflect this
majority. Also, given that the presidential and parliamentary terms are now
the same length and are synchronised, then all else being equal there is no
opportunity for ‘mid-term’ elections that return an Assembly majority
opposed to the president.
Duverger, Semi-presidentialism and the Supposed French Archetype 259
That said, the ambiguities of the 2000 reform remain. The death or
premature resignation of the president may mean that the presidential and
assembly terms go out of sync. Unless they are re-synchronised by a
dissolution, the threat of mid-term cohabitation would then re-emerge. The
2000 reform also raises more general questions about the president’s power
to dissolve the Assembly. If the president dissolves the Assembly, then
unless the president resigns at the same time as the dissolution, which would
be a high-risk strategy, or unless the president dissolves the Assembly again
immediately after the next presidential election,9
the terms of the two
institutions will be out of sync and the threat of mid-term cohabitation
would re-emerge. Thus, there is a sense in which the quinquennat reform
reduces the likelihood of Assembly dissolution. Finally, even though the
synchronisation of the electoral calendars reduces the risk of cohabitation, it
does not rule it out altogether, nor does it rule out the risk of the Assembly
elections returning no overall majority for the president, again introducing
one of Duverger’s scenarios where the president is likely to be less powerful
than is usually the case. Indeed, in Romania, which is the only other
European country with synchronised elections, cohabitation has been
avoided, but minority government has been quite common (see Table 3).
Overall, Duverger’s schema of presidential/parliamentary relations helps us
to understand variations in (or modalities of) presidential influence in the
French case. That said, it is unclear how useful it is comparatively. Firstly,
FIGURE 4
DUHAMEL’S SCHEMA OF PRESIDENTIAL/PARLIAMENTARY RELATIONS IN
FRANCE
Source: Duhamel (1995: 125). Updated by the author.
260 R. Elgie
concepts such as cohabitation have no resonance in parliamentary-like
presidential countries. In Ireland, there were some tensions between the
Labour president, Mary Robinson, and the Fianna Fa´ il government during
one period of cohabitation in the early 1990s, but the circumstances were not
comparable to the French case. The same point can be made about Austria,
Iceland and Slovenia. In countries such as these, the power of the president
scarcely varies. The president is simply weak whatever the relationship
between the president and parliament. Instead, the prime minister’s influence
varies as a function of the relationship between the prime minister and
parliament. Secondly, it is unclear that minority government necessarily
weakens the president as Duverger suggests. Cindy Skach (2005) argues the
opposite. What she calls ‘divided minority government’, where ‘neither the
president nor the prime minister, nor any party or coalition, enjoys a substantive
majority in the legislature’ (Skach 2005: 17), can, she argues, ‘predictably
lead to an unstable scenario, characterised by shifting legislative coalitions and
government reshuffles, on the one hand, and continuous presidential
intervention and use of reserved powers, on the other’ (Skach 2005: 17–18).
In contrast to Duverger, Skach argues: ‘The greater the legislative immobilism,
governmental instability, and cabinet reshuffling resulting from the minority
position of the government, the more justified or pressured the president may
feel to use their powers beyond their constitutional limit, for a prolonged
period of time’ (Skach 2005: 18). This scenario may explain the power of the
Ukrainian president, despite the long periods of minority government in the
country. In short, Duverger’s variable provides a useful explanation of why
presidential power varies within France, but it is not clear that it helps to
explain the politics of semi-presidentialism generally.
What does this institutional mapping exercise tell us about European
semi-presidentialism? What does it tell us about Duverger’s comparative
analysis of semi-presidentialism given that the universe of semi-presidentialism
has increased well beyond the six West European countries with which
Duverger was acquainted?
The first point to be made is that Duverger’s schema for the study of semipresidentialism
is problematic. He fails to specify his dependent variable
clearly. His measure of the president’s constitutional powers is contestable.
His conceptualisation of the founding context lacks rigour. His conclusions
about the importance of the president’s relationship with the parliamentary
majority may apply well to France, but there are limits to its comparative
application. In short, Duverger deserves great credit for first introducing the
concept of semi-presidentialism, but, as it stands, his schema for studying
semi-presidentialism is fundamentally flawed.
