1 Introduction: Interdependence in a Complex World

The dominance of International Organisations in the production of global metrics has not only penetrated the transnational social and policy fields; numbers have become an integral part of the fabric of International Organisations themselves. However, little is known about the ways in which global processes of quantification are reconfiguring not only the field of the global governance of education per se, but also—and crucially—the international organisations that have arguably brought it to existence. Metrics have infiltrated not only organisational cultures and the environments these organisations inhabit; crucially, they are reshaping the ways International Organisations co-exist, compete and survive in an increasingly quantified yet uncertain world.

Recent decades have seen fervent activity by International Organisations to build working collaborations and broad alliances for finding ‘global solutions’ to ‘global crises’. Financial investment in these collaborations is increasing and so is hope: If only we had known, we could have acted. Given the moral dimension that these new indices of progress have taken, as well as the enormous human and environmental cost of their failures, there is growing recognition that the interplay of International Organisations in transnational governance has led to the production of knowledge that is de-contextual, standardised, comparable and even at times seen as universal, given its heavily moral undertones and the global nature of the discursive agendas it has given rise to.Footnote 1

Building on international relations (IR) theory, science and technology studies (STS), and using theoretical strands from organisational sociology, as well as the field of the social studies of metrics, this chapter examines the interrelationships of International Organisations (IOs) in constructing the global metrological field. As is well-known by now, IOs have been central to processes of standardisation, de-contextualisation and performance management through numbers; as a result, they have been instrumental in commensurating, and therefore transforming global education governance.

Thus, a central focus of this chapter is the—concomitant with the lure of numbers, albeit less spectacular—recent moves of large IOs not only to establish collaborative partnerships through connections with governments and local agencies, but also crucially with one another. The encoding of data processes and organisational cultures that these collaborative endeavours require (in order for data to be shared and co-produced) allows a comprehensive analysis of the workings of quantification for education governance. In other words, the examination of the interplay of IOs is a unique opportunity to open, rather than stack yet another ‘black box’ in the field of global monitoring (Bhuta, 2012).

Despite the renewed prominence given to the need for alliance-building by IOs, collaboration has always been central to their operation, since they have traditionally needed to work closely with governments, NGOs and the private sector. Yet, the complexity of ‘wicked’ problems, ‘donor duplication’, resource-pooling and data overload have become some of the most common reasons that IOs are increasingly compelled to work together. Indeed, most major global strategies, such as the Millennium Development Goals (2000–2015), the Sustainable Development Goals or major education testing regimes, such as the OECD Programme for International Student Assessment (PISA), are collaborative endeavours, dependent on pooling of resources and expertise. This chapter focuses on understanding the rise of universality of knowledge-making in transnational governance by focusing on the ways that IOs learn from one another. In the making of numbers, how do they negotiate financial resources and knowledge controversies? How do they actively produce collective sense-making (Weick, 1995) and issue-framing strategies (Baumgartner & Jones, 1993)? How much do we know about their expert networks and their collaborations? Ultimately, if rating and ranking practices are a ‘zero-sum’ game for the assessed, how much do we know about the rules of the game for the assessors?

Empirically, the chapter examines two separate cases from the field of education. Education policy, both in the global South and the global North, has increasingly been dependent on the measurement of its performance for the improvement of human capital. Education can be a productive vantage point, since assessment and quantification of performance have a very long history in the field. It is a key element in the newly emergent well-being and sustainability strategies that have prevailed the statistical governing project post the global pandemic. Education is closely congruent with the efforts to use ‘softer’ data sets for calculating the social. Last but not least, it is one of those policy areas that large IOs like UNESCO, the OECD, the European Commission and the World Bank have invested large amounts of data and expertise from the mid-twentieth century.

The chapter begins with a short review of the literature of the politics of quantification; it then moves on to a consideration of the theoretical underpinnings of the analysis and continues with the presentation of the two case studies under examination. Finally, it is concluded by a discussion of international organisations, interdependency and metrology in the field of transnational education governance and beyond.

2 ‘Governing by Numbers’ in Transnational Governance

Scholarship on the role of numbers in governing societies has been abundant and has attracted multiple fields of study, including sociology, history, political science, geography, anthropology, philosophy, STS and others. Prominent authors have written lucidly about the role of numbers in the making of modern states and the governing role of measurement regimes in various areas of public policy and social life (Alonso & Starr, 1987; Desrosieres, 1998; Espeland & Stevens, 2008; Hacking, 1990, 2007; Porter, 1995; Power, 1997; Rose, 1999). Similarly, anthropologies of numbers suggest that ‘our lives are increasingly governed by – and through – numbers, indicators, algorithms and audits and the ever-present concerns with the management of risk’ (Shore & Wright, 2015, p. 23; see also influential work by Merry, 2011; Sauder & Espeland, 2009; Strathern, 2000). Further, important insights and perspectives on indicators in particular come from STS (Bowker & Star, 1999; Lampland & Starr, 2009; Latour, 1987; Saetnan et al., 2011), including actor network theory (Latour, 2005). Finally, there is a small but growing body of studies relating to specific uses of indicators and quantification in transnational governance contexts (for example, Bogdandy et al., 2008; Palan, 2006; Martens, 2007; Fougner, 2008; Bhuta, 2012).

Nonetheless, despite the burgeoning number of publications on the global ‘governing by numbers’, our understanding of the relationship of the politics of measurement and the making of transnational governance is less well-examined; as Djelic and Sahlin-Andersson (2006) suggest, due to the fluidity and complexity of the intense cross-boundary networks and soft regulation regimes that dominate the transnational space, transnational governance is a particularly productive field of enquiry on the role of numbers in governing. This lack of attention could be due to disciplinary boundaries; for example, scholars of IR and international law have not paid much attention to the field so far, although there is a rise in some interesting literature of the role of numbers in global political economy (for example, Fougner, 2008; Martens, 2007; Palan, 2006).

What are the properties of numbers that would suggest such a central role in the production of transnational governance? By contrasting numbers to language, Hansen and Porter (2012) suggest that, although it took scholars a long time to recognise the constitutive nature of discourse, we are now well aware of the role of language in shaping reality. However, they suggest that numbers are characterised by additional qualities that make their influence much more pervasive than words: these elements are order; mobility; stability; combinability; and precision. By using the example of the barcode, they lucidly illustrate ‘how numerical operations at different levels powerfully contribute to the ordering of the transnational activities of states, businesses and people’ (2012, p. 410). They suggest the need to focus not only on the nominal qualities of the numbers themselves but also, according to Hacking, ‘the people classified, the experts who classify, study and help them, the institutions within which the experts and their subjects interact, and through which authorities control’ (2007, p. 295).

It is precisely on international organisations as data experts that this chapter focuses upon; following the literature on the capacities of numbers to both be stable yet travel fast and without borders, the chapter sheds light on what Latour called ‘the few obligatory passage points’ (1987, p. 245): in their movement, data go through successive reductions of complexity until they reach simplified enough state that can travel back ‘from the field to the laboratory, from a distant land to the map-maker’s table’ (Hansen & Porter, 2012, p. 412).

