Abstract
As the policy field of education has become increasingly internationalized over the last two decades, international organizations like the Organisation for Economic Co-operation and Development (OECD) play an increasingly decisive role in the diffusion of knowledge, monitoring, and research in global education policy. Although the OECD lacks any binding or coercive governance instruments in relation to states, or ability to provide material incentives for compliance, it has nevertheless successively expanded its influence regarding education. From a perspective of social constructivism, we argue that the transmission of ideas and information generated through ratings and rankings can be viewed as a crucial governance tool for the influence of an international organization (IO). Our article seeks to analyze how the OECD uses large-scale education assessments to promote the economically based idea of human capital and related learning techniques in education policy – thus influencing national education systems. Furthermore, the OECD and its distinctive approach of soft governance through putative hard fact may become a role model for other IOs, both in the field of education and beyond.
Introduction
In contrast to trade, security, or environmental policy, education displays no vital need for international cooperation to achieve an overall better policy outcome. There is no obvious underlying dilemma calling for multilateral coordination as in other fields, leading to a lack of interdependencies which would otherwise encourage states to cooperate with one another (Porter and Webb, 2008: 43). Education simply exhibits no semblance of zero-sum logic, no necessity for the increase of educational outcome in one state to coincide with a decrease in another. Rather, education is considered a core responsibility of the state in order to generate social coherence, economic independence, and national identity (Nagel et al., 2010: 15). As its prerogative, education policy serves as a means to educate its people in civic rights and duties, enable them to succeed in the labor market, and teach national history and languages. Accordingly, this policy field seems to be traditionally dealt with exclusively on the national level only.
Nevertheless, in the wake of ongoing globalization processes and worldwide competition, education policy is indeed becoming increasingly internationalized (Deacon, 2007; Mundy, 2007). School and university training is considered to play an important role in contributing to the further economic, scientific, and social development of progressively intertwined societies. National education systems have to respond to new challenges posed by the emergent global knowledge economy (Robertson, 2005). From the early 1990s on, there has been an observable, steady increase in international exchange and cooperation, particularly through IO activities in global education policy (Martens et al., 2007). A striking peculiarity of this current trend is the prominence of comparative quantitative evaluations of national education systems and the rising numbers of countries participating in such international large-scale assessments (Benavot and Tanner, 2007). It seems fair to say that education policy has developed into an arena of increasing international coordination.
Expanding education into the international arena can be analytically grasped as a complementary process of two developments. On one hand, international institutions, like IOs, are directly mandated by states to administer and facilitate multilateral cooperation in the field of education that states could not otherwise realize on their own. IOs carry out these tasks by establishing rules, procedures, or standards for monitoring and reporting governance indicators and mechanisms (Arndt and Oman, 2008; Davis et al., 2012; Von et al., 2012). Historically, state actors themselves have been the main driving force behind the formation of international initiatives in education (Fulge et al., 2016). On the other hand, these IOs may exceed their predefined mission and produce unintended consequences, lock-in effects, or path dependencies for state actors. In education, IOs autonomously expanded their portfolio and successively became important players in the field without having a strong predefined mandate for doing so (Martens et al., 2007). Among these, the Organisation for Economic Co-operation and Development (OECD), the International Monetary Fund (IMF), and the World Bank are some prominent examples. By becoming proactive in education, they have not only shaped the international field of education but also reinforced their position as influential and autonomous actors vis-à-vis their member states.
In this article, we focus on the OECD as a prime example of an IO that has constantly expanded its expertise and influence in education policy (Henry et al., 2001; Martens and Jakobi, 2010). It can be regarded as an example of a de jure powerless IO, but one which has gained regulatory influence (De Francesco, 2016). In our contribution, we argue that the OECD influences national education policies by means of soft governance through hard facts; that is, by changing its own strategy toward comparative quantifiable assessments in education policy, it augments its leverage over states. Such governance strategies could possibly become a role model for other IOs, particularly the OECD’s governance by numbers technique (Erkkilä et al., 2016; Espeland and Sauder, 2007). In other words, the OECD has become a central node in the network of international education politics, allowing us to now speak of an OECD-fication of global education policy.
