Meeting development goals: evidence from the Civil Society Education Fund

ABSTRACT In recent years, the Civil Society Education Fund has supported national education civil society coalitions (NECs) in low-income countries so that they put pressure on governments and donors to implement the Education for All agenda and the Millennium Development Goal on education. This article draws on literature on global governance as well as on an extensive evaluation of the CSEF to explore the actual contribution of this initiative to the activity of NECs. The article highlights the achievements and shortcomings of the CSEF and includes a set of practical recommendations on the role of global civil society in international development processes.


Introduction
Between 2000 and 2015, the Education for All (EFA) programme has established the main goals of global educational development. These goals were approved in the World Education Forum held in Dakar in 2000, and afterwards were coordinated and monitored by a consortium of international donors led by UNESCO. National and global civil society organisations were strategic in the definition of the EFA agenda, as well as in the monitoring and evaluation of progress towards the EFA goals. Because of this reason, key international donors established funds to support the promotion of civil society coalitions that advocate for Education for All in low-income countries, including the socalled Civil Society Education Fund (CSEF). 1 The CSEF was established in order to build the capacities of civil society organisations to advocate for EFA in a great variety of lowincome countries (Archer 2010). The CSEF is a global initiative run by different stakeholders mainly in Africa, Asia and Latin America, that eventually provided support to National Educational Coalitions (NECs) comprising teachers unions, NGOs and community associations. This paper reflects on the global governance of educational development on the basis of a systematic analysis of the CSEF initiative. The initial sections of the paper elaborate on relevant research debates and on our theoretical orientations to the global governance of education theme. Afterwards, the organisation of the CSEF and the methodology of the study are briefly presented. The following sections analyse the functioning of the CSEF between 2009 and 2011 according to a set of normative criteria that can be applied to global governance affairs (cf. Biersteker 2015). The final section links the findings back to the theoretical framework as a way to make some general claims and define policy recommendations. Main sources of data come from the official internal evaluation of the CSEF that was commissioned by the Global Partnership for Education (GPE) in 2012 to our team [name deleted to maintain the integrity of the review process]

The global governance of education
In addition to international organisations such as UNESCO, the World Bank or the OECD, non-state actors such as global teacher unions, international NGOs, social movements and activist networks have become significant voices in the field of global education policy (MacPherson 2016). Observers agree that these political actors coordinate their action in a fluid yet influential way that also builds new forms of legitimation. Political scientists normally label this pattern as the governance shift, which metaphorically associates the steering of collective affairs with the art of driving a ship.
The point is that there is not a single driver of global governance but rather that governance is impacted by a number of factors, including: political actors with diverse interests and preferences (including state and non-state actors), hierarchies (e. g. national and local governments and NGOs, but also international organisations, transnational corporations, transnational advocacy coalitions), market developments (e. g. financialisation, international trade, innovation), as well as networks, including new advocacy and professional networks and communities of experts.
International relations studies have produced a very rich intellectual debate on global governance since the very inception of the League of Nations. This debate has focused on the potential and the actual advantages of international law and international organisations, the aggregate outcomes of allegedly self-serving national foreign policies, the making of an international society, and the management of the world economy.
Noticeably, however, the relevance of global governance and, in particular, the role of non-state actors, are underappreciated themes in these studies (Jorgensen 2010).
With regard to the education sector, Mundy and Murphy (2001) aptly foresaw more than a decade ago that public debates and advocacy around education have gone global. While these authors highlighted the new proliferation of transnational non-state actors who were interested in education and engaged in innovative patterns of interaction, later discussion has also started to appraise the consequences of this social change. So far, international comparative research on campaigns supporting EFA throughout the world has noted that cohesive and well-connected coalitions often manage to influence decision-making and agenda-setting processes, while weaker coalitions are only able to articulate a set of more or less coherent policy demands. Authors have also noticed the complementarity of the delivery and advocacy functions of NGOs (Mundy et al. 2010;Rose 2011;Strutt and Kepe 2010).
