Policy decisions characterised by their stability

Contrary to most Western countries where nuclear policy programmes were considered or launched, France never experienced any nuclear policy reversal. In this respect, it appears as a real outsider, sharply contrasting with most other countries. In 1978, Austria renounced any use of nuclear energy before the first plant was even switched on. In the same year, a Swedish referendum decided on a progressive phasing out of the use of nuclear energy. Later reversals occurred after the accident at Three Mile Island, which prompted the United States and the United Kingdom, pioneers in nuclear development, to stop their plant construction programme, and at Chernobyl, after which referendums dictated the end of Italian nuclear electricity production (1987) and a moratorium on the construction of new nuclear plants in Switzerland (1990). The subsequent rise of Green parties and their participation in governments led to further policy reversals in the 1990s, notably in Belgium (1999) and in Germany (2000) where votes led to a progressive phasing out. The 2000s saw a series of new reversals in favour of the nuclear industry, leading some observers to announce a “nuclear renaissance” (see Findlay Reference Findlay2011, 1 for book examples). The Belgian, American, Italian, British and German governments took measures to revive or extend their nuclear industries, while the Swedish moratorium on the construction of new plants was lifted in 2009. Most recently, the Fukushima accident resulted in a new series of reversals in Germany, Japan and Italy, where decisions were taken to phase out the use of nuclear energy.

Against this background, French nuclear policy distinguishes itself by its stability. French nuclear policy can be traced back to 1945, when the Commission for Atomic Energy (CEA) was created as a specialised nuclear research centre. This decision reflected three motives: a scientific ambition to deeply involve the French research community in nuclear research; an interest linked to the potential for civilian use of nuclear energy for electricity production in the context of post-war reconstruction; and the objective to develop nuclear weapons in order to avoid a new military defeat and to gain independence from the United States and the Soviet Union (Hecht Reference Hecht1998).

In 1951, future Prime Minister Félix Gaillard proposed the first five-year plan for the development of nuclear energy. This plan was adopted a year later by the National Assembly despite the Communist MPs’ reluctance for any military use of atomic power and the desire of rightist parties to prevent the participation of communist sympathisers in the programme. This plan gave birth to the first nuclear plant in 1956 and to the development of a second one, built in 1959. Framed in terms of independence and industrialisation, the electricity production programme did not rely on existing American technology, but on new French technology (using graphite) developed by the CEA. EDF, the electricity company nationalised in 1945, succeeded in setting up a collaboration with the CEA on the first nuclear plant and was able to commercially deploy electricity produced by nuclear energy starting in 1966.

The association and balance between the CEA and EDF structured the nuclear policy field in the 1960s – crucial years for French nuclear policy (Hecht Reference Hecht1998). While the CEA, along with the army, was focused on French nuclear independence and the production of plutonium for the nuclear weapons programme, the EDF primarily sought to intensify electricity production, and thus favoured the more efficient and robust American technology based on pressurised water reactors over graphite-moderated reactors. In 1969, the government decided to split the military and the civilian wings and allowed the EDF to use its preferred technology. This decision directly resulted in the building of a new series of nuclear plants. The commitment to this industrial perspective occurred before the oil crisis of 1973–1974, but the crisis reinforced it. The “Messmer plan”, adopted in 1974, scheduled the construction of 13 new plants before 1980, 50 before the middle of the 1980s and 200 before 2000. This overweening programme was only partially realised given the over-estimation of the growth in electricity consumption in France (Bataille and Galley Reference Bataille and Galley1999; see also Guillaumat-Tailliet Reference Guillaumat-Taillet1987). In 1973, electricity needs for the year 2000 were predicted to be 1,000 terawatt-hours (TWh). In 1980, the forecast changed to 650 TWh, and then 500 TWh in 1983. Finally, in 2000, the actual demand for electricity was around 450 TWh.