3 Theoretical Frame and Key Intermediary Concepts

The chapter follows a ‘constructivist-institutionalist’ approach (Smith, 2009), as it works with Lagroye’s definition of governing as ‘a set of practices which participate in the organization and the orientation of social life’ (1997, p. 25). Thus, it builds on the premise that far from being a system composed uniquely of ‘national’ and ‘transnational’ bodies, governing the transnational is an ‘institutional order’ made up of all the actors who participate in the construction and institutionalisation of global problems (Smith, 2009). In turn, transnational ‘governing’ is conceptualised as those ‘assemblages of apparatuses, processes and practices’ that make governing happen (Clarke & Ozga, 2011).

As already suggested, a considerable body of research has already focused on the work of IOs in transnational governance. Yet, this research has often seen them as monolithic institutions, or actors with similar interests in a similar context, without attention to the complex set of realities that bring them together and apart over time (with notable exceptions of course, see for example Cram, 2011). IOs are often also seen as internally stable—this means that divisions of authority, institutionalised norms, expectations and values are thought to be commonly shared by all actors within an IO. Nevertheless, ‘most of the time, […] at least some of the actors within an IO will be seeking to change at least some of its institutions, whilst others will work to retain their stasis’ (Jullien & Smith, 2010, p. 4). The examination of actor alliance formation and mobilisation is hence vital in order to understand these relations—both upstream, i.e. the setting of rules and problem framing, as well as downstream, namely the application and maintenance of rules among the actors who are all engaged in competitive relationships (Jullien & Smith, 2010). Indeed, some of this actor mobilisation and alliance-building is achieved not internally but through networking with other IOs.

Thus, one of the key concepts that mobilises this argument is the notion of ‘political work’ (Smith, 2009). When one studies political work, institutions themselves are not the objects of study per se; rather, the focus of the investigation is on the continual cycle of institutionalisation, deinstitutionalisation and reinstitutionalisation of ideas and values within the organisation in question. The study of quantification as a policy instrument can become a particularly fruitful context for such an analysis as one can examine ‘political work’ as those processes that engender the construction of new arguments and the activation of new alliances; subsequently, they either produce change or reproduce institutions, namely actors’ rules, norms and expectations (Jullien & Smith, 2010).

Before moving on, our attention needs to be directed to two intermediary concepts, those of the ‘field’ and ‘knowledge controversies’. To start with the latter, Barry (2012) uses the notion of ‘political situation’ to explain the ways that STS could have been misguided in their definition of knowledge controversies as conflicts that relate principally to a clash of scientific evidence and ideas. Instead, he suggests that ‘the significance of a controversy needs to be understood in relation to a shifting and contested field of other controversies and events that have occurred elsewhere and at other times’ (Barry, 2012, p. 324). Whereas STS initially mostly focused upon the ‘black box’ of science by looking at issues of credibility, objectivity and reliability (Shapin & Schaffer, 1985), it then moved on to the analysis of public knowledge controversies, where expert knowledge would clash with public, lay knowledge (Wynne, 2003). Yet, Barry argues that despite the growth of transnational standardisation processes, the issue of knowledge controversies has not been addressed either by the IR or the STS literature, as if the simplification of data (and the consensual expert practices it involves) decreases rather than increases the possibility of knowledge disputes and failures. However, it is widely known that achieving transnational standards is infinitely difficult; political contestation often gets submerged and hidden behind the need to appear as collaborative and open to partnerships. Knowledge contestations are then seen as an impediment to the formation of collaborations. In fact, it appears that it is precisely in the knowledge controversies that one has to focus upon, if one aims to understand the very process of simplification and the exclusion of unwieldy or awkward data (or awkward experts for that matter). To return to Barry then, ‘what the concept of political situation captures, is how the significance of a controversy is not so much determined by its specific focus, but needs to be conceived in terms of its relations to a moving field of other controversies, conflicts and events, including those that have occurred in the past and that might occur in the future’ (2012, p. 330).

Second, the chapter suggests the need to examine the interplay of IOs as they construct the ‘global metrological field’. Emanating from physics, the notion of field has been used in the social sciences in order to broadly refer to actors’ relational topographies. Nevertheless, it is often reduced to merely looking at specific geographical and relational spaces. Yet, as Djelic and Sahlin-Andersson also suggest (2006), such a conceptualisation of fields misses a vital ingredient from the way fields operate; that is an understanding of the field as a field of power. Drawing on Bourdieu, transnational governance appears as a field of actors who constantly negotiate and push their own agendas forward; according to Bourdieu (1993), the logic of positionality is what gives the notion of the field meaning. In other words, the positions occupied by the different agents in the field, their advances and withdrawals relate to their efforts for distinction within this field as an expression of their professional, educational, or other interests. Meanwhile, the structure of the field is neither static, nor does it change in any systematic way. On the contrary, it is endlessly reformulated, according to the agents’ struggles for recognition and improvement of their situation. Agents use the force of their economic, social, cultural or epistemic capital to raise their game and advance their front. It is the relational nature of these advances that gives the field its explanatory significance. Thus, following Bourdieu, the chapter uses Djelic and Sahlin-Andersson’s idea of fields as ‘complex combinations of spatial and relational topographies with powerful structuring forces in the form of cultural frames or patterns of meaning’ (2006, p. 27). An examination of the interplay of IOs in the rise of the global metrological field is therefore necessary, as it is vital to examine transnational governance not only as a field of numbers or as a field of actors, but also as both.

Thus and to conclude, the chapter adopts a constructivist standpoint by focusing on the social and political conditions that influence the production of numbers, adopting the ontological position that their existence is not organic but rather the product of the interconnectedness of IOs, as outlined above. It examines IOs for whom this transnational game exists (‘what keeps them running’ as Bourdieu would put it) and even national actors who just utilise it as an instrument in their local political battles.

4 The Case of the OECD as a Europeanising Actor: The Rise of the Interdependency of the OECD with the European Commission

How do international education agendas look like from the perspective of national education systems? The answer is that they look broadly similar, without much attention being paid to the source of expertise and policy recommendations, as long as they fit with the local policy agendas and direction of reforms. Indeed, the empirical analysis in this section broadly builds on previous research (Grek, 2009, 2012, 2020) that suggested that the European Commission (EC) and OECD recommendations are often received at the national level as homogeneous. In order to understand this increasing trend, the empirical investigation moved beyond top-down accounts of the mere and one-directional transfer of policy from the international to the national, towards more attention to the interaction and mediation across ‘levels’ and actors. In terms of methods, the empirical analysis focuses on the examination of policy discourse as well as the interviews of 15 actors from both the Commission and the OECD, as well as other relevant research agencies; the interviews focused on the actors’ role in processes of coordination (conferences, meetings, project work), their interactions with other actors within and beyond their organisations and other relational ties that link them and others through channels of flow of data, ideas and/or material resources. The policy actors interviewed have had positions of power and significant decision-making leverage: they had first-hand experience and participation in meetings and debates between the Directorate General Education and Culture (DGEAC) and the OECD in regard to the financing and conduct of large international assessments.

Hence, this chapter examines how the OECD became a dominant education policy actor as a result of its deliberate and systematic mobilisation by the European Commission which found in the OECD not only a great resource of data to govern (which it did not have before) but also a player who would be pushing the Commission’s own policy agenda forward, albeit leaving the old subsidiarity rule intact. As I will show, testing is important here because it produces numbers and consequently ratings and rankings; once the OECD had created PISA’s (Programme of International Student Assessment) unprecedented spectacle of comparison in European education, no system could remain hidden any longer. The field of measurement became instantly the field of the game.