Some authors claim that global advocacy networks and NGOs widen the space of democracy to the extent that they are committed to human rights (Dryzek 2010;Tandon and Brown 2013), and particularly to the right to education (Gaventa and Mayo 2009), in a more reflexive and creative way than public organisations. However, other experts blame these agents for lack of democratic control and for exercising subtle forms of domination beyond the official spaces of policy engagement (Kamruzzaman 2013).
Recent studies have suggested that global civil society has a dual character that leads to both emancipation and regulation simultaneously. This feature is particularly noticeable in the educational terrain (Mcpherson 2015;Menashy 2016). In the present paper we try to expand this argument by looking at a wider set of evaluation criteria (Biersteker 2015).

Theoretical orientation
In order to define global governance, this article draws on the International Political Economy tradition, which links the concerns of international relations experts with concepts of conflict, power and economic institutions. From this perspective, global governance is an interactive order that conveys authority in multiple sites.
-In global governance scenarios, agents influence each other mutually, but not evenly since influence happens under the constraints and endowments patterned by social structures. The underlying assumption is that any agent not only impinges on others' strategies but is also influenced by them. Moreover, agents use a set of economic, political and ideological technologies to pursue their interests. They are equipped with varied (sometimes uneven) sets of instruments depending on their history, their position in the world order, the resources available to them, their traditional relations, etc. These are structural processes insofar as they have been constructed prior to the particular interaction under analysis (Biersteker 2015;Kahler and Lake 2003).
-Global governance conveys authority, guidance, and compliance concerning collective activities. Some agents are more authoritative in that they can make a difference in the rules of the game, or they can set the guidelines of official policy-making (Biersteker 2015). Notably, the location of authority can shift over time and may be the outcome of transformations within the nation state, for instance, the reinforcement of the executive branch with regard to the legislative chambers.
-Global governance takes place in multiple sites, normally structured as nonnested, complex geographical scales by the strategies and collaborations of the agents included. It is global, regional, national and local, and in fact operates through several geographical scales at the same time (Lingard and Rawolle 2011). Agents provoke this complexity by looking for the better site(s) to launch their strategies. Since they can operate locally, nationally and globally simultaneously, these scales are not nested, that is, it does not make sense to assume that the local is subsumed under the national, and the local and the national are subsumed under the global (Kahler and Lake 2003).
A strand of political theory has discussed the potential to democratise global governance.
In other words, some democracy theorists have availed of studies of global governance to challenge the taken-for-granted assumption that democracy can only take place at the national level. In their view, it is necessary to explore the potential for both local and global democracy since there are not normative reasons to exclude these other scales of policy-making from democratic tenets. Similarly, it is not only elected governments who must be held accountable for democratic action, but organisations, corporations, networks, associations and communities operating at a supra-national scale too as far as they have an influence in citizens' lives. In this vein, global governance should tend towards cosmopolitan democracy, which is defined as "an attempt to generate democratic governance at a variety of levels, incorporating different spheres of politics" (Archibugi and Held 2011, 441).
The CSEF can be actually seen as an attempt to promote the global governance of education in the direction of cosmopolitan democracy. Our goal here is to convey what can be learned in terms of the global governance of public assets such as education, and whether the CSEF had an impact in democratising global education policy. To this end, we draw on Biesrteker's (2015) proposal to appraise the effects of the CSEF on democratic practice of global governance by looking at the following dimensions: inclusiveness, representativeness, adaptability, effectiveness, efficiency and fairness.
Each of these dimensions is defined in the relevant sections below.
Drawing on this perspective, we want to make a methodological and a substantive point regarding the contribution of the CSEF to global governance. As to methodology, global governance should be analysed through the strategic interaction of political actors. That is to say, it is necessary not only to identify who are these actors, but also which strategies they implement, which resources (both material and cognitive) they avail of, and how their action shapes geographical scales and constitutes complex institutional assemblages. These outcomes are sometimes transitory, but this is not the only pattern that can be observed empirically. Often, the outcome of the interplay of actors is the shaping of institutional assemblages that entail important contradictions, but may be stable over time. Through the CSEF, the use of international funds aiming to strengthen civil society networks in the educational terrain has arguably created one of these assemblages. In addition, such assemblages normally transform the relationship between institutions and territory because political interaction takes place between actors who operate at different geographical scales.

