Towards an education approach à la finlandaise? French education policy after PISA

In this article, we address whether international student comparisons have changed the dynamics of French secondary education policy. We focus on the increasingly significant impact of the Organization for Economic Cooperation Development (OECD)’s Programme for International Student Assessment (PISA) on France, a country previously known for its aversion to international comparisons and its turbulent relationship with the OECD. We argue that not only are transnational pressures – in our study the perception of PISA – crucial determinants for the fate of potential reform measures, but also the capacity of the state to transform its education system and take corrective measures. Along these lines, we also examine the role of historically embedded guiding principles of education, in the French case most notably that of equality (égalité). We focus, in particular, on efforts of French policy-makers to emulate elements of the recent ‘PISA champion’ Finland.


Introduction
In this article, we analyse to what extent international comparative assessments of student performance have changed the dynamics of French secondary education policy. In addition to the long tradition of evaluation of French schools (see Pons 2011), comparative performance assessment has increasingly gone international over the past 10 years. These transnationalization processes in the field of education have enabled international organizations (IOs) to invent and apply various stimuli for reforms at the national level. Besides promoting a common discourse, coordinating joint activities, and financing reform measures, IOs have used the tool of comparative student assessment to shape and influence education policy at the national level. Currently, two major organizationsthe Organization for Economic Cooperation Development (OECD) and the International Association for the Evaluation of Education Achievement (IEA)are particularly active IOs in comparative assessments of education achievement and performance. While the IEA has conducted various studies focusing on mathematics and science skills Third International Mathematics and Science Study (TIMSS) and reading literacy Progress in International Reading Literacy Study (PIRLS), the OECD (see Henry et al. 2001) has recently become the most influential international driver of comparative assessment with its Programme for International Student Assessment (PISA) (see Martens, Rusconi, and Leuze 2007). The PISA study is the largest international comparative evaluation of the skills of 15-year-olds and covers reading, mathematical skills and scientific literacy. Recently, PISA andto a lesser extentthe other mentioned evaluations have triggered diverse reactions and reform measures in education systems around the globe. This applies both to countries with far-above-average results such as New Zealand (see Dobbins 2010) and Finland (Grek 2009) as well as those with less satisfactory results such as Germany (see Niemann 2010) and Switzerland (Bieber 2010), which have taken various measures to address issues such as performance disparities, equality and school autonomy.
France has participated since the first PISA round in 2000 and great attention has been dedicated in the academic literatureboth in France and abroadto explaining the strengths and weaknesses of the PISA methodology as well as disparities in outcomes (Goldstein 2004;Rémond 2006). However, the concrete impact of PISA and other comparative assessments on domestic policy-making in France has only received little scholarly attention (see Pons 2009a, 2009b). We aim to overcome this gap in the literature by first looking at France's rocky relationship with international comparative assessment and then examining its response to the PISA study. We argue that France, a country historically known for its more introverted approach to education 1 (see Meuret 2003a;Meuret and Duru-Bellat 2002), has increasingly engaged in processes of international policy learning, leading to an ever more dynamic interplay between the international and national levels of policy-making. The first two PISA rounds (2000,2003) essentially confirmed the results of previous studies: the performance of French pupils is at best in line with the international average (see Table 1; see Rémond 2006 for an analysis of the French PIRLS results). Particularly striking, however, is the deterioration between PISA 2003 and 2006: while the French pupils came in place 13 in the overall PISA results in 2003, they only were ranked 21st among 34 OECD member states in 2006. The very recent PISA 2009 results also did not reflect any major improvements. Above all, the French results for reading skills have significantly deteriorated between PISA 2000 and 2009 (see Table 1).
Thus, to what extent has the increased problem pressure resulting from international comparative assessments, in particular the PISA study, impacted French secondary education? To theorize the interplay between international pressures and the domestic institutional context we examine political-institutional factors which may facilitate or impede education reforms. We argue that not only are the public  Reading  505  13th  496  14th  488  19th  496  16th  Mathematics  517  10th  511  13th  496  19th  497  14th  Science  500  12th  511  10th  495  21th  498  19th  3-area average  507  12th  506  13th  493  21th  497  17th reception and perception of international comparative assessments crucial determinants for the fate of potential reforms, but also the capacity of the state and education system to channel increased problem pressure into concrete reform measures. In other words, to what extent is the state able to implement reforms and take corrective action? The domestic policy context can substantially influence the direction and speed of national reactions to international stimuli. To examine policy change and/or inertia, we combine two theoretical approaches. We focus, on the one hand, on the institutional opportunities and constraints in the political and education system, reflected by the number of veto players and the degree of centralization (Ganghof 2003;Tsebelis 1995). On the other hand, we incorporate the notion that ideas, norms and identities can also impact reforms, especially in the field of education (see Fearon and Wendt 2002;Nagel, Martens, and Windzio 2010). Thus actors may cling to historically entrenched principles of education despite internal or external pressures for change (see Martens et al. 2010). In the following, we are therefore interested in how international pressures are 'digested' by the political system and how policy outcomes are shaped by political constraints and education policy norms and traditions. We first briefly outline the French education and political systems with a focus on actors, structural aspects as well as historical traditions and guiding ideas. We then elaborate on the role of comparative assessment in France, including both PISA as well as forerunner studies. In light of the institutional and political background described belowhigh state capacity for action, high level of mobilization of reform adversaries, scepticism towards international comparisons (see below)we expect a rather restrained reaction to PISA and similar international comparative assessments. We assess this assumption by looking at the current education policy reforms which have been facilitated to a considerable degree by transnational education governance. In the conclusion we turn back to the explanatory framework and reflect on the interplay between international reform stimuli and national responses. The French education system: between continuity and adaptation The French system bears several remarkable characteristics. The school system is horizontally structured and the secondary level is divided into two phases (Hörner and Many 2010, 239). Following the five-year elementary school, all pupils attend a comprehensive collège for four years (until age 15). Approximately 95% of each age group completes the collège by passing a uniform national examination (brevet). In the second segment of secondary school the majority of French pupils attends the lycée, while a small minority begins an apprenticeship (apprentissage). A distinction is made here between general and technical or vocational schools. 2 After completing two years at the lycée d'enseignement professionnel it is possible to obtain a vocational certificate (CAP), a vocational diploma (BEP) or after four years a vocational university entrance diploma (Hörner and Many 2010, 253-54). The last school year prepares pupils for the baccalauréat, a centralized national graduation examination, which entitles graduates to attend a university or grande école.