4.1 International Comparative Assessments: the OECD’s International Adult Literacy Survey (IALS) and PISA

Indeed, international comparative testing has become the lifeblood of education governance in Europe and globally. It is more than simply a project of measurement; rather, it has become part of consistent efforts to restore legitimacy and trust between populations and their governments. As Hall contends, ‘building legitimacy requires potential users in the process, as well as technical experts. The most important role of indicator sets may be in framing the issues and defining the problems, rather than suggesting the solutions’ (2009).

The governance of international comparative testing reflects these values. Project boards usually work in conjunction with a large range of consortia of international partners and technical advisors (statisticians, media specialists and, interestingly, philanthropists); they also consult with a vast array of different actor groupings, such as academics, private companies, policymakers, associates, country correspondents, regional working groups and others. Regular training courses are delivered as well as seminars, and regional, thematic and global conferences. Although all these initiatives suggest sustained efforts to include and create consensus with the greatest number of stakeholders possible, the role of experts remains central: before they acquire a more ‘public’ and visible face, tests are being discussed, negotiated and indeed fought over among field experts for a long period of time.

The case of the OECD as a knowledge producer for education governance is particularly interesting because, unlike the EU, it has neither the legal instruments nor the financial levers to actively promote policy-making at the national level within member nations. Nonetheless, through ranking exercises such as the ‘Education at a Glance’ annual reports, the Indicators in Education project (IALS), through PISA and its national and thematic policy reviews, its educational agenda has become significant in framing policy options not only at the national but also, as it has been argued, in the constitution of a global policy space in education (Grek & Lingard, 2023; Lingard et al., 2005; Ozga & Lingard, 2007). This raises the question—what transformed the OECD into one of the most powerful agents of transnational education governance? What are the qualities of the OECD’s expert work that maintain the organisation as a highly trusted source of education policy recommendations both prior to and post-COVID? Martens (2007) has contributed substantially to this discussion suggesting that the ‘comparative turn’—‘a scientific approach to political decision making’ (2007, p. 42)—has been the main driver of OECD success. Through its statistics, reports and studies, it has achieved a brand which most regard indisputable; OECD’s policy recommendations are accepted as valid by politicians and scholars alike, ‘without the author seeing any need beyond the label “OECD” to justify the authoritative character of the knowledge contained therein’ (Porter & Webb, 2004).

Drawing on Marten’s (2007) ideas, we can see that by now there is a taken for grantedness about education indicators, despite all the commentary asking for contextualisation in their interpretation (e.g. Nóvoa & Yariv-Mashal, 2003), and this is indicative of the way in which they have become an accepted part of the contemporary educational policy lexicon across the globe, within and well beyond the OECD, and of their growing significance to the work of the OECD itself since the 1980s. Despite its ups and downs, and the supposed demise of its glamour, PISA continues to account for a large chunk of the Education Directorate’s budget inside the OECD. One could suggest that the OECD’s greatest impact has been in relation to its Indicators agenda, including PISA, and its role in constructing a global educational policy field through governance by comparison (Martens, 2007; Ozga & Lingard, 2007). Indeed, Antonio Nóvoa argued, ‘comparing must not be seen as a method, but as a policy … the expert discourse builds its proposals through “comparative” strategies that tend to impose “naturally” similar answers in the different national settings’ (2002, p. 144). Although that might be too stark a claim, and although comparison can be both (there are certainly good epistemological reasons for comparative research that owe nothing to policy), it is still important to acknowledge the sustained power of comparison as a governing technology, especially when governing is done at a distance and through the use of ‘soft power’.

Thus, a brief historical analysis would show that there has been a range of such studies that the OECD has been organising since the early 1990s, the majority of which were adult literacy studies to start with, followed by the delivery of the most successful one, PISA and PIAAC, the Programme for the International Assessment of Adult Competencies. The first International Adult Literacy Study (IALS) was the first and largest international comparative testing regime of its kind. Conducted from the early 1990s, IALS was innovative, as it was the first time ever that an international comparative dimension was added to the construction of a literacy survey instrument. Thus, it heralded a new era in the construction and evolution of international comparative studies, as for the first time ever it gave international testing a comparative dimension, where measurement against other countries’ performance offered unprecedented visibility and thus exposure. As it was an original and new endeavour, slowly at the start but increasingly later on, IALS boosted confidence in the construction of measurement tools of this kind, increased their persuasive power in regard to their validity and transparency and created substantial revenues to the research agencies administering them. Finally, and perhaps above all, it created a circle of like-minded expert communities, who found in these studies a platform for promoting the problematisation of specific issues, their institutionalisation through their exchanges and the setting up of the study, as well as their legitimation, in the form of advice to failing countries, once the results were published.

Following the successful IALS endeavour, the Programme for International Student Assessment (PISA) became a major instrument in providing data for the European education systems almost from the start. The international dimension of the survey, which overrides the boundaries of Europe to compare student performance in countries as diverse as the United States, Greece and Indonesia, gave PISA a particularly significant weight as an indicator of the success or failure of education policy. While always testing reading, mathematical and scientific literacy, its innovative dimension—and part of its interest as a governing device—lies in the fact that it does not examine students’ mastery of school curricula, rather the focus is on an assessment of young people’s ability to practically apply their skills in everyday life situations. The focus on ‘real-life’ circumstances and on students’ capacity to enter the labour market with core skills, such as literacy and numeracy, has taken PISA’s focus of interest away from less explicit educational aims that resist measurement (e.g. democratic participation, artistic talents, understanding of politics, history, etc.), towards a more pragmatic view of education’s worth: ‘its relevance to lifelong learning’ (OECD, 2003). Finally and perhaps most significantly, a key feature of PISA is:

its policy orientation, with design and reporting methods determined by the need of govern- ments to draw policy lessons. (OECD, 2003, no page numbers)

Hence, this is not simply a testing regime—it is constructed and operates under a clear and specific policy framework, which is to be adopted by the participant countries if they are to improve their future PISA assessments and thus improve their standing in attracting economic and human capital investment. In other words, the involvement of the OECD with the steering of education policy in participant countries does not stop with the publication of the PISA—or whichever study’s—results; on the contrary, this is perhaps where it begins. Expert groups write expert reports, analysed and taken forward by other national and local experts, while the Commission expert committees are also on board in order to keep the game in sight and keep it running.

This is the kind of status that the OECD acquired with the conduct of large international tests; the seal of unequivocal and trusted expert truth. In other words, OECD not only produces evidence quickly and effectively but also digests it and offers it to policymakers in the format of policy solutions. In a sense, if we are used to accounts of European policymaking as slow, cumbersome and ‘coming from nowhere’ (Richardson, 2001, p. 21), the OECD bypasses these obstacles in four key ways; first, it defines the limits of the possible by suggesting what can be measured, hence what can be ‘done’; second, it carries no political jurisdiction, therefore, it carries no external threats to national policymaking, as perhaps the Commission or other EU institutions might have done; it now has the experience, networks and the technical and material resources to speed up the policy process so that it can show ‘results’ within the usually short timeframe that policymakers are in power; and last but not the least, it carries all the ‘right’ ideological messages for education systems in the twenty-first century—that is, it connects learning directly to labour market outcomes and economic growth.

