sign in sign up
<<
Home , French Fifth Republic, Henri Emmanuelli, Politics of France

SOM01p.qxd 7/9/08 23:02 Page 41

1 2 3 dynamics and 3 4 ministerial careers in the French 5 Fifth 6 7 8 Christopher Kam and Indridi Indridason 9 0 11 12 13 Introduction 14 has operated under two constitutional regimes since the end of the Second 15 World War, the purely parliamentary Fourth Republic (1946–58) and since 16 then the semi-presidential Fifth Republic. The immediate cause of the Fourth 17 Republic’s demise was the open refusal by the French military to countenance the 18 of , but the Algerian Crisis was itself rooted in the structural 19 weakness of the Fourth Republic: the country’s deep-seated social cleavages 20 combined with the Fourth Republic’s permissive proportional representation 21 electoral system to produce legislative assemblies that were too fragmented and 22 polarized to maintain stable cabinets. The chronic cabinet instability of the Fourth 23 Republic thus left French civil servants and military officers in the regions and 24 colonies to their own devices, unconstrained by and ultimately resentful of civilian 25 authority. The crisis was averted by de Gaulle agreeing to return to 26 power, but he did so on the condition he be given a free hand to draft a new 27 constitution. This draft constitution was put to referendum on 28 September 1958, 28 and endorsed overwhelmingly by the French electorate. The Fifth Republic came 29 into force on 4 October 1958. 30 The Fifth Republic is an innovative fusion of presidential and parliamentary 31 government. The Constitution outlines a dual executive composed of a President, 32 popularly elected since 1962, and a Prime Minister who has the confidence of the 33 majority of the . Article 8 of the Constitution establishes the relation- 34 ship between the two executive officers, empowering the President to appoint the 35 Prime Minister and, on the advice of the Prime Minister, other cabinet ministers. 36 Article 8 also allows the President to terminate the Prime Minister’s appointment 37 when the latter tenders the government’s resignation.1 Similarly, the President 38 terminates the appointments of other cabinet ministers on the Prime Minister’s 39 advice. In addition to the powers of appointment and dismissal provided by Article 40 8, the President is granted the power to chair meetings of the Council of Ministers 41 (i.e. the cabinet) (Article 9) and dissolve the (Article 12). 42 Finally, the Constitution designates the President as the ‘guarantor of national 43 independence, territorial integrity and observance of treaties’ (Article 5) and 44 commander-in-chief of the armed forces (Article 15). These last two articles allow 45 the President to dominate the conduct of French foreign policy. 46 SOM01p.qxd 7/9/08 23:02 Page 42

42 The selection of ministers in The Prime Minister’s powers flow from Articles 20 and 21 (Chagnollaud and 1 Quermonne 1996: 337–8). Article 20 empowers the government ‘to determine and 2 conduct the policy of the Nation’, whilst Article 21 places the government under 3 the direction of the Prime Minister. In constitutional theory, then, the Prime 4 Minister recommends cabinet ministers to the President, sets governmental policy 5 and determines the boundaries of departmental competencies (Elgie 1993: 10). The 6 chief constraints on these Prime Ministerial powers are, first, that the ministerial 7 appointments require the President’s approval, and secondly, that ministerial office 8 is incompatible with parliamentary office (Article 23), i.e. ministers cannot sit in 9 parliament and, should they hold a parliamentary seat, must resign it before taking 0 up ministerial office.2 11 There are several classes of ministers within the government. The lowest rank 12 is comprised of junior ministers, Secrétaires d’Etat. Above Secrétaires d’Etat one 13 finds three classes of ministers, Ministres d’Etat, Ministres délégués and Ministres 14 (Chagnollaud and Quermonne 1996, 295–8).3 The first of these titles is an honorific 15 reserved for prominent or long-serving politicians, and its usage has declined over 16 time. Ministres délégués are ministers who are responsible for a specific policy 17 or function, often within the bailiwick of a larger portfolio (e.g. European Affairs 18 within the Foreign Affairs portfolio). Ministre is the more typical title, identify- 19 ing the holder as responsible for a specific portfolio. French cabinets averaged 20 36 ministers (all ranks) between 1968 and 2002 (16.5 Ministres, 4.5 Ministres 21 délégués, 2.4 Ministres d’Etat, and 12.5 Secrétaires d’Etat), but have also grown 22 slightly larger as time has passed. Indeed, while much was made of Sarkozy’s 23 intention to reduce the size of the cabinet, Fillon’s second cabinet (June 2007) grew 24 to 32 members, his first having included only 21 members. 25 26 27 The informal rules of the game in French politics 28 Thanks to the bipolar nature of the French party systems and the important 29 role of pre-electoral agreements, French are usually formed within a 30 couple of days of an election (Thiébault 2000). A strict reading of the Constitution 31 suggests that the Prime Minister controls the management and direction of the 32 cabinet. The President, after all, cannot appoint or remove ministers without the 33 Prime Minister’s initiative and, moreover, the President is required to do so upon 34 the recommendation of the Prime Minister. Things work differently in practice, 35 however. The crucial variable is whether the President’s party (or coalition) 36 controls the National Assembly. A President who enjoys the support of a parlia- 37 mentary majority is the cabinet’s dominant political figure, in charge of its 38 composition, membership and political direction. In 1976, for example, Giscard 39 d’Estaing created a post for a junior minister of industrial affairs without informing 40 the Prime Minister. Ministers have also been appointed by the President against 41 the advice of the Prime Minister (Safran 1998). For example, in 1974, Giscard 42 d’Estaing appointed Jean-Jacques Servan-Schreiber against Prime Minister 43 Chirac’s wishes (Knapp and Wright 2006). 44 Of course, the membership of the cabinet is typically the result of consultation 45 between the President and Prime Minister (Wright 1989: 86). However, these 46 SOM01p.qxd 7/9/08 23:02 Page 43