This nuclear policy was not contested in Parliament until 1997, and anti-nuclear stances were kept outside political institutions (Nelkin and Pollak Reference Nelkin and Pollak1981; Kitschelt Reference Kitschelt1986; Rucht Reference Rucht1994). However, the 1970s were marked by an upsurge in opposition to nuclear plants. Protests ran from local action to massive demonstrations, with violent confrontations with the police reaching a peak of mobilisation in Creys-Malville against the Superphénix fast-breeder reactor in 1976 and 1977. The anti-nuclear protests persisted in the early 1980s, with only limited policy results. The building of nuclear plants was briefly suspended after Mitterrand’s election in order to allow for an assessment and a policy debate in parliament. The only cancellation, in 1981, in the construction of a new nuclear plant was that planned in Plogoff. The decision to slow down the construction of new plants was only weakly related to popular protest, and more greatly owing to the overproduction of electricity by the EDF.

The 1998 decision to stop the Superphénix plant, a subject we will return to in greater detail below, was part of the agreement of the government coalition between the Socialist, Communist and Green parties. This was highly symbolic for opponents of nuclear energy given the numerous demonstrations this experimental plant had provoked over two decades. Nevertheless, the level of nuclear energy production remained unaffected, and even kept on growing. In the 2000s, despite a continued positive annual balance in the electricity trade,Footnote ² the surplus has been shrinking, and imports are regularly needed in winter because of the decreasing availability in French plants resulting from aging and maintenance issues (Delion and Durupty Reference Delion and Durupty2010, 173). After a 15-year pause (1990–2005), these trends motivated the decision to extend the service life of existing plants and to build several EPR reactors in France. This new development phase was revised downwards in 2009 in light of stagnating energy needs, but was not seriously called into question after the Fukushima accident.

A policy of major importance

The unique continuity of a pro-nuclear policy has generated the second largest set of nuclear plants in the world (58 reactors) behind the United States (104). Between the beginning of the programme and 2010, French investments in nuclear research and development, as well as in nuclear facilities, are estimated to have reached over 83 billion Euros.Footnote ³ As a result, France was able to reduce (until the mid-1980s) and then stabilise its oil imports, as well as triple its energy production since the 1970s, despite the decline of domestic natural resources (coal, gas and oil). Notwithstanding, scarce domestic natural resources, the level of dependence on foreign energy has regularly decreased, falling below 50%, and France is able to export electricity (Guillaumat-Tailliet Reference Guillaumat-Taillet1987; Hadjilambrinos Reference Hadjilambrinos2000). Nuclear energy represents 80% of French energy production and almost the same proportion of French electricity consumption, which makes France the leading country using this energy worldwide.

The significance of the French nuclear programme should also be evaluated with respect to the central importance of the state-owned nuclear sector to French industrial policy. The main justification for this persistent state control is the French concept of mission de service public, which implies guaranteeing equal and universal access to electricity – objectives defined in the 2000 law on electricity. Yet, the giant (partly) state-owned companies are also a pillar of French industrial and commercial policy. The state has favoured concentration in the energy sector and encouraged competition between major groups.Footnote ⁴ This “policy of champions”, and the technological advantage gained through the almost continuous expansion of the nuclear industry, fostered the emergence of high-performing and competitive groups that are successfully following a strategy of internationalisation (Chick Reference Chick2007; Delion and Durupty Reference Delion and Durupty2010). Most notably, Areva and EDF are the leading groups worldwide in the nuclear energy sector and, respectively, the first electricity producer and provider.

How can we explain this French exception? The resolute involvement in the nuclear option may not plausibly be fully explained by a lack of natural resources. Other countries suffering from an incapacity to meet a strong industrial and domestic demand with domestic gas, oil, mines or renewable energy sources – Germany in particular – made distinct policy choices. In most cases, phasing-out decisions resulted from the successful mobilisation of anti-nuclear movements and parties, along with a hostile climate of opinion (Hatch Reference Hatch1986; Jasper Reference Jasper1990; Jopke Reference Jopke1993; Baisnée Reference Baisnée2001). Was public opinion more supportive in France than in other countries? Was anti-nuclear contestation less mobilised? Or, were public and ecologist preferences not channelled into the process of policymaking? And, if so, what made such an absence of democratic policy feedback regarding such a crucial political, economic and industrial issue possible?

A pro-nuclear public mood?

Public contestation over nuclear energy is considered a crucial factor for policy reversals, as well as a potential constraint for the future development of nuclear infrastructures. As Findlay (Reference Findlay2011; see also Ramana Reference Ramana2011) puts it, “many of the factors are matters of contention and controversy, most obviously the question of the economic feasibility of new nuclear build and the challenges of nuclear waste disposition. The fact that these issues generate controversy creates uncertainty for governments, industry and private financiers considering investing in nuclear energy. This uncertainty itself acts as a brake on a nuclear energy revival”. Public opinion is sometimes considered to be overwhelmingly pro-nuclear in France. The continuity of French nuclear policy could indeed have been favoured by broad popular support.

Testing this hypothesis involves estimating the level of support for nuclear energy and its long-term evolution, which is a tricky enterprise given the constraints of data on attitudes towards nuclear energy (Jasper Reference Jasper1988). After gathering as much data as possible to assess this level of public support, we must acknowledge the highly heterogeneous nature of our data set in terms of frequency, timing and response categories. Some items were asked about only once or twice, while response categories and the wording of questions changed over time – with different questions leading to different answers, due to strong framing and priming effects.Footnote ⁵

In order to deal with data heterogeneity, we resorted to the “public mood” indicator developed by Stimson in his study of liberalism versus conservatism in the United States (Reference Stimson1999). Based on the “Dyadic Ratio Algorithm”, the measure of public mood has been used increasingly to create a single public opinion indicator from heterogeneous data sets (Baumgartner et al. Reference Baumgartner, De Boef and Boydstun2008; Stimson et al. Reference Stimson, Tiberj and Thiébaut2010). The “public mood” is an aggregate measure, based on survey marginals. By capturing the shared variation of different series loading on the same dimension over time, this measure offers an opportunity to neutralise bias induced by the wording of questions. Following Stimson’s guidelines, we included in our indicator of public mood towards nuclear energy all questions asked at least three times, which allowed a clear and symmetric distinction between positive and negative attitudes towards nuclear energy (see Appendix).Footnote ⁶ We used the ratio of positive feelings to the sum of positive and negative feelings for each question to compute a “nuclear support index” that estimates the relative level of nuclear support in France.Footnote ⁷ The French “public mood” towards nuclear energy was computed using the software WCalc, and is based on 302 estimates of the “nuclear support index” from 24 different question wordings collected from 13 sources (surveys and barometers). The data obtained, depicted in Figure 1, covers 39 years from 1975 to 2012.

Figure 1

Level of support to nuclear energy in France 1975–2012 (French “mood”, with confidence interval).Source: Eurobarometer surveys; EDF nuclear barometer; CREDOC barometers; IRSN barometers; “The French and Science” barometer; French elections surveys (Panel electoral français; Trielec); IFOP, IPSOS and Sofres surveys (own analyses).

This new indicator delivers several interesting conclusions. The first striking observation is the absence of majority support for nuclear energy among French citizens across most of the period considered. Despite the bias of our support index towards more support (since this indicator measures the proportion of supportive respondents among those who positioned themselves, this proportion would be substantially lower when calculated in relation to the whole sample), our index exceeds 50.0% only over the 1975–1987 period, with values situated between 47.9% and 60.9% and averaging 55.1%. This timing is surprising, as it is set against the background of the strongest anti-nuclear mobilisations in French history.

The Three Mile Island accident in 1979 does not seem to have had any short-term impact at this aggregate level (as the early 1980s mark a peak in support for nuclear energy). However, this support declined over the 1980s, especially after the Chernobyl accident in 1986. This trauma caused a rapid collapse, with a drop by 18.8 points in the public mood within five years. Interestingly, perhaps owing to the fading memory of Chernobyl, or because of the increasing visibility of alternative framings, this trend was reversed in the 1990s and was marked by a progressive and incremental increase until 2000 – paradoxically by the mid-term of the Jospin government, which involved the Green Party for the first time.