The power of ideas in the realm of soft governance
Whether intended or unintended, consequences of the globalization of education policy for states are clearly visible: national policy makers and other stakeholders in education can hardly ignore initiatives from the international level, which are widely perceived to have clout by the public. These international impetuses frame problems and priorities and challenge domestic policies, politics, and traditional conceptualizations of education. Central to this regard is the role of IOs and their ability to gain autonomy and authority. Often equipped to set agendas, prepare and shape decisions or foster implementation, IOs are more than the sum of their member states’ interests (Koremenos et al., 2001).
While it is well understood that international institutions matter, the way in which they actually impact states’ behavior remains the subject of a contested theoretical debate. Especially in an asymmetrical, top-down setting without enforceability, it is pivotal to provide an explanation of how the ‘top’ exerts influence on the ‘down’. In addition, asymmetries can also be constituted differently, if, for instance, one party possesses information which other parties do not. In the absence of command and control, IOs make use of their ability to produce information and knowledge to generate influence and exert soft governance (Conzelmann, 2008: 44). In consequence, soft governance cannot be equated with traditional hierarchical steering.
IO soft governance implies that although IOs are set up by states and consist of state delegates, they are able to develop their own positions, ideas, or dynamics because of intra-organizational networks and interactions that cannot be fully controlled by any principals (Hawkins et al., 2006). Despite the provision of a clear mandate on how to act, IOs can go beyond their previously defined roles and generate new aims and administrative activities that exceed their initial purpose and scope. With time, IOs even act against the interest of their founders, thereby exerting influence back onto their member states and beyond, given the potential for agency slack (Koenig-Archibugi, 2006). This may be accomplished when IOs – or more specifically, their bodies (e.g. secretariats, departments, and working groups) – embrace a topic and develop their own ways of dealing with issues: instead of simply carrying out what their member states urge them to do, IOs follow their own interests and agendas (Barnett and Finnemore, 2004).
For such governance to be possible, an IO needs to be accepted as a ‘cognitive authority’ in the given policy field (Broome and Seabrooke, 2012, 2015). Authority can be granted by the state principals (e.g. through conventions, constitutions, and legal acts) or acquired through action, expertise, and routines. Furthermore, authority of an actor lacking coercive powers is strongly linked to the aspect of legitimacy. An indicator of an IO’s legitimacy is its reputation; IOs with ‘good’ reputations, that is, a reputation for being rational and impartial, are accepted as legitimate sources of information and advice to a great extent due to the fact that they exhibit apolitical and technocratic expertise (Barnett, 2002: 113). The impact of IO outputs ‘is inseparably bound up with judgments about the reputation of that institution’ (Sharman, 2007: 30). The perceived legitimacy of an IO leads others to follow its recommendations.
How then do IOs govern? In general, ‘governance […] encompasses the activities of governments, but it also includes the many other channels through which “commands” flow’ (Rosenau, 2005: 46). Thereby, a command can be understood as any token that implies a behavioral adaptation in the sense that something should be done (or omitted) (Niemann, 2014a, 2014b). This understanding of governance presents IOs (and other inter- and transnational actors) as having the ability to create, diffuse, and implement rules, norms, and standards through means of soft governance rather than through binding legislation understood as hard law (Abbott and Snidal, 2000). Hence, the perspective of governance entails a shift from states to a multiplicity of regulatory actors, from hard to soft law, and from formal to informal rules (Mingst, 1999: 93).
Social constructivism offers explanatory strength in assessing IO governance capabilities by highlighting the ability of IOs to promote the normative value of a certain norm as legitimate and worthy of acceptance. IOs do not revert to a set of predefined prescriptions to influence others’ behavior but convince the addressees to do (or omit) something by sound arguments rather than sticks or carrots. This discursive power of IOs, which relies on their moral authority, is depicted as ideational rather than material as it refers to influencing states and others to adhere to IO outputs. In this sense, IOs disseminate certain norms and constrain behavior that is not in accordance with a promoted norm by discursive means (Joachim et al., 2008: 11) by reverting to mechanisms such as shaming or prestige rather than coercion or material incentives (Manners, 2009). Hence, this type of soft governance can also delegitimize a certain behavior by establishing the understanding that it runs counter to an aspired higher goal. In this case, behavior is stigmatized as socially unacceptable or undesirable.
