Building a theoretical framework to understand the role of aid in achieving the education Millennium Development Goals in fragile states

This thesis aims to build a theory for understanding the role of aid in achieving the education Millennium Development Goals (MDGs) in fragile states. In so doing, it responds to claims that both educational research (see e.g. Cohen et al., 2000), and the economic literature on aid and international development (see e.g. Deaton, 2008), are insufficiently grounded in theory. In finding a methodological voice for this thesis, I distinguish between three research paradigms: positivist, interpretive and critical theory. I ask whether theory is essentially a positivist project, better suited to quantitative methods and to the natural sciences. I argue for a 'mixed-method' approach, proposing that when qualitative methods generate data that are subjected to a stronger process of generalisation – including comparison between data derived from qualitative and quantitative methods, and from macro and micro level analysis – then that evidence may be sufficiently strong to underpin theory. I use a four step process to build theory: (i) categorising data into domains for analysis, (ii) hypothesising linkages between these domains, (iii) investigating these hypotheses through assessing the evidence supporting them, (iv) organising hypotheses into a theoretical framework. To assess the strength of evidence in support of each hypothesis, I use an instrument to ‘grade the evidence’, based on a threefold assessment of method, observer bias and corroboration. I include evidence from new research conducted for this thesis, including: a portfolio analysis of 145 DFID education projects in fragile states (1991-2007), and an analysis of primary data collected for the 2008 DFID ‘Education Portfolio Review’. The findings of this research confirm a potential relationship between aid inputs and education outcomes in fragile states. Positing that this relationship might work through intermediate financing and institutional effects, it finds weak evidence for the former, but stronger evidence for the latter. With both aid and non-aid inputs (e.g. diplomacy, military engagement), external inputs appear better at supporting existing incipient reform than generating that reform, suggesting that donors should adopt a more modest and opportunistic approach to aid, as opposed to deploying a ‘transformational’ blueprint (Easterly, 2009). The inter-dependence between aid inputs and non-aid inputs points to the importance of deploying instruments within a single approach to strategy and possibly delivery. There is relatively strong evidence for ‘pre-conditions’ for successful interventions – proposed here as political will, community ownership and security / stability – whereas evidence for conventional proxies of ‘aid effectiveness’ is weak relative to the importance generally ascribed to it. The evidence linking education and social stability is mixed, and weakly researched in developing country contexts – potentially significant for critical theorists who question the wisdom and motives of donor governments investing in education to counter radicalisation. I conclude by assessing whether the theory generated has validity or utility. I assess the theory against five key characteristics of theory: empirical grounding; explanatory power; predictive power; utility; verification / falsification. I conclude that my theory has explanatory power and utility, but that claims to generalisability are weak, given the importance of context. The thesis and its product (the ‘theory’) provide a framework that advances our understanding of the relationships between aid and education outcomes in fragile states. It tests the evidence base for these proposed relationships and, notwithstanding limits of generalisability, offers a narrative and framework with practical utility for future research, policy development and programming.