Cabinet dynamics in the 43 1 examples show that during periods of unified government it is the President who 2 has the upper hand. The premiership in these periods takes on the character of a 3 presidential deputy, and certainly Presidents at the head of unified governments 4 have not shied away from sacking their Prime Ministers. De Gaulle set the 5 precedent early on by sacking his first two Prime Ministers, Debré and Pompidou. 6 Pompidou’s case is especially illustrative in that he was sacked after just having 7 led the Gaullists to a landslide victory at the legislative elections of 1968 (Time 8 1968). Pompidou’s transgression was to hint at his succession to the 9 (Wright 1989), and his subsequent sacking reveals an important maxim of French 0 politics under unified government, namely, that Presidents dismiss Prime Ministers 11 who are either too unpopular or too popular. As Elgie (1993: 1) notes: 12 13 When things go well, the President often receives the credit. When things 14 go badly, the Prime Minister usually takes the blame. If things go very badly, 15 and the President starts to be criticised, then the Prime Minister is replaced. 16 If things go very well and the Prime Minister starts to be praised, then the 17 Prime Minister is also replaced. 18 19 Things change markedly under cohabitation, the Prime Minister becoming the 20 dominant player in the cabinet and in domestic politics more broadly, the President 21 retiring into his domaine réservé of defence and foreign affairs. Prime Ministers 22 do not have untrammelled power to appoint, dismiss or reshuffle ministers during 23 cohabitation (divided government), however. Elgie (1993: 50) notes, for example, 24 that Mitterrand vetoed Chirac’s appointment of Léotard as Minister of Defence 25 in March 1986. Now, Léotard did get into cabinet as minister of Culture and 26 Communication, and Chirac managed to replace Jacques Fournier, a Mitterrand 27 ally, with Renaud Denoix de Saint Marc, a gaullist, as head of GSG (the French 28 Secret Service). These examples indicate that even during cohabitation Prime 29 Minister Chirac’s ability to appoint and dismiss ministers was constrained 30 by President Mitterrand. Stevens (1992: 103) suggests that in these situations 31 reshuffles will take the form of a bargain between the two executives. 32 The dual executive nature of the Fifth Republic would seem to place French 33 ministers in the position of having to serve two masters, especially during periods 34 of cohabitation. The safety valve in the system is a convention of ministerial 35 autonomy – which is itself less a constitutional than a tacit recognition 36 of a particular set of incentives and constraints. First, the Fifth Republic’s dual 37 executive facilitates this sort of ministerial autonomy because it allows ministers 38 to circumvent the Prime Minister by appealing to the President directly (Elgie 39 1993: 32–3). Secondly, many French ministers have their own power bases as 40 heads of party factions (courants).4 In the Parti Socialiste (PS), for example, 41 Chevènement, Rocard, Mitterrand, Strauss-Kahn, Jospin, Fabius, Hollande, and 42 Dray were all ministers and factional chieftains (Knapp 2004: 167). Ministers have 43 also been able to amass independent political capital via the French device of cumul 44 des mandats, that is, the accumulation of other elected offices. Many deputies and 45 ministers seek mayoral or other positions to secure a power base upon which to 46 solidify their political careers.5 Thirdly, French ministers are provided with the SOM01p.qxd 7/9/08 23:02 Page 44

44 The selection of ministers in Europe institutional resources to operate independently. French ministers have cabinets 1 of advisers loyal to them personally, head highly centralized departments and are 2 less subject individually to parliamentary scrutiny than ministers in purely parlia- 3 mentary systems (Thiébault 1994: 140–1). Finally, mutual self-interest induces 4 ministers to keep to their own portfolios, a tradition labelled cloisonnement 5 or ‘compartmentalization’ (Stevens 1992: 104; Elgie 1993: 30–1). 6 7 8 Ministerial selection 9 French Prime Ministers are not restricted to choosing ministers from the 0 membership of the National Assembly, and frequently look to the civil service, 11 academia or the private sector for ministerial recruits. This flexibility means that 12 the set of people from whom French Prime Ministers select their ministers is 13 neither strictly defined nor fully observed. Indeed, one observes only the people to 14 whom the Prime Minister has offered (and who have accepted) cabinet office. This 15 selection bias complicates efforts to make accurate statements about ministerial 16 selection in France. Even if we observe a high proportion of lawyers among French 17 ministers, for example, it remains possible that the proportion of lawyers is just as 18 high in the (unobserved) set of those eligible for cabinet office, and that practising 19 law is statistically independent of cabinet membership. 20 We require some other frame of reference with which to put our data on 21 ministerial selection in context. One comparison to draw is between the Fifth 22 Republic’s first cabinet, Debré I (1959–62), and its successors. Debré I was 23 dominated by politicians who had built their careers under the parliamentary 24 institutions of the Third and Fourth , and so comparisons between Debré 25 I and later cabinets provide a sense of how the complexion of the French cabinet 26 has changed over time and of how it has responded to the transition to semi- 27 presidential government. We focus on three ministerial characteristics, academic 28 training (at the Ecole Nationale d’Administration (ENA), in particular), gender and 29 political experience prior to obtaining ministerial office. Other traits are almost 30 certainly relevant to ministerial selection, but these three characteristics afford us 31 some basis for inferring whether longitudinal changes in the make-up of the French 32 cabinet reflect the changing importance of certain selection criteria or simply 33 mirror changes in the composition of the recruitment pool. 34 Table 3.1 shows the distribution of these traits across French cabinets. The 35 figures for the Debré cabinet in the first row are followed by the figures for the 36 remaining 30 cabinets clustered in two ways. First, we group the governments 37 by periods of left or right government. Secondly, to provide a more balanced 38 picture of long-term development, we group governments into 15-year periods. To 39 maintain comparability across cabinets and to avoid double-counting individuals, 40 we confine our sample to the initial membership of each cabinet and record data 41 only for an individual’s first ministerial office.6 42 43 44 45 46 SOM01p.qxd 7/9/08 23:02 Page 45 . e e h e t c h l

1 i t n b w e f a N i t o 2 r d e e s e h p t r 0 1 4 7 6 1 0 3 0 1 7 9 t s e x

3 i 6 2 3 6 5 4 9 9 2 3 3 e l b n i 2 e m e t v f s 4 e a e o l h M h e t o 2 5 7 9 7 0 0 8 0 0 5 4 n s o t ...... s a s o t l 4 9 7 1 0 2 0 1 5 9 5 0 a h a e 1 1 N 1 1 2 2 1 2 6 e g c r y i e n r a e h

7 i m r m w n e o e o , r p c i s f

8 y t x n t c i r 9 8 9 5 4 9 3 2 2 7 1 e i e r . . . s ...... t i e i a r e 3 o 4 2 4 5 4 3 3 0 3 2 1 l