However, even when the mood indicator reached a peak in 2000, it remained under 48.0%. Further, the 2000s saw a new low in the public mood towards nuclear energy, with values between 44.0% and 48.0%. The subsequent increase in support observed in 2012 may seem surprising given the aftermath of the Fukushima accident, in the context of the diminishing support of public opinion for nuclear power plants in most countries (Bonneval and Lacroix-Lasnoë Reference Bonneval and Lacroix-Lasnoë2011; Ramana Reference Ramana2011), the phasing out decided by four of France’s direct neighbours and the reflections on the safety of the French plants. As shown elsewhere, however, this development can be understood with respect to the reframing of nuclear energy, which goes hand in hand with its stronger presence in the media (Brouard et al. Reference Brouard, Gougou, Guinaudeau and Persico2013). The members of the executive branch and the leaders of their party have successfully emphasised the jobs and the industrial advantages linked to the nuclear sector, but also its merits in terms of carbon dioxide emissions. Faced with this right-wing offensive, the ecologists did not manage to impose their framing in terms of risks. It is still too soon to assess the long-term effects of the Fukushima-related debates, but we should notice that, in 2012, our indicator of support for nuclear energy points to a sharp rise.

The public mood indicator computed allows us, for the first time, to estimate the climate of opinion towards nuclear energy among French citizens and its evolution over almost four decades, with a correction of bias linked to the wording of questions. This series indicates that French public support for the production of nuclear energy is far from broad, and only with exception reaches the majority of respondents. This ambivalent, sometimes sceptical, climate of opinion historically found its expression in a relatively strong social mobilisation against nuclear energy (Nelkin and Pollak Reference Nelkin and Pollak1981; Kiersch and von Oppeln Reference Kiersch and von Oppeln1983; Kitschelt Reference Kitschelt1986; Jasper Reference Jasper1990; Rucht Reference Rucht1994; Kriesi et al. Reference Kriesi, Koopmans, Duyvendak and Giugni1995; Baisnée Reference Baisnée2001). Stemming mainly from the ecologist movement, groups like Greenpeace and local collectives, such as “Stop Golfech”, “Tchernoblaye” or the initiatives in the network “Sortir du nucléaire”, this contestation took various forms ranging from petitioning and mobilising to mass demonstrations or spectacular actions. French anti-nuclear mobilisations were vigorous in the 1970s after the Three Mile Island and Chernobyl accidents, after Germany’s decision to phase out in 2000 and in reaction to the decision to build the first EPR plant in Flamanville in 2005.Footnote ⁸ Looking at public opinion data and at the comparative history of anti-nuclear social movements thus reveals a gap between stable and persistent policy choices over time and weak and fluctuating public support for these choices. How can such a gap be explained with respect to policy decisions of primary political, economic, social and industrial importance?

A non-salient issue?

The disconnection between the preferences of the median voter and the decisions of nuclear policy in France could be rendered possible by a low salience of this issue in the public sphere. In the case of United States nuclear energy policy, Baumgartner and Jones (Reference Baumgartner and Jones1993) have shown how crucial issue salience in the media is for understanding policy change. Faced with the absence of a “most important problem” barometer, we explored this possibility on the basis of a study of media coverage of nuclear energy problems, with a focus on Le Monde.Footnote ⁹ The high inter-media consistency observed in media studies (Eilders Reference Eilders2002; Vliegenthart and Walgrave Reference Vliegenthart and Wlagrave2008) justifies using this newspaper as a proxy for the French media agenda. Le Monde is generally acknowledged as the leading paper in France over the entire post-war period, and its agenda-setting role has been shown to find expression in other media – the daily Libération, the weekly magazines Le Point and L’Express and the news on the public TV channel France 2 – taking up Le Monde’s issue priorities (Blanchard Reference Blanchard2010, 325–334). Figure 2 provides an overview of the salience of nuclear energy in Le Monde between 1977 and 2012, measured as the number of articles published on this topic per year.

Figure 2

Attention to nuclear energy in Le Monde 1977–2008 (number of articles per year).Sources: Index Le Monde; Factiva search engine.