In strong contrast to decentralized systems, for example the American or German system, all education content is determined by the National Ministry of Education (Ministère de l'Éducation Nationale, MEN). 3 Moreover, the French school system is almost exclusively publically financed and all primary and secondary school teachers are employed by the central government, making the education ministry the country's largest employer (MEN 2010a). Although students are increasingly granted the possibility to select between different specializations, the same school curriculum defined by the Bulletin officiel de l'Éducation nationale (BO) is taught at all schools. The strong governmental steering is also reflected in the French policy with regard to school choice, as children have been required to attend the school assigned by the government's carte scolaire, a geographical map with school districts.
Long before the emergence of international comparative assessments, the French government sought to mitigate various perceived weaknesses of the system with an array of corrective measures aimed at decentralization, eliminating inequality and promoting inclusion. For example the Lois de décentralisation of 1982 and 1983 triggered a process of very restrained decentralization (Mallet 2006). Simultaneously, attempts were made to localize the nationally defined educational objectives by granting additional authority to the directors of the school administrative districts (recteurs d'académie) (Bundesministerium für Bildung und Forschung [BMBF]-German Federal Ministry of Education 2003, 104). Education Minister Chevènement aimed to further 'democratize' the education system by increasing the number of baccalauréat graduates (Duru-Bellat and Kieffer 1999;Hörner and Many 2010;Larue 2003). Besides setting a target of 80% baccalauréat graduates per age group, 4 the government also introduced a baccalauréat professionnel to ensure that technically inclined pupils can complete the lycée with a vocationally oriented diploma. Moreover, so-called zones d'éducation prioritaires (ZEP) were established in disadvantaged socio-economic areas and received additional government funding. This measure was aimed at increasing the number of lessons. However, its success is questionable (see Meuret 1994;Moisan and Simon 1997). In the following years various other reforms were introduced, which point to an expanded role of the central government in education policy, e.g. the national council for school evaluation 5 created in 2000 to provide public annual performance and quality reports on individual schools and the entire school system. During the same timeframe a national monitoring system as well as standardized tests for pupils of the third, sixth and tenth grade were introduced (see Pons 2011) and carried out by the evaluation department of the MEN Direction de l'évaluation, de la prospective et de la performance (DEPP). 6 Secondary school teachers were also obligated to undergo uniform training at the Institut Universitaire de Formation des Maîtres after completing university studies (BMBF 2003, 104).

Predicting policy change: facilitators and constraints
In formal terms, policy change in France is currently facilitated by two central factors. The semi-presidential system allows for phases of majority rule (fait majoritaire), in which the president and prime minister are from the same political party holding a parliamentary majority, as well as cohabitation 7 during which the executive is headed by members of opposing parties (see Ismayr 1997, 15). While in phases of cohabitation we can expect a particularly difficult consensus-building process, resulting in frequent presidential vetoes and/or watered-down reforms, majority government constellations (fait majoritaire) offer favourable conditions for legislative output and swift policy change. Precisely this is the case since 2002, i.e. the timeframe in which international comparative assessment has received increased public attention and triggered far-reaching reforms in neighbouring countries (see Martens et al. 2010).
The high institutional capacity for policy change is also enhanced by the strong centralization of the education system. Contrary to Germany, in which policies are made by individual states (Länder), or the extremely decentralized American system, the French central government holds the decisive steering authority at all levels of education. The strong centralization is consistent with the typically French tendency for political uniformity and central government steering based on the logic of the L'une et indivisible République. Regardless of trends towards regionalization and the strengthening of sub-national administrative structures in recent years, the French education system is still also regarded by the OECD and other observers as highly centralized (OECD 1994;see Corbett 1996, 19). Hence, French education policy embodies the principle of political concentration described by Katzenstein (1976, 15), which is reflected in the high capacity for state executive action.