9 h e t v o P p 1 p o a p x n ) 0 h l e a s a r g l l e c e n a a n t 11 i o c i l s i o n r i t i t i e a n n n l t i e

12 i i o s m n i d p i e e m 2 8 8 5 1 8 9 6 0 9 7 1 e n r l c ...... b 13 t i l a n s 9 3 1 7 7 7 4 8 5 2 2 1 s a a c e e r i 2 1 M c 1 1 1 i t o e n r l i 14 t e s e n i g r p i n n a x i s i

15 f e c v y o m i a e e t r s f i h h

16 ) v t l e o l o r a s l t o g a l a o h e e a i p a n

17 y c g t h d c o l t o a a n e n n i s t t a l e 4 7 0 8 4 9 7 0 2 4 o e n u g m n ...... i 18 c a s c e e r o u 5 0 7 3 1 4 4 9 5 3 5 0 o r r s c r e y o r 1 e e L ( o C a c i p

19 e p r x p ( e m o r e l s g a 20 l n l r y e a t r o e a m a t h m c a u c s i t 21 s l i t w r n e i o n a l s e e c i o ) s e o

22 e r s m y o h s m p i t t o h a u h t t i r u h a t 23 d l , a c o p c 8 2 0 1 0 3 0 4 0 6 8 6 e r . . . . i ...... t y e n e n n l a r i c 5 1 0 3 5 3 1 3 5 1 3 5 b r e e 24 d u r a P 6 P s ( 6 5 7 7 6 6 6 7 5 8 5 r d l 0 t i F e 0 s t

25 f m 1 n n i o o n u o S c e t o e 26 . c e e c m 7 5 0 8 0 6 8 0 1 9 3 p ...... v c s n o u a r 2 0 6 3 1 0 4 4 5 6 8 3 n

27 e h e e i d 1 2 1 1 2 1 1 3 W t i r d r s e e i a 28 e w n p p t i x l e x o 5 6 6 4 1 0 4 0 9 3 1 d ...... s m e P e a n

29 9 a n 6 8 0 6 6 . 0 0 5 2 4 1 l n ; E y b o u I 1 2 2 1 1 4 1 2 1 1 2 1 r e a o l d i a c o c

30 t t a i r n s t n c a e i g i e e i l c r k r g A 9 6 4 5 1 6 4 0 0 5 31 2 e o u m e a . . c ...... t a p d N p m 8 a 0 3 2 7 5 0 0 9 3 2 i n l x u E E 1 b 0 2 1 1 1 2 2 5 2 1 2 e

32 d r e n c a n r p y d a e r

33 e p r y a 1 6 6 1 5 8 3 x t 5 e n e 2 8 7 8 9 0 8 9 7 fi n a 0 ...... d s 6 . 9 34 2 e t a e . 5 8 3 5 5 5 3 . n 5 ...... 0 u h m e 4 . 6 . e . 3 5 0 5 1 0 9 . o r a T g 1 35 5 i h 4 1 2 2 1 3 1 2 2 . u l . t , 3 s r i 1 – – – – – – – – 6 e d – a n s – 9 2 1 6 1 6 8 3 –

36 w n r e p 9 2 5 6 8 7 9 8 8 9 7 s e e u ...... d 5 0 e 9 e s v o . . 4 4 5 8 5 3 5 3 . t r n e ...... 37 . o r o 1 5 6 a a r h . . 4 4 1 7 5 0 0 9 . f g t g e e 9 D 6 1 1 2 2 1 2 1 2 2 k r

38 e n h c i e r t t a a h t n t b s o

39 I I u I c l s I n I o i i I I y a t c t I i I c o n 40 l u y I n I e . a e r o v i o l b I t r r i s r o I é p b i o r n , t a g u l

41 s u p e h a g a i f d r e a o p f é c t r b e m C B c r I d a a u a s u a a I o i 42 é c i J l c – – I d d l R F v i r t B o I I n E – d o – s a i – – a

43 R e v – I P u u I I e I a t d 1 I s I o o r c – . n e a I t n n I o y n d d u a i 3 n o I

44 e d i i i o c e r r e o T d n r é i n p e p l r s b i a a e b p : r i a r a l r f r s a l a u e o e p b c i b f r 45 l b t m m t c e r e a s p e a a a r h o a o a o o o u h x s o E T D R T N C e P P a T M B C C R B 46 J SOM01p.qxd 7/9/08 23:02 Page 46

46 The selection of ministers in Europe Ministerial selection and the ENA 1 2 The ENA is one of several Grandes Ecoles which serve as training grounds for 3 France’s administrative and political elite. What sets the ENA apart from its 4 competitors and what we find useful, however, is the fact that the ENA was 5 founded only in 1945; hence its first cohort of 100-odd graduates (known as 6 énarques) was just entering the ministerial recruitment pool at the advent of the 7 Fifth Republic.7 In contrast, many of the other Grandes Ecoles are over a century 8 old, and so the number of their graduates in the ministerial recruitment pool can 9 be assumed to have been stable throughout the Fifth Republic. Using one of these 0 schools as a comparative base (we chose the Institut d’Etudes Politiques de 11 (IEP) founded in 1872) allows us to assess just how quickly the énarques 12 penetrated the cabinet and how relatively advantageous an ENA has 13 been to ministerial aspirants. 14 It took some time for a critical mass of énarques to build up in the recruitment 15 pool and enter the cabinet. Indeed, no énarque held a portfolio in Debré’s initial 16 cabinet, though one, Giscard d’Estaing, was the junior finance minister and was 17 soon promoted.8 By the next wave of cabinets, however, 12.5 per cent of ministers 18 were énarques, almost the same percentage as IEP graduates. Even so, it would 19 take another 10–12 years for the first cohort of énarques to reach a retirement age 20 of 65 and for their number in the recruitment pool to stop growing. This suggests 21 that énarques enjoyed a success rate in securing ministerial office in the 1970s that 22 was out of all proportion to their numbers in the recruitment pool. The fact that 23 énarques and IEP graduates remained in parity in the ministry in the 1990s 24 indicates that this advantage has subsided, though given the ENA’s small size 25 (producing less than a fifth of the number of graduates as the IEP) it remains very 26 large.9 27 28 Ministerial selection and gender 29 30 Article 23 makes ministerial office incompatible with parliamentary office, but in 31 doing so it frees French Prime Ministers to look outside parliament for their 32 ministers. This removes two barriers to ministerial office that operated under the 33 Fourth Republic: the need for ministerial aspirants to be selected as a candidate 34 for parliamentary office and to win a seat in the National Assembly. These barriers 35 have proved difficult for French women to overcome, and so one might expect 36 women’s cabinet representation to have outstripped their parliamentary represen- 37 tation in the Fifth Republic.10 This did not occur. Debré’s initial cabinet contained 38 no female ministers, with Nafissa Sid Cara joining the cabinet only as a junior 39 minister via a reshuffle. The situation remained largely unchanged through the 40 1970s and 1980s. The breakthrough (relative to past levels of female cabinet and 41 parliamentary representation) came in 1991 with the elevation of Edith Cresson to 42 the premiership, the first woman to occupy the post. Even so, only 24.8 per cent 43 of new ministers were women over the period 1991–2005 and the percentage 44 of women in the French cabinet remains below parity, achieving a maximum of 45 34.4 per cent of the posts in Fillon’s present cabinet. 46 SOM01p.qxd 7/9/08 23:02 Page 47