Our data suggests considerable variation in the visibility of the nuclear issue in the media. Attention to this issue was consistently high by the end of the 1970s when continuous anti-nuclear protest and the Three Mile Island accident in 1979 provided fuel for the news media. In 1977, on average, more than two articles per day were devoted to nuclear energy, often on the front page. This level of importance decreased and remained, during periods of routine coverage, at a non-negligible level of about 100 articles a year. Occasionally, nuclear energy gained particular prominence in the media, notably on three occasions. First, as expected, the Chernobyl accident in 1986 attracted much attention, with more than one article per day, and an intensity of coverage still above average in 1987. A new peak in attention occurred in 1996–1997, thanks to a combination of factors: the commemoration of Chernobyl in 1996, a controversy on the nuclear fuel reprocessing plant of La Hague shaped by the publication of a scientific study in 1997, a spectacular and professional campaign carried out by Greenpeace France (Baisnée Reference Baisnée2001), a debate opened by the Green Party and their ministers about the dismantling of the fast-breeder reactor Superphénix, as well as German discussions regarding the future of nuclear energy. More recently, media attention paid to nuclear energy has risen since 2005, with a peak in 2007 in the context of building EPR power plants after a long hiatus, the debate regarding nuclear dissemination (often related to the Iranian nuclear programme), global warming and the potential role of nuclear energy in the reduction of carbon dioxide emissions, energy costs and availability, the “nuclear renaissance” and the related opportunities for the French nuclear industry. Attention remained high in 2008, notably owing to incidents at two French fuel cycle facilities – the uranium waste treatment plant of Tricastin and the Cerca research reactor fuel fabrication facility in Romans. The data presented in Figure 2 do not cover the most recent years, but nuclear energy certainly acquired a historic level of visibility in the months following the Fukushima accident and during the 2012 presidential campaign (Brouard et al. Reference Brouard, Gougou, Guinaudeau and Persico2013).

Thus, attention to nuclear energy has gone “up and down” (Downs Reference Downs1972), but even in periods of routine coverage, nuclear energy is present in the media. In the period 1977–2012, at least five critical moments opened windows of opportunity to politicise the nuclear issue, which was important because several of them, in particular those in the second half of the 1970s and after Chernobyl, corresponded to phases of fading public support for nuclear energy. In addition, if the second half of the 1980s was not marked by any peak in media attention, there was a clear trend towards more visibility, coupled with growing distrust by citizens. The hypothesis of low salience of nuclear energy issues can thus not explain why the predominantly negative attitudes of French citizens towards nuclear energy have had virtually no impact on the course of French nuclear energy policy.

Vote-related incentives to contest nuclear energy

A reversal in French nuclear energy policy would have required an agent implementing an anti-nuclear policy. Parties’ incentives to respond to citizens’ skepticism towards nuclear energy may be shaped by the preferences among their voters. Figure 4 shows that the distribution of preferences of nuclear energy differs substantially among the supporters of left and right parties.Footnote ¹² While the majority of the electorate of parties on the right and far-right (between 50% and 70%) supports nuclear energy, this proportion is much smaller among Communists and Socialists (between 40% and 50%) and especially among ecologists (20%). The level of nuclear support among the supporters of all four traditional governing parties has been eroded since the beginning of the 1980s. For both rightist parties, the level of nuclear support decreased from three-fourths in 1980 to a slight majority in 2005. After 1986, among Socialist Party (PS) sympathisers, the proportion of nuclear supporters fell and remained below 40%. Voters of the Communist Party (PC), traditionally among the most pro-nuclear, are no exception to this trend. The decline of this party has been accompanied in the last 15 years by the affirmation of other far left and Trotskyist parties, whose sympathisers are as opposed to nuclear energy as the ecologist parties. In sum, the decline in support for nuclear energy affects all parties’ constituencies. Coupled with mass protest against nuclear energy and the growing social, political and economic implications of this policy, this may have generated electoral incentives to advocate for and promote a policy reversal, especially in the 1980s.

Figure 4

Level of nuclear support among party leaners in France, 1980–2005 .Sources: Sofres; “The French and Science”; IFOP.

The long depoliticisation of nuclear energy in France: office – and policy-related incentives (1946–1989)

The growing skepticism of French voters was not followed by any realignment, as all French parties have depoliticised nuclear energy in their official positions and in their electoral campaigns.Footnote ¹³ This depoliticisation results in the low salience of matters of nuclear energy policy in party discourse and in the broad consensus on a pro-nuclear stance involving all parties. The first point may be illustrated by the marginality of nuclear energy in the electoral manifestos of the main parties between 1958 and 1988 when compared with the further relevant issues of “farmers” and “culture”, as measured by the Comparative Manifesto Project (Figure 5).