Particularly striking for outsiders is the seemingly paradoxical co-existence of the deeply entrenched and state-guaranteed principle of equality and the de facto elitist character of the system (OECD 2010a). On the one hand, the state is expected to provide the same educational opportunities to all regardless of socioeconomic standing by means of centralized steering and control. Furthermore, various system features promote education equality and strong performance, for example the wide availability of pre-school, no tracking of pupils in secondary education, small average class size and high public expenditures (see Duru-Bellat, Mons, and Suchaut 2004;Duru-Bellat and Suchaut 2005, 188;Schlicht, Stadelmann-Steffen, and Freitag 2010). On the other hand, internal and external studies reveal that the education system is based on a myth of equal opportunity and to a large extent geared towards forming national elites (Baudelot 2009;Duru-Bellat 2006;Meuret 2000;OECD 2010). Along the same lines, it is well-known in France that the social status and income of parents are a decisive variable for educational performance (see Meuret and Morlaix 2006;Moisan and Simon 1997).
However, the historically embedded strong role of the state in education policy has also led to situations in which well-intentioned reforms aimed at dismantling bureaucracy, increasing autonomy, competition and decentralization and thus 'less government' are interpreted as an assault on equality. Reform proposals are often instinctively regarded by left-leaning political forces as an affront to educational equality, despite the high selectivity of the existing structures. Moreover, conservative political forces are often accused of élitisme républicain, i.e. an excessive tolerance for inequality and its alleged reproduction within the school system (see Baudelot and Establet 2009, 10). The reform capacity of France is also aggravated by the strong mobilization capacity of reform adversaries. The threat and reality of nationwide public strikes reduce the state's means for taking action despite favourable institutional prerequisites for reforms. Finally, France's traditional aversion to international comparative assessments can be viewed as an additional impeding factor. As shown below, France had previously dedicated little attention to international comparative assessments and 'governing by numbers' (see Meuret 2003a;Pons 2009a, 2009b) and can perhaps be regarded as the country most sceptical of the PISA methodology. Based on these facilitating and inhibiting factors, to what extent has transnational education governance impacted secondary education policy in France?
France's rocky relationship with international comparative assessment Even before PISA, France can look back on a tradition of international education policy cooperation. In 1994 it took a step which most other large countries preferred to avoid by mandating the OECD to review its entire education system (Corbett 1996, 17). Despite praise for the universal access to pre-school education, the tradition of social integration, and the newly established zones d'éducation prioritaires, the OECD lamented that a large share of education participants fails to pass the baccalauréat, even though the French system is based on the idea of equality. Moreover, the OECD criticized the fact that the principle of fraternité ('brotherhood', i.e. solidarity) is overshadowed by strong competitive pressures and thatdespite the alleged guiding principle of liberté (freedom)few possibilities for individual self-development existed. A particular point of critique was the lacking individuality in teaching and pedagogical methods (OECD 1994; see also Corbett 1996). Most importantly, the OECD (1994) criticized the highly centralized and inflexible governance of the system despite the recent decentralization and the allocation of more autonomy.
However, little significant policy change resulted from the OECD's review activity. During the same timeframe, the OECD's means of influence on French education policy were significantly diminished as a result of the International Adult Literacy Survey (IALS), also conducted by the OECD. The results appeared to show that three-quarters of the French population had levels of literacy too low for them to perform normal everyday tasks such as reading a newspaper, writing a letter, or understanding a short text (see Guérin-Pace and Blum 2000). As a result, the French Education Ministry withheld the results from the public and became increasingly disengaged from the OECD, as from a French viewpoint ILAS shed major doubts not only on the objectivity and credibility of the OECD, but also on the methodology applied in international standardized testing (ibid 2000).
Thus, precisely while the OECD was expanding its international comparative assessment activities, the French public and policy-makers had taken a highly skeptical stance towards comparative assessment and voiced accusations of cultural bias, inadequate statistical methods and overly simplified indicators (Guérin-Pace and Blum 2000; see also Pons 2009a, 2009b;Pons 2011). These doubts over the partiality and adequacy of international assessment have indeed persisted into the 'PISA' era. Firstly, critique has been expressed that the PISA study is ideologically biased, as it evaluates education from an economic perspective (OECD 2004b; see also Henry et al. 2001). Like the ILAS study, various observers have also purported that PISA bears a bias towards English-speaking countries and Anglo-American education concepts 8 (Duru-Bellat and Suchaut 2005, 182;Rémond 2006, 76) and is thus inadequate for France. 9 Moreover, various observers have criticized PISA for not covering secondary education in its entire breadth (e.g. Dohn 2007;Fuchs 2003;Hermann 2005;Kraus 2005;Ladenthin 2004). And most importantly, serious issues arise as to whether the products resulting from different education systems are at all comparable (see Duru-Bellat and Suchaut 2005, 182).