Cabinet dynamics in the French Fifth Republic 47 1 Ministerial selection and political experience 2 3 One reason why women’s cabinet representation remains anchored to their lower 4 levels of parliamentary representation is that the recruitment pool for ministers 5 remains dominated by parliamentarians. Article 23 notwithstanding, 60–75 per 6 cent of French ministers have prior experience as a deputy or senator. This 7 far exceeds the proportion of ministers who arrive only with experience in 8 local politics, as adviser in a ministerial cabinet or after having held an extra- 9 parliamentary party position. Of course, these positions are not mutually exclusive, 0 and many ministers with parliamentary experience were also mayors, cabinet 11 advisers or party executives. This fact hardly dilutes the essential message of these 12 data: parliamentarians claim two of every three ministerial positions. As Gaxie 13 noted in reaching a similar conclusion to ours, winning a national election (i.e. 14 a seat in the National Assembly) is almost a necessity for advancement to the 15 cabinet in the Fifth Republic (Gaxie 1986: 66). In addition, ministers who come 16 to the cabinet from politics (rather than from the civil service) also tend to secure 17 more politically important posts (Gaxie 1986: 68–9). 18 The Fillon cabinets deserve a special mention with regard to political experi- 19 ence. Fillon’s first cabinet included among others a former socialist minister, 20 , as well as Hervé Morin of the Union pour la Démocratie 21 Française (UDF) who had lent his support to Bayrou on the first ballot in the 22 preceding presidential election. The composition of the first Fillon cabinet was 23 perceived to be a strategic move to split the opposition and neutralize arguments 24 that Sarkozy’s right-wing agenda needed to be balanced by a left-leaning legis- 25 lature in the upcoming legislative election. 26 27 Ministerial career patterns 28 29 Taking ministerial tenure as a metric of regime stability, Huber and Martinez- 30 Gallardo (2004) argue that the standard view of a chronically unstable Fourth 31 Republic being replaced by a stable Fifth Republic is overdrawn. This is not 32 because ministers in the Fourth Republic managed to accumulate more experience 33 than one might expect given the regime’s high level of cabinet instability, but 34 because the Fifth Republic’s stability has not extended to ministerial longevity. 35 Our data echo this message, but there are two points on which we diverge. First, 36 whereas Huber and Martinez-Gallardo suggest that the initial surge in ministerial 37 experience and stability in the early years of the Fifth Republic has declined 38 over time, our data show no such trend. Secondly, our data indicate that the Fifth 39 Republic has succeeded in replacing the statesmen of the Fourth Republic with its 40 own stable core of ministrables. 41 Huber and Martinez-Gallardo argue that the relationship between ministers’ 42 portfolio experience (time spent in a specific office) and ministers’ cabinet 43 experience (time spent in any cabinet post) provides important information on the 44 cabinet’s dynamics. Low levels of portfolio experience combined with high levels 45 of cabinet experience, for example, suggest a cabinet of musical chairs with core 46 ministrables regularly exchanging portfolios. The distinction between portfolio SOM01p.qxd 7/9/08 23:02 Page 48

48 The selection of ministers in Europe experience and cabinet experience is a useful one, but in focusing on just the 1 ten most prominent portfolios Huber and Martinez-Gallardo ignore the experience 2 that ministers accumulate as they climb up the ministerial pecking order. A full 3 picture of ministerial careers requires that one tracks both the lateral move- 4 ment of ministers across portfolios and their vertical movement in the ministerial 5 hierarchy. We use a three-tiered ranking of ministerial positions to track the 6 vertical movement of ministers. The lowest rank is comprised of junior ministers, 7 Secrétaires d’Etat. For the ranks above Secrétaires d’Etat we use Warwick and 8 Druckman’s (2006) portfolio salience measures to divide portfolios into two sets, 9 one containing the twelve most salient ‘major’ portfolios, the other the remaining 0 ‘minor’ portfolios.11 As Ministres d’Etat tend to hold major portfolios, we count 11 them as a major portfolio in cases where the title appears without a portfolio. 12 Similar reasoning leads us to count Ministres délégués as minor portfolios unless 13 there are clear indications to the contrary (e.g. as Ministre délégué 14 auprès du ministre de l’économie et des finances, chargé du budget in Mauroy’s 15 first cabinet). This leaves us with three ranks of ministers: junior ministers, minor 16 ministers and major ministers. 17 With this ranking system in place we can measure first the time spent in any 18 specific post, secondly the time spent at a given rank, and thirdly the overall length 19 of ministers’ careers, from the time they first enter the government until they depart 20 from their final posts. We present these data in Table 3.2, broken down by decade 21 over the course of the Fifth Republic so as to be comparable with Huber and 22 Martinez-Gallardo’s data. The average number of months ministers spend in any 23 specific office is between 10 and 15, with no evidence of systematic increases or 24 decreases during the Fifth Republic, not even with the alternance of 1981. Much 25 the same is true of ministerial careers, which range only between four and five 26 years on average. One pattern that we observe in the data is that ministers spend 27 about twice as long at a given rank as they do in any specific post, and further that 28 their careers span twice as long again as the average time they spend at a given 29 rank. This pattern suggests that a typical ministerial career in the Fifth Republic 30 involves a minister gaining 24–30 months’ experience in two different posts before 31 advancing in rank to spend an additional 24–30 months at two more posts. 32 This picture of a typical ministerial career is somewhat misleading, however. 33 Notice that the overall means in Table 3.2 are high relative to the decade-by-decade 34 means. This suggests the presence of a coterie of experienced ministers scattered 35 36 37 Table 3.2 Average time in the French government 38 Decade Months in post Months in rank Months in government 39 40 1960s 14.3 26.6 57.2 41 1970s 10.2 24.6 48.9 42 1980s 12.6 25.0 51.8 1990s 12.8 26.7 64.6 43 2000s 13.4 22.0 54.0 44 45 Overall 15.3 28.2 58.3 46 SOM01p.qxd 7/9/08 23:02 Page 49