The second point is that it is little use searching for an archetypal semipresidential
regime. Countries with semi-presidential constitutions operate
in very different ways. The consequence of this observation is that semipresidentialism,
as defined here, should not be used as a single explanatory
variable. For example, it would be methodologically nonsensical to compare
Duverger, Semi-presidentialism and the Supposed French Archetype 261
countries with parliamentary, presidential and semi-presidential constitutions
as regards, for example, their respective democratic performance. Even
if parliamentary and presidential countries were sufficiently homogeneous to
treat each as a single category, the same is not true for semi-presidential
countries. As Cheibub (2006) has demonstrated, any such study may tell us
something about the relative performance of parliamentarism vs. presidentialism,
but it cannot tell us anything meaningful about semi-presidentialism
as a whole. The variation within this category is simply too great.
That said, we can still operationalise semi-presidentialism, but when we
do so we need to distinguish the variation with this category. We may
distinguish between two types of semi-presidentialism: president-parliamentarism
and premier-presidentialism (Shugart 2005; 2006). We may
distinguish between three types of semi-presidentialism: parliamentary-like,
dual presidential/prime ministerial, and presidential-like semi-presidentialism
(Elgie 2005). We may use a continuous measure, such as Siaroff’s (2003)
measure of presidential power, to capture the variation within semipresidentialism.
Whatever the preferred distinction, we can then explore the
effect of institutional variation. For example, do semi-presidential countries
with president-parliamentary regimes perform worse than countries with
premier-presidential regimes as Shugart seems to suggest? Do semipresidential
countries with higher Siaroff scores perform worse than semipresidential
countries with lower scores? We may decide to explore the effect
of variation within semi-presidentialism solely within the universe of semipresidential
countries as in the above examples. Otherwise, we may wish to
compare the effects of variation within semi-presidential countries to other
types of countries. For example, we may compare the performance of
presidential countries solely with semi-presidential that are presidentparliamentary.
We may decide to compare the performance of both
presidential countries and president-parliamentary countries with premierpresidential
countries. These research strategies flow directly from the
acknowledgement of variation within semi-presidentialism (see Figure 5).
The notion of variation within semi-presidentialism is entirely consistent
with Duverger’s work. His schema for the comparative study of semipresidentialism
may be problematic, but one of his insights was the idea of
systematic variation. For instance, despite the logic of his definition and its
implication that a semi-presidential country was one where the president
must have ‘quite considerable powers’, Duverger consistently classed
countries such as Austria, Iceland and Ireland as semi-presidential, even
though political practice is undoubtedly parliamentary (Duverger 1980:
166). Writing about these countries in his political science textbook, he
states ‘these three countries have semi-presidential constitutions but operate
effectively like parliamentary systems’ (Duverger 1996: 504). In other words,
Duverger viewed countries such as Austria, Iceland and Ireland as having
parliamentary-like semi-presidential systems, whereas France had a more
presidential-like semi-presidential system.
262 R. Elgie
The idea of variation within semi-presidentialism means that there can be
no archetypal example of semi-presidentialism. So, France should no longer
be associated with any such epithet. At best, France could be a typical
example of particular type of semi-presidentialism – perhaps as an example
of the dual presidential/prime ministerial type or, in Shugart’s terminology,
as an example of premier-presidentialism – but not a typical example of
semi-presidentialism as a whole. The universe of semi-presidentialism is so
varied that no country should be considered as an archetypal semipresidential
country.
After Duverger: The Recent Study of Semi-presidentialism
and the French Case
In his work Duverger was concerned with explaining why the president’s
constitutional powers were not an accurate predictor of actual presidential
power. He was also concerned with explaining why presidential power varies
over time within one country. This is why he focused on the importance of
the president’s relationship with the parliamentary majority. Recent work
on semi-presidentialism has addressed issues that are closely related to
Duverger’s concerns. This work has also illustrated the importance of
identifying variation within semi-presidentialism. At the same time, the
recent work has had a much narrower focus than Duverger’s overarching
concern for presidential power generally. This work has also operationalised
the notion of variation within semi-presidentialism much more systematically.
Some of this work has incorporated France into the research
agenda. However, there is plenty of room for this recent work to be applied
to France much more systematically.