The Civil Society Education Fund
The CSEF aims at providing support to the work of civil society education coalitions, also known as national education coalitions (NECs), so that they can fully engage in the promotion of EFA goals in those countries where they operate. NECs are networks of civil society actors that articulate local NGOs, teachers unions, women's organizations, international NGOs, parents associations and grassroots' organizations. According to the CSEF project, developing an active, well-organized and well-articulated civil societylinked together regionally and globally -that advocates the right to education and progressive policy change is one of the best ways to ensure that all children will have access to relevant and quality education [name deleted to maintain the integrity of the review process].
In its first stage, the CSEF project operated during two years, between mid-2009 and mid-2011, and US$17.6 million was allocated to it. The project has been funded through the Education for All Fast Track Initiative (EFA-FTI), currently known as the Global Partnership for Education, and has been implemented by the Global Campaign for Education (GCE) under the supervisory role of the World Bank. The GCE, the biggest and most active civil society transnational network advocating EFA, was chosen for its central role in the EFA movement and its capacity to agglutinate key civil society actors.
As [name deleted to maintain the integrity of the review process] explain: "The GCE was set up in the late 1990s, in the run up to the Dakar Conference, with the objective of pushing for an ambitious EFA agenda. It brought together several International NGOs (Oxfam, Action Aid, Global March for Labour) and Education International (the global federation of teachers unions). With the passage of time, the GCE evolved into a multi-scalar organization by promoting and strengthening the role of civil society advocacy coalitions operating at the national and regional level.
In the context of these coalitions, very different types of organizations work together to put pressure on national governments, donors and international organizations to honour financial and political agreements to deliver high quality education to all." At the global level, the GCE acted as the executing agency and hosted a secretariat to have overall oversight of three regional CSEFs established in Africa, Asia and Latin America. The regional coalitions, which were already members of the GCE (African Network Campaign for Education For All in Africa, Asia South Pacific Association for

Basic and Adult Education in Asia and Latin-American Campaign for the Right to Education [Campaña Latinoamericana por el Derecho a la Educación] in Latin
America), hosted sub-secretariats that were in charge of supporting NECs within this process in different ways. Three regional funding committees were established in each region, each one made up of credible individuals from across the region, to make decisions regarding the allocation of funds. And three financial management agencies were identified (Oxfam GB in Africa, Education International in Asia, Action Aid in Latin America) to ensure sound financial management. In the period analysed, the CSEF supported NECs in 45 countries as a way to enable them to advocate the EFA goals more effectively and to ensure an active role in GPE related policy processes.
The design of the CSEF clearly separated strategic management from funding management as a way to avoid the interference of international donors and international NGOs in the policy-making process of the beneficiary organizations, but also as a way for the GCE and the regional organizations to focus on their programmatic and strategic role. At the operational level, the CSEF has been structured, on the one hand, around objectives and activities that are linked to the GCE and primary processes: namely, educational advocacy for EFA goals and, related to it, the necessary training, research and technical capacity building initiatives to ensure efficient educational advocacy. On the other hand, the CSEF has followed long-term objectives of a different nature and that are usually less visible, such as the organizational strengthening of the coalitions (through capacity building and other means) and the development of models for self-sustainability.
These support processes consist of financial management, human resources management, communication, and monitoring and evaluation levels. Diagram 1 visually depicts the organization and workflow of the CSEF.