Despite such criticism of the PISA methodology, there are indeed several reasons to assume that PISA may more profoundly affect French policy-making than previous endeavours. First, PISA is not only the largest cross-national standardized evaluation scheme, but also the most politically acknowledged, even in France. Second, PISA constitutes a significant advancement beyond previous attempts at comparative assessment (Goldstein 2004) and provides the most comprehensive comparative data ever available. Third, it has provided a basis for tighter inter-organizational synergies between the OECD and EU, the latter of which has taken the PISA results as the basis for its peer-learning activities and for 'governing the European education space by the numbers' (Grek 2009, 33). PISA filters out examples of 'best practices' of high-ranked countries and identifies weaknesses of low-ranked participants. This puts countries under pressure to improve their systems and/or adapt to 'winning models'. Fourth, the OECD's policy recommendations have taken on an authoritative character (Grek 2009, 25) and tend to be more frequently accepted as valid by a broader spectrum of politicians and scholars than previous endeavours. Finally, PISA focuses on education output and how students are capable of applying knowledge and skills learning in school for their future working life and society. Thus PISA correlates well with the Lisbon Strategy of the European Union, which aims to make the Europe the most dynamic knowledge-based economy in the new century and has substantially affected the education reform discourse in France (see Aghion and Cohen 2004).
Analysing France's weak PISA performance With regard to the results, not only the below-average performance of French students strikes one's eye, but also the noticeable deterioration between The results show that among the OECD countries in particular Finland and the Netherlands, several English-speaking countries (Canada, New Zealand, Australia and Ireland) and various Asian countries (Japan, South Korea, Singapore in 2009) perform significantly better than France. The study also revealed that the number of strong-performing students has not decreased (contrary to the Japanese case), rather that the number of students with learning difficulties has increased between the PISA rounds (OECD 2010b; Rollot and de Verges 2007). The PISA studies have also shed light on additional less flattering features of the French education system. First and as indicated above, the system produces a high number of underachieving or failing pupils and, at the same time, the performance of the French elite tends to be lower than those of peer countries (Meuret 2003b, 96). Additional points of critique are not only the high failure rate in the baccalauréat examination, 10 but also the fact that only 93% of pupils reach the last school gradethe preparatory class for the baccalauréat. 11 In addition, the study reveals that the number of pupils who do not feel at ease in school is twice as high in France as the OECD average (OECD 2007). Accordingly, the participating pupils also gave a comparatively poor evaluation of the quality of teaching. A large number of French youths believe that they are not sufficiently supported and encouraged by their teachers during the learning process (Grenet 2008;OECD 2001). For example only 43.4% of French 15-year-olds claimed that their teachers actively and individually support them during the learning process (OECD-average 64.6%) (OECD 2002, 392; see Meuret 2003b for a similar conclusion).
The academic debate triggered by the poor French performance has revealed that France has fallen short on two major fronts: the pedagogical approach and educational equality. First, it appears that there is not sufficient leeway for individualized and tailor-made teaching and learning methods (see Rémond 2006, 79). According to Grenet (2008) the PISA results also indicate that the dominant pedagogical approach excessively promotes the acquisition of passive knowledge, which is then reproduced in standardized tests. In other words, that French pupils have difficulties in expressing critique or dealing with situations, which require independent analytical thinking (2008; see also Baudelot and Establet 2009, 26-7;Rémond 2006). In this context, the OECD also criticizes that the lessons are too monotonous and teacher-centred and that there is an absence of methodological diversity (Sérès 2008). Furthermore, the results point to a high level of fear and low self-confidence among pupils. French pupils also often refrain from providing an answer to questions which require a personal opinion or independent analysis, instead of taking the risk of giving a false answer. 12 In addition, a large share of secondary students views their own school achievements and performance pessimistically (see Rémond 2006). This pedagogically oriented interpretation of the weak PISA results is also underscored by Meuret (2003, 99). Compared to Scandinavian and Anglo-Saxon countries, various institutional characteristics of the school system (e.g. learning climate, level of discipline, support for weaker students) are relatively favourable in France. This applies also to extra-school factors such as expectations, pressure from family and the social environment, which are not regarded as major obstacles to performance. Instead, Meuret (cautiously) traces the poor performance back to the attitude of teachers towards students and a large degree of homogeneity among them with regard to values and teaching practices (see also Dubet and Martucelli 1996). He concludes that teachers, who give strong consideration and attention to all students, are less frequent in France than in English-speaking countries (Meuret 2003b).
Secondly, the PISA results reveal that socio-economic status is still a key factor in explaining the success or failure of individual participants in the education system. The PISA study shed light on significant disparities based on socioeconomic backgroundin a country that attaches great importance to social equality (see also OECD 2010)and the fact that the school system may reinforce inequalities through widespread practices such as redoublement (grade repeating). 13 In particular, the Finnish, South Korean and Canadian PISA results show that overall performance correlates positively with a low disparity between the best and worst performing pupils. However, France has a high number of advanced pupils and at the same time a large share of particularly weak performers (Baudelot and Establet 2009, 38-44). This performance disparity and France's alleged inability to cope with school underperformance and underprivileged students were subsequently highlighted in the widely publicized study Que vaut l'enseignement en France? (What is teaching worth in France?) (Forestier, Thélot, and Emin 2007), hence increasing pressure on policy-makers to take corrective action.