Nonetheless, despite the numerous and in-depth analyses of the OECD’s education measurement work, how can the OECD continue to be such a powerful player in education governance in Europe? As some of the people who work there might have argued, the OECD Education Directorate staff who are based in Paris take a few decisions, if any; the OECD, as they argue, is no other than the participant countries and the national actors and experts sent to the OECD committees and meetings. Thus, how accurate is to examine the emergence of this new policy arena by simply focusing on this single international actor? Not entirely accurate, as the following interviewee suggests:

So around 2003–04, we [OECD and Commission] started becoming far more involved. Meetings all over the world, I don’t know how many countries I visited but what is important is that the Commission is there.... The European member states should see that the Commission is there because one of the criticisms of the Commission since all this started was that we didn’t take into account all the good work of the OECD. Which was wrong but they said it. The way of showing them was to actually be there – not an empty chair. (EC4)

Indeed, although the Commission and the OECD had been leading quite separate ideological paths, a new ‘love affair’ between the two organisations began emerging—this relationship would gradually strengthen and eventually become the sine qua non for the governing of European education systems. Another interviewee was even more eloquent in his discussion of this flourishing relationship:

We used to have great competition between the two institutions [OECD and the EC] which was that they were research-based, we were policy-based. And we needed that. They needed the policy aspect to mobilise the European consciousness ... it was in their interest working with us ...We had some differences but we are working closer and closer together, we are very very good friends now, there is no conflict. (EU3)

As evident in the following quote, the relationship soon became more than one of influence:

When the OECD started speaking about TALIS [survey on teachers] it attracted the attention of the member states that all this is very good but it is expensive. ... So I managed to convince my Director General of supporting (the OECD) with an awful lot of millions of euros. And I went back to the OECD with that message and said that of course if we pay we want influence. (EC7)

On the other hand, OECD actors appear also as quite open to the Commission, stressing from their own point of view, the reasons that the DG Education would work closely with them:

First of all I think we’ve been very lucky that on the Commission side, that they’ve given a lot of emphasis to skills recently and they have this ‘New skills for new jobs’ initiative and so I think we were fortunate that the work that we decided to do on PIAAC corresponded extremely well with their areas of interest and research priorities... So they made a direct contribution, an actual contribution to the international costs and also eventually agreed to subsidise EU countries, the cost that they had to pay as well to the OECD. So we got just a block of direct funding and indirect funding to countries that they then had to pay us for the international costs. That made a big contribution in financial terms and therefore of course enhanced interest in the project. (OECD3)

Another OECD actor also suggested the way that the relationship, rather than hostile, has been much closer recently, in fact ‘hand in hand’:

We have the same perceptions like other international organisations that it is important that we work together and that we avoid duplication of effort and that we know what the other organisations are doing and that there are often occasions that jointly we can do more than what we can do individually. I think we were always aware of that but I think that has become increasingly important that we work hand in hand. And inevitably because we have some common goals. The OECD has had for some time its own job strategy, the Commission has its own employment strategy and its Lisbon goals and there is a lot of overlap. So, I think it is quite normal that we can cooperate on a lot of areas. (OECD5)

However, there is also a reverse side to the coin. Even though the interdependency and collaboration between the two organisations increased, more often than not these exchanges take place in a competitive field, where the delivery of studies and the collection of education statistics is not a choice anymore, but a necessity. Conflict and tensions can run deep:

The main reason is that they are competitors and both in scientific and in financial terms it is getting more and more difficult to conduct these surveys. There was a message from member states to the OECD and the IEA – get together, sit down and discuss it and do it. Now, 6 months later, we all come together and we ask what was the result of that meeting and the answer was that we didn’t find a date. They don’t work together because they don’t like each other. (EC9)

Interviewees also describe internal conflict within international organisations and their departments, for example, within the OECD itself. The following quotation describes the conflict between the Centre for Educational Research and Innovation (CERI) and the Directorate of Education, similar to the kinds of processes Jullien and Smith (2010) describe when they discuss IOs as internally unstable institutions, rather than the opposite:

They live in different worlds – the same floor at the OECD but in different worlds. They don’t like each other – one is more research-based, the other one more indicators and data, surveys. One is more reflection, the other one is more publicity, the charts – different traditions, the same director. (EC12)

Finally, another account that describes the conflict and competition for securing contracts for education research in Europe comes from another interviewee, a key member of staff of one of the Commission’s research agencies:

I think because the OECD is very much looking for member states’ subsidies and grants and financial support for each separate research activity, they are also keen in showing that they do something unique and innovative in order to get such funding. And so then in a way they are in competition with us. An example is they did a recent policy review which is called ‘Learning for Jobs’ which basically deals with VET. And they didn’t invite us to some national expert groups and so on that are in development—and they did very little use of our work because they wanted to do something that was different and specific so that they could sell it to the member states – this is my interpretation, of course. But I think that there is this kind of competition, differentiation between European institutions because we are in competition for funding. (EC3)

The quotations presented above suggest that the descriptions of a field of actors who come together harmoniously and in partnership to work on certain agendas might be misleading. On the contrary, they highlight the need to also focus our attention on those meetings that never happen, as well as those actors who are consistently not invited to expert meetings. They direct us to an understanding of a field, which is riddled with internal and external competition for funding, especially in times of reducing national budgets in an era of crisis. What is important to note is that the story of the collaboration of the OECD with the European Commission is neither a story of smooth consensus and collaboration nor merely a story of struggle and competition: rather, it is the unfolding story of continuous and increasing interdependence of the two organisations in the production of governing knowledge. The latter is both strategic and technical and functions as a way of creating the conditions for a universal and consensual way of amalgamating data and policy priorities, as the next empirical case will discuss.

5 From Dissensus to Universality: The Case of Narrative-Making in the SDG4

The previous section focused on an analysis of the ways quantification increased the interdependence of international organisations through the construction of a single measurement field; we analysed the specific political work that PISA and other international assessments achieved by bringing together actors around common ideas and measurements, even in instances of conflict and disagreement. This section will move this discussion further, in order to show how this increasing interdependence in the construction of a single global metrological field in education has created the conditions for the emergence of common, universal ideas (and ideals) of what education is for and what it should achieve. For this analysis, we will turn our lens to the production of the Sustainable Goal 4, the education SDG that has proclaimed to ‘ensure inclusive and equitable quality education and promote lifelong learning opportunities for all’ (UN, 2023). Specifically, we will focus on a discourse analysis of two crucial documents that emerged in the preparatory stages of putting together the SDG4: these are the 2014 Muscat Agreement and the 2015 Incheon Declaration.

Using discourse analysis of these documents, I will discuss how, through a series of major events and the publication of pivotal texts, such as ‘declarations’, large global ‘agreements’ and ‘frameworks for action’, the work of measurement does not only bring actors a lot closer together than ever before (as discussed in the previous section), but is also inscribed, materialised and made plausible by the production of a powerful—however ambiguous—rhetoric of development, equality, democracy, universality and morality. If statistical data is all about the measurement of possibility, the construction of universal narratives of progress fosters plausibility; narratives bring coherence and give meaning to informal and fragmented global governing spaces. In the following sections, I will examine how old and well-established ideas around global development and educational equity and progress are getting new momentum through the use of language that re-frames them as global goal-setting endeavours. The aim of this analysis is also partly conceptual; the examination will show how the emergence of the SDG4 narrative requires the work of storytelling to reach out to wider audiences, appeal to local contexts and sentiments and therefore reinforce the narrative in a continuous cycle of bolstering the reach and appeal of the targets themselves.