Cabinet dynamics in the French Fifth Republic 49 1 amongst a large group of ministerial transients; only when this long-lived coterie 2 is collected together, not dispersed across the decade-by-decade subgroups, does 3 it noticeably affect the overall means. This view is buttressed by evidence set out 4 in Figure 3.1, which shows the number of posts held by ministers through their 5 careers. Most ministers hold only a single post, and the median number of posts 6 held is just two, but about 10 per cent of ministers hold five or more posts in their 7 careers. A similar picture emerges from Figure 3.2, which shows the length 8 of ministerial careers in the Fifth Republic. The career length distribution is highly 9 skewed to right with most careers lasting less than 50 months. Nevertheless, 0 some 20 per cent of ministers have careers that are at least twice as long. The 11 ministers in this class are among the most prominent figures of modern French 12 politics: Giscard d’Estaing, Chirac, Faure, Raffarin, Rocard, Bérégovoy, Couve 13 de Murville, Cresson, Juppé, Sarkozy, and Fabius to name a few. 14 It is worth considering the career path of one of these long-lived ministers in 15 detail. ’s career serves as a good example both because Chirac built 16 his career entirely within the confines of the Fifth Republic and because he served 17 in so many capacities.12 Chirac’s ministerial career began on 6 April 1967 with his 18 appointment as Secrétaire d’Etat and it was only after his fourth stint as a junior 19 minister that he was promoted to Ministre délégué. Eighteen months later, in 1972, 20 Chirac was given his first major portfolio, becoming Ministre de l’agriculture 21 et du développement rural. Chirac then served briefly as Ministre de l’intérieur in 22 1974 before securing the premiership itself. 23 On resigning the premiership on 25 August 1976, Chirac’s career trajectory 24 takes a turn that would be considered unusual in many . Chirac 25 26 40 27 28 29 30 30 31 32 33 t n

34 e c 20 r e

35 P 36 37

38 10 39 40 41 42 0 43 1 2 3 4 5 6 7 8 9 10 11 44 N posts held over career 45 46 Figure 3.1 Number of offices held by Ministers in the French Fifth Republic SOM01p.qxd 7/9/08 23:02 Page 50

50 The selection of ministers in Europe

40 1 2 3 4 30 5 6 7 8 t n

e 9 c 20 r e

P 0 11 12

10 13 14 15 16

0 17 0 50 100 150 200 250 300350 400 450 18 Months in government 19 20 Figure 3.2 The length of ministerial careers in the French Fifth Republic 21 22 returned to parliament, secured the chairmanship of the Rassemblement pour 23 la République (RPR), won and gave up a seat in the European Parliament, and 24 become of Paris (an office he retained until 1995). When he returned to 25 government on 20 March 1986, Chirac returned as Prime Minister. Chirac’s second 26 term as (and France’s first experience with cohabitation) lasted two years 27 until socialist electoral gains forced him out of office. Even this was not the end 28 of Chirac’s political career: Seven years later, on 7 May 1995, he was elected to 29 the presidency. What we find instructive here, and quite unique to French politics, 30 is the capacity of a small subset of French ministers to build up and fall back on 31 alternative bases of political power so that they can leave and return to ministerial 32 office, repeatedly and over long periods of time. 33 How typical is Chirac’s career? How many ministers fall by the wayside after 34 holding just one or two government posts for every Jacques Chirac who manages 35 to move from portfolio to portfolio while climbing steadily up the ministerial 36 pecking order? Table 3.3 helps us answer this question. For each minister we know 37 what position the minister held just previously (i.e. p–1) and what position he held 38 next (i.e. p+1). Cross-tabulating every minister’s past and future positions as in 39 Table 3.3 thus provides an overall sense of ministerial advancement and mobility 40 in the Fifth Republic.13 The bottom row of Table 3.3 indicates that any given point 41 in time we can expect 36.9 per cent of French ministers to be exiting the 42 government, 17.4 per cent to be acting as Secrétaires d’Etat, 24.1 percent to be 43 holding minor portfolios, and 21.6 per cent to be holding major portfolios. Three 44 things stand out in Table 3.3. First, between 35 and 40 per cent of all ministers are 45 out of government within two periods, and, except for ministers in major portfolios, 46 SOM01p.qxd 7/9/08 23:02 Page 51