PISA as a facilitator of education policy change? How have French education policy-makers reacted to the sub-optimal PISA results and the ensuing academic and public debate? And to what extent have the PISA study and international assessments in general substantiated concrete policy measures? Altogether, the public debate on the PISA study remained relatively discrete and primarily limited to academic and OECD representatives based in Franceat least after the first two rounds Pons 2009a, 2009b). Political figures tended to display their satisfaction that not only the general level of education had demonstrably increased since the 1960s, but also the number of baccalauréat graduates (see Baudelot andEstablet 1989, 2009;MEN 2009). Particularly noticeable are the efforts of various French observers and governmental officials to shed doubt on the accuracy of the PISA study. To this end, two argumentative strategies have been employed. As indicated above, some observers have argued that the PISA methodology bears various biases to the disadvantage of France. 14 Second, the sub-optimal results are also frequently traced back to several peculiarities of the education system, which perhaps have a negative impact on French performance. One argument relates to the widespread practice of redoublement ('doubling', repeating a class, staying back) in France. It is assumed that French pupils are at a structural disadvantage in international comparisons, because almost half have repeated at least one class. Here indeed lies a large discrepancy. While by international average 5-10% of students have repeated a class (in Finland 2.8%), this figure amounts to approximately 40% in France (see Meuret 2003b; see also Duru-Bellat and Suchaut 2005). According to this argument, the weak overall results of France can be explained by the fact that a large portion of French 15-year-olds have not reached the same level of academic progress of participants from other countries, in which classes are less frequently repeated (see Grenet 2008;Meuret 2003b). In fact, the results of French youths, who have not repeated a grade, are at the level of the top PISA performers, while the results of those who have repeated at least one grade, lie far below the OECD average. 15 Hence, it is agreed among both policy-makers as well as academic observers that inequality is reinforced by structural features of the system. Nevertheless, at the beginning of the decade the education ministry and its Evaluation Department (DEPP) attempted to keep the PISA results far away and hidden from the public. 16 Moreover, the DEPP lamented lacking support and interest from upper ministerial cabinet members (see Pons 2009a, 16, 2009b, 18, 46). The withholding of the results can also be traced back to efforts of the then newly elected Education Minister Jack Lang to appease French teachers and leftists, who in part perceived the OECD as a neo-liberal organization lacking legitimacy (Pons 2011). Furthermore, French labour unionswith the exception of the Confédération française démocratique du travail (CFDT)had not taken a clear stance on PISA in the early 2000s (Mons and Pons 2009b, 35).
After the end of the cohabitation phase (2002), the education ministry initially triggered a debate on 'ideal' education policy and governance (gouvernance idéale). Together with the high level of youth unemployment, the Lisbon Strategy and the resulting focus on the economic value of education, the PISA study belongs to a bundle of factors which contributed to the national debate on the crisis of the education system (Meuret 2007) and awoke the spirit of reform of diverse political forces. In this context, the so-called Commission on the National Debate on the Future of Schools (Commission du débat national sur l'avenir de l'École) was established in 2003 under the leadership of the French education expert and former director of the Haut Conseil de l'évaluation de l'école, Claude Thélot. The resulting Thélot Report addressed, among other things, the weaknesses of the education system in dealing with underperforming students and provided various reform proposals, based on the principles 'educate, teach, integrate and promote' to optimize the French education system (see Thélot 2003; see below). 17 However, for the period after 2006 Mons and Pons (2009b, 47) speak of a rupture, i.e. a break with the previously 'lax' handling of the below-average PISA results, and a new education policy activism and alarmism drivento a significant extentby transnational comparative governance. The new reactionary approach can be traced back, in particular, to the continual and significant deterioration of the French PISA results (see Table 1) as well as the increasingly widespread notion of education as a crucial factor behind economic competitiveness (Aghion and Cohen 2004). It is particularly noticeable in this context that the PISA study and other international comparisons, such as PIRLS, have been increasingly politically exploited in the past five yearsby both major political camps. While centre-left politicians and the closely associated labour unions (e.g. Solidaires Union Syndicale (SUD)) interpret the high performance disparity as a symbol of the political failure of the centre-right, President Sarkozy and his adherents have strived to legitimize their own education policy reforms by referring to PISA (Mons and Pons 2009b, 47).