FIGURE 5
OPERATIONALISING THE STUDY OF SEMI-PRESIDENTIALISM
Duverger, Semi-presidentialism and the Supposed French Archetype 263
In his work, Oleh Protsyk has adopted the strategy of examining only
countries with semi-presidential constitutions and has focused on institutional
variation within semi-presidentialism to explain variation in political
outcomes. In his 2005 study, he aimed to explain why cabinet formation
outcomes vary across countries in Central and Eastern Europe (Protsyk
2005). He argues that variation is the result of whether a country has a
president-parliamentary form of semi-presidentialism or a premierpresidential
form, though he notes the importance of other variables too,
such as whether or not the party system was clientelistic and whether or not
presidential and parliamentary elections were concurrent. In his 2006 study,
he focused on presidential/prime ministerial conflict and aimed to explain
why it varied from one country to another (Protsyk 2006). Again, he focused
only on semi-presidential countries and again he hypothesised that the level
of conflict varies as a function of whether there was a presidentparliamentary
or a premier-presidential form of semi-presidentialism. In
fact, he finds that this variable did not explain the level of conflict very well,
but that other factors, such as the level of party system development, the
ideological positioning of the prime minister and the level of support for the
prime minister in the legislature were much better predictors of intraexecutive
conflict.
Both of these studies have great potential to be applied more widely and
for France to be included in the study. For example, in his 2006 work
Protsyk uses secondary sources to determine the level of presidential/prime
ministerial conflict. There are plenty of equivalent sources that could
identify periods of greater and lesser degrees of conflict in the French case.
Indeed, Duverger’s empirical work would provide a good starting point for
coverage of the first 30 years of the Fifth Republic at least. The explanatory
variables are also relevant, especially if the general level of party system
development were to be substituted for the level of party institutionalisation
as proxied perhaps by the date that French parties were most recently (re-)
formed.
Another recent study is by Neto and Strøm (2006). Their aim was to
explain why the level of non-partisan ministers varies across countries.
They hypothesised that the level of presidential powers and the prime
minister’s electoral prospects explained the variation. They captured the
variation within semi-presidentialism by using Shugart and Carey’s (1992)
measure of the president’s legislative powers. They hypothesised and found
that the greater the president’s powers, the more non-partisan ministers
were included in the cabinet. The situation in France neatly fits their
findings.
The research strategy adopted by Neto and Strøm (2006) differed from
the one adopted by Protsyk (2005; 2006) in two key respects. They used a
continuous measure to capture variation within semi-presidentialism rather
than Protsyk’s dichotomous measure. In addition, their universe was not
simply semi-presidential countries, but also parliamentary countries with an
264 R. Elgie
indirectly elected head of state. This shows how semi-presidential systems
can be compared with other types of similar systems. In her work, Margit
Tavits has also adopted this case selection strategy. She wishes to explain
why actual presidential power, rather than the president’s constitutional
power, varies across this set of countries and she identifies a number of
proxies to capture the extent of presidential power (Tavits 2008). She uses
quantitative and qualitative analysis to show that, contrary to commonsense
expectations, direct election does not account for variation in
presidential power across Europe. Instead, presidential power varies more
as a function of the party-political opportunity structure.
The works by Neto and Strøm (2006) and Tavits (2008) include France as
one of the countries in the model. Like Duverger, their work is comparative
and France is a suitable case for consideration. Unlike Duverger, though,
this work starts from a theoretical basis and then tests hypotheses on the set
of suitable cases. In Duverger’s work, there is always the sense that France
comes first. The French case is the main focus of attention and a
comparative model is then derived from the French case. This inductive
strategy is understandable, but it may be the cause of some of the problems
with Duverger’s research design. The recent deductive literature incorporating
semi-presidentialism suggests that theoretically driven models can
provide more robust cross-national explanations of the fundamental issues
with which Duverger was concerned – presidential power, presidential/
prime ministerial conflict etc. – than studies that focus on France alone or
on studies that extrapolate from the French case. In short, the study of
France and the study of semi-presidentialism generally would benefit, first,
from the incorporation of France into comparative studies and, second,
from the development of theory-driven comparative work that includes
France.