Diagram 1 here
The national educational coalitions drew on these resources provided by the CSEF in order to develop new instruments for advocacy, conduct research, underpin the training of their stakeholders, increase their human resources and strengthen their profile in the media. Although the variation between countries was noticeable, Table 1 highlights the most common activities undertaken by the CSEF-funded NECs between 2009 and 2011. Table 1 here

Methodology
In order to capture the intricacies of global governance, in 2012 and 2013 we collected data through interviews and document analysis at the national level in seven national education coalitions (see Appendix 1), and at the regional and global levels (see Appendix 2). Basically, the interviews asked questions about national politics, the emergence of national educational coalitions, the articulation of advocacy with research and training, and the outcomes of this advocacy. In-depth case studies were conducted in Bolivia, Cambodia, Mongolia, Sierra Leone, Senegal, Mozambique and Malawi. This purposive sample included countries located in different continents, more and less propitious political contexts, long-term and recently established NECs, varying areas of expertise and interest and both big and small grants. Interviews in each country covered multiple stakeholder types such as NEC members (board, secretariat), teachers' unions (independently of whether they are members of the coalition or not), other civil society organizations, national CSEF coordinators, international donors (bilateral and multilateral), education policy-makers and decision makers, parliamentarians, media representatives, and education scholars. Thus, a total number of 126 interviews triangulated the voice of NEC members and non-members in order to depict a nuanced picture.
We also conducted 33 interviews with key global and regional actors in the governance and management of the CSEF. Specifically, at the global level, we interviewed members of the GCE board, GCE secretariat, and GPE secretariat, and, at the regional level, we interviewed representatives of the CSEF Regional Funding Committees, the CSEF regional secretariat and the regional finance agency. Interviews with key global and regional players have been an important data source to evaluate the architecture and governance of the CSEF.
Beyond interviews with key national, regional and global players, the evaluation relied on secondary data sources (including NEC completion reports, CSEF progression reports, a GCE survey on the CSEF, and GCE/regional organizations narrative reports).
The document equivalence technique (i.e. comparing the global discourse with the national discourses in relation to the fund) was used to empirically triangulate the effects of the CSEF at the national level.
On the matter of outcomes, a final point on attribution is necessary before proceeding. While we do, in some sections, discuss a range of discrete outcomes, we understand the CSEF not as being the only causal factor but rather as being a key contributor. That is, in addition to tracing each of the supranational links portrayed in Diagram 1, when considering effects at the national level, we analysed the data collected in terms of how the CSEF affected NECs' institutional setting and how that, in turn, enabled certain advocacy strategies and organizational processes, which subsequently fed into their legitimacy, involvement, and policy influence. Put differently, we identify the ways that the CSEF-as one factor among many-contributed to the strategies, processes, achievements of NECs (Klugman 2011;Mayne 2008) using analytic tactics such as process tracing, contribution analysis, and the most significant change approach (White and Phillips 2012). This point on attribution is important because the CSEF was introduced into a complex assemblage where multiple processes and pressures overlapped and impinged on one another. Table 2 summarises the main findings of the evaluation of CSEF 2009-2011 with regard to Biersteker's (2015) criteria. This section draws on the evidence collected by [name deleted to maintain the integrity of the review process] in order to evaluate the achievements and shortcomings of the CSEF in these terms. Table 2 here

Findings
An inclusive system of governance should engage all the affected populations. The system would be also representative if these populations were able to express themselves on an official basis (Biersteker 2015). In general, the CSEF was quite inclusive, although its representativeness was somehow problematic.
On the one hand, the official strategy openly aimed at engaging all the NECs. A diverse array of political actors was involved in the coalitions. Since all of them could tailor their strategies to the challenges, contexts, and needs they faced, the CSEF clearly allowed national actors to make their own decisions with regard to advocacy, research, and training, thus underpinning their own capacity. "We valued that what [NECs] wanted to do was also in line with regional and national campaigns.... I wouldn't say that coalitions were pushed into doing that [budget tracking], we'd presented them a menu of things where we describe how, you know, if what you're doing is fitting, it does fit in this international campaign, that's great and then these or those are available too, like for example Global Action Week things, or One Goal Campaign... some coalitions took them up with energy and others did not, ahh, but they were not penalised in any way or pressured..."(Interview with a member of the CSEF-Asia) In general, national coalitions were able to adapt the general strategy to their own terms.