Figure 5

Attention to nuclear energy, farmers and culture in French party manifestos.Sources: Comparative Manifestos Project; French electoral manifestos.

This comparison highlights the modest attention to nuclear energy within the four main parliamentary parties. Nuclear energy is mostly far less salient than other issues of similar, or even smaller, scope, be it in the manifestos of the main left parties or those of the Gaullist party – the latter of which is the one that historically pushed the expansion of nuclear energy. Beyond this low salience, the depoliticisation of nuclear energy has also resulted, for a long time, in the convergence of the main French parties towards pro-nuclear stances. The enduring misalignment between citizens’ preferences on a salient issue on the one hand and electoral supply and policy decisions on the other cannot be understood without taking into account all motivations of party representatives, who are not only concerned with attracting votes, but also with policy- and office-related considerations (Müller and Strom Reference Müller and Strom1999).

The interest of elected representatives in channeling the popular preference for a policy reversal varies considerably by party. The ideology or the political project of each party goes along with distinct policy preferences. The conservative parties anchored in the Gaullist tradition lean towards nuclear energyFootnote ¹⁴ given their commitment to national independence and their proximity to representatives of the nuclear sector. On the other side of the political spectrum, the PC is traditionally the most fervent advocate of nuclear energy,Footnote ¹⁵ which is congruent with its proximity to EDF’s dominant trade union, CGT, the ideological belief of the PC in the power of science and technology to transform the world, its productivist orientation and the priority set by this party to improve the standard of living, job creation and preservation. The pro-nuclear stance of the PC may also be linked to the firm commitment of the USSR in massive programmes of nuclear energy development, and to the high number of Communist scientists involved in French nuclear research, most notably the first director of the CEA, Frédéric Joliot-Curie. The political programmes of other parties, such as the PS, the Centre Party (UDF) and the National Front (FN), have no such clear and straightforward implications regarding their positions on nuclear energy.

In this respect, and provided that the pro-nuclear policy was implemented by the rightist coalition involving the UDF,Footnote ¹⁶ a policy reversal was most likely to emanate from the PS. The only party opposed to nuclear energy was actually the small Unified Socialist Party, which partly merged with the PS in 1974. The PS is traditionally close to the only anti-nuclear mainstream organisation, the trade union CDFT (Garraud Reference Garraud1979; Nelkin and Pollak Reference Nelkin and Pollak1980), and even closer to anti-nuclear protest than the PC, with Socialist activists heavily involved in national and local protest. However, this party did not act as the influential ally anti-nuclear movements would have needed to reverse the course of French nuclear energy policy. Rather, the PS representatives were more concerned with accommodating the pro-nuclear stance of the main coalition partner, the PC, than with responding to anti-nuclear voters. The interplay of policy- and office-related motivations and their consequences for French nuclear policy choices is particularly striking when looking at the critical juncture of nuclear expansion and political alternation (1973–1985).

The second part of the 1970s was a key moment for nuclear energy policy, with an acceleration in the speed of construction of new plants and a quick rise in nuclear electricity production. The development of mass anti-nuclear protest and the Three Mile Island accident could have fostered the politicisation of nuclear energy in the electoral and parliamentary arenas and, at that point in time, contestation by the parliamentary opposition or the decision to phase out could have had huge consequences on the scope of the French nuclear programme. However, parties did not respond to this challenge with a politicisation of nuclear energy policy, but instead with a strategy of issue avoidance. This is in particular the case with the PS, which faced intense party-internal debates between pro- and anti-nuclear positions and the necessity of accommodating the pro-nuclear PC as the only possible partner for coalition building. Already at the 1973 legislative elections, the Communist and Socialist parties had defended a common manifesto that was heavily pro-nuclear, acknowledging the key role played by the CEA in French research policy and defending state support for the development of nuclear power plants run by a public nuclear industry. Focused on the objective of seizing power, thanks to an alliance with the PC, the PS ignored the peak of anti-nuclear protest in France and centred its 1978 campaign on other issues. After the leftist alliance came close to winning a majority, the likely perspective of winning the 1981 presidential election made alternation and office even more the priority.

