Education, Research, and Inter-Generational Poverty: Addressing the Challenges to 2015 and Beyond

By Alison Girdwood, DFID.

Over recent years, the scale of resources put into education in developing countries has increased significantly, and is likely to continue to increase.  But research on education in developing countries has, on the whole, remained primarily descriptive and small in scale, investigator driven and mono-disciplinary.

If the challenges of improving quality and equity in international education are to be met, then new ways of commissioning and undertaking education research will be necessary.  In this piece, I explore the drivers for a new approach to research, and some of the possible reasons for the current situation.

Education research – a perspective from one donor (i)

Education has long been a priority for DFID, and by 2015 it will be the largest sector for DFID bilateral expenditure.  The ambition in education programmes is high – to help countries consolidate their education systems, and to contribute to the aspirations of partner governments to provide all children with a quality education and the opportunity of a productive life.  To achieve this level of ambition, donors and the governments / partners they work with need the best possible evidence on both the ‘what’ and the ‘how’ of educational reform.

From analysis undertaken as part of the DFID Bilateral Aid Review in 2010, we know that most of our education programmes rate the evidence base which supports them as ‘low’.  DFID currently spends roughly comparable amounts in its programmes on education and health – but the quality and scale of the research evidence available is very different.  Development interventions are complex, assuming social and behavioural change, and implicitly requiring change to deeply-embedded political structures, with no ‘easy’ solutions.  This is fully recognised.  But so is the fact that the evidence supporting complex interventions is often quite weak.  We seek now to deepen the evidence base in education. The task is challenging.

Policy-relevant research in education

It is only comparatively recently that the lack of strong policy-relevant research (the what and the how) in education has been identified as problematic. This seems to stem from the fact that for many years the international ‘architecture’ for education has been advocacy driven – andthis orientation has subsequently been reflected in a considerable proportion of the research literature (this is clearly a generalisation, and there are of course exceptions to the picture outlined in this blog piece). For example, international literature on Education for All (EFA) was initially designed to galvanise action and to elicit funding commitments – and it focused primarily on conceptualising an idealised framework for commitments, and quantifying shortfalls in both financing and (primarily) access and achievement.

In 2000, MDG2 and supplementary analyses focused attention primarily on access – as an initial step – and generated a large literature on access issues.  These sought to demonstrate the scale of the problem, but only minimally to experiment with solutions. The major global approach to a coherent EFA implementation structure, the Education for All Fast Track Initiative (now the Global Partnership for Education), was specifically designed to fill the policy, funding, data and capacity gaps to achieve the Education for All vision – but was designed with a ‘lean’ secretariat, and minimal analytical capability (and no capacity to undertake research – only now being addressed).

The situation is now changing.  From a donor perspective, the key drivers of the demand for change in education research have included the following:

• Low levels of learning at the foundation level: Consistent  findings emerging from a range of comparatively new – and increasingly widely-applied – techniques to assess the quality of learning have demonstrated conclusively that the current level of learning in schools in many countries is unacceptably low.       Some of the underlying causes of these shocking statistics may be complex to unravel and to address, although some useful attempts have been undertaken.      However, the pressure for effective and reliable methods to improve  the quality of learning is high (and increasing).
• Accountability: Accountability frameworks have strengthened – both donors and  national governments are held increasingly accountable by their      constituencies, whether these are the parliaments, citizens and tax-payers  who approve the financing of international aid; or the electorates,  communities and parents in countries where national surveys (eg those conducted by Pratham / ASER and UWEZO). For both, evidence that learning gains have been minimal shows that opportunities and resources may have been wasted, and demand for change is high.
• Policy-relevance and the use of research:  Accountability for the outcomes of research funding have also strengthened – and research itself is under scrutiny, for example through the Research Excellence Framework in the UK, which requires evidence of research impact.  Work  undertaken on the use of evidence in policy-making, and, for development policy, ODI’s RAPID framework have increased awareness that research which does not engage with policy-makers and respond to their needs is unlikely to be used. For donor-funded research, close relationships between the need of policy makers in the south and the research funding from the north is seen as essential.
• Saturation   with descriptive research:  As stated in my introduction, there is a sense of the growing ‘saturation’ with a generation of primarily descriptive research following the international target setting conferences of 1990 and 2000.
• Fragmentation  and scale:  The fragmented nature and small-scale of much of the literature  has meant that it has often been difficult to aggregate findings against the high level questions facing the sector.
• Nature of  education research: There has been  little significant multi-disciplinary or longitudinal research – although  educational change is, by the nature of its effects, both  multi-disciplinary and long-term. Current research rarely draws  effectively on the wealth of knowledge, approaches and methods in other  disciplines, or from mainstream education research in other contexts.  A richer set of disciplinary ideas and the methodologies may yield new insights and ‘solution-oriented’ research.
• Experimental  and Quasi-experimental studies:  There has recently been an increase in the number of experimental and quasi-experimental studies.  This is an important change, but the initial range of subjects was comparatively a-theoretical and input-dominated. To date, inter-disciplinary approaches and perspectives remain lacking.  As yet (though changing) there has not been enough emphasis on synthesis, contextualisation and replication of findings.
• Theory generation and impact evaluation: Knowledge has been tentatively advanced on some important issues, but there has been no emphasis on a strategic approach to theory and/or knowledge generation – studies have rarely been able to address the ‘black box’ of the  classroom and the process of learning.      
• New technologies: New techniques and means of gathering information through social media and mobile technologies, which can provide real time information and share information offer promising opportunities for future analysis.  These are gradually being used in experimental and ‘real-time’ approaches – but their promise will be enhanced with focus on the policy questions which require research effort. 

The way forward:  It is clear that research should be long-term in nature, and collaborative, if it is to provide the knowledge which policy-makers need.  The demand for policy relevant knowledge will increase, and these demands are beginning to be met – but changes in approach and the training of young (or new) researchers will be needed.  Some promising examples exist, for example, recent emphasis on theory-based impact evaluation – which uses a rigorously designed counterfactual to attribute causality, supplemented with explanatory mixed method and contextual studies – offers great potential, particularly with the scale of many sectoral interventions, and the generation of theory and questions for investigation.  An increasing emphasis on well-designed longitudinal research is also likely.   

(i) My comments arise from work recently undertaken within DFID, and are made informally. These comments do not represent DFID’s formal position – DFID’s approach to education research has yet to be finalised. 

Dr Alison Girdwood is a Humanitarian Evaluation Adviser at UK Department for International Development (DFID). Email: a-girdwood@dfid.gov.uk