A meta-analysis of practice, policy and theories

Development Policy18 Nov 2011A. Helmsing, Sietze Vellema

The discussion in this blog is about shaping platforms for knowledge generation, exchange and utilisation. Our input identifies the function of a knowledge platform as using existing research, in practice, policy and academia, for a meta-level analysis informing strategic choice making as well as evaluation and learning. This function requires flexibility in the form and processes wherein teams contrast and weigh different theories and pieces of evidence. We think that researchers are trained to guide a platform towards meta-level analysis. Researchers no longer hold the monopoly in primary research. Hence platforms open a space for linking all forms of data collection to a joint meta-analysis.

Our perspective is based on a DPRN-project (Development Policy and Review Network) about value chains, inclusion and endogenous development. Development-driven interventions and policies in value chains have adopted a strong orientation towards environmental and social standards and linking smallholder farmers to business. The related activities focus, for example, on enabling compliance with standards, organising women in units supplying a firm, or training farmers in good agricultural practices. As researchers, we observed that the inner workings and governance of value chains and their articulation with endogenous development are not automatically favourable for smallholder farmers, workers or small and medium enterprises.

A one-sided focus on standards and compliance with them, captured under upgrading, is one piece of the puzzle, and from a developmental perspective maybe not be the most significant one despite its intentions. In practice, however, value chains have been reduced to a tool. This was a good starting point for a process linking policy, practice and research. Where policy and practice spoke about a value chain ‘approach’, referring to an often implicit intervention theory, research considered value chains to be an interesting phenomenon that requires cross-disciplinary research, referring to a set of social theories that are usually not linked.

Current analytical and interventionist value chain approaches have their limitations. These limitations are in part related to a shared exclusive focus on the logic or functionality of value chains per se. In research this can be seen by a global value chain school, claiming an independent theoretical approach exclusive for value chains rather than embarking on a theoretically plural and integrative approach to a socially embedded and multi-dimensional phenomenon.

In policy and practice this exclusive focus is reflected in a large numbers of organisations claiming to do value chain development, thereby attributing all kinds of development outcomes to a single approach or instrument. The network of researchers we worked with in the DPRN-project, which still exists, contrasted such a one-sided perspective with a variety of theoretical frameworks and with a variety of value chain – context configurations. We tried to shift attention from framing value chains as a development strategy to the conditional nature of pro-development outcomes. We concluded that a one-sided approach to value chains fails to trace the causal processes that produce development outcomes. Explanations and interventions purely derived from the internal logic of a value chain or production network do not suffice.

The separate worlds wherein we operate give rise to conflicting interpretations. Hence, language is vital for making exchanges in a knowledge platform fruitful. But there is more to language than would seem at first sight: each actor uses his/her own frameworks for understanding concrete realities. This easily leads to a situation where umbrella labels, such as value chains, hide the differences in perspectives. We consider the task of a knowledge platform to recognise the distinct frameworks in policy and practice as well as in scholarly work, and to use this for a systematic exchange or confrontation in agenda-setting processes.

Consequently, we consider the function of a knowledge platform to provide a forum for evidence-based discussion on and testing of the theoretical underpinnings of policy, practice and research. This requires integrating theoretical work from different academic domains and triangulating this with the different interventions theories in policy and practice. Hence, a platform will benefit from finding the contrasts. For the testing of theories, a knowledge platform can strongly benefit from anchoring (action) research in the activities of practitioners (e.g. NGOs, businesses or their partnerships) experimenting, piloting or implementing interventions.

The task of research is to complement the (implicit) theoretical assumptions in policy and practice with scholarly theories and with evidence from other cases. A knowledge platform will be able to flourish if the coordinators purposefully compose research teams drawing together different disciplines and combining a variety of empirical evidence. Moreover, we learned that processes aiming at consensus from the start are not able to realise an in-depth discussion on assumptions.

We also concluded that the context dependency of development impact necessitates cross-case investigations. Precisely the interaction between the internal and external dynamics may form an important point of action for desirable development processes, complementary to an exclusive emphasis on development interventions through value chains per se. We tried to make exchanges evidence based, that is to say grounded in well-documented actual experiences and practices, as formulated by government officials, related by NGOs, as told by companies and as reported by researchers. A knowledge platform seems to well equipped to assemble different types of experiences, lessons learned, and (anecdotal) insights that inform a more systematic comparison across contexts.

We identified a precise function for a knowledge platform, namely combining various forms of data collection in a process leading towards a meta-analysis, which is grounded in on-going interventions, actions and initiatives. This will work when a knowledge platform focuses on multi-dimensional processes, which have an internal logic but where the outcomes are strongly context dependent. The results of a knowledge platform then are an in-depth understanding of the internal logic of interventions, a comparative analysis of contextual processes co-shaping development outcomes, a way of coming to grips with the issues of replication and scaling of interventions, and an approach to use theory-testing as a practice in evaluating development interventions.

More information:
Development impacts of value chain interventions – a knowledge agenda