The common platform of PS and PC had to be defined on a common denominator: state intervention in the economy through nationalisation and an active industrial policy provided the ground for a coalition agreement. From the point of view of each party, politicising nuclear energy may have attracted more votes but would have undoubtedly undermined a left-wing coalition and thereby render impossible the 1981 alternation. This calculus sheds light on the strategy pursued by both parties. On the one hand, the PS adopted a soft stance: limiting the nuclear programme to the nuclear plants already under construction, taking time for a national debate, holding a referendum on nuclear energy development and enacting legislation guaranteeing the transparency of nuclear plant security. The socialist candidate and future President François Mitterrand also promised a moratorium on the construction of any new nuclear plants. On the other hand, the manifesto presented by the PC was the first to keep silent on nuclear strategy and, since then, this party has mostly avoided the topic.

Policy-related arguments went against a phasing out of the nuclear programme as well. In the 1970s, François Mitterrand steadily emphasised the importance of remaining independent and even promoted the idea of turning back to the French graphite technology. A nuclear phasing out would have also undermined the prospects of the macro-economic policy planned by the French left. Nuclear energy was not only seen as an illustration of the merits of state-owned companies for building and exploiting large infrastructures, but also as a source of cheap energy for future industry development and a sector employing thousands of employees. In addition, most nuclear plants were already under construction, or nearly finished, by 1981, so a drastic phasing out would have been very costly. The most expensive part of the nuclear programme had already been realised, and a phase out would have created considerable stranded investment to which the costs of searching for and funding alternative energy sources would have been added. The corresponding funding would not have been available for the nationalisation policy strongly desired by the coalition of the left.

As long as the entry costs into the French political system were high enough to forbid any anti-nuclear party being pivotal, the PS had incentives to stick to nuclear energy. After the left coalition won the presidency and a legislative majority in 1981, only symbolic pay-offs for anti-nuclear activists were delivered – for example, the criticisms against the undemocratic nuclear policymaking implemented by the rightist government (e.g. the location of the nuclear plants) before 1981, the six-month moratorium and the cancellation of the Plogoff nuclear plant. These symbolic pay-offs challenged neither the principle of using nuclear energy nor the building of new plants after 1981. Indeed, even two years after the Chernobyl accident, the Socialists presented to the 1988 national elections the most pro-nuclear platform ever: “Which country is the only one that succeeded in developing nuclear energy in order to overcome the scarcity of our natural resources? It is ours, because it succeeded in organizing a 20-year plan, which no private company would have been able to do”. Given the skepticism that prevails in French public opinion, this situation generates a misalignment between the preferences expressed by citizens and party representatives. In 1982, the “French and Science” barometer indicates that around 70% of citizens opposed to nuclear energy sympathised with pro-nuclear parties. This figure remained high in the following decades, reaching nearly 90% in 1989. The relevance of this explanation of the continuity of the French nuclear policy is underscored by the contrast with the patterns stemming from another critical juncture.

Changing coalition-related incentives with limited policy consequences: the lack of policy opportunities (1989–2012)

If the ecologist movement had run in the French national elections since the presidential contest of 1974 (gaining 1%), it became a relevant player only from the late 1980s. This situation changed from 1989, with the Greens’ breakthrough at the municipal elections in March and at the European elections in June. In 1992, for the first time, ecologists were successful in the 1992 regional elections, which allowed them to be represented in a decision-making institution for the first time. With 14.3% of votes, more than 200 ecologists entered the French regional councils. The distribution of seats and political parties located the two ecologist parties as pivotal players in many regions, as they formed regional coalitions with either right or left parties and provided external support for formal minority governments (Brouard Reference Brouard1999). The 1993 legislative and the 1995 presidential elections confirmed the ecologists’ electoral strength. Despite the instability of their organisation and of their leadership, the Green party became increasingly relevant and strategic. The rise of this party was associated with an emphasis on new issues, among which nuclear energy was particularly important. In the 1990s, the Greens were the only party addressing the environmental issue, with concerns regarding nuclear risks and decision procedures judged “undemocratic” and too centralised. Issue avoidance and consensus regarding nuclear energy in electoral competition were suddenly broken by the emergence of the Green Party. As shown in Figure 5, this party devotes a higher share of its programmes to nuclear energy. More importantly, its concerns regarding long-term pollution and industrial risks implies strong anti-nuclear policy preferences.