Thus, in Senegal the NEC decided to stand for the rights of the poorest and most marginalised population. In Cambodia, the coalition aimed at strengthening its members' advocacy capacity and ability to facilitate community engagement in quality education in selected provinces, in addition to compiling a community engagement pilot manual and encouraging replication by education NGOs and other interested parties. In Mongolia, the priority was widening the circle of official consultation of educational policy-making [a reference has been deleted to maintain the integrity of the review process] On the other hand, these NECs did not have a say in all the areas of activity. In the context of global governance, effectiveness focuses on whether a governance arrangement is able to provide public goods that are not delivered at the domestic level, either by local institutions or other international organizations (Biersteker 2015). The point is that Table 1 clearly indicates that most NECs managed to use the funds to undertake the activities they were expected to. However, it is remarkable that their research capacity was quite disparate, since some NECs went further than other ones.
Notably, the CSEF has clearly had a visible impact in reinforcing the visibility and the recognition of the NECs. Many governments, audiences and international organisations have actually recognised the NECs in a number of countries as key educational stakeholders, to the extent that coalitions sit on many governmental committees, and are included in various consultation and decision-making platforms of governmental actors and donors. Table 3 shows that NECs were invited to a variety of official meetings. Most of them were fully recognised as partners in the national commissions dealing with EFA in each country (A). A similar number of NECs took part in the development, endorsement, appraisal and monitoring of the Education Sector Plan (B). Little variation is visible regarding the number of NECs active in annual joint sector reviews (C) and having full access to education sector plans and sector policy documentation (D), while only two coalitions had only partial access (E). Table 3 here The 'net' effect in terms of prestige and recognition has to be considered against the different political contexts in which NECs develop their work. In countries like Cambodia or Senegal, where CSOs were traditionally seen by the government as an obstacle for policy making, the increase in political recognition in such bodies as the Education Sector Working Group meant a significant impact achievement for civil society and the CSEF in particular.
Adaptability lies in the capacity of global governance schemes to accommodate change (Biersteker 2015). Certainly, the evaluation of a short period does not ground a sound conclusion in this regard, but a few pieces of evidence sketch the potential of the CSEF to respond to some emerging challenges. In fact, the work chart depicted in Diagram 1 maps out a complex balance between two logics of social action. On the one hand, teacher unions and multi-stakeholders' coalitions focusing on EFA were gathered so as to put pressure on policy-makers. On the other hand, the regional and the financial agencies were established in order to carry out an official project defining clear lines of command and mechanisms for accountability. As illustrated by the quote below, the coexistence of these two logics of action led to an uneasy transformation of the relationship between the regional coalitions and the NECs -and in fact some NECs start regarding regional coalitions as donors to whom they were accountable: "One thing is to have a network where members generally contribute to the network by, for instance, paying membership fees, because the coalition members are collectively committed to this collective political project, beyond the interests of their particular coalition… This is a social movement or network logic, versus the donor logic, where coalitions receive funds from the network (…) The relationship gets reversed, there begins to be a tension that undermines the political alliances that should be there..." (Interview with a member of the GCE Board) In the evaluation research, some voices expressed concern with the repercussion of this complex architecture for social movements, and vindicated the transformation of the fund into an open learning platform. Clearly, this was not the official position. Besides the aforementioned constraint for accountability, it is plausible to interpret that the very possibility to express this tension, captured by the quote below, opens some room for the CSEF to accommodate change. "We would like to stay on campaign, and advocacy for all and so on and so on. Not to become something receiving money, distributing money (...)" (Interview with a member of the GCE Board) Efficiency is the capacity of a governance scheme to provide these public goods at the lowest possible cost (Biersteker 2015). In this vein, and despite it is hard to assess to what extent the CSEF was efficient in quantitative terms, most NEC reports mentioned some indirect evidence of an economic use of resources, not only because the very institutional capacity of the NECs depended on the staff they could hire with CSEF funds, but also because many of these NECs also reported an increasing capacity to attract volunteers.