Thus, both the social creation of common knowledge as a standard in a policy field and the role of IOs in shaping international discourse are essential for soft governance (Abbott and Snidal, 1998: 5). Since IOs utilizing soft governance rely on their function as advisors and opinion leaders, one key element is the role and dissemination of ideas. 1 The central argument in this respect is that the proliferation of ideas and ideational change in turn promote policy change. Ideas serve as a cognitive framework for interpreting an issue, identifying something as a problem, and rendering suitable solution strategies. Consequently, how agents act in the world is strongly determined by how they perceive their environment. Furthermore, the perceptions of the actors’ environment are neither stable nor fixed. Perception, in the first instance, is a matter of interpretation.
Ideas serve as ‘cognitive filters through which actors come to […] conceive of their own interests’ (Hay, 2011: 69). First, ideas shape the definition of an issue as a problem. In this regard, the reinterpretation of a policy in the light of a new idea reveals that something is wrong. Second, by identifying something as problematic, ideas can also indicate goals – that is, a more desirable policy. Third, suitable means for accomplishing the (new) goal are communicated through ideas. Ideas are not just tools in the hands of strategic actors (Lieberman, 2002: 699), they need agents to be disseminated. IOs act as these disseminators or broadcasters of ideas (Djelic and Sahlin-Andersson, 2006: 17) and aim to ‘nurture people’s identities, helping them to construct their fundamental values which, in turn, shapes their beliefs and interests’ (Béland and Cox, 2011: 9). IOs (and other institutions) help to define what (domestic) stakeholders want and provide them with the justification for why they want something. In this regard, an IO (and its staff) can be a ‘transfer entrepreneur’ (Nay, 2012) in that it determines and disseminates policy goals that shape national policy frameworks, policy instruments, and policy contents.
Two interlinked modes in exerting soft power can be derived from the work of Mahon and McBride (2009): the inquisitive and the meditative mode. The inquisitive mode involves the gathering of information regarding a specific issue. In this context, the ability to define something as a problem becomes pivotal. This means IOs first create a common shared knowledge by providing information, which was otherwise not accessible. The collected data are then interpreted against the background of views and ideas within the IO. Pure information is transformed into substantial knowledge (Barnett and Finnemore, 2004) as a basis for further (soft) governance actions. It is deemed essential to back arguments with empirical evidence, as they must be proved to be conclusive and sound. In addition, the meditative mode addresses direct contributions to the policy discourse. IOs make recommendations to their members (and beyond) on the basis of publicized information and findings about best practices in a certain policy field, and consequently lobby for them (Martens and Jakobi, 2010). This lobbying can take different forms, for instance, as recommendations which illustrate directly how to act in a policy field. More indirectly, recommendations can also emphasize the behavior of a peer actor in order to serve as a blueprint.
Overall, soft governance by IOs is characterized by discursively constraining the frame of appropriate behavior. This governance technique aims at increasing or decreasing the legitimacy of a certain behavior, and therefore tries to influence a policy outcome indirectly. IOs are able to govern their member states because they possess the authority to orient action and create social reality (Barnett and Finnemore, 2004, 2005) and, with the tool of ‘naming and shaming’, IOs can generate immaterial, rather than traditional, costs for behavior that is defined as not desired (Hafner-Burton, 2008; Joachim et al., 2008). IOs aim to frame a common understanding of the issue at stake and define goals for policy making by increasing or decreasing the legitimacy of a certain norm, policy, or behavior (Nay, 2014).
OECD as knowledge broker: empirical observations
How did the OECD mature into a reference point for education policy and how was it able to apply its influence at the national and global levels? In the following part, we trace empirically how the OECD has become a knowledge broker and norm entrepreneur in education policy over the last two decades. We use the Programme for International Student Assessment (PISA) as a prime example of the IO’s practice of exerting soft governance. From the year 2000 onward, PISA tests have been conducted triennially with the results published in the following year. Thus, by now we have had not only the results of six PISA circles but almost 20 years of worldwide testing through this OECD regime.