The PS and, above all, the PC simultaneously saw their electoral support shrinking and felt pressure from the Green Party electorally, given the attractiveness of the Greens in the leftist constituency; politically, with new issues entering the electoral agenda; and strategically, with the Greens becoming a pivotal player in potential left coalitions. This development changed the incentives to politicise nuclear energy. In short, the PS faced the challenge of limiting electoral losses, dealing with new issues and maintaining the ability to lead a left coalition, which would probably have to include the Greens. PS representatives acknowledged the necessity of signalling their willingness to form a coalition with the Green Party and their openness to their policy proposals. These considerations regarding an anticipated coalition with the Greens probably explain why pro-nuclear arguments vanished from the PS manifesto in 1993 and why the PS substantially revised its discourse in the 1997 legislative campaign, adopting a much more negative tone towards nuclear energy: “We will change the course of French energy policy by adopting a moratorium on the building of nuclear plants, by increasing the incentives to energy saving and to the development of alternative energy sources. We will close the fast-breeder reactor Superphénix”.

Nuclear policy was, for the first time, affected by an effective Green pressure in 1997–1998, when Greens entered the plural left coalition (1997–2002) with the PS and the PC. The government decided to close the experimental plant running the fast-breeder reactor Superphénix. Yet, this reversal was mainly symbolic and did not hinder the connection of new nuclear plants to the electricity grid and a peak in nuclear electricity production. The promised moratorium on nuclear plant construction was never voted on, even if it was de facto implemented. This is not necessarily linked to Greens being in government, as the nuclear development plan was already fully realised and no nuclear plant reached its projected end of life during this period. Nevertheless, as soon as the right-wing coalition came back to power in 2002, the nuclear programme was immediately re-launched with the decision to build a new plant in Flamanville, followed by a second one in 2009 in Penly. If the Greens did not reduce the importance of nuclear energy, they probably precluded the growth of the French nuclear programme when they were in office. Given the highly constraining legacy of past policy choices, downsizing the nuclear programme would have been more costly in 1997 than in 1981: the discrepancy between political time and policy time gave the Greens little opportunity to further challenge French nuclear policy.

The logic induced by the changing coalition-related incentives has persistent effects. The nuclear issue became contentious during the last two waves of presidential and parliamentary elections. In 2007, presidential candidate Ségolène Royal and the PS manifesto announced for the first time a desire to decrease the proportion of nuclear energy by increasing the share of renewable energy of total energy consumption by 20% by 2020 (and 50% in the longer term). The conservative party candidate, Nicolas Sarkozy, reacted to this proposition by stressing the merits of nuclear energy in terms of carbon dioxide emissions and comparative industrial advantage. As a result, this issue was hotly discussed during the TV debates involving the opposing candidates, which took place between the two rounds of the presidential election. The visibility of questions related to nuclear energy was even stronger in the run up to the 2012 national elections, which took place in the aftermath of Fukushima and in the context of the decisions of several countries to phase out the use of nuclear energy. The PS took a more critical position and promised to reduce the share of nuclear energy in electricity production by 25% before 2025, to open a debate over energy policy, to set a moratorium on the construction of new plants until the end of this debate, to close the aging Fessenheim plant and to give up the construction of another EPR in Penly. Public controversies resulted from the decision of the conservative party UMP and of the executive to politicise their pro-nuclear stance and to frame nuclear energy in terms of employment, purchase power and reliability. As a result, the issue seems to be absorbed by the traditional left–right cleavage, with parties on the left (except the PC) being more opposed to nuclear energy than parties on the right (Brouard et al. Reference Brouard, Gougou, Guinaudeau and Persico2013). According to our theoretical perspective, under these political conditions, democratic responsiveness is more likely – within the limits of the policy legacy. The 2012 political alternation was associated with three policy decisions: the oldest nuclear plant (Fessenheim) will be shut down in 2016–2017 despite the recent authorisation of the extension of its life span; the nuclear plant under construction (Flamanville) will be finished and connected to the grid; and the newest project (Penly) is cancelled. Nevertheless, any long-term and effective policy effect remains to be observed.