Essay 1

The Pathway to Progress on SDG 4 Requires the Global Education Architecture to Focus on Foundational Learning and to Hold Ourselves Accountable For Achieving It

Girindre Beharry

selective photography of green leaf plant

Galileo Galilei (1632, pp. 120–121)

"But surpassing all stupendous inventions, what sublimity of mind was his who dreamed of finding means to communicate his deepest thoughts to any other person, though distant by mighty intervals of place and time! Of talking with those who are in India; of speaking to those who are not yet born and will not be born for a thousand or ten thousand years; and with what facility, by the different arrangements of twenty characters upon a page! [] Let this be the seal of all the admirable inventions of mankind […]."


Today, nine in ten children in Low Income Countries (LICs) cannot read with comprehension by their tenth birthday
1. In other words, they are functionally illiterate, this after decades of declarations and initiatives by the global education community to improve the quality of basic education. This being th e degree zero of our collective aspirations, there is understandably a sense of malaise about the effectiveness of the global education architecture in helping countries address what has been termed the “learning crisis.”

Note: This essay was first published in the International Journal of Education Development in April 2021.

In the last many years, the education community has sought to respond to the perceived deficiencies in the architecture in the form of well-intentioned partnerships, specialized financing facilities, commissions, committees, platforms, initiatives, and forums 2 But these attempts to “fix” the architecture have yet to demonstrate meaningful success: learning levels are persistently low (Le Nestour and Sandefur, forthcoming), and positive deviants hard to find 3.

As a committed partner to the global education agenda, I believe the opportunity is ripe to re-energize the education community by showing meaningful results in the next few years. To do that we should focus on a few objectives, work in countries that share those objectives, go at them with all that the global community has to offer, monitor progress regularly, and hold ourselves collectively accountable for progress. My submission is that one priority objective ought to be addressing Foundational Literacy and Numeracy (FLN) in LICs, particularly in Sub-Saharan Africa (SSA).

I propose FLN as a priority because it is critical for any meaningful progress on the wider Sustainable Development Goal (SDG) agenda. It is also concrete and measurable enough to be both actionable and provide a much-needed metric against which to hold ourselves collectively accountable. This is necessary because the primary actors in global education are currently significantly less than the sum of their parts, prone to general calls to action, but lacking strong incentives for focus and results—a tendency exacerbated by inadequate demand from developing countries themselves for specific outcomes. This paper provides an analysis of the insufficient leadership exercised by the global education architecture, and lays out a set of proposals for these institutions and the sector as a whole to make sure the next decade is one where FLN takes its rightful place in the global education agenda.

2. A short and incomplete history of tall and unfulfilled aspirations

1990s: The Jomtien Declaration of Education for All emphasizes the need to improve literacy because “literacy is a necessary skill in itself and the foundation of other life skills” (UNESCO, 1990, p. 6). It sets out the need for precise floor metrics: “such that an agreed percentage of an appropriate age cohort […] attains or surpasses a defined level of necessary learning achievement” (p. 5).

2000s: The Dakar Framework for Action includes a commitment to improving quality in education and ensuring the achievement of learning outcomes by all in literacy and numeracy (UNESCO, 2000). The seminal 2002 Global Education Monitoring Report (GEMR) report already acknowledges the difficulty in monitoring learning (UNESCO, 2002). GPE's precursor, the Education For All Fast Track Initiative (EFA FTI) is established in 2002. The 2010 evaluation of EFA FTI concludes: “the FTI has remained a weak partnership, with weak accountability, and has not delivered the "compact" to which it refers” (Cambridge Education, Mokoro and Oxford Policy Management, 2010, p. 11). Following the evaluation, the organization evolved into GPE.

2010s: The World Bank’s 2011 Education Strategy posits: “The overarching goal is not just schooling, but learning” (2011, p. 1). It notes the alarmingly low levels of learning, starting early. The Bank’s proposed response is to focus on increasing accountability and results, and to support education reforms that promote learning outcomes. The Brookings Global Compact on Learning report (Perlman Robinson, 2011) again notes that students are in school but not learning; that there are no agreed-upon metrics for tracking learning; and makes the case for a focus on basic literacy and numeracy in school.

Other key milestones from the last decade include the Education Commission’s report, which laments that “despite the known and increasing benefits of education, the world today is facing a global learning crisis” (2016, p. 29). This was followed by the World Bank’s World Development Report (2018) which reprised the theme of a learning crisis.

To review past exhortations on the need to improve learning is to invite reflection on our collective failure to be anywhere near ending illiteracy. The failure does not primarily reside in the quality of the ideas: many of the current prescriptions are not terribly dissimilar from past ones: focus on foundational literacy, improve the quality of assessments, address proximal and distal system issues, hold each other accountable, etc. Why do we then chronically underdeliver on those ambitions? Of course, there is a limit to what the education aid architecture can do; education is local: aid accounts for only 2% of education spending in Lower Middle-Income Countries (LMICs) and 18% in LICs (World Bank & UNESCO Global Education Monitoring Report, 2021). However, the aid architecture surely bears its share of responsibility: in addition to funding, it provides guidance to countries and influences domestic budgets, so the nature of its support should come under critical examination. Is it providing the right guidance, does it have appropriate feedback loops on its performance, and does it hold itself accountable?

3. The global education architecture and its discontents

Burnett (2019) identifies six deficient aspects of the architecture: (a) global leadership, including prioritization; (b) norms and standard setting; (c) knowledge generation and dissemination; (d) monitoring of performance; (e) provision of accountability; and (f) inadequate finance.

A quick mapping of the responses offered by the global education community to these problems suggests that their primary preoccupation has been (f) inadequate finance. Before exploring these responses, it is worth examining whether this aspect of the architecture deserves the spotlight. While the current levels of education spending fall short of what is needed to achieve SDG 4, particularly for LICs, it is less clear that large increases in either donor aid or domestic financing are feasible. First, as noted above, the magnitude of donor aid is currently minimal relative to domestic spending in all but LICs, and the scope for increasing this funding in the context of the pandemic is limited. The value-add of donors in LICs and, even more so, in LMICs, is rather to offer support to countries to improve their education systems by sharing technical expertise. Countries, which provide the bulk of education funding, also have limited fiscal space to increase education spend, except as their tax-to-GDP ratio improves, and their economies grow.

What are some of the education architecture’s responses to the problem of inadequate funding? These fall broadly into two groups: advocacy, and new instruments, both of which have had limited effect. On the advocacy side, the Education Commission’s Learning Generation report estimated that an increase in education finance from $1.2 to $3 trillion a year by 2030 is needed, with Official Development Assistance (ODA) increasing from $13 billion in 2015 to $49 billion a year by 2030 (Education Commission, 2016). In fact, ODA has remained well below this level. As for the instruments, several of these appear to be based on the premise that donors, dissatisfied with the effectiveness of current bilateral and multilateral instruments, will find financing instruments specializing on particular groups of countries, e.g., facing emergencies (Education Cannot Wait), or those graduating from concessional lending (International Finance Facility for Education); or promises of a tighter link between funding and results (Education Outcomes Fund) to be more attractive, and thus increase education ODA. Although that might yet change, there has hardly been a rush to support these new instruments, and they remain marginal to Development Assistance to Education (DAE), just as DAE is marginal (except in LICs) to domestic financing. These new instruments provide increased competition for scarce funds from the same few education donors—in 2019, the US, UK and Norway provided nearly half of ODA to basic education.4 The case for more aid funding can certainly be argued, but there is no state of the world in which goals for donor spending commonly advocated can plausibly be met. Aid for education will always face the unwelcome question of prioritization.

There has been little global conversation about (b) norms and standards, and more about (c) the deficiency of knowledge generation and dissemination, often phrased as a dearth in “global public goods.” The education sector is contrasted unfavorably with the health sector in that regard—in education research receives much less ODA than in health (CGD, 2019). As a foundation, our initial foray into the global education arena was to focus on this piece of the puzzle: what is the evidence, what works, what can we learn from positive deviants? We continue to invest in knowledge generation and dissemination. But the conclusion we came to is that there is little demand for these public goods. There is an insufficient but growing literature out there on the cost-effectiveness of rigorously-evaluated interventions, there are compendiums of promising innovations in education, there is a growing body of research on the complex system dynamics at play in the sector, the Global Partnership for Education (GPE) has launched a Knowledge and Innovation Exchange, and the UK’s Foreign, Commonwealth & Development Office (FCDO) and the World Bank jointly launched a Global Education Evidence Advisory Panel, all of which seek to fill the gap in know-how. But there remains a yawning gap between the knowledge that has been produced and what donors and countries choose to do. My throw-away conjecture, that I will not seek to defend here, is that (a) the questions answered by the research community are rarely the questions asked by policymakers; and (b) there is no true demand for knowledge because there is no sense of urgency about solving problems.

The balance of the global education community’s responses seems to address Burnett’s (a): the global leadership vacuum. In Burnett’s working definition (2019), the exercise of leadership is primarily about the ability to prioritize. I propose a friendly amendment: leadership is actually the exercise of three of his six functions: (a) prioritization (of certain goals), (d) monitoring of performance (towards these goals), and (e) provision of accountability (to achieving the goals). The exam question then is: does the education aid architecture provide such leadership?

3.1 Prioritization

Launched in 2000, the education Millennium Development Goal (MDG) 2 had a very sharp focus. While problematic, it spurred global action to expand primary school completion and, by that metric, could be argued as successful. Fifteen years later, SDG 4 is conversely all-encompassing: it is hard to find an education-related objective that is not included in one of its 10 targets and 43 indicators (UIS, 2020a), which span the entire spectrum from universal access to pre-school education to “education for sustainable development and global citizenship.”

Until very recently, attempts to prioritize have primarily come from bilateral aid agencies. USAID’s 2011 education strategy (USAID, 2011), had a sharp focus on early grade reading—the more recent strategy for basic education is more expansive in its ambitions for young children, with early childhood education, numeracy, and social-emotional skills featuring alongside literacy (USAID, 2018). And the UK’s FCDO (previously DFID) has declared various areas of focus over the years, more recently girls’ education, disability, and learning (UK Government, 2018; UK Government, 2019). Despite trying hard to nudge multilateral organizations to adopt their priorities, bilaterals’ declarations of focus primarily affect their own programs.

Prioritization seems entirely impossible for member-state organizations like UNESCO that seek to serve the needs of all members: the SDG-Education 2030 Steering Committee it convenes reflects this in its composition and in the fact that it devotes attention to all 10 SDG 4 targets for all countries, which can only mean a shallow overview at best. When prompted to focus, UNESCO's reaction has been to oppose, or offer terribly unconvincing responses.5

Nor is focus easy for constituency-based partnerships like the GPE. GPE has some focus: of the aid agencies, their funding is more oriented to primary education, to LICs, and, as of December 2020, to learning outcomes (GPE, 2020a). But their planning and consultation process is for the education “sector” as a whole. And their partnership structure means that they need to cater to the divergent interests of all their stakeholders. While there continue to be voices in the Board seeking focus, that is not where the Board lands. In recent months there have been Board level discussions at GPE on whether to adopt a compelling “rallying cry” (GPE, 2020b), i.e., a sharper articulation of its overarching objective. It is unclear whether this was meant to be a useful slogan going into their 2021 replenishment, or a genuine attempt at sharpening the goal the Partnership would hold itself accountable for. In any case, the Board was unable to agree on a focus and abandoned the endeavor.

Prioritization is also challenged by some as anti-democratic attempts to relitigate the SDGs (Archer, 2019). Another common angle is that since education is a right, rather than prioritize particular goals, the task ahead is to make sure the programs needed to attain all SDG 4 targets are adequately funded. The key weakness in this line of reasoning is the enormous discrepancy between the scale of the financing needs which have been calculated—e.g., the Education Commission’s estimate that spending would need to rise from $1.2 trillion to $3 trillion a year in 2030 (Education Commission, 2016)—and the resources that Ministries of Finance and donors can realistically draw upon. Conversely, I would argue that the right to learning, which starts with acquiring foundational skills, is a powerful argument for prioritization.

Others agree to the need for some prioritization but have very different ideas of what those priorities ought to be. A non-exhaustive list of current topics includes: early childhood development (ECD), girls’ education, universal free secondary schooling, socio-emotional and other “21st Century” skills, and short-term skilling for workplace readiness. Who is to argue that these are not entirely desirable goals? All of them find their rightful place among the SDG 4 targets, but countries will need to consider which ones to prioritize depending on the maturity of their education systems.

Perhaps most importantly, it is even less feasible for the global education actors to focus when the countries they seek to support do not express any strong interest in such focus. I will return to this shortly. Because of this, attempts by global actors to prioritize an agenda frequently leads to the following outcome: countries go along with what they know to be the “donor agenda” because funding and technical assistance flow from it, but domestic energies are directed somewhere else entirely.

Why does it matter that the sector lacks clear priorities? The reality is that prioritization is happening all the time by virtue of the fact that there is simply not enough money to go around to meet all the objectives. Which means every agenda is underpowered and progress is grindingly slow. Getting any of the SDG 4 objectives accomplished will be extraordinarily difficult and cannot be achieved by devoting inadequate attention, ingenuity, and resources to it. It is perhaps a little too hopeful to believe that there will be massive funding redirected to education from national or aid budgets when countries are being hammered by a pandemic. Note that LICs and LMICs already allocate more of their government budget to education than richer countries. If anything, we are likely facing a world where maintaining current levels of aid and domestic funding would be deemed a success.

The real choice we are making is between a tacit form of prioritization and an explicit one. There is no question that countries and aid agencies will always end up doing something that is a balance between things that are important to do, things that are possible to do, and things that are urgent to do, but if there were an explicit prioritization framework, the hidden costs of poor decisions would be made obvious.

We do not have a nice framework that factors in equity and returns considerations, so I instead lay bare the observations and beliefs that make me advocate for the aid architecture to do all it can to help improve foundational literacy and numeracy in SSA/LICs as one of its top objectives:

 Practically all SDG 4 goals depend on the achievement of FLN. Without FLN Mali, say, cannot possibly ensure that by 2030, all girls and boys complete free, equitable and quality primary and secondary education leading to relevant and effective learning outcomes. FLN is a gateway skill: there is no leapfrogging one’s way to twelve years of schooling for all girls without it.

 The current levels of learning are catastrophic; e.g., two percent of Malian children meet minimum proficiency levels for early grade reading (UIS, 2020b).

 Most poor students today can “access” some form of schooling, but they drop out disproportionately because schools fail them; by focusing on universalizing FLN, we are sure to primarily address the poor and marginalized.

Universalizing quality FLN means we are addressing the learning of children in and out of school.

 While it would be satisfying to point to convincing causal evidence on the impact of FLN skills on later outcomes, this is still an area that lacks rigor, and deserves increased research attention (Evans and Hares, forthcoming). However, even without proof that FLN will have the largest impact on long-term outcomes, building the foundation is the only route to more advanced skills.

The case for focusing on FLN in LICs/SSA is that they are behind on all attainment, achievement—nearly 9 of 10 children aged 10 in SSA are not able to read with comprehension (World Bank, 2019a,)—and equity metrics. They have had to very quickly hire masses of teachers to meet the MDG 2 targets at the cost of quality—the proportion of trained teachers at the primary level in SSA fell from 84% in 2000 to 64% in 2017 (UIS, 2020c); and, uniquely, SSA faces massive student population growth for the next many decades. Countries like Madagascar, where the population of 28 million, is projected to rise to 100 million by 2100 (United Nations, 2019a), and the cohort of new students entering primary school every year is meant to increase through to 2093 (United Nations, Department of Economic and Social Affairs, Population Division, 2019b).

I hasten to add that improving FLN, or its World Bank formulation of improving reading comprehension by age 10 (World Bank, 2019b), is not a new goal plucked out of thin air. Indeed, they are mere variations or expressions in lay terms of the very first two SDG 4 indicators: SDG 4.1.1(a) and (b) for the initiated. Equally importantly, this is not to indicate that education stops at the acquisition of literacy, or that we should reify particular literacy or numeracy metrics: there is a substantive conversation to be had in each country on its quality aspirations for basic education.6 But the reality of it is—once again—that there is no bypassing the step of getting FLN right.

There are few case studies of substantial system-wide improvement in learning outcomes in LICs and LMICs, so the few cases of improvement are of great interest. Take the case of the municipality of Sobral, in the state of Ceará in Brazil. While Ceará has the fifth lowest GDP per capita in the country, the municipality of Sobral has the highest ranking in the National Education Index (Rodrigues & Loureiro, 2020). The results were obtained entirely endogenously to the system. The biggest takeaway from Sobral, apart from the predictable ingredient of political will, is their relentless focus on the achievement of literacy by the end of second grade(Crouch, 2020a). Education in Sobral does not stop at the second grade; like every school system they have much larger aspirations for the education of their students, but they understood that they could not achieve any of those if they did not achieve early grade literacy. And because of that focus, they were able to examine not the “sectoral” dysfunctions at large, but the system dysfunctions as they relate to this objective, which provides a sharper diagnosis, for instance, on the instructional practices and the suite of system levers (textbooks, teacher training, coaching, assessments) that sustain good practice (Crouch, 2020b; Loureiro et al., 2020).

The contrast with the health sector is striking. A focus on reducing neonatal, infant, and under-5 mortality in LICs does not especially provoke heartburn. These are useful markers: while it is perfectly understood that health does not stop mattering at age 5, it is also obvious that there is no thriving without surviving. These metrics elicit pointed responses, much more so than broad “health sector plans” can. Finally, these indicators are often markers of the health of the health system more broadly.

Far from being a relitigation of SDG 4, securing FLN is a precondition for its attainment. To take the SDG 4 goals very seriously we must recognize that they cannot be achieved without solid basic education; and to take the invitation to “reach the furthest behind first” (UN, 2015, p. 7) literally we must advocate for a concentration of donor energies on SSA and LICs.

However, if there is such a strong case for prioritizing FLN in LICs, a reasonable question is then: why is it not at the top of the domestic education agenda in those countries?

First, policymakers in LICs rarely prioritize FLN because there is no electoral demand for quality primary education—see for example Harding and Stasavage (2013). The few governments that have decided to afford it priority have done so because of the personal conviction of policymakers—as outlined by Crouch in his review of the case studies of Sobral in Brazil, Puebla in Mexico and Kenya, political motivation was crucial to the focus on FLN (2020b). That the Government of India recently launched an FLN mission (Dhawan, 2020) is the product of the conviction of a few bureaucrats, not something that is in response to electoral demand. Governments face parental and popular pressure on more tangible things and thus more readily respond to issues like free secondary schools, and workplace readiness. FLN is fundamental to good secondary schooling and to skilling, but that perspective is hard to maintain.

Second, what makes it difficult for policymakers to prioritize FLN is that many think it is somehow solved. A recent survey by the Center for Global Development (CGD) of some 900 LIC and LMIC education bureaucrats shows that policymakers’ perception of levels of learning vary highly in accuracy, and that in most countries the estimations are optimistic relative to the actual levels of learning (Crawfurd et al., forthcoming). In other words, policymakers are not aware they have a problem to solve.

Third, there is the common view, based on long historical precedence, that the primary purpose of an education system is to produce a highly-trained elite who will be captains of industry, run government, and power economic growth. Thus, a system where only a select few make it through the end of secondary school, and even fewer are admitted to university, is perceived to be exactly what the country needs. In this model, the pursuit of broad human capital development, and even something as basic as literacy, is not considered to be a priority (Muralidharan, 2019). This view of the world is underpinned by the belief that the system is meritocratic, and that the best and brightest, irrespective of their family circumstances, will rise to the top, to the benefit of all. This is of course not true—young people in Sub-Saharan Africa whose head of household did not complete primary education are ten times less likely to get to tertiary education than those whose parents attained at least a secondary education (Darvas et al., 2017).

3.2 Performance monitoring

The second component of global leadership is monitoring performance. How is the education aid architecture monitoring performance today? The reference points for global7 monitoring are UNESCO Institute for Statistics (UIS)’s publications (e.g., the SDG 4 Databook) and the GEMR.

The 2019 edition of the UIS SDG 4 Databook (UIS, 2019) was interesting mostly because it was highly problematic: not only were there very few data points for SDG 4.1.1 a and b in SSA (which together ought to give a good sense of learning in early grades), but in some cases the data from the same country did not make sense from year to year. While a new reporting protocol has removed some of the inconsistencies from the data on the UIS website (UIS, 2020b), for many countries in SSA there are no data points at all in the last 10 years for early grade reading, while for others the only available data is shockingly old—for example, the latest data point for Botswana is for 2011.

UIS is hampered by the fact that LICs/LMICs may not conduct quality assessments with any established periodicity, often because donors support them inconsistently; report on data sometimes coming from entirely different assessments, so that year-to-year comparison is meaningless; and rarely participate in cross-national assessments, particularly for early grades. But, perhaps more importantly, who is the client for UIS data? If this data were feeding into an accountability structure, an alarm would be raised about the fact that, at a global level, we cannot answer very basic questions: which countries are making progress on FLN, which are not, which are on track to meeting their SDG targets and which are not, which countries stand out that we ought to learn from, where should resources be targeted based on need? If these questions are not being begged, it is because there is no clear, high-powered recipient that can act on the data UIS produces.

The Global Education Monitoring Report is meant to, well, monitor. But because they are meant to report on all 43 indicators for every region, and because they are in good measure reliant on UIS data, it is hard to make sense of what these reports tell us, besides that we are behind on everything, repeatedly. And monitoring on its own may not be particularly helpful if not also accompanied by an examination of why progress is slow or quick. The primary problem, however, is that, like the UIS, the GEMR does not feed into a process by which the data is taken to heart and corrective action is taken as a result; in other words, it is not tied to an accountability mechanism.

3.3 Accountability

The third leg of the global leadership stool is accountability. The term itself is a source of great angst in the education sector, often because accountability is equated—wrongly—with blaming teachers. I am unapologetic about the use of the term in the following sense: if we, the global education community, are collectively committing to something, and are to take that commitment seriously, it is incumbent on us to take periodic stock of progress, to reflect on the reasons why we are or aren’t making any, and to alter the course of our action as required. Moving education outcomes is incredibly hard; it is even harder when we are not learning and adapting our work. The lack of accountability today does not translate into immediate consequences for anyone—not technical agencies, not funders, not philanthropies, not governments, not NGOs, not CSOs—except for students. They are the ones whose chances in life are denied because we are not willing to take a critical look at what we are doing and striving to do better.

It is not true to say there is no accountability at all. Indeed, each individual aid agency tracks their projects and reports internally, and countries have their own feedback loops, particularly pass rates for high-stakes exams. Not to single them out, but if the World Bank already correctly identified the problem of learning in its ‘Learning for All – Education Strategy 2020′ as far back as 2011, how does it rate itself against the organizational performance, outcomes, and impact metrics it signed up for back then, and what has happened as a result of not achieving them? Tellingly, the Learning at Scale Project funded by us-- an attempt by researchers to unearth and examine exemplar projects that improved learning meaningfully at scale—failed to find a single World Bank or GPE project that met the researchers’ inclusion criteria (Piper & Dubeck, 2021). It is hard to convince oneself that the aid agencies hold themselves to account particularly rigorously.

Do we as a collective hold ourselves accountable? Even less so. First, there are few collective fora to discuss progress. In theory the UN SDG-Education 2030 Steering Committee could play that role, but in its current form it is not set up for anything like reviewing progress except in the broadest possible sense, and it certainly does not energize collective action for specific objectives. The Committee is in the process of reimagining itself; this presents an opportunity for it to think through fit-for-purpose structures to strengthen performance monitoring and mutual accountability.

The GPE is on paper an ideal platform for such collective reflection and collective response. After all, it focuses on LICs and on basic education, and it has all the partners at the same table: if there is lack of progress there is no one else to point to but the organizations represented at the GPE Board. But GPE’s Results Framework, for instance, disappointingly only tracks the “proportion of developing country partners (DCPs) showing improvement on learning outcomes (basic education),” (GPE, 2019, p. 1) and reports on only about a third of countries in their portfolio. More curiously, the GPE Board does not seem keen to take stock of progress on a regular basis, say, to look at the progress in getting even basic data about learning levels, or look at the bottom 20 countries and ascertain whether they are making sufficient progress on proximate indicators of learning. One concern is that this would put countries on the spot, which is not the spirit of the partnership. Yet this is an unconvincing argument: this could be an opportunity to review the work that the various agencies do in support of countries and hold them accountable. An uncharitable view would be that no one wants to be held accountable: countries may stand to lose funding if they don’t tell a good story; donors need to report back to their domestic constituencies and also show their tax dollars are being put to good use; so a low-expectations low-performance equilibrium obtains. GPE is in the thick of a governance review, and of developing a new strategic plan and results framework: these present a timely opportunity to strengthen its accountability structure.

The current global architecture is exquisitely positioned today to know next to nothing, let alone have a meaningful conversation about whether we are making sufficient progress on even such a foundational objective of SDG 4 as that of ensuring basic literacy.

4. Getting a few things done well

Here then is the state of things: (a) there is a dependency among SDG 4’s objectives: if the foundation is weak—as is the case in LICs—it is to fool ourselves to pretend that we can meet objectives such as 12 years of schooling for every girl by 2030; (b) improving the quality of basic education is very hard, and cannot be accomplished by facile prescriptions: it requires rigor, serious system-wide effort, and persistence; (c) there is not enough money from domestic and international sources to support all SDG 4 objectives more than symbolically; (d) the outlook for aid and domestic budgets is bleak and to live in the hopes of a financial manna in the middle of a brutal pandemic is to be excessively hopeful; (e) if we are to make meaningful progress, the aid community needs to prioritize among the many competing objectives—I make the case for FLN in LICs as a priority—and hold itself collectively accountable for progress; (f) neither countries nor constituency- or member-based partnerships are willing or capable of prioritizing or holding themselves accountable. We are stuck. In the past many months, there have been a number of attempts to “do something”: the creation of the Global Education Forum, the Education Commission, the Multilateral Education Platform, a new GPE strategy, and now revamping of the SDG-Education 2030 Steering Committee. My submission is that unless these efforts aim squarely at the problems of prioritization, performance monitoring, and accountability, their impact will be minimal at best.

Past confabs have skirted these difficult issues, and collaboration efforts end up being loose commitments to partner better, and to advocate for more money for education (which is admittedly needed for LICs). Because the incentives faced by actors in the global architecture are so powerful, because there is no real sense of crisis about the ‘learning crisis’, and because sectoral leadership is so distributed, I don’t see room for big shifts in the architecture, or grand bargains that address the problems of prioritization, of monitoring progress, and of holding oneself accountable. The experience in the health sector indicates that even high-powered initiatives (IHP+, UHC2030), which seek to strengthen collaboration and accountability among agencies in pursuit of joint objectives, meet with limited success: the incentives within aid agencies to pursue their own agenda and show results trump those inviting collaboration between agencies; and agreements at head-of-agency level is never sufficient to get their highly-decentralized operations to follow suit, unless internal incentives are made strong enough.

I would passionately like my pessimism to be proven wrong. In the meantime, my proposed solution is to cut the Gordian knot by building on what we have, inviting tactical shifts by a few actors, leaving the architecture as it is, and side-stepping the vexing prioritization issue by simply working with countries and agencies already persuaded of the need to prioritize FLN. The opportunity is seeing greater convergence of late of a few major actors around FLN as a priority objective: the World Bank’s Foundational Learning Compact (FLC), seeks to support “accelerator” countries in their bid to improve FLN; USAID has been the agency most sharply focused on “all children reading” in the last decade and continues to be a prominent investor in this area; UNICEF launched a “mission-approach” to FLN. There is also the cumulated knowledge of many local and international NGOs which have worked on this particular problem for a decade or more. While the FLC is a promising nugget to build from, it will require the same ingredients of maintained focus, performance monitoring and accountability structures to motivate real progress. Here is my wish-list for the pieces that these and other partners could contribute to make sure we stay on task.

My invitation to policymakers in LICs is:

  •  To those who have chosen to focus on FLN as a priority and decided to work with global education agencies on improving it, to make sure there is an honest annual review of progress.
  •  To those that have not, to perhaps take another look at the data for your country, not the exam scores, and if you don't have a learning assessment, make sure you introduce one and ask for support from the development partners. Decide whether you are satisfied with the levels of learning obtaining at the end of grade 3, knowing that those students who have not achieved mastery of those skills by then likely never will; and knowing that secondary schooling will thus remain the privilege of the fortunate.

My invitation to GPE is:

  •  To leverage its current governance review to make good on its desire for greater accountability. Or, cutting to the chase, how about reviewing at every Board meeting priority outcome indicators, particularly FLN, for a set of countries?
  •  To ensure that each Joint Sector Review reports back on the proximate indicators for priority SDG 4 objectives, including FLN.
  •  Ensure that countries receiving funding from GPE establish robust assessment systems, as per its own requirements.

My invitation to UNESCO is:

  •  To avail themselves of the current effort to restructure the SDG-Education 2030 Steering Committee so it can fill in the leadership gap. A sub-committee could, for instance, focus on LICs and a set of priority indicators as identified by UIS and the GEMR, which includes FLN (Montoya & Antoninis, 2019); use the data collated by UIS and the GEMR to shed light on progress regularly; and be high-powered enough in its composition so that the review from the committee triggers action by countries and the global actors supporting them.
  •  To fund and fundraise for UIS properly. We cannot monitor SDG 4.1.1 without data. Until countries produce robust data from their assessment systems, we need creative ways to make collective sense of learning outcomes.
  •  To rally its regional networks behind the Global Education Monitoring Report’s ability to monitor progress on FLN in SSA and LICs on a regular basis.
  •  To support the International Institute for Educational Planning (IIEP) to develop specific capacity in country to interrogate their education systems with a view to improving FLN.

My invitation to the World Bank is:

  •  To treat their institution-wide commitment to reducing learning poverty by half by 2030 with utmost seriousness. It will require a serious effort to align Bank operations against those objectives. This is not a given: the Bank has objectives at a regional level that compete with their corporate objectives, and its operations are highly decentralized.8 It will also be incumbent on the World Bank to equip staff with the right know-how.
  •  To partner actively with others, especially those organizations that have been sharply focused on FLN in the last decade; we need to make use of the assets we already have.
  •  Equally importantly, if this is to have the level of effort and persistence over the decade that it will require, to conduct honest, regular reviews of progress with governments at the highest level and with partners, and to report on progress.

My invitation to bilateral donors is:

  •  To fund FLN adequately, particularly in SSA. Bilateral aid in education is larger9 than multilateral aid, and bilateral agencies have more latitude to prioritize. A recommitment from USAID to their “all children learning” agenda would be incredibly welcome, as would seeing FCDO deepen their country work in support of learning.
  •  To support research and evaluation. Again, bilaterals are uniquely able to direct funding to public goods, which have far greater reach than their programming (e.g., the evaluation of Tusome).
  •  To use their voice in multilateral forums not only to foster “greater collaboration” but to promote the review of progress and collective accountability against agreed metrics.

My invitation to CSOs and NGOs is:

  •  To use their powerful voices to not only advocate for more spending on education, but to hold countries and the global aid architecture accountable for the collective promises made over the years to improve learning outcomes, starting with FLN.

My invitation to other actors in the global architecture is:

  •  To join this collective venture to address FLN, or SDG 4.1.1(a) and (b). This is a very hard task: there are precious few case studies of LICs having done so at any scale. It will require focus, persistence, know-how, rigor, ingenuity, political savvy, and accountability. The incentives of the global aid architecture to maintain that focus and persistence are weak, and organizations can contribute in myriad ways to keep us collectively on our toes.
  •  To offer a better pathway for progress, one that retains contact with fiscal realities, if you do not like this one.

Our Global Education Program at the Bill & Melinda Gates Foundation emphasizes foundational learning in primary grades. We support efforts to improve the availability and quality of learning assessment data, to identify barriers to educational access and learning, and to conduct research on effective instructional practices, including using educational technology. We also support efforts to measure progress, celebrate successes, and challenge education decision-makers when commitments are not realized. Within the means of our program, we look forward to supporting these shifts and our partners as we collectively work towards meaningful results in the next few years.


Archer, D. (2019, October 25). Opinion: The World Bank shouldn't be setting global goals for education. Retrieved February 3, 2021, from Devex:

Burnett, N. (2019). Invited Essay: It’s past time to fix the broken international architecture for education. International Journal of Educational Development, 68, 15-19. doi: 

Cambridge Education, Mokoro & Oxford Policy Management. (2010). Mid-term evaluation of the EFA Fast Track Initiative. Final report volume 1 - Synthesis report. Executive summary. Retrieved from

CGD. (2019). The state of education aid finance [CGD Analysis of CRS/DAC Data - unpublished powerpoint]. Washington D.C.: Center for Global Development.

Crawfurd, L., Hares, S., Khan, A., & Sandefur, J. (forthcoming). Understanding the education aid architecture: The view from recipient governments. Washington D.C.: Center for Global Development.

Crouch, L. (2020a, July 9). How to rapidly improve learning outcomes at system level? Retrieved February 3, 2021, from Global Partnership for Education:

Crouch, L. (2020b). Systems implications for core instructional support lessons from Sobral (Brazil), Puebla (Mexico), and Kenya. Retrieved from RISE:

Darvas, P., Gao, S., Shen, Y., & Bawany, B. (2017). Sharing higher education’s promise beyond the few in Sub-Saharan Africa. Retrieved from

Dhawan, A. (2020, August 3). NEP can help make India a knowledge-based economy. Retrieved February 3, 2021, from Hindustan Times:

Education Commission. (2016). The learning generation: Investing in education for a changing world. Retrieved from

Evans, D. K., & Hares, S. (forthcoming). Foundational literacy and numeracy investments: Examining the evidence. Washington D.C.: Center for Global Development.

Galilei, G. (1632). Dialogue concerning the two chief world systems. (S. Drake, Trans.) Modern Library Inc.

GPE. (2019, April 11). GPE results framework for 2016-2020. Retrieved February 10, 2021, from

GPE. (2020a, December 3). Board decisions. December 2020. Retrieved February 3, 2021, from

GPE. (2020b, June 11). Decisions of the Board of Directors. Retrieved February 3, 2021, from

Harding, R., & Stasavage, D. (2013). What democracy does (and doesn’t do) for basic services: School fees, school inputs, and African elections. The Journal of Politics, 229-245. doi:

Le Nestour, A., & Sandefur, J. (forthcoming). Africa’s literacy boom, 1940-2000. Center for Global Development. Washington D.C.: Center for Global Development.

Loureiro, A., Di Gropello, E., & Arias, O. (2020, July 9). There is no magic: The formula for Brazil’s Ceará and Sobral success to reduce learning poverty. Retrieved February 3, 2021, from World Bank Blogs:

Montoya, S., & Antoninis, M. (2019, December 19). Benchmarks: the forgotten commitment for accelerating progress towards SDG 4. Retrieved February 3, 2021, from

Muralidharan, K. (2019). Reforming the Indian school education system. In A. Banerjee, G. Gopinath, R. Rajan, & M. Sharma (Eds.), What the economy needs now. Juggernaut. Retrieved from

OECD CRS. (2019). Creditor Reporting System (CRS) [dataset]. Retrieved February 1, 2021, from oecd.stat

PASEC. (2020). PASEC 2019: Qualité des systèmes éducatifs en Afrique subsaharienne francophone [Quality of education systems in francophone sub-Saharan Africa]. Retrieved from

Perlman Robinson, J. (2011, June). A global compact on learning: Taking action on education in developing countries. Retrieved from

Piper, B., & Dubeck, M. M. (2021). Structured pedagogy literature review: A how-to guide. Retrieved from

Rodrigues, D. C., & Loureiro, A. (2020). Achieving world-class education in adverse socioeconomic conditions: The case of Sobral in Brazil (English). Retrieved from

UIS. (2019). SDG 4 Data Book. Global education indicators 2019. Retrieved from

UIS. (2020a). Official list of SDG 4 Indicators. Retrieved February 3, 2021, from

UIS. (2020b). Sustainable Development Goals 1 and 4: 4.1.1 Achieving at least a minimum proficiency level in reading in Grade 2 or 3 [dataset]. Retrieved January 28, 2021, from

UIS. (2020c). Education: Percentage of trained teachers by teaching level of education [dataset]. Retrieved August 18, 2020, from

UK Government. (2018). DFID Education Policy: Get children learning. Retrieved from

UK Government. (2019, September 24). PM steps up UK effort to get every girl in the world into school [Press release]. Retrieved February 3, 2021, from

UN. (2015). Transforming our world: The 2030 agenda for sustainable development. Retrieved February 3, 2021, from

UNESCO. (1990). World declaration on education for all. Retrieved from UNESDOC Digital Library:

UNESCO. (2000). The Dakar framework for action: Education for all: meeting our collective commitments (including six regional frameworks for action). Retrieved from UNESDOC Digital Library:

UNESCO. (2002). Education for All: is the world on track? EFA global monitoring report, 2002. Retrieved from UNESDOC Digital Library:

UNESCO. (2019, December 11). Bringing the international education community together/World education blog. Retrieved February 3, 2021, from

United Nations. (2019a). World Population Prospects 2019, Volume II: Demographic Profiles. Madagascar. Retrieved February 3, 2021, from

United Nations, Department of Economic and Social Affairs, Population Division. (2019b, August). World Population Prospects 2019, Online Edition. Rev. 1. [dataset]. Total population (both sexes combined) by single age, region, subregion and country, annually for 1950-2100 (thousands). Medium fertility variant 2020-2100. Retrieved February 3, 2021, from

USAID. (2011). USAID education strategy: Opportunity through learning. Retrieved from

USAID. (2018). U.S. Government strategy on international basic education. Fiscal years 2019-2023. Retrieved from

World Bank. (2011). Learning for All: Investing in people's knowledge and skills to promote development. World Bank Group Education Strategy 2020. Retrieved from

World Bank. (2018). Learning to realize education's promise. Retrieved from

World Bank. (2019). Ending learning poverty: What will it take? Retrieved from

World Bank. (2019, October 17). New target: Cut “Learning Poverty” by At Least Half by 2030. Retrieved February 3, 2021, from

World Bank. (2020, June 2). India - Strengthening Teaching-Learning and Results for States operation (English). World Bank Group. Retrieved February 3, 2021, from

World Bank & UNESCO Global Education Monitoring Report. (2021). Education finance watch 2021. Retrieved from