PISA aims to evaluate and compare education systems by testing the skills and knowledge of a nationally representative sample of 15-year-olds, an age in which students in many countries may start to make the transition into adult life. PISA tests focus on assessing students’ competencies in reading, mathematics, and science in the context of everyday situations. The number of participating countries gradually increased from 32 (of which four were not members of the OECD) in 2000; taken together, around 80 states all around the globe have participated regularly or occasionally in the study.
PISA as the OECD’s tool for global governance
Since the mid-1990s, the OECD has emphasized the production of human capital as a counteraction to the emerging effects of globalization (Henry et al., 2001: 45). Its thorough educational means were seen as a precondition to succeed on the global market. From the human capital perspective, education is an investment for which public authorities are responsible (Resnik, 2006). This perspective is largely borrowed from the Anglo-American ideological education framework, which focuses much more on the dimension of economic usability of education than other traditions. Education is defined as a driving force for growth and the OECD is committed to improving the quality, equity, efficiency, and effectiveness of their member countries’ education systems (OECD, [2010] 2011). The OECD’s understanding of education outcomes is closely coupled to the utility of generated knowledge. The value of knowledge depends on its utility in other areas (e.g. applied research, technological progress; Mangez and Hilgers, 2012). The emphasis on generating human capital became the cornerstone of the OECD’s education framework, from which it proactively pushes strategies and recommendations for intensifying the competitiveness of its member states through education.
The primary turning point in the OECD’s education activities was characterized by a shift from ‘discursive contributions to education policy’ (Martens and Jakobi, 2010: 15) to the gathering of empirical quantitative comparative data. It was this comparative turn which the OECD took gradually, starting in the early 1990s and boosting it to become an important player with discursive power in education policy: whereas in the past it has focused on each state individually acknowledging differences and idiosyncrasies, it now decisively compares states with each other and against standardized criteria. Such direct comparisons put states under greater pressure to reform their systems because the OECD’s statistics on education performances have become more easily accessible and interpretable by politicians, the media, and the wider public. (Martens, 2007: 40)
Most importantly, the OECD advanced its status as an influential education IO by designing, managing, and conducting PISA tests. Although from the side of the IO, it is argued that ‘[a]ll the OECD provides is evidence, analysis, and advice of the type PISA presents’ (Schleicher and Zoido, 2016: 375), its assessment tool is more than a database on education information: it has become a political instrument for the OECD. The IO and its staff not only solve methodological problems and technical issues of data collection but also make and promote decisions on how to perceive and present education goals. As Bloem (2016) has shown, while the OECD in the beginning of PISA still kept a rather neutral position as a mere producer of education data, it increasingly applied PISA as a tool to govern the international discourse on education.
While PISA is continuously expanding, there are still states that have not been scrutinized by the OECD’s education flagship. Non-participation is not to be equated with irrelevance, however. Even tiny states, like Grenada or Vanuatu, are influenced by PISA where it comes to reforming their own education systems or designing similar national evaluation mechanisms. For instance, in Vanuatu, the PISA results were used as a source of data to justify reforms and provoke policy debates. In Grenada, the PISA model proved helpful in improving the national education system regarding curriculum, the organization of education governance, and in shaping the qualification and preparation of teachers. 2 Thus, the study reaches well beyond the core group of OECD member states, influencing the education performance of emerging and developing economies. PISA is an internationally applied framework for education performance and is depicted as the ‘harbinger of changes in both the political frameworks and the educational objectives’ (Meyer and Benavot, 2013: 10).
Thus, today domestic education systems are evaluated on a global perspective. The OECD not only assesses voluntarily participating member countries in secondary education but is also arguably extending its range of influence gradually by expanding its participatory base. In 2015, for example, the IO started PISA for Development of middle- and low-income countries (Addey, 2017); in addition, it has run the PISA for Schools program for the last few years, allowing individual schools to be measured against both state and non-state entities in the database. Sellar and Lingard (2013) argue that the expansion of the OECD activities in education includes widening the scope of assessment by measuring a broader set of competencies, increasing the scale by covering more countries, and, thus, enhancing the explanatory power for policy makers and educators.
In general, the international acceptance of testing regimes is associated with key ideological forces that emphasize the globalization of national and international cultural, economic, and political structures (Kamens and McNeely, 2010). Hence, PISA reflects the demand of states to make education outcomes internationally comparable. At the same time, the comparative PISA study contributes to a global concept of ‘good education’: national education systems are evaluated against a predefined set of benchmarks concerning which factors produce the best performance outcomes.
Why does PISA seem to be an attractive tool for soft governance? PISA is easily accessible and useful for both experts and a wider public audience (Martens and Niemann, 2013). Experts can extract detailed quantitative information from the encompassing assessment data, establish mathematical correlations between diverse items, and produce policy recommendations on the basis of PISA data. At the same time, the broader public can learn how national education systems perform, are able to compare them in a general sense based on the provided league tables, and find out why some are better than others by reading further publications provided by the OECD. The general focus of the process has been on comparability, and data are processed in order to be interpreted in terms of ‘the development of common prisms’ (Woodward, 2009: 66). Since PISA has been conducted regularly for six cycles, these rankings are also compared over time to demonstrate if and to what extent a country has improved.
The normative advantage of the OECD’s governance
Although the OECD does not have any legal means or fiscal capacities to force states into compliance with its policy recommendations (Carroll and Kellow, 2011; Woodward, 2009), states nonetheless follow OECD advice (Breakspear, 2012) – or at least feel the need to justify decisions against these recommendations. However, the intuitive expectation for a strong correlation between domestic reforms and performance in the international education test does not hold true empirically. As we have shown in previous work, countries that perform similarly in PISA are not inevitably interchangeable with respect to their reform reactions to PISA (Martens and Niemann, 2013). It has been outlined that negative press coverage on countries’ performance in PISA would also reflect the negative public opinion about national education policy (Dixon et al., 2013). Taking domestic media reception as an indicator for the saliency of an issue in national politics, we observed that the plain performance of a country only has some minor influence on the extent and nature of national responses to international education studies: In regard to particularly poor results, the likelihood of it becoming a substantial issue of public discourse increased in countries such as Germany, Spain, Austria and Mexico. However, other countries at the lower end of the PISA league, like the US, Poland and Portugal, did not experience a public outcry reflected in the media coverage. As an overall trend, the worse the rank, the more differentiated the reactions are. (Martens and Niemann, 2013: 320)
Since other mechanisms must be accountable for explaining the impact of PISA, it seems worth taking a closer look at the soft governance activities of the OECD. The OECD made use of naming and shaming by evaluating national education systems, comparing them to other states, and deriving indirect implications from inferior educational performances. PISA possesses considerable influence on education policy through use of the media by ascribing the status of ‘winners’ and ‘losers’ to participating countries (Grek, 2009; Porter and Webb, 2008: 47). Obviously, no state, and in particular no industrialized OECD member state, wants to be labeled as an underperformer when measured by international education standards that define how well a state is prepared for future economic challenges. By making these publicly visible, the OECD is potentially able to stimulate national discourse in low-performing countries on the necessity of reforms or to reinforce the proven-as-successful policy paths in high performing countries.
At the same time, the OECD’s PISA showed which education policies seemed to produce better performance outcomes (in PISA) and those which do not (Bieber et al., 2014). In highlighting ‘best practices’, the OECD uses PISA to urge national policy makers to look across borders in order to identify international education policies worth implementing (Niemann, 2014a). This could be identified with regard to several OECD analyses which provided information on a multitude of education issue areas: for instance, early childhood education (OECD, 2011b), teachers’ training and working conditions (OECD, 2013), using new information technologies (OECD, 2007), evaluation and measuring of education performances (OECD, 2008), and school governance (OECD, 2011a). In short, the OECD had a sweeping portfolio on education policy data and presented it to enable policy orientation.