Against this background, one can observe since the 2006 PISA study two distinct developments. First, the increased interest in international comparative assessments has resulted in the visible development of a 'culture of international comparison'. The French government has attempted to institutionalize a result-and evaluation-based approach with a stronger reference to international comparisons than in the past (see Mons and Pons 2009b, 63;Pons 2011, 62). Particularly noteworthy is not only the appointment of Xavier Darcos, the former French ambassador to the OECD, to education minister, but also the increasingly frequent public appeals for the targeted application and analysis of international comparative assessments (see e.g. Forestier et al. 2007). 18 This is reflected in explicit statements of the president as well as Xavier Darcos (see Darcos 2007;Jacob 2008;Mons and Pons 2009b). Second, PISA and related comparative assessments have apparently facilitated the development of a stronger 'reform advocacy coalition' in France (see Sabatier 1987). The weaknesses exposed by PISAlacking pedagogical accountability and lacking educational equalityhave provided a stronger basis for the linkage of reform objectives of the French centre-right and centre-left, the former of which aims above all to boost performance, accountability and pedagogical autonomy, while the latter aims to reduce social inequalities. This linkage of policy goals has also been facilitated by the increasing mobilization of some of the OECD's previous biggest criticsthe labour unions. In particular the CFTD and UNSA have recently drawn on PISA as a rhetorical tool for pressing for greater education equality (CFDT 2010;UNSA 2010). The broader reform coalition has also been strengthened by support from the DEPP, which has distanced itself from its previous critique of the PISA methods and instead increasingly advocated concrete policy reforms (see Mons and Pons 2009b), as well as High Inspectors for secondary schools, who have increasingly legitimized their desire for teaching reforms and more pedagogical autonomy with the PISA results.
As previously mentioned, the French government has favourable institutional conditions for producing policy change, in particular since the end of the last cohabitation-based government constellation in 2002. And the status quo of the secondary education system should be alarming to the French government, not only due to the above-mentioned reasons. Diverse indicators also show that France is among the OECD forerunners, when it comes to state expenditure for education. For example, in 2007 the expenditure per secondary-level student in France amounted to 9303 US-Dollars and was thus higher than the OECD average of 8006 US-Dollars (OECD 2009). The French government thus spends more (and more) money for education (MEN 2003), but only achieves below-average results by international comparison. Precisely this high expenditure has led to favourable conditions both for French schools and pupils. With 10.4 students per teacher, France has one of the lowest teacher-student ratios in secondary education in the OECD (BMBF 2003, 186). Too little time in class is also not an issue: while 15-year-old Finnish pupils have an average of 846 h of lessons per year, this figure amounts to 1036 h for French pupils (OECD 2009, 314; see also Duru-Bellat and Suchaut 2005, 187). Moreover, France is among the OECD countries with the largest offer of remedial education for weak performers. Against this background the PISA results offer education reformers an ideal argument: France spends a tremendous amount of money on education, which is reflected in the high number of teachers and the diverse state support and monitoring measures, but only obtains below-average and, in part, deteriorating results. This view is also underscored by Meuret (2003b), who examines the organization of the education system as well as additional variables such as parental support and school discipline, and concludes that the institutional parameters of French secondary education are comparably favourable.
The past 10 years can indeed be regarded as a phase of strong reform dynamics in French education policy. On the basis of the proposals of the Thélot Commission and the grand public debate on education in 2003 and 2004, in which teachers, students and parent representatives as well as education policy-makers participated using new media technology, the so-called Fillon Law (Loi Fillon)named after the then education minister and current prime ministerwas proposed with specific reference to the French PISA results (Mons and Pons 2009b;Pons 2011) and France's contribution to the Lisbon Process, i.e. making the European Union the most competitive and dynamic knowledge-based economic area in the world (Weber 2005). Moreover, with targeted measures aimed at promoting equal opportunities and school democratization, policy-makers drew on the arguments of Baudelot and Establet (2009, see above) to garner support from part of the progressive left. Passed by the National Assembly in spring 2004, the law pursued two primary goals: to reduce the number of school drop-outs and increase the general level of education of French students. Besides the proposed reform of the baccalauréat procedure (see below), the following goals were to be achieved by 2010 (Loi Fillon 2005): (1) Increase the number of baccalauréat graduates from low-income families by 20%.
(2) Increase the number of girls in scientific and technical tracks (filières) by 15%.
(3) Increase the number of teachers in advanced training by 20%.
(4) Increase the share of pupils who reach level B1 19 in the first foreign language (in most cases English) by 20%. (5) Increase the number of pupils learning German by 20%. (6) Increased monitoring of the acquisition of basic knowledge (French, basic math skills, modern means of communication, at least one foreign language, basic skills in humanities).
However, the Fillon Law did not fundamentally transform French education policynot least because important reform components were retracted after intensive and enduring student and teacher strikes. Here, the desire to uphold the guiding principle of educational equality appears to have been a significant reform obstacle. For example, the aspired reform of the baccalauréat, which was aimed at easing the student workload, was rejected by a significant number of students. The reform was supposed to reduce the number of examined subjects from 12 to 6 and part of the final result was to be based on performance during the past three school years and not only on the examination result. The students, who are the main sufferers of the centrally organized final examinations, wished for the most part to maintain the previous rigorous testing procedure (Die Zeit 2005) and forced education minister Fillon to reconsider the project. Concerns were expressed that teachers would not be able to objectively and neutrally judge students' performance without reference to their social background. Student representatives also argued that only the central government is capable of guaranteeing objectivity, anonymity and thus the education ideal of equality. Furthermore, it was feared that the stronger incorporation of individual schools and teachers into the examinational procedure would reveal details on the place of residence and thus the social status of the pupils' families (ibid). The decentralization of the testing procedure and the resulting greater autonomy for schools was thus regarded as an attack on the principle of égalité, even if the state was aiming to relieve students from the rigorous testing procedure characterized by intense cramming and memorization.