In addition, many NECs were able to to attract significant additional funding. In the period analysed, the total amount generated by all the coalitions amounted to almost ten million USD. Table 4 presents the evolution of the funding generated by the coalitions outside the CSEF, on a yearly basis. We distinguish, on the one hand, between externally generated funding received from the government, private organizations and other aid agencies and, on the other hand, internally generated funds in the form of member fees, fund-raising activities, etc. Not all the coalitions could attract funding, whether internal or external, at the same level, since in the analysed period half of the NECs received external funding, and less than half of them generated part of their funds internally. Table 4 here In general terms, the evidence presented so far shows that the CSEF has contributed to building stronger and more credible advocacy coalitions. Nonetheless, the potential for the NECs to maintain their activity ultimately depends on their capacity to enact a virtuous cycle between their primary processes (i.e. capacity building, advocacy, research and training), support processes (i.e., human resources management, internal communication, monitoring and evaluation, etc.), and concrete policy outcomes. In some cases, training and capacity building have proved to be the missing link in this cycle. However, where the NECs have faced more recurrent problems is in terms of support processes. We reflect below on how a mismatch between primary and support processes has impeded the CSEF to achieve its full potential.
To start with, the GCE did not have the appropriate human resources in place when the CSEF project started. The tensions between the World Bank and the GCE l e d t o i rregular and late funding disbursements, which created cash flow problems and put pressure on project execution and final spending.
Nevertheless, the main challenge the CSEF faced during the project cycle was the high turnover of managerial positions at all levels and at key points in the project cycle, in addition to a lack of leadership and coordination capacity from the CSEF global secretariat.
Various reasons for this turnover were identified in the interviews: (1) insecurity because of unclear sustainability of the programme, (2) disparities and delays in salary payments and, (3) workload and external pressure. High turnover also affected organisational memory and the impact of learning, skill-sharing, and staff training.
The overarching monitoring and evaluation for the whole of the CSEF-funded processes was also agreed upon too late, once the project had already started. The framework covered a broad range of items and allowed for the collection and publication of important data related to project implementation. However, the framework was also very complex and time-consuming, and the lack of central guidance contributed to the fact that not all members followed the framework with the same level of consistency. Subsequently, changing the requirements around monitoring and evaluation procedures led to confusion and to additional workload among the CSEF staff. Lastly, several interviewees complained of problems with the information technology system, and their difficulty communicating up-to-date information to all members, information which was relevant for monitoring and evaluation.
The CSEF poses important questions regarding to whom the national coalitions and the regional committees were eventually accountable. In fact, the whole institutional design included two conflicting logics of political action. Interestingly, whilst official programmes respond to explicit projects which are allegedly supported by public resources according to some specific conditions, civil society normally launches collective action to pursue some very meaningful but not always precise goals on the grounds of the present capabilities of unions, non-profit organisations and community movements. Integrating these two sets of expectations is neither automatic nor easy.
According to the organisational chart of the CSEF, national coalitions had to report to regional committees who also had to report to the incumbent financial agency either in Africa, Asia or Latin America. This procedure became quite confusing for most stakeholders insofar as some representatives sat on both the boards of national coalitions and the regional committees. In addition, the regional committees played a contradictory set of roles, because they had to provide support to the national coalitions at the same time as they observed whether the activities of these coalitions were appropriate.
Although the issue did not become a conflict of interest, it was a real problem when many interviewees attempted to figure out how the whole organisation worked [name deleted to maintain the integrity of the review process].
In fact, since the World Bank did not adopt a proactive role, the GPE secretariat eventually became a de facto supervising entity. As a consequence, interviewees reported that the GPE ultimately failed to support some important activities related to the CSEF. "There was no enabling engagement from the Bank or even from the FTI to ensure that what the CSEF wanted to accomplish was supported by them in their processes. I think there was a huge sort of gap between the stated aims of this project and its ownership by the partnership that approved this project. The bank is a partner, the FTI's secretariat, it supports the partnership, the other donors are partners, and yet none of them own this agenda enough (...) And I think that really stems from this model of the Bank, 'Here is the money, we deliver it, then you report back, and as long as there is no corruption, you know, our reputational risk is reduced, and we are not worried'." (Interview with a member of the GCE) In addition, the time and effort required by the monitoring and evaluation mechanisms prevented regional bodies to focus on political and strategic action, such as the organization of advocacy campaigns and research at a regional level: "If you ask me, I think it went a little over board, you know, these are the tensions, because in some ways I think a balance needs to be struck between how much energy sapped out the coalitions, to deal with, attending to financial and administrative nittygritty... versus the energy that goes into the substantive part of the work... It's a constant tension, there's a constant tension in that." (Interview with a member of the CSEF -Asia) In this vein, decentralisation became a further challenge to map out a clear picture of the whole system. In coherence with the organisation of the GCE, each world region decided on its own system to monitor and evaluate the activities sponsored by the CSEF. Therefore, since it was extremely difficult to identify a common approach to accountability in the three world regions, it was not possible to conclude that all the coalitions had been responding to the same accountability requirements [name deleted to maintain the integrity of the review process].