By identifying models of ‘what works’ as well as providing periodical reviews and comparative datasets, PISA results put pressure on national policy makers (and other stakeholders in the education administration and civil society) to improve national education policies and to reflect the identified best practice models in order to be competitive in globalized knowledge economy (Niemann et al., 2017). Through the generation and interpretation of PISA data, the OECD also highlights specific features of particular education systems that are appropriate for improving human capital production. For instance, school autonomy in combination with increased accountability measures is seen as integral to enhance effectiveness in secondary education (OECD, 2011a). Based on the knowledge derived from the PISA results, the OECD makes references to peer countries which should be converted to domestic education systems to boost academic performance. By doing so, the IO is able to set global agendas and diffuse policies in a variety of fields among OECD members and non-members alike (Ougaard, 2010).
Basically, with PISA benchmarking, the OECD generates normative pressure for lagging countries to implement educational policies consistent with better performing countries and best practices (Sellar and Lingard, 2013). The OECD makes resources and expertise available for other actors, introduces networks of experts, provides forums for coordination and advice, and acts as a surveillance and evaluation organization regarding the adoption of commitments, agreements, and the like. Hence, reporting procedures of the OECD can ‘provide ammunition crucial for “naming and shaming” techniques or for lobby campaigns by domestic coalitions’ (Conzelmann, 2008: 36). PISA was not (merely) an activity of the OECD to conduct research on education and to provide general information but was also designed and conceived as an instrument to support decision making (Mangez and Hilgers, 2012: 196). Hence, the OECD does not stop at knowledge fabrication, it seeks to disseminate this knowledge as policy advice.
Soft governance through hard (PISA) facts
The last decade witnessed comprehensive education reforms in several countries. It remains uncontested that international institutions have had an impact, but it remains an open question as to how they manage to exert influence. By using standardized tests as a means of measuring educational outcomes, large-scale assessments like PISA allow for the quantification as well as comparison of education systems across countries, regions, and even individual schools. These comparisons are used to identify strengths and weaknesses of education systems, which can prompt and shape educational reform processes. This trend has been observed in many countries, wherein school reforms were initiated following the publication of international assessment results (Lingard and Grek, 2007; Martens et al., 2014).
While some states reformed their education systems with clear reference to PISA, others did not. In fact, about 50% of all PISA participating countries initiated reforms in direct response to PISA (Breakspear, 2012). Take Germany as an example: it has been argued that the reforms in the German education system would not have occurred to such a great extent without the OECD’s PISA study (Münch, 2009). While it was contested that PISA (or the OECD) invented any new reform steps for the German education system, it is largely accepted that PISA was a catalyst for introducing overdue reforms. Accordingly, PISA in Germany became the object of intensified research (Niemann, 2016; Tillmann et al., 2008). While the awareness of potential German education deficits rose in the mid-1990s, PISA actually triggered a landslide of education reforms in Germany. While the country was below average in PISA 2000, it constantly improved and was among the group of states that performed above average in PISA 2012 (Prenzel et al., 2013). Overall, the diagnosis by PISA was not a one-shot observation but an identification of a systemic problem embodied in the German education system which was designed in such a way as to strengthen the already privileged students while neglecting disadvantaged students. This led to massive equity issues (Allmendinger and Leibfried, 2003), and since then the German policy discourse on education reforms has regularly addressed how socio-economic factors predetermine education performances (Niemann et al., 2017).
The political reactions in Germany to PISA were equally remarkable. Almost instantly after the publication of the first PISA results in December 2001, comprehensive education reforms were introduced in Germany to improve the educational outcome. German policy makers began to look beyond national borders, evaluating other education systems against the background of performance and aspects worth copying (Bundesministerium für Bildung und Forschung [BMBF], 2003). The focus was on early education in order to create a better basis for academic performances and on the advancement of socio-economically disadvantaged students (e.g. students with an immigrant background). Furthermore, and directly derived from the example of PISA, a stronger focus on the outputs of the education system and corresponding evidence-based policy making took place built on empirical evidence to guide policy decisions. In this context, comparative education research was massively promoted by launching several assessments to review performances by implementing comparative tests in schools.