Cabinet dynamics in the French Fifth Republic 51 1 Table 3.3 Career advancement and mobility among French ministers. Vertical axis: 2 position at p – 1. Horizontal axis: position at p + 1. Cell entries are row 3 percentages, with the raw numbers shown in parentheses 4 Secrétaire Minor Major 5 Outside d’Etat portfolio portfolio N 6 7 Outside 36.7 22.2 26.7 14.3 490 8 (181) (108) (131) (70) 9 Secrétaire d’Etat 37.4 34.9 17.6 10.1 278 0 (104) (97) (49) (28) 11 Minor portfolio 38.2 5.2 32.3 24.3 288 12 (110) (15) (93) (70) 13 Major portfolio 35.3 2.1 16.2 46.5 241 14 (85) (5) (39) (112) 15 Total 36.9 17.4 24.1 21.6 1297 16 (479) (226) (312) (280) 17 18 exit from the government is the most likely destination. Indeed, 14 per cent of all 19 French ministers (181 of 1,297) are out of government after serving in a just single 20 post. Secondly, alongside this picture of transience is one of stability: Ministers 21 who still hold a post (i.e. p+1 is not ‘outside’) are most likely to be in a position 22 of the same rank. This is especially true of holders of major portfolios, 46.5 per 23 cent of whom retain a high-level cabinet position (though perhaps not the same 24 one). Thirdly, fewer than 25 per cent of junior ministers or minor portfolio holders 25 at any given point in time move on to a higher office. The corollary of most 26 ministers exiting or remaining in place is that only a minority of French ministers 27 advances in office. 28 29 30 Departing from government 31 The short spells that French ministers spend in any specific post are largely due to 32 frequent government terminations. What tends to happen is that every 18 months 33 (on average) the Prime Minister brings his or her government to a formal end. 34 Sometimes the Prime Minister stays in place, and the government is remade within 35 days with many of the same ministers reappointed. On other occasions, however, 36 the Prime Minister is replaced and a wholly new cabinet is formed. This process 37 is often triggered by election results of some sort, and between presidential, parlia- 38 mentary, cantonal, municipal, and European elections there is almost always an 39 election for a French Prime Minister to respond to. Some recent examples show 40 the range of conditions under which this process of government termination and 41 reformation unfolds: 42 1. The Socialist victory at the June 1997 parliamentary elections. In 1997, 43 Chirac dissolved the National Assembly to trigger early parliamentary elections. 44 Chirac’s gambit backfired, however, and the PS and their allies won the election. 45 Faced with a hostile parliamentary majority, Chirac was forced to accept Alain 46 Juppé’s resignation and install as the new Prime Minister. SOM01p.qxd 7/9/08 23:02 Page 52

52 The selection of ministers in Europe 2. The Presidential and parliamentary elections of May–June 2002. Following 1 his re-election and Jospin’s resignation, Chirac appointed an interim government 2 led by his own party’s Jean-Pierre Raffarin. Notwithstanding his party’s success 3 in the legislative election the following month, Raffarin tendered his government’s 4 resignation, but Chirac immediately reappointed Raffarin and the government was 5 reformed. 6 3. The regional elections of March 2004. The Union pour un Mouvement 7 Populaire (UMP) suffered a significant setback at the March 2004 regional 8 elections. Raffarin responded by tendering his resignation, but again Chirac 9 reappointed Raffarin to the premiership and the government was reformed, though 0 with many personnel changes. 11 4. The referendum to ratify the EU constitution in May 2005. Chirac and the 12 government championed the ratification of the EU constitution and, following its 13 defeat, Raffarin tendered his resignation for a third time. This time Chirac accepted 14 and appointed as Prime Minister. 15 The first and fourth cases involve the constitutional replacement of one govern- 16 ment by another. The second and third cases, on the other hand, are effectively 17 cabinet reshuffles in the French style. In all these cases, however, ministers’ 18 appointments are formally terminated. Consequently, it is not surprising to learn 19 that, of the 1,307 ministerial posts in our dataset, 1,131 (86 per cent) were ended 20 by a government termination. A more accurate account of ministerial attrition in 21 France requires tracking how many ministers are reappointed to the government 22 after these terminations. We do this in Table 3.4 which lists the six ways in which 23 the ministerial appointments in our data set end: A cabinet termination – followed 24 by reappointment or not, a reshuffle from one post to another exclusive of those 25 that take place as a result of termination, the end of an interim appointment, 26 resignation and death. Most ministers (54.2 per cent) are reappointed following a 27 government termination, with two-thirds of these reappointments seeing the 28 minister take up a different office in the government. Another 8 per cent of cases 29 are shuffled independently of a cabinet termination. A further 32.4 percent of 30 ministers are not reappointed. In many of these cases, the ministers are effectively 31 sacked, but in others the ministers are not reappointed because their govern- 32 ment has suffered an electoral defeat (see example 1 above). The remaining 33 appointments are ended because their holder dies, has completed an interim 34 appointment or resigns. 35 36 Table 3.4 Reasons for the end of French ministerial appointments 37 38 N % 39 40 Cabinet termination without reappointment 423 32.4 Cabinet termination followed by reappointment 708 54.2 41 Reshuffle 105 8.0 42 Resignation 60 4.6 43 End of interim appointment 6 0.5 44 Death 5 0.4 45 Total 1307 46 SOM01p.qxd 7/9/08 23:02 Page 53

Cabinet dynamics in the French Fifth Republic 53 1 Reasons for resignation 2 Resignations are only a small fraction (4.6 per cent) of all ministerial exits, but 3 because they are often connected to important political events they deserve 4 attention. Policy differences are one reason for resignation, and several cases can 5 be cited, e.g. Pierre Sudreau’s resignation as Minister of Education in 1962 in 6 protest against the direct election of the President and Jean-Pierre Chevènement’s 7 resignations, first as Minister of Research and Industry in 1983 in protest against 8 Mitterrand’s changed course on economic policy and, then, from his post as 9 Minister of Defence in 1991 in protest over the Gulf War. A resignation on the 0 grounds of policy disagreement is also a means by which ministers distance 11 themselves from the government and establish themselves as independent electoral 12 actors. Rocard’s resignation from Fabius’s government in April 1985 is one 13 example of this sort of resignation, as he intended to be a Presidential candidate 14 for the left had Mitterrand retired. 15 Ministers are also forced from office by policy failures, though again there is a 16 significant difference in the way these events are handled in France as opposed to 17 Britain. In France, Article 68-1 of the Constitution makes ministers criminally 18 liable for their actions in office. Article 68-1 has only been invoked once to date, 19 in 1999 when Laurent Fabius, Georgina Dufoix and Edmond Hervé were charged 20 with manslaughter on account of having failed in their official capacities to protect 21 France’s blood supply from HIV contamination in the mid-1980s.14 All three were 22 out of office by the time the trial occurred, and so one can only conjecture as to 23 how the process might have affected their ministerial tenures; Hervé and Dufoix 24 have not held ministerial office since. 25 Resignations have also been guided by informal rules adopted by government 26 leaders. Bérégovoy initiated the principle that ministers placed under investigation 27 were required to resign from their post, and this principle has remained in place 28 with Bernard Tapie resigning in 1992, Alain Carignon, Gérard Longuet and Michel 29 Roussin in 1994, Dominique Strauss-Kahn in 1999 and Renaud Donnedieu de 30 Vabres in 2002.15 Other prime ministers, e.g. Fillon, have insisted that ministers 31 who run for election resign if they fail to win a seat. Alain Juppé, minister of the 32 environment, transportation and energy and number two in Fillon’s cabinet, was 33 forced to resign after losing the Bordeaux constituency in 2007. 34 Personal scandals provide a third ground for resignation, and on this front we 35 do not lack for examples. Financial scandals are especially frequent in France, 36 and a party-financing scandal in the 1990s that encompassed almost every polit- 37 ical party suggests that some of the problems are systemic rather than personal 38 in nature.16 It is not just the number of ministers allegedly involved in corrupt 39 activities that is startling, but their stature: Dominique Strauss-Kahn, a former 40 finance minister, Alain Juppé and Pierre Bérégovoy, former , and Chirac, 41 himself, have all been accused of financial wrong-doing. One should not come 42 away thinking that French ministerial careers are especially vulnerable to personal 43 scandals, however. They are, in fact, fairly resistant: Of the 39 scandals we 44 documented involving sitting ministers, only 13 led to resignation.17 45 46 SOM01p.qxd 7/9/08 23:02 Page 54