5.1 The 2014 Muscat Agreement

As indicated in the Introduction, the global governance of education from the start of the twenty-first century was characterised by the coexistence of multiple, and sometimes overlapping arguments about the role of education in society: these arguments were not always harmonious or conflict-free. On the contrary, there have been significant power asymmetries and competing expectations in relation to aims and objectives of the policy priorities in discussion, as well as the decision-making architecture towards their realisation.

In more detail, since 2000, the global education agenda had been informed by two separate sets of goals; these were on the one hand, the Education for All (EFA) goals, established in Dakar (WEF, 2000) and on the other, the MDGs. Importantly, both sets of goals were associated with a specific decision-making architecture and with different communities of practice. Therefore, both agendas emerged in parallel (interestingly the loci of power were two cities: New York for the EFA and Paris for the MDGs) as a result of the interaction of different groups of actors, who relied on particular consensus-making scripts. This is significant in relation to the production of narratives, since the UNESCO-led EFA negotiations faced competition by the MDG education-related goals, only for UNESCO to ‘surrender’ in the face of a losing battle. While the EFA agenda (and especially, the so-called Dakar goals) was very much the product of consensus, carefully crafted by the global education community and reflected the multiple priorities of education agencies while also allowing civil society to make a meaningful contribution, this was not the case with regard to the MDGs, which viewed education in much more narrow terms and focused exclusively on universal primary education.

Since the negotiation of the SDGs was approached by different agents as an opportunity to put an end to this duality of education agendas (given that disagreement might have meant an exclusion of education from the SDGs), the re-alignment of the EFA agenda with the MDG education ‘camp’ required crafting a new set of education targets. Thus, it was through a new agreement on the goals that a break-through was found: this was the Muscat Agreement, signed in May 2014. The document was approved at the World Education Forum 2015, with the expectation that it would become an integral part of the global development agenda to be adopted at the UN Summit in New York City in September 2015—i.e. the SDGs.

Indeed, the Muscat agreement, signed by a large number of education ‘ministers, leading officials of multilateral and bilateral organisations, and senior representatives of civil society and private sector organisations’ (p. 1), was the result of the Global Education for All (EFA) meeting in Oman under the auspices of UNESCO’s General Conference on ‘Education beyond 2015’. It is obvious, even from the very first sentences of this document, that the Agreement and thus the reason for this large gathering of education actors from around the world, was not a new development; rather, it is another meeting in the long line of efforts to achieve ‘Education for All’. In fact, the document not only does not shy away from its past, but also is bolstered by the fact that this appears by now an established and well-trodden path, and one that the EFA ‘movement’ had established:

We acknowledge that the worldwide movement for Education for All, initiated in Jomtien in 1990 and reaffirmed in Dakar in 2000, has been the most important commitment in education in recent decades and has helped to drive significant progress in education. (GEM, 2014, p. 1)

Here, we see that the narrative-building begins through the construction of a shared agenda and a ‘movement’ that should be not specific to some actors versus others, but that is ‘worldwide’ and that is marked through important, similar events, in other places and times: in Jomtien in 1990 and Dakar in 2000. As a result, the text here gathers the pace and progress of past events that have prepared it but also asserts EFA as a significant locus of decision-making in the field. In addition, it also hails the Muscat meeting as a milestone in the line of such agreements and gatherings.

However, the tone quickly shifts and offers an olive branch, as according to the document, neither of the two separate goal-setting ‘movements’ has achieved their aims:

Yet we recognise that the Education for All (EFA) agenda and the education-related Millennium Development Goals (MDGs) are unlikely to be achieved by 2015…More than 57 million children and 69 million adolescents still do not have access to effective basic education. In 2011, an estimated 774 million adults, of whom almost two-thirds were women, were illiterate….At least 250 million children are not able to read, write or count…Gender inequality is of particular concern, as only 60% of countries had achieved gender parity at the primary level and 38% at the secondary level by 2011. (GEM, 2014, p. 1)

The use of evidence in narrative-making is a powerful rhetorical tool in creating the necessary epistemic and measurement context for launching new decisions and commitments. Startling is also the change of mood here: from the positive and encouraging collective work that has led to this moment (i.e. the Muscat meeting in 2014), numerical evidence is used to show that these efforts still leave a lot to be desired. Therefore, the narration of numbers sets the stage and the mood as one of continuous crisis and emergency: there is urgent need for new action to be taken. Above all, the script is using a certain logic of appropriateness (what is moral and ethical to do) in order to suggest that such evident crisis needs a united policy front, not one riddled with conflict and separation. Such a discourse of consensus-building is core in the production of the narrative in the Muscat Agreement: this is a story about earlier disunity and failure, versus a future of universality and achievement.

However, the part failure of past efforts does not deter the authors of the text to pace the rate of change; it is precisely the urgency of the situation that further strengthens the commitment to not only achieve the targets previously set, but also set new, even more aspirational ones:

Therefore, we recognise that there is a strong need for a new and forward-looking education agenda that completes unfinished business while going beyond the current goals in terms of depth and scope, as well as to provide people with the understanding, competencies and values they require to address the many challenges that our societies and economies are facing. (GEM, 2014, p. 2)

The Muscat Agreement constructs a narrative that builds on three pillars: first, it clearly spells out that the EFA has been a force of change with a history of over 25 years, the gathering and commitment of key education actors from local, national and international levels; second, the achievement of—at very least—a technical and robust measurement agenda that can offer a fairly concise picture of the levels of educational inequality around the globe; third, the need to unify efforts by both education communities (EFA and the MDGs) in order to have education established as an SDG target in its own right.

Additionally, as the section on ‘Vision, principles and scope of the post-2015 education agenda’ shows, it works on defining and reaffirming the place of education in—what is slowly emerging as—a global agenda that places sustainable development at its core: it achieves that through outlining the main principles of the group, as well as specifying what the targets for achieving these principles should look like. Interestingly, this is the set of principles that the Muscat participants agreed upon; in summary,

  1. 1.

    ‘We reaffirm that education is a fundamental human right…’;

  2. 2.

    ‘The post-2015 education agenda should be clearly defined, aspirational, transformative, balanced and holistic, and an integral part of the broader international development framework…Education must be a stand-alone goal in a broader post-2015 development agenda and should be framed by a comprehensive overarching goal, with measurable global targets and related indicators…’;

  3. 3.

    ‘We affirm that the post-2015 education agenda should be rights-based and reflect a perspective based on equity and inclusion, with particular attention to gender equality and to overcoming all forms of discrimination in and through education…’

  4. 4.

    ‘We stress that the full realisation of the post-2015 education agenda will require a strong commitment by both governments and donors to allocate adequate, equitable and efficient financing to education…accompanied by strengthened participatory governance, civil society participation and accountability mechanisms… as well as improved planning, monitoring and reporting mechanisms and processes’ (GEM, 2014, p. 2, my emphasis).

We see that there are three primary concerns outlined above; these relate to first, reaffirming the place of education as a human right, therefore, connecting closely not only this agreement but also the emergence of the SDG education agenda as a whole with the culture, tradition and institutional identity of UNESCO. This is an important move, as by 2014, multiple other actors, such as the OECD and the World Bank, had also become key education policy-trendsetters globally, and their perspectives on education did not always coincide with those of UNESCO and the EFA movement. As a result, the Muscat agreement indirectly specifies who the key organisation behind the new post-2015 agenda should be.