Conclusion
France is often treated as the example of semi-presidentialism par excellence.
This article has demonstrated that semi-presidential countries operate in
many different ways. Therefore, the idea of an archetypal semi-presidential
country, whether France or any other, is misplaced. That said, the concept
of semi-presidentialism captures a particular constitutional arrangement
and the identification of this concept was one of Maurice Duverger’s great
contributions to political science. (Hopefully, constitutional lawyers will
catch on soon). Moreover, within the category of countries with semipresidential
constitutions, there are identifiable types of semi-presidentialism.
There are ways in which variation within semi-presidentialism can be
captured systematically. Again, Duverger recognised this diversity and
found ways of expressing it, although his inductive approach is less
successful at explaining cross-national variation than more recent deductive
models. Although not an archetypal example of semi-presidentialism,
Duverger, Semi-presidentialism and the Supposed French Archetype 265
France is a natural case for inclusion in semi-presidential studies. The study
of semi-presidentialism would benefit from theories being tested on as wide
a range of countries as possible, including France. The inclusion of France,
and as many comparable countries as possible, in semi-presidential studies
can only make the conclusions drawn about this regime type as generalisable
as possible. This strategy will also further increase interest in Duverger’s
concept of semi-presidentialism, a concept whose time has come, not just
apparently but definitely.
Notes
1. Subsequent editions of his political science textbook updated existing work, rather than
offering a new analysis.
2. Depending on where the boundaries of Europe are drawn, then there could now be 20
European semi-presidential countries following the October 2007 referendum in Turkey that
approved the direct election of the president. The first direct election is due take place no
later than 2014 or 2012 if the reduction in the president’s mandate from seven to five years is
applied to the incumbent president rather than the first directly elected president.
3. It might also be noted that Moldova was semi-presidential until 2000 when a parliamentary
system was established.
4. For the West European countries, there was a high degree of reliability between Shugart and
Carey’s (1992: 155) scores and the author’s calculations for this article. In the French case,
we disregarded Article 40 because it does not refer specifically to a bill introduced by the
president. Given Shugart and Carey also seem to disregard this article in their scoring, we
assume that they do intend to impose this restriction.
5. The reliability of some of Siaroff’s scores must be questioned. For example, in Ireland the
president does not have the right to veto legislation, yet Siaroff’s measures indicate that this
is the case. In Ireland, the president has the right to send a bill to the Supreme Court for its
constitutionality to be judged. This is the equivalent of Article 61 of the 1958 French
Constitution rather than Article 10, which outlines the French president’s limited veto
power. So, Siaroff’s score for Ireland should be 2, not 3, bringing it rightfully back into his
category of countries where political practice is parliamentary.
6. Siaroff scores 1 for direct election. So, all semi-presidential countries will score at least 1.
7. The recent Turkish shift to semi-presidentialism is similar to the French case in both aspects.
8. Translations by the author.
9. In addition, the constraint imposed by Article 12 means that the previous dissolution must
not have taken place in the preceding 12 months in order for this strategy to be possible.
References
Cheibub, Jose A. (2006). ‘Mixed Systems and Democratic Performance: Do Popularly Elected
Presidents in Parliamentary Systems Matter?’, paper presented at the annual meeting of the
American Political Science Association, Philadelphia, 31 August–3 September.
Duhamel, Olivier (1995). ‘Pre´ sident, premier ministre, gouvernement. Les diffe´ rents cas de
figure’, in Nicholas Wahl (ed.), La France pre´sidentielle. L’influence du suffrage universel sur la
vie politique. Paris: Presses de Sciences Po, 121–37.
Duverger, Maurice (1970). Institutions politiques et Droit constitutionnel, 11th ed. Paris: Presses
Universitaires de France.
Duverger, Maurice (1978). Echec au roi. Paris: Albin Michel.
266 R. Elgie
Duverger, Maurice (1980). ‘A New Political System Model: Semi-presidential Government’,
European Journal of Political Research, 8, 165–87.
Duverger, Maurice, ed. (1986). Les re´gimes semi-pre´sidentiels. Paris: Presses Universitaires de
France.