54 The selection of ministers in Europe Conclusions 1 2 The Fifth Republic is a hybrid of parliamentary and presidential government. The 3 introduction of a dual executive has certainly changed French cabinet dynamics. 4 This is especially noticeable during periods of unified government, when the Prime 5 Minister and the government are driven by the President’s political agenda. Thus, 6 we observe prime ministers handing in their resignations and their governments 7 terminated to protect the President. In other ways, however, the Fifth Republic has 8 left longstanding cabinet dynamics intact. French ministerial careers have 9 remained stubbornly parliamentary in nature, a repeatedly vacated and reoccupied 0 seat in the National Assembly (or failing that, in the ) remaining the primary 11 power base of most French ministers. And French ministers, at least those who 12 hope to become part of the regime’s core set of ministrables, must develop 13 alternative power bases. French ministerial careers are not long, but they are 14 flexible and the government permeable; provided a ministrable remains politically 15 active, in parliamentary, municipal or European politics, a French minister can 16 repeatedly enter and exit government. 17 18 Notes 19 1 An English translation of the French Constitution of 4 October 1958 can be obtained 20 online from the National Assembly’s website, http://www.assemblee-nationale.fr/ 21 english/8ab.asp (last visited 9 March 2007). 22 2 Zarka (2000) is a useful guide to the constitutional arrangements and norms of 23 France’s semi-presidential regime. 3 Garde des sceaux (Keeper of the Seals) is another ministerial title, but one that is held 24 alongside the justice portfolio. Debré’s initial cabinet also listed four ministres 25 conseillers, all of whom were indigenous political leaders of major French colonies. 26 Once these colonies became independent, the title disappeared. We exclude these 27 ministres conseillers from our analysis. 28 4 This is a more recent development; prior to Mitterrand’s presidency ministers tended to emerge from the civil service rather than from party politics (Thiébault 1994). 29 About 35 per cent of the ministers since 1959 have had background in the civil service 30 (Lacam 2000: 115). 31 5 Prime ministers Mauroy (Lille in ), Mollet (Arras in Pas-de-Calais) and Fabius 32 (Grand-Quevilly in Seine-Maritime) all held mayorships, and among ministers Deffere 33 ( in Bouches-du-Rhône), Soldani (Draguignan in ), and Chevènement (Belfort) held mayorships. Article 23 forbids members of the government from 34 simultaneously holding another public office, so ministers must resign from any other 35 such office before assuming government office. 36 6 At this stage of the analysis we exclude junior ministers. 37 7 If the first cohort of énarques graduated in 1948 (coursework at the ENA takes three 38 years) at about 25 years of age, they would have required some 10 years of profes- sional and political seasoning before they were eligible for ministerial office. (Only 39 two ministers out of the 314 in our full sample assumed office before they were 35; 40 Jacques Soustelle and François Mitterrand were 33 and 31 years old, respectively, on 41 first assuming ministerial office.) This would mean that the first cohort of énarques 42 would have been eligible for ministerial office in 1958. Although our sample contains 43 ministers during the Fifth Republic, some ministers may have first entered office during the Fourth Republic. 44 8 Giscard d’Estaing was also an alumnus of the Ecole Polytechnique, another of the 45 established Grandes Ecoles, however. 46 SOM01p.qxd 7/9/08 23:02 Page 55