Second, the Agreement sets a clear demand for the way forward: education must be a stand-alone goal and not be subsumed by other goals in the SDG agenda. The Muscat Agreement is a key narrative script in—momentarily at least—unifying a vastly conflicted field, that had seen two parallel streams of work emerging globally and often in opposition to each other. Narrative-building starts from three commonplaces for education communities: first, the line of similar events and global meetings in the last; second, the challenges of disagreement and of creating some form of consensus in a really complex and conflicting field; third, the growing crisis of education inequalities. There is a clear message in this narrative that highlights the need to move away from fragmentation towards bringing the two different ‘movements’ together, in an effort to ensure a singular place of education in the SDG agenda, and not its subsumption within other policy areas and goals.

However, where is it that the education community should now move to? What kind of governing instrument can ensure unity and create a universal agenda, accepted by all participating actors? As the document continues, goal-setting becomes a key narrative practice and takes centre stage in the story, since the signatories appear to universally agree that this should be a technical exercise, focused on a pre-defined and well-specified measurement and monitoring agenda, with clear accountability mechanisms and generous funding from donors and governments. The mixing of accountability and financing indirectly connects the two as interdependent. Interestingly, the next section in the agreement moves on to do something quite extraordinary; it sets a number of nominal targets without numerically specifying them:

We support “Ensure equitable and inclusive quality education and lifelong learning for all by 2030” as the overarching goal of the post-2015 education agenda.

We further support the translation of this goal into the following global targets, for which minimum global benchmarks and relevant indicators will be identified/ developed:

Target 1: By 2030, at least x% of girls and boys are ready for primary school through participation in quality early childhood care and education…. (GEM, 2014, p. 3)

The list of targets continues with seven targets in total, all of which begin with the time framing of ‘by 2030’. They all set specific targets without, however, specifying numerically what the goal should be: in other words, this is a list of ‘targets’, outlined using language, decontextualised by aspiring them to be applicable globally, yet with no specific numerical inscriptions assigned to them. This practice highlights the ‘target-setting’ in itself is a narrative-building practice as it creates ‘narrative scaffolding’ for the policy stories to be told—stories of improvement and mobilisation but also stories of urgency (‘by 2030’). What is unique in this case is the fact that this scaffolding is so pervasive that it allows creating a numerical narrative, even without the use of specific numbers—just notional percentages (to be agreed) of an imagined world ‘by 2030’. This is how numerical narratives construct universal agendas of education progress, despite the absence of real numbers quantifying the goals.

Finally, the Agreement ends by explicitly outlining its support to UNESCO to act as the lead organisation for the facilitation of this agenda, in addition to reaffirming the significance of ensuring that the SDG framework has ‘a strong education component’ (GEM, 2014, p. 3). Although the ambiguity of such non-numbers is startling, what is of interest is the ways numbers still operate as the negotiation instrument for agreeing on a common agenda. Thus, we see the ways that the Muscat Agreement, through its carefully crafted script becomes the governing locus where organisational, epistemic and political struggles manage to settle. Its significance is evident, by the documents that succeeded it; first of which was the Incheon Declaration.

5.2 The 2015 Incheon Declaration

The Education 2030 Incheon Declaration was published in the World Education Forum, in Incheon, the Republic of Korea, from 19–22 May 2015. According to the document, ‘over 1,600 participants from 160 countries’ took part; the Forum was organised by UNESCO, ‘together with UNICEF, the World Bank UNFPA, UNDP, UN Women and UNHCR’ (UNESCO, 2016, p. 5).

The narrative-building in the Incheon Declaration begins from the commonplace of the education emergency that nations are faced with. Nonetheless, it also offers, for the first time, the marrying of the two previous initiatives, in constructing one education goal in the SDG agenda. This is what came to be known as the SDG4—Education 2030 (hence the double-barrelled name):

The world has made some remarkable progress in education since 2000, when the six Education for All (EFA) goals and the Millennium Development Goals (MDGs) were established. Those goals were not, however, reached by the 2015 deadline and continued action is needed to complete the unfinished agenda. With Goal 4 of Transforming our world: the 2030 Agenda for Sustainable Development – ‘ensure inclusive and equitable quality education and promote lifelong learning opportunities for all’ (hereafter referred to as SDG4- Education 2030) – and its associate targets, the world has set a more ambitious universal education agenda for the period from 2015-2030. Every effort must be made to guarantee that this time the goal and targets are achieved. (UNESCO, 2016, p. 22, emphasis in the original)

The document moves on to explain the ‘broad consultative process’, ‘facilitated by UNESCO’, which took place to arrive to the SDG4-Education 2030 targets and further expands on the membership of the decision-making body to now also include a range of actors, such as the OECD; the Global Partnership for Education (GPE); civil society; the teaching profession; and the private sector. Therefore, the Incheon Declaration further stabilises the narrative of a universal and aspirational motto of ‘education for all’ by announcing a single strategy and by adding new, crucial actors to the mix of stakeholders agreeing to work together to achieve them; notably, the OECD and the private sector.

Similar to the Muscat agreement, the Declaration is structured around different sections; namely, these discuss ‘vision, rationale and principles’; ‘the global education goal and its associated seven targets and three means of implementation’; ‘governance, monitoring, follow-up and review mechanisms’; and finally, ‘financing and partnerships’ (UNESCO, 2016, p. 24). Although the Declaration begins by referring to the ‘old’ instruments of establishing principles and values in universal education (‘treaties, conventions, agreements and protocols, as well as international instruments, such as recommendations and declarations’, p. 31, ibid.), it swiftly shifts ground to set a new normal for building global education initiatives. We see a substantial narrative change here towards a transformation to a whole new governing logic, where monitoring, data and accountability are not only important but also in fact an indispensable tool for the strategy:

In implementing the new agenda, the focus should be on efficiency, effectiveness and equity of education systems…Furthermore, to ensure quality education and conditions for effective education outcomes, governments should strengthen education systems by instituting and improving appropriate, effective and inclusive governance and accountability mechanisms; quality assurance; education management information systems; transparent and effective financing procedures and mechanisms; and institutional arrangements, as well as ensure that robust, time and accessible data are available. (UNESCO, 2016, p. 32)

This—importantly—is not only a narrative outlining the policy direction, but rather it also offers new meaning around the governance processes themselves. According to this new narrative, targets should not be open-ended and aspirational declarations any longer; instead, they have to be ‘specific and measurable’ and ‘country-led’ (p. 35)—as such, it proposes both the new heroes of the story (the country government as the key players) but also requires a specific moral to the SDG story, one formulated through precise targets. The Incheon Declaration changes the narrative from previous story-making (the Muscat Agreement, for example) and suggests that just goal-setting in broad terms will not be enough: instead, there is a need to establish specific targets which will have to be monitored through regular cycles of reporting and accountability: ‘this requires establishing intermediate benchmarks (e.g. for 2020 and 2025) through an inclusive process, with full transparency and accountability, engaging all partners so there is country ownership and common understanding.’ More explicitly, ‘intermediate benchmarks can be set as quantitative goalposts for review of global process vis-à-vis the longer term goals’. Finally, ‘intermediate benchmarks are indispensable for addressing the accountability deficit associated with longer-term targets’ (ibid., p. 35).