Taking a look at the actual impact of PISA beyond Germany, we can identify a plethora of different reactions. Like Germany, Denmark was shocked by its first PISA results, particularly since Denmark’s Scandinavian neighbors performed much better in the first round. Substantial changes toward increased national assessment procedures and support for disadvantaged students, however, were only implemented after an in-depth international review (Egelund, 2008). After the surprising results spurred broad public debate, numerous studies analyzing these results were conducted, bringing policy recommendations to the fore as of 2003 (Egelund, 2008: 250).
Although Switzerland placed in the upper tier of the PISA league table in all testing rounds, about 20% of its students placed within the two lowest competence levels in reading literacy. Consequently, the results enforced already existing attempts of innovation, which accelerated the famous reform project, ‘HarmoS’, that finally harmonized the 26 different cantonal school systems. Swiss policy makers adopted a majority of PISA-based OECD recommendations for secondary education, such as social equity, school autonomy, and quality assurance, within only a few years (Bieber et al., 2015).
With regard to PISA’s impact on Japan, Takayama (2008) shows that PISA greatly influenced Japan’s education discourse and policy reforms. A perceived crisis in education policy erupted in the late 1990s in Japan as publications showed great shortcomings in the education system. In this climate, the results from the 2000 PISA round fell on fertile ground (Takayama, 2008). The Japanese government utilized the PISA results as an external source of legitimacy for highly sensitive policies and reforms (Takayama, 2008: 401).
After the publication of the 2000 PISA results in 2003, the Ministry of Education in Israel used the momentum to create a task force for educational reforms. Feniger et al. (2012) argue that the ministry had already been working on reforms before PISA brought in the first internationally comparative rankings on the Israeli education system. The newly formed committee emphasized reforms that clearly carried the thumbprint of PISA recommendations in terms of ‘managerialism and a globalistic approach’, including the goal to improve the country’s performance in international rankings (Feniger et al., 2012: 329).
Reverse reactions appeared in the United States, where a relatively low performance position in the league tables compared to other advanced economies did not necessarily lead to public or political responses (Martens and Niemann, 2013). Only with the 2009 PISA study, when the Chinese demonstrated extraordinarily good results, did PISA become central to education discourse. This was viewed essentially as a new Sputnik shock.
Similarly, the Chinese lead in PISA was interpreted as an omen of China overtaking the United States in its economic output.
In other countries, reactions to PISA vary between testing cycles and professions. For example, the British education system had already been substantially reformed in the 1980s, and standardized testing programs were already commonplace. England performed well in the first round of PISA in 2000 but dropped in later rounds. Although results did not improve significantly, reactions to PISA were moderate, and the British government employed a ‘pick-and-choose’ strategy to adopting OECD recommendations (Knodel and Walkenhorst, 2010). In Finland, reactions vary by status groups. Being among the top of the first two PISA testings, the government argued that comprehensive schooling should not be altered, rather that the country may serve as a model for others while Finnish teachers, as professionals, complained about the meager investments in education and the poor appreciation of the education system as a whole (Rautalin and Alasuutari, 2007).
Apart from case studies of single countries, systematic comparisons of policy changes in a larger number of countries are scarce. By looking at different aspects of accountability and assessment practices, Teltemann and Klieme (2016) showed that the use of standardized assessments increased throughout OECD countries between PISA 2000 and 2009. Likewise, the use of assessment for purposes of comparison between schools increased in many OECD countries. Other policies, such as school inspections and accountability in the form of tracked achievement data, show more mixed patterns of change between different rounds of PISA.
Taken together, the OECD seeks to frame how education should be organized. Thereby, the OECD empowers national stakeholders and interest groups to prompt domestic education reforms by providing sound empirical arguments for their views while others’ positions are eclipsed. As noted in Germany and Switzerland, the PISA findings were used to introduce assessment procedures and measures of accountability. In particular, employers’ organizations gained leverage in this regard by referring to PISA data while other groups against these measures, such as teachers’ unions, lost discursive power (Niemann, 2014a). They were no longer able to block these reform streams because of the indirect yet influential power of PISA. Hence, the OECD was indirectly able to change the power constellations in national education politics since the international organization strengthened those actors whose preferences and beliefs were identical to the OECD’s program (Armingeon, 2004). Actors who held different positions in education that were not supported by the empirical evidence presented by PISA lost power in terms of argumentative leverage.