A Finnish reform recipe? Although the PISA results do not offer a patent remedy to increase the education performance of French students, the frequent references of high-ranking governmental representatives to the Finnish education system in recent years are particularly noteworthy. According to various observers the Finnish recipe for its PISA success lies primarily in the comprehensive teacher training, the flexible and diverse pedagogical approaches, in the school support offers, and in the highly integrative pedagogical culture (Bruneel 2008;Jacob 2008;Robert 2008;Thélot 2005). Finland's success thus has inspired a rather idealized 'Finnish PISA hypothesis' 20 among French policy-makers that high scores of schools correlate with their degree of autonomy and their targeted efforts to promote education equality through social integration. 21 According to the Sarkozy government, the reasons for the sub-optimal performance of French 15-year-olds lie not in the PISA methods, rather in the lacking pedagogical freedom and autonomy for schools and teachers (Sarkozy 2007), a view also supported by academic observers (see, e.g. Meuret 2003b).
These circumstances have in turn facilitated a convergence of interests between those pushing for more pedagogical autonomy, flexibility and result-oriented accountability and those pushing for more equality and the more effective integration of weak performers. In other words, from a French perspective, the Finnish model can be characterized by its emphasis on social integration, adaptability and flexibility (see Robert 2008), which in turn has provided solid arguments for both political camps.
Against this background and in view of the clear deterioration of the French PISA results, the French government began to take an alarmist stance and a new education minister, Xavier Darcos, was mandated with the implementation of a broad range of secondary school reforms 'with a Finnish touch' (Jacob 2008). For example, schools are to be empowered to pursue their own education policy strategies and pedagogical approaches and granted more financial autonomy for the implementation of government-promoted reforms. Furthermore, the school year is to be redesigned into two semesters based on the Finnish high school system. Along with that, the allegedly segregation-promoting carte scolaire (see also Sarkozy 2007) is to become somewhat more flexible so that students from outside a school district may also be admitted to schools. The reforms also promote the principle mixité sociale (social diversity), which is increasingly regarded as an educational policy recipe for success in view of the Canadian and Finnish PISA results (Bruneel 2008;Sarkozy 2007; see also Duru-Bellat and Marin 2010). Moreover, additional teaching staff is to be deployed to 'problem schools', in order to develop innovative projects to alleviate social conflicts and overcome performance disparities. According to the slogan 'lycée à la carte', the Darcos reforms envisioned 6 h of elective subjects in addition to 31.5 obligatory lessons each week. In addition, 3 h of special education were to be introduced for slow learners andas a sign of the stronger market orientation of the education systemeconomics was to become an obligatory subject for all pupils. However, subjects such as history and geography were to become optional in the last school year, which was harshly criticized by intellectuals (France24 2009). Altogether, many of these measures also intended to limit redoublements, i.e. repeating classes, which had a particularly negative impact on France's PISA performance (see Duru-Bellat and Suchaut (2005, 190) for the negative impact of grade repeating on equality).
Although other reform objectives were to relieve students of the heavy workload, increase targeted learning opportunities, grant additional autonomy to schools, and provide additional support to schools with a high share of weak-performing pupils, nationwide demonstrations and strikes broke out in late 2008. The protesters students, teachers and labour unionscriticized above all the looming elimination of 13,500 teachers' jobs resulting from a simultaneous overarching reform of the public sector. Furthermore, the teachers argued that the planned additional special lessons would be a too large burden on teachers and pupils and would contradict the ideal of equal education opportunities (i.e. in this context equal attention for all pupils) (Cody 2009).
During the political process, it also became apparent that strong centralization is not necessarily always a reform-promoting force, rather may also trigger public backlash. Thus, not only the proposed policy content was regarded as objectionable, but also the political process leading to the reform proposals became a matter of debate. After an unparalleled wave of aspired and in part achieved reforms (e.g. in the pensions system, higher education system, public service and in labour, immigration and fiscal policy) the reform opponents argued that the government had made unilateral decision without respecting the demands of students (Der Spiegel 2008). In other words, the strict separation between state policy-making and society addressed by Katzenstein (1976) came to bear. This resulted in a situation in which the protesters were less interested in the substantive aspects of the reform, rather in what they regarded as an excessively state-centred decision-making process. Compared to the more corporatist political process in Germany, the Netherlands or Austria, for example, organized interests in France are not well integrated into the early phases of the policy-making and planning process (Quittkat 2006, 41). The highly state-centred policy-making process frequently leads to situations in which strikes, from viewpoint of reform adversaries, are the only practical means of influence on policies already decided by the government.