The Incheon Declaration continues the incremental changes pushed by the Muscat Agreement, by offering a measurement-led programme of education governance: the monitoring agenda is not only essential, measurable and country-driven, but it also has to be based on a governing architecture with reporting mechanisms at regular intervals through the establishment of intermediate benchmarks. As is commonplace when declaring such substantial shifts in narrative-building, this passage quickly pivots to dramatic language of continued crisis and failure to deliver equitable education for all:

Despite significant progress since 2000, an estimated 59 million children of primary school age and 65 million adolescents of lower secondary school age…were still out of school in 2013…At least 250 million primary-school-aged children, more than 50% of whom have spent at least four years in school, cannot read, write or count well enough to meet minimum learning standards’. (UNESCO, 2016, p. 36)

The critical turning point that the education emergency has taken requires the drawing up of four different sets of indicators to outline policy priorities and organise the measurement goals: these are specified as global (a small set of globally comparable indicators for all SDGs); thematic (a broader set of globally comparable indicators proposed by the education community); regional; and national. Although this differentiation of indicators appeared here as based on levels of government only, it is by now well-documented that eventually, it became a qualitative distinction; in other words, much more emphasis has been given to the global indicators (versus all the other sets) precisely because of the comparability element and the fact that they are part of the SDG framework.

Finally, in terms of ‘implementation modalities’ (p. 57), national governments are seen as having the ‘primary responsibility’ or ‘regulating standards, improving quality and reducing disparity’ (p. 57), following a ‘whole of government’ approach to education: ‘Country-led action will drive change’ (p. 60). Interestingly, the document highlights the need for ‘regional coordination’, too, by suggesting to focus on ‘such aspects as data collection and monitoring, including peer reviews among countries; mutual learning and exchange of good practices; policy-making; dialogue and partnerships with all relevant partners; formal meetings and high-level events; advocacy and resource mobilisation; capacity-building; and implementation of joint progress’ (p. 61). Thus, the document not only establishes a framework for delivering a measurement agenda, but it also creates the expectation that national governments deliver on this agenda and that they do so through peer pressure mechanisms and comparisons with their neighbouring countries and globally.

Therefore, discursively at least, another interesting feature of the new global education narrative in the Incheon Declaration is the repeated emphasis on the need for capacity-building in relation to statistical expertise, as well as the ‘need for sustained, innovative and well-targeted financing and efficient implementation arrangements’. In fact, the signatories of the Declaration state that the SDG4 targets and policy priorities are explicitly promoted as needing to become part of existing national education policies, plans and processes. It is strongly advised that efforts to realise SDG4 commitments should not result in parallel or separate plans and processes:

SDG4 policy commitments do not exist outside of existing national policies, planning, management and monitoring processes and mechanisms. Rather, existing country-led systems, processes and mechanisms should be supported or strengthened to ensure better alignment/adaptation with global commitments’. (UNESCO, 2016, p. 9)

To conclude, it is evident that the SDG4 is not exclusively a performance monitoring agenda. It uses a strong narrative built around it, in relation not only to the need for measurement towards achieving the priorities set (described almost exclusively in the language of different sets of indicators), but also for the new agenda to be seen as necessary, ethical, participatory and universal.

6 International Organisations: Interplay and Interdependence in the Making of the Global Metrological Field

Through the analysis of the collaboration between the OECD and the European Commission, as well as the co-production of education indicators for the SDG4, this chapter has evidently shown how IOs do not constitute ‘centres of calculation’ independently from one another. Increasingly we find that they need to collaborate in the production of global education metrics. However, according to Merry (2011), IOs are not significant merely in terms of their knowledge production capacities, be they combined or separate. By examining specifically the role of indicators in transnational governance, Merry elucidates the governing effects of numbers themselves. Consequently, if we consider IOs central in the production of knowledge, we can infer that their operation—as the knowledge gatherers, controllers and distributors—must have crucial governing impact (2011). These effects empower IOs and set them in a complex and ever-evolving power game for influence and resources. Through an examination of the interplay and interconnectedness of IOs’ data apparatuses, it is precisely this power game and its rules that this chapter tries to cast light upon. Indeed, Shore and Wright argue that, ‘while numbers and “facts” have both knowledge effects and governance effects, it is also important to consider how these are produced, who designs them, what underlying assumptions about society shape the choice of what to measure, how they deal with missing data, and what interests they serve’ (2015, p. 433).

In light of this chapter’s empirical analysis and in the tradition of the seminal work of Barnett and Finnemore (1999), we need to question the International Relations’ conceptualisation of IOs as passive entities which merely distribute ‘principles, norms, rules, and decision-making procedures’, as the more economistic, rational-theory analysis would have seen them to be. Instead, building on sociological institutionalism, Barnett and Finnemore see IOs as powerful agents which have ‘power independent of the states that created them’. Thus, they are purposive actors (Cox, 1992, 1996; Haas, 1992): ‘they define shared international tasks (like “development”), create and define new categories (like “refugee”), create new interests (like “promoting human right”’), and transfer models of political organization around the world (like markets and democracy)’ (Barnett & Finnemore, 1999, p. 699).

However, given the prominence of IOs in IR literature, it is surprising how little attention has been given to the interplay, organisational overlaps and mutual dependencies of IOs. As this chapter has shown, rather than state-bound, IOs are increasingly dependent on other IOs to operate. For example, we find that new IOs are usually founded by other IOs, rather than member states (Shanks et al., 1996). In addition, staff mobility in IOs is very high: ‘a large part of staff …is employed on fixed term contracts which generally run up to three years with the possibility, but not the obligation, of renewal’ (Ringel-Bickelmeier & Ringel, 2010, p. 525). In fact, the case of the OECD is particularly interesting, since it has ‘annual turnover rates sometimes as high as 40 per cent for certain staff’ (Ringel-Bickelmeier & Ringel, 2010, p. 526). The ‘revolving doors’ of IOs may suggest that staff often move between them, or even occupy multiple positions at the same time.

Hence, organisational interplay matters. Although, as Brosig (2011) suggests, IOs are dependent on states, the case of the rise of the global education policy field shows clearly that IOs do perform operations that states cannot and will not perform—in fact, most of them were founded in order to operate as cross-governmental diffusers of knowledge and norms. Barnett and Finnemore are again helpful in suggesting that cooperation between IOs may create mutual dependency, a situation that IOs would normally be seen to want to avoid (1999). Nonetheless, given the complexity of transnational governance and the technological advances of the last decade, we are facing a different situation altogether: IOs cannot and do not act independently to solve major social problems and challenges. Hence, and as the case of the education policy arena has shown, we see IOs as increasingly mobilising their resources through their interaction with other IOs with comparable knowledge producing abilities and interests—an IO’s success may be seen as its power and influence over a larger regime of organisations that work towards specific policy directions, rather than through their complete insularity and autonomy (Raustiala & Victor, 2004). In addition, as we saw above, IOs are characterised by highly mobile workforces; what does this increased actor density and fluidity suggest about the coordination of measurement practices? Indeed, it appears that states ask for the collaboration of IOs as it is seen as a way of increasing efficiency, resource-pooling and coordination of their agendas—the example of the ways that European Commission’s DG Education and Culture was in effect compelled to work with the OECD because of efficiency concerns by the member states, is a good one here (Grek, 2009, 2014).