Conclusion: a dissonant orchestra of IOs in education policy?
We have argued that the OECD is able to shape domestic education policies by a soft governance strategy of legitimizing and delegitimizing national education systems and practices. This strategy is based on comparative empirical data on the performance of domestic education systems. The provided examples illustrate that the OECD diffuses its ideas regarding education policy to the national level. At the same time, the OECD, first and foremost by the means of its PISA initiative, also defined a standard for assessing education performance through comparative, large-scale assessments. PISA became the most prominent and most comprehensive international education assessment to date.
Our findings also provide some implications for theorizing in the realm of social constructivism. First of all, the definition and dissemination of specific ideas, in our case in the field of education, is interlinked with sound reasoning based on data. Ideas become perceived suitable in framing political action if they correspond with empirical observations presented as hard facts. Ideas do not float around and get eclectically chosen by stakeholders; rather, they successively diffuse throughout the system by being linked to measurement and, eventually, become a cornerstone in the mind-set of actors. Second, an important caveat for explaining influence of IOs’ ideas on domestic policy making is that ideas work mostly through intermediaries. In this regard, IOs make an offer regarding their ideational framing. Actors on the national level need to be convinced by the argument of the IO as beneficial or legitimate and make use of it in their pursuits to introduce a change (e.g. in terms of reforms).
The field of global education policy is today populated by several IOs (and other international non-governmental institutions) constituting a dense net of actors competing for influence. Although various single studies on the ‘usual suspects’ (e.g. on the UNESCO, the OECD, and the World Bank) exist, we know little about the universe of education IOs as a whole. Of all 495 internationally operating IOs in the Correlates of War dataset, 20 IOs can be identified as working intensely in the field of education (Martens and Niemann, 2017). More scholarly work is needed on how these IOs interact with and influence both each other and states. In combination with our research result that the OECD is powerful in influencing states toward reforming their education systems, the subsequent research question arises whether the OECD has become a role model for other IOs. Since the OECD is often considered one of the most influential IOs in global education, it seems plausible to assume that other IOs are eager to copy or imitate the OECD’s approach and foster their own significance. In sum, a next step may be to identify if and how other education IOs operate in regard to education.
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
Author biographies
Dennis Niemann is post-doctoral researcher in the project “The global development, diffusion and transformation of education systems” at the Collaborative Research Centre 1342 “Global Dynamics of Social Policy” at the University of Bremen. He holds a PhD in Political Science from the University of Bremen. His research interests include the internationalization of education policy, qualitative methods, and the role of international organizations in global governance. He has published on soft governance influence of international organizations and on domestic policy-making and on recent internationalization processes in the field of education policy in international journals such as German Politics, Publius: The Journal of Federalism, and European Journal of Education.
Kerstin Martens is Professor of International Relations at the University of Bremen, Germany. Her research interests include social and education policy in international and comparative perspectives, international large scale assessment in education, international organisations, NGOs and qualitative methodology. Since 2018, she heads the research project on “The global development, diffusion and transformation of education systems” as part of the Collaborative Research Centre 1342 “Global Dynamics of Social Policy“. Previously, she headed a research project on “Internationalization of Education Policy”, a 12-year project financed by the German Research Foundation. She has published widely on the above mentioned research areas, including leading international journals such as Journal of International Relations and Development, Policy Studies, Global Society and European Journal of Education. Recent volumes she co-edited include, Mechanisms of OECD Governance – International Incentives for National Policy Making?, Oxford University Press 2010, Internationalisation of Education Policy, Palgrave Macmillan 2014, and Education in Political Science – Discovering a Neglected Field, Routledge 2010. She holds a PhD in Social and Political Sciences from the European University Institute, Florence, Italy.