Cabinet dynamics in the French Fifth Republic 55 1 9 In the absence of exact numbers on the number of graduates from each school we 2 cannot say precisely how much more likely an énarque was to be selected as a minister 3 than an IEP graduate, and we should keep in mind that over time the number of ministers who have attended both schools has also grown. 4 10 Matland’s (1998) data show that only 4.3 per cent of National Assembly deputies were 5 women in 1980, and only 10.9 per cent in 1997. As of October 2007, 13.3 per cent of 6 French deputies were women (http://www.assemblee-nationale.fr/12/tribun/gs2.asp#P 7 -1_0 (visited 7 October 2007)). 8 11 The twelve major portfolios are: the premiership, finance, interior, justice, foreign affairs, education, defence, budget, agriculture, labour, industry and planning. 9 12 An English translation of Chirac’s biography can be found at http://www.elysee.fr/ 0 elysee/anglais/the_President/biography/biography.39706.html (visited 20 March 2007). 11 13 It may seem counter-intuitive not to use ministers’ present positions (i.e. p) in the 12 cross-tabulations – but by definition a minister cannot have been outside the govern- 13 ment at both p–1 and p. Of course, cross-tabulating ministers’ immediate past and future positions blinds the analysis to cases in which an individual moves from a 14 low-ranking position at p–1 to a higher position at p, and then down again at p+1. 15 Empirically, however, this is a minor issue: in only 17 cases did a government member 16 at p–1 assume a higher office at p followed by a lower office at p+1. Similarly, in only 17 16 cases did a minister move from a higher office to a lower office and then back to a 18 higher office. Thus, these irregular career trajectories represent just 2.5 per cent of the data. 19 14 Fabius, Dufoix and Hervé were Prime Minister, Social Affairs Minister and Secretary 20 of State for Health, respectively, in Fabius’s government. See, e.g., Bosia (2007), 21 Chalaby (2004) and British Medical Journal (1998). 22 15 See http://www.rfi.fr/actufr/articles/049/article_26173.asp (visited 6 February 2007). 23 16 The scandal resulted in a number of convictions of ministers or former ministers, including Henri Emmanuelli of the PS, François Léotard of the Parti Républicain (PR) 24 and the UMP, Pierre Méhaignerie of the Centre des Démocrates Sociaux (CDS) and 25 Gérard Longuet of the PR. 26 17 Half the scandals involved financial misconduct. Many instances of financial mis- 27 conduct come to light only after the ministers have left office. Of 43 ministers placed 28 under legal investigation for financial misconduct between 1990 and 2000, only five were in office at the time (Adut 2004). 29 30 31 Note on sources 32 33 Our data on ministerial appoints are derived from Les gouvernements et les 34 assemblées parlementaires sous la Vème République 1958–2004 (Paris: 35 Assemblée nationale, July 2004). Biographical data come primarily from Who’s 36 Who in France, though for ministers who held office early in the Fifth Republic, 37 Benoit Yvert’s Premiers ministres et présidents du conseil: histoire et dictionnaire 38 raisonné des chefs du gouvernement en France, 1815–2002 (Paris: Perrin, 2002) 39 and websites at the Académie française (http://www.academie-francaise.fr/) and 40 Ordre de la Libération (http://www.ordredelaliberation.fr/) proved helpful. 41 Information on political scandals is taken from Keesing’s Contemporary Archives. 42 43 References 44 45 Adut, A. (2004) ‘Scandals as Norm Entrepreneurship Strategy: Corruption and the French 46 Investigating Magistrates’. Theory and Society 33: 529–578. SOM01p.qxd 7/9/08 23:02 Page 56

56 The selection of ministers in Europe Amorim Neto, O. and Strom, K. (2006) ‘Breaking the Parliamentary Chain of Delegation: 1 Presidents and Non-partisan Cabinet Members in European Democracies’. British 2 Journal of Political Science 36: 619–643. 3 Bosia, M. J. (2007) ‘From Contention to Engagement: French Social Science and the 4 Politics of HIV/AIDS’. French Politics 5: 96–105. 5 British Medical Journal (1998) ‘Ex-ministers to Face Trial in Tainted Blood Scandal’. 6 7154, 1 August 1, 317: 302. Butler, D. (1994) ‘French Ministers Face Public Trial in HIV Blood Affair’. Nature 370, 7 6487: 243. 8 Chagnollaud, D. and Quermonne, J.-L. (1996) Le gouvernement de la France sous la Ve 9 République, nouvelle édition. Paris: Librarie Arthème Fayard. 0 Chalaby, J. K. (2004). ‘Scandal and the Rise of Investigative Reporting in France’. 11 American Behavioral Scientist 47, 9: 1194–1207. 12 Conley, R. S. (2006) ‘From Elysian Fields to the Guillotine? The Dynamics of Presidential 13 and Prime Ministerial Approval in Fifth Republic France’. Comparative Political 14 Studies 39, 5: 570–598. 15 Dogan, M. (1979) ‘How to Become a Cabinet Minister in France: Career Pathways 16 1870–1978’. Comparative Politics 12, 1: 1–25. 17 Elgie, R. (1993) The Role of the Prime Minister in France, 1981–91. New York: St Martin’s Press. 18 Gaxie, D. (1986) ‘Immuables et changeants: les ministres de la Vème République’. 19 Pouvoirs 36: 61–78. 20 Huber, J. D. and Martinez-Gallardo, C. (2004) ‘Cabinet Instability and the Accumulation 21 of Experience: The French Fourth and Fifth Republics in Comparative Perspective’. 22 British Journal of Political Science 34: 27–48. 23 Knapp, A. (2004) Parties & the Party System in France. A Disconnected ? 24 New York: Palgrave Macmillan. 25 Knapp, A. and Wright, V. (2006) The Government and , 5th edition. 26 London: Routledge. 27 Lacam, J.-P. (2000) La France, une République de mandarins? Les hauts fonctionnaires et 28 la politique. Paris: Éditions Complexe. Lewis, E. G. (1970) ‘Social Background of French Ministers, 1944–1967’. The Western 29 Political Quarterly 23, 3: 564–578. 30 Matland, R. (1998) ‘Women’s Representation in National : A Comparison of 31 Democracies in Developed and Developing Countries’ Legislative Representation’. 32 Legislative Studies Quarterly 23: 109–125. 33 Safran, W. (1998) The French Polity. New York: Longman. 34 Stevens, A. (1992) The Government and Politics of France. New York: St Martin’s Press. 35 Thiébault, J.-L. (1994) ‘The Political Autonomy of Cabinet Ministers in the French Fifth 36 Republic’. In Laver, M. and Shepsle, K. A. (eds) Cabinet Ministers and Parliamentary 37 Government. New York: Cambridge University Press, 139–149. 38 Thiébault, J.-L. (2003) ‘France: Forming and Maintaining Government Coalitions in the 39 Fifth Republic’. In Müller, W. C. and Strøm, C. (eds) Coalition Governments in Western Europe. Oxford: Oxford University Press. 40 Time (1968) ‘A Sudden Parting: How Pompidou Was Fired’ 19 July. 41 Warwick, P. V. and Druckman, J. (2006) ‘The Portfolio Allocation Paradox: An 42 Investigation into the Nature of a Very Strong but Puzzling Relationship’. European 43 Journal for Political Research 45: 635–665. 44 Wolfreys, J. (2001) ‘Shoes, Lies and Videotapes: Corruption and the French State’. 45 Modern and Contemporary France 9, 4: 437–451. 46 SOM01p.qxd 7/9/08 23:02 Page 57

Cabinet dynamics in the French Fifth Republic 57 1 Wright, V. (1989) The Government and Politics of France, 3rd edition. London: 2 Routledge. 3 Zarka, J.-C. (2000) Fonction présidentielle et problématique majorité présidentielle / 4 majorité parlementaire sous la cinquième République (1986–1992). Paris : LGDJ. 5 6 7 8 9 0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46