Duverger, Maurice (1996). Le syste`me politique franc¸ais, 21st ed. Paris: PUF.
Elgie, Robert (1999a). ‘The Politics of Semi-presidentialism’, in Robert Elgie (ed.), Semipresidentialism
in Europe. Oxford: Oxford University Press, 1–21.
Elgie, Robert, ed. (1999b). Semi-presidentialism in Europe. Oxford: Oxford University Press.
Elgie, Robert (2005). ‘A Fresh Look at Semi-presidentialism. Varieties on a Theme’, Journal of
Democracy, 16:3, 98–112.
Elgie, Robert (2007). ‘What is Semi-presidentialism and Where is it Found?’, in Robert Elgie
and Sophia Moestrup (eds.), Semi-presidentialism Outside Europe. London: Routledge, 1–13.
Elgie, Robert, and Sophia Moestrup (2008a). ‘The Impact of Semi-presidentialism on the
Performance of Democracy in Central and Eastern Europe’, in Robert Elgie and Sophia
Moestrup (eds.), Semi-presidentialism in Central and Eastern Europe. Manchester: Manchester
University Press.
Elgie, Robert, and Sophia Moestrup, eds. (2008b). Semi-presidentialism in Central and Eastern
Europe. Manchester: Manchester University Press.
Laski, Harold J. (1944). ‘The Parliamentary and Presidential Systems’, Public Administration
Review, 4:4, 347–59.
Linz, Juan J. (1994). ‘Presidential or Parliamentary Democracy: Does It Make a Difference’, in
Juan J. Linz and Arturo Valenzuela (eds.), The Failure of Presidential Democracy. Baltimore,
MD: The Johns Hopkins University Press, 3–87.
Martins, Ana (2006). ‘The Portuguese Semi-Presidential System. About Law in the Books and
Law in Action’, European Constitutional Law Review, 2, 81–100.
Metcalf, Lee Kendall (2000). ‘Measuring Presidential Power’, Comparative Political Studies,
33:5, 660–85.
Mitchell, Paul (2001). ‘Divided Government in Ireland’, in Robert Elgie (ed.), Divided
Government in Comparative Perspective. Oxford: Oxford University Press, 182–208.
Neto, Octavio Amorim, and Kaare Strøm (2006). ‘Breaking the Parliamentary Chain of
Delegation: Presidents and Non-partisan Cabinet Members in European Democracies’,
British Journal of Political Science, 36, 619–43.
Price, Don K. (1943). ‘The Parliamentary and Presidential Systems’, Public Administration
Review, 3:4, 317–34.
Protsyk, Oleh (2005). ‘Prime Ministers’ Identity in Semi-Presidential Regimes: Constitutional
Norms and Cabinet Formation Outcomes’, European Journal of Political Research, 44, 721–48.
Protsyk, Oleh (2006). ‘Intra-Executive Competition between President and Prime Minister:
Patterns of Institutional Conflict and Cooperation under Semi-Presidentialism’, Political
Studies, 54, 219–44.
Shugart, Matthew Søberg (2005). ‘Semi-presidential Systems: Dual Executive and Mixed
Authority Patterns’, French Politics, 3:3, 323–51.
Shugart, Matthew Søberg (2006). ‘Comparative Executive–Legislative Relations’, in
R.A.W. Rhodes, Sarah A. Binder and Bert A. Rockman (eds.), The Oxford Handbook of
Political Institutions. Oxford: Oxford University Press, 344–65.
Shugart, Matthew S., and John M. Carey (1992). Presidents and Assemblies. Constitutional
Design and Electoral Dynamics. Cambridge: Cambridge University Press.
Siaroff, Alan (2003). ‘Comparative Presidencies: The Inadequacy of the Presidential, Semipresidential
and Parliamentary Distinction’, European Journal of Political Research, 42, 287–
312.
Skach, Cindy (2005). Borrowing Constitutional Designs: Constitutional Law in Weimar Germany
and the French Fifth Republic. Princeton, NJ: Princeton University Press.
Tavits, Margit (2008). Presidents with Prime Ministers: Do Direct Elections Matter? Oxford:
Oxford University Press.
Duverger, Semi-presidentialism and the Supposed French Archetype 267