Fairness consists of equal treatment in the distribution of resources as well as equal recognition of the voice of participants. As to resources, since the CSEF responded to the proposals of the individual NECs that participated voluntarily in the CSEF, fairness was not at all an issue. However, some coalitions asked for specific resources to foster coordination with local communities. Others openly complained for the quality of the feedback that the regional bodies provided to them. Apparently, their vision of a two-way dialogue did not easily fit within the on-going activity of the CSEF, as illustrated by the following quotes: "I realise that our current main shortcoming has to do with evaluation and monitoring. In our view, a coalition which aims at being really national should be much more present on the ground. It means that the organisation has means of transport to get to places like Tambacounda, Kédougou or Ziguinchor, where people live. The organisation has to meet and help the local activists on a regular basis, because they often lack the necessary expertise. Definitely, the CSEF should cater to this need. " (Interview with a member of the NEC-Senegal, authors' translation) "[The regional body] should have a section on monitoring and evaluation ... of the projects (…) They don't have -the evaluation is external, like somebody coming from the outside, but they don't have internal evaluation, even that when you come here and you put your report, (somebody reads it, but there is no one for whom it is their organizational responsibility to read and act on the content of the report. There needs to be a position and process.) Like we have a research unit and an advocacy unit, I think they should have a M&E unit" (Interview with a member of the NEC-Cambodia)

Discussion and concluding remarks
Global governance takes place in an interactive order that conveys authority and other political resources in complex geographical settings featured by several geographical scales (or non-nested spatial hierarchies of sites). Although political theory cannot establish a general principle of global democracy, the democratic component of the observed modes of governance can be partially appraised in accordance to some widely accepted criteria, including inclusiveness, effectiveness, accountability and fairness, among other criteria (Biersteker 2015). These criteria suggest how to assess both the intended and the emergent effects that interactive orders bring about. In the end, while the CSEF clearly represents a step forward for efforts at creating cosmopolitan democracy (Archibugi and Held 2001, 441), a number of lessons and challenges to be addressed in future similar initiatives can be elaborated.
The main takeaway of the evaluation of the CSEF for the 2009-2011 period is that a virtuous combination of international recognition plus improved capacity, research, and advocacy, enabled NECs, first, to be recognized by governments, second, to gain access to key spaces of agenda-setting and decision-making, and third, to contextualise the global EFA goals in meaningful terms for national political actors. Crucial to this cumulative circle is the energizing effect of advocacy underpinned by research.. National coalitions launched more effective strategies when they became capable to find significant evidence of a number of issues (e.g. actual compliance with the official educational budget, attention to discriminated groups such as girls or minorities, observation of teachers' labour conditions). The deployment of communication strategies also contributed to these effects in most countries. Additionally, the NECs that were supported by the fund managed to multiply their human resources via volunteer work and their financial resources by becoming more appealing to national donors.
Once NECs are recognised, their proposals can be more easily taken into account within education policy processes, or enter the political agenda and become possible assets for candidates in elections and i n campaigns launched by diverse social movements. However, it must be kept in mind that formal presence in policy spaces does not guarantee significant participation and influence, though greater levels of visibility and recognition are a necessary condition for achieving higher levels of impact. Recognition can also have a positive effect on medium-term sustainability since it raises their credibility and makes donors more willing to fund their activities. Once a new actor is fully accepted, the actor's influence is likely to be durable and to increase because it can entrench itself in the structure of political interplay over time.
However, one question remains: have these advocacy activities accelerated progress toward educational goals? On the one hand, we cannot claim causality insofar as other indirect factors may have played a role, particularly if their intervention is difficult to observe. On the other hand, Table 1 shows that many coalitions put pressure on policymakers and legislators. In addition, in fifteen countries the CSEF supported coalitions and the education budget increased substantially. Thus, we can consider that at least this support contributed to that progress.. This study has also evidenced a number of challenges and considerations that must be taken into account in the future to improve global governance arrangements like the CSEF, which work across multiple scales and depend on the coordination and collaboration of multiple actors. These include: -The need to define and to disseminate the overarching monitoring and evaluation framework ahead of time, to inform those who will be evaluated in a timely fashion, and so that program implementers can appropriately respond to those frameworks before it is time to be evaluated. At the same time, it is important to balance the design and application of the monitoring and evaluation framework with the need of national coalitions to design strategies that respond to their local contexts.
-Prevent discrepancies between the funding and implementing agencies (in this case, the World Bank and the GCE), since these conflicts can have important consequences for the availability and timing of funds for national actors.
-Improving functionality requires clarifying lines of accountability and perhaps separating the roles of support and evaluation/monitoring, in order to avoid dysfunctional relationships between national actors and, for example, their counterparts in regional or international entities.
-At the design stage, it will be important to think about how to enhance the representativeness of national coalitions/actors, so that they are not only included in global governance processes, but they can contribute to shape these processes from the beginning. Relatedly, it will be essential to include protocols in global governance designs that allow the stakeholders to challenge the fairness of resource allocation and the processes on which the global governance arrangement depends.
-Programme sustainability and continuity will be essential for many reasons, one of which is to reduce turnover in key staff positions.
-In terms of communication, improvements can be made when it comes to instituting systems and processes for sharing information, successes, strategies, etc. across countries and regions.
Ultimately, we need to learn from the analysed initiative because global governance arrangements are still evolving and because these arrangements can have real impact.
Moreover, such arrangements will be extremely necessary if the world is to have any chance of meeting the new international development agenda, as embodied in the Sustainable Development Goals. As emphasized by this paper, in moving towards these and other goals through global governance arrangements, it is not only the outcomes that matter, but also the democratic character of those arrangements.

Advocacy
Most NECs engaged in lobbying activities such as meeting policy-makers (in ministries of education and finance as well as parliaments), convening roundtables, making petitions (by issuing memoranda, sending public letters to authorities and delivering research reports to parliaments), putting pressure on budget decision-makers and contributing to legislative work (29 coalitions reported contributing to legislative changes such as promoting new education laws and launching new policies)

Training
Training initiatives focused on advocacy, lobbying and communication strategies (25 NECs); education budgeting and funding (18 NECs), networking, partnership building, resource mobilization (9 NECs), national education policies, the right to education, intercultural dialogue, decolonisation, non-formal education, extracurricular education, rural education, inclusive education and community participation in education. Member organizations, non-member union representatives, local government representatives, school boards and councils, teachers, journalists and members of parliament benefited from these initiatives.

Human resources
The staff of the 7 case studies increased from 121 to 195 between 2009 and 2011

Communication
In their final reports 39 NECs mentioned their participation in mass media in order to raise public awareness of the importance of education. Mostly, they participated in radio and television debates, announcements, interviews, round tables, talk-shows and news. Some of them managed to get the support of celebrities. 21 NECs reported to have launched a website. 16 NECs produced a regular newsletter. All of them used at least one of the following instruments of public communication: producing their own posters, flyers, brochures, booklets on good practices in education, policy briefs and digests; disseminating copies of education acts; issuing user-friendly Education Watch reports; and circulating grass-roots stories. 32 NECs joined the GCE Global Action Week, 12 World Teachers' Day, and the International Literacy Day. The CSEF also sponsored the One Goal for Africa campaign during the 2010 World Football Cup. Source: [name deleted to maintain the integrity of the review process]