Due to the intensity of the strikes and the fears of an escalation of violence in urban areas, the reform was initially put on hold and a new education minister appointed. However, in early 2009 a new initiative known as Lycée pour tous (upper secondary school for all) was proclaimed in which students, teachers and parents were publically consulted about their ideas with regard to the future of secondary education. The new education minister Luc Chatel then presented a new concept for the reform of the lycées later that year, which no longer aimed to modify the baccalauréat procedure. The reform proposal not only strictly separated education reforms from the broader public sector reforms, but also focused primarily on substantive and pedagogical issues. Its primary aim is to assure a stronger correlation between the qualifications acquired by the pupils and the demands of the knowledge economy of the twenty-first century (Chatel 2010). Hence, these reform components were supportedwith reservationsby some labour unions, in particular the CFDT. The measures also aim to support weak-performing students and to design classes in a more flexible and individual manner and bear unmistakable parallels to what is perceived as the Finnish recipe for success (see OECD 2011; Rautalin and Alasuurtari 2009). Specifically, the current reforms of the lycée provide two hours of personalized special lessons in small groups and individualized systematic support and development tracking for all pupils. In addition, integrated class lessons (enseignements communs) (MEN 2010b(MEN , 2010c are to be expanded as well as voluntary additional lessons during vacation to support students wishing to avoid repeating a class. Moreover, students are now provided the possibility to select specialized courses (filières) later than previously, while two 90-min lesson units (enseignements d'exploration) are being introduced in the first upper secondary grade to prepare students for graduation. Even though various aspects of the reform package are still disputed and are being further negotiated, the government was able to assert itself this time, so that the reform is being gradually implemented since late 2010.

Conclusions
The analysis has shown that the PISA study has brought new momentum to French education policy, despite an earlier period of detachment from the OECD. International comparative statistics and evaluations have not only stimulated broad public debate on the shortcomings of the French system and provided examples of foreign best practice, but also justified concrete education policy reforms. This applies, in particular, to the most recent reforms, which purportedly align the French education system with the demands of a competitive knowledge-based economy. In this context, the allusion to the Finnish model has also provided a window of opportunity for a convergence of policy ambitions between right-and left-wing political forces.
As a result, the orientation towards Finnish pedagogical methods and education policy instruments is unmistakable: classes and lessons are to become more flexible and tailor-made to pupils' needs, while schools are to acquire more autonomy with regard to pedagogical methods and the support offer is to be further expanded. Moreover, the reforms are aimed at reducing the high performance disparity, the number of drop-outs and redoublements.
As discussed in the introduction, domestic political institutions can explain the reaction of a country to transnational processes and in particular the results of international comparative assessments such as PISA. In the French case, we assumed a reserved, moderate reaction to the PISA study due to two reasons. On the one hand, the clear parliamentary and governmental majority (i.e. no cohabitation constellation since 2002) and the strong concentration of power in the central government provide favourable conditions for governmental action. On the other hand, in France there is a profound scepticism towards policies, which are perceived as neo-liberal and potentially endanger the guiding principle of educational equalityeven if various studies, including PISA, demonstrate that the education system tends to produce and/or reinforce inequality (see Duru-Bellat 2002).
This assumption appears to have been confirmedwith several reservations. After enduring protests, important components of the Fillon reform of the baccalauréat had to be retracted, and after long-lasting strikes a watered-down version of the original Darcos reform of upper secondary education (lycée) is coming into force. Seen from the outside, the reactions of the reform opponents may appear to be contradictory: the concrete reform measures of the government were interpreted by some leftist unions in an almost reflex-like manner as symptomatic of a 'neo-liberal virus', forced 'Americanization' and as an affront to the principle of égalité (Sud Education 2009), even though they were in fact borrowed from or inspired by Finland and other countries with a high degree of social and educational equality (see Robert 2008). In other words, one might argue that some reform adversaries are clinging to a principle which has not been realized and to a policy framework, which may be detrimental to its realization.
However, beyond formal institutions and guiding principles, other factors also appear to have been crucial. First, one must take into account the specific institutional framework in which evaluations are processed. The already existing French 'evaluative state' which is embodied by the DEPP and marked by a high degree of bureaucratization and centralization (see Pons 2011) provided an institutional basis for knowledge distribution among policy-makers and channelling evaluation outcomes into concrete policy measures. Second, the linkage of policy goals à la finlandaise between centre-right (flexibilization, pedagogical autonomy, accountability) and centre-left (equality promotion, support for underperformers) played an equally or more significant role than formal political institutions.
Despite its sceptical stance towards international comparative studies, France is today eagerly aiming to improve its education performance. While the country first tried to ignore its mediocre PISA results, it today cannot hold against the pressures of growing transnational governance in the field of education. Like in several other countries (see Niemann 2010 on Germany, Bieber 2010 on Switzerland), PISA has thus provided an additional window of opportunity for overcoming the backlog of overdue and deep structural reforms.
18. The study includes the results of a comprehensive evaluation of the French school system by the Haut Conseil de l'évaluation de l'école and argues for a stronger orientation of French education policies towards international comparisons. 19. Reference figures of the European Council for modern foreign languages. 20. Duru-Bellat and Suchaut (2005, 193) warn that the assumed cause-effect relationship between good performance and school autonomy has not been fully established and may blend out other factors. 21. See Simola (2005) for a more cautious interpretation of the Finnish miracle and the university applicability of its education model elsewhere.