The concept of organisational interplay is not entirely new to IR: there has been some stimulating work that has examined the interplay of ‘international regimes’ and consequent attempts to produce typologies (Gehring & Oberthür, 2009; Raustiala & Victor, 2004). Nonetheless, regimes lack precisely what Barnett and Finnemore (1999) suggest above: agency (Rittberger & Zangl, 2006). However, even when IR theory has acknowledged IOs constitutive nature as actors, there are other problems. By examining treaty regimes, for example, Young suggested two typologies for organisational interactions: nested and overlapping institutions (1996). But, as Brosig (2011) suggests, ‘research on regime complexes in which relations between institutions are of such density has indicated that disentangling them would compromise the collective character these regimes have acquired’. In addition, most of IR theory that has examined treaty regimes has done so from a rational-theory perspective, one that would explain the interactions as serving specific IOs interests and benefit calculations (Galaskiewicz, 1985; Oliver, 1990; Van de Ven, 1976).

Nonetheless, even when IOs are assigned with agency, asymmetries and power relations are only explained on the basis of rational, interest-based behaviour. However, as the example of the European Commission’s collaboration with the OECD has shown (Grek, 2009, 2014), material resources do not always explain organisational interaction; IOs may actually be very well-off but lack the knowledge and expertise, even legitimacy to promote specific policy agendas. To use the same example again, the notion of subsidiarity would suggest that for the European Commission, the OECD could act as a mediator of its own policies in the member states. In other words, DG Education and Culture lacked the legitimacy to enter national policy spaces; OECD, as an expert institution, did not. On the other hand, organisations like the OECD may well have both the resources and the expertise, but could be lacking in policy direction and influence.

Therefore, it is evident that although important scholarship in the fields of IR, organisational sociology and the social studies of quantification exists, little has it enlightened us about the politics, processes and practices of the interdependence and interplay of IOs in the field of the production of global metrics. On the one hand, IR theory has emphasised the role of IOs in transnational governance; initially through an examination of treaty regimes, and later with an emphasis on IOs influence in power play, the field is dominated by rational, interest-based theoretical perspectives. Thus, it has failed to examine qualities of IOs that relate to their constitutive powers as independent, yet interconnected, actors in the shaping of global policy agendas through their expert knowledge work. On the other hand, organisational sociology, although rich in its intellectual history of competing views about how organisations work, has not as yet examined closely the role of numbers in reshaping organisational behaviour. The insistence on separating the internal from the external organisational ‘lifeworlds’ fails to take into account precisely what this chapter has shown that numbers are able to do: that is, diffuse boundaries and set IOs in a more complex and fluid reality. Finally, studies of quantification, although coming from a wide variety of disciplinary perspectives, have largely focused on the role of numbers as agents in themselves; there has been little, if any, attention, to the political work of the actors that organise these processes, that shape and are shaped by them. The chapter has shown how a focus on the interdependency of international organisations in the production of education metrics and narratives is a useful notion for explaining the slow build-up of a global governing infrastructure, made up by a variety of actors and using education data as its connective tissue.

However, if this growing interdependence of IOs on transnational education governance has significant effects, what do these effects look like? How can we trace the ways that these new collaborations and exchanges change the global education policy field? This chapter examined two major collective declarations as the production of a universal narrative of progress that brings all relevant actors together as participants of a single technocratic exercise. As the discourse analysis of these policy documents shows clearly, although the myth-making of education utopias where ‘no-one is left behind’ has always been prevalent in education, what quantification has managed to deliver is the transformation of vague and broad premises into a technical programme of prescribed numbers and targets, crafted and co-signed by a range of actors. In other words, it enveloped former utopian political aims with the material inscriptions of numbers, modelling therefore not only what such education futures should look like, but also how to achieve them. Thus, the chapter showed how numerical narratives are used as the material building blocks of a governing infrastructure where all major IOs and other types of actors converge. Thus, numerical narratives become vital components of specifying ‘who should do what, and how, when and why they should do it in order to address policy dilemmas’ (Kaplan, 1986, p. 770).

Indeed, it is the potential of such numerical narratives to create coherence and consistency of message and structure that makes them particularly necessary as the material underpinnings of the epistemic infrastructure of global education governance. As Ricoeur suggests, ‘the plot or narrative…groups together and integrates into one whole and complete story multiple and scattered events, thereby schematising the intelligible signification attached to the narrative taken as a whole’ (1984, p. 10). The intelligibility of events, actors and decisions is of particular significance in global education governance, since the multiplicity of fora, projects and actors renders the field often unknown even to those who are active participants in it. Thus, they do not only create coherence but also create logic and, as we saw from the examples above, through the use of ‘shocking’ numbers of failure they offer compelling and passionate accounts of complex phenomena.

Additionally, I have also discussed the ways that stories and narratives in global education policy also depend on creating a crisis discourse; thus, logos (data), pathos (emotion) and ethos (values) are closely intertwined to create calls for unity and action. Data and numbers therefore become the engines of universality: they are not only the valuable resource that allows actors to understand—even feel—the emergency, but also through the dominant instrument of goal-setting, metrics are also offered as the vital component of establishing universal ideals of education futures. Here, I follow closely Boswell et al.’s conceptual contribution to the study of narratives, which has stressed the cognitive dimension of knowledge claims made. Thus, we see quantified targets as taking centre stage in delineating the nature and scale of the problem, in constructing causality by offering arguments that appear comprehensible and convincing, and importantly, in appearing themselves as the only viable option for a way forward (Fischer & Forester, 1993); the example of the Muscat Agreement that outlines what the targets and hence policy priorities should be, without specifying numerical figures, is a telling example of this.

Perhaps more importantly, however, we have seen how narratives use goal-setting and numerical targets in order to create bridges and find compromise between otherwise competing and opposing interests and world views. Narratives impose coherence on complex and messy political realities—and they do so predominantly by selecting ideas and events that are organised by the chosen plot while excluding others. The field of education and especially the case of the SDG4 is very rich in such a history of education communities being at ‘war’ with one another, with substantial and enduring differences in relation to both the architecture of governance of the global education policy space, as well as the policy content itself. The scripting of the Muscat agreement is a case in point here: it allowed, after a very long time, the crafting of a narrative that created enough space and shared targets for both communities to align themselves with, especially under the threat of the exclusion of education from the SDGs as a stand-alone goal. In such a context, goal-setting appears not only as significant instrument for the scripting of the story, but also almost as the necessary pre-condition that brings actors together; for if there is one common frame, that is goal-setting as the one globally accepted norm of organising policy work.

To conclude, this chapter focused on the materialities and intertwinements of numerical data, actors and discourse. In doing so, it manifested their vital work in creating actor interdependence and ultimately the universality of datafied education governance, as the only viable means of achieving consensus and a technical/political equilibrium. It therefore showed the multiple ways that, instead of the supposed contextualisation of knowledge that Mode 2 purported as the new ways of producing expertise, it was indeed the construction of universal numerical narratives that gave heart and soul to the global education governance game. The following chapter will contest another key principle of the Mode 2 paradigm, namely interdisciplinarity. Through an exploration of the developments that led to the establishment of the SDG4, it will show that rather than interdisciplinary knowledge; it was mono-disciplinarity that led the way and especially the production of a particular kind of knowledge: that is, economic knowledge and its very specific disciplinary and epistemological assumptions.