The long walk to evidence-informed development: arriving at irreversible policy changes

By Laurenz Langer


In last month’s blog we presented a theory of change for evidence-informed development in an effort to explain how we think UJ-BCURE as a capacity-building programme for policy decision-makers is linked to the bigger objective of improving the livelihoods and capabilities of people living in poverty. The theory of change highlighted that capacity-building is but one input required to establish an effective evidence-to-policy ecosystem. We claimed that evidence-informed development would come about as a result of an institutionalised culture of evidence use practiced by all actors within the development system. In this blog, we will follow-up on the possible implications such an institutionalisation of evidence use as an approach to designing development policies and programmes might carