Widening the Evidence Base in Education: The Urgent Need to Expand ‘What Counts’ to Deliver Better Policy Making

In this blogpost, which is part of NORRAG’s blog series #TheSouthAlsoKnows, Sophia D’Angelo argues that the evidence base in educational policymaking must be widened to include voices of the Global South.

Using evidence to inform policymaking in the Global South presents many challenges. All too often, policymakers find themselves relying on research that is either inaccessible, incomprehensible or irrelevant to their circumstances (Education.org, 2021).

In evidence synthesis, academic literature is held up as the gold standard, but academic education papers are dominated by ‘Anglo-Saxon’ (i.e. the United Kingdom, Canada, the United States, Australia or New Zealand) contexts and authors in the Global North (OECD, 2025).

Currently, the hierarchy of knowledge privileges studies published in international academic journals, the vast majority of which are produced by scholars in the Global North. Research by Brun, Saab, and Shuayb (2024) highlights that 86% of single-authored, English-language studies on refugee education are written by Global North scholars, even when the research focuses on Global South contexts. Our experience gathering evidence for syntheses on Accelerated Education Programmes—delivered primarily in low- and middle-income contexts—echoed these findings.

This imbalance is not just academic—it directly influences which policies are implemented and whose perspectives are legitimised. Considering these limitations, where are policymakers to turn—beyond academic literature for contextually relevant insights?

As a former Deputy Director of Ghana’s Ministry of Education put it in a discussion with Education.org in 2023: “Most of us in the policy department are often groping in the dark when promulgating education policies because we rarely have relevant data to support our actions.” This is a sentiment we hear consistently when working with civil servants throughout the region.

Yet, valuable knowledge—generated by NGOs, think tanks, teacher networks, governments, and even researchers—often sits on shelves or desktops, unrecognised by the international research community. This creates a policymaking environment where global ‘solutions’ are imposed without adaptation to local realities.

A more inclusive approach is not only necessary but possible. Education.org’s LIFTED framework (Locally Inclusive Framework for Transforming Evidence in Education Decision-Making), which is designed to ensure evidence syntheses capture a broader and more representative knowledge base, evolved from our leadership of an International Working Group (IWG). Established in 2023, the IWG brought together 27 organisations—including the African Union, APHRC, Echidna Giving, Education Endowment Foundation (EEF), ESSA, Gates Foundation, GPE, INEE, Jacobs Foundation, PACE, Porticus, SUMMA, UNHCR, USAID, World Bank, and others—to create and test guidance on finding, classifying, and appraising a wide range of evidence.

Our latest synthesis on accelerated education demonstrates the potential impact of including appraised non-academic sources, identifying evidence from 70 countries—41 of which would have been excluded if only academic literature were considered. Expanding what counts as valuable evidence is particularly crucial for marginalised groups. For instance, the number of sources providing evidence on girls’ education increased tenfold from 8 to 83 when non-academic sources were included. Similarly, the number of sources on forced migrants and emergency contexts quadrupled, from 16 to 64.

Notably, incorporating non-academic sources also elevates voices from the Global South. While only 26 academic sources in the synthesis were authored by at least one institution based in the Global South, this number increased fourfold when non-academic sources were included. Despite this progress, of course, significant disparities remain—two-thirds of the non-academic sources identified still lacked any Southern authorship. Addressing these structural imbalances is essential to achieving true equity in educational research.

As Dr. Lucy Heady, CEO, Education Sub Saharan Africa (ESSA) said in an International Working Group session in 2024, “We’ve seen that the gender balance was better in grey literature compared to publications in indexed journals. Focusing only on internationally published articles becomes an equity issue.”

Expanding the evidence base enables a greater understanding, not only of what’s effective but why, how, and for whom. But doing so requires a mindset shift from international organisations, funders, and researchers. Policymakers in the Global South do not have the luxury of waiting for peer-reviewed studies; they need actionable, context-specific insights now. The OECD’s (2025) report, Everybody Cares About Using Education Research Sometimes, notes that while 99 organisations curate education research repositories, the majority focus on traditional academic outputs, failing to incorporate diverse knowledge sources. The status quo is not just inefficient—it actively undermines education reform efforts in the places that need them most.

Widening the evidence base is not about lowering standards or inclusion for inclusion’s sake; it is about recognising that valuable knowledge exists in multiple forms and that the incorporation of non-academic literature—government reports, community-led evaluations, and local case studies—helps improve, not only the relevance of insights derived from synthesis work but the global education discourse, making it more responsive to the needs of those on the ground.

This issue is even more pressing in an era where artificial intelligence (AI) is increasingly being used to aggregate knowledge and generate insights. AI models are trained on existing data, which is overwhelmingly biased toward Global North perspectives. If the underlying knowledge base remains exclusive and unrepresentative, AI-driven insights will further entrench epistemic inequalities rather than democratising knowledge. As AI tools become more central to decision-making, it is critical that they draw from a diverse, contextually relevant, and inclusive evidence base. Otherwise, we risk automating and perpetuating the very biases we are striving to dismantle.

The time for change is now. Without such a shift, we risk perpetuating an inequitable system where those most affected by education challenges are systematically excluded from shaping their own solutions.

The Author:

Sophia D’Angelo, Director of Research, Education.org. Contact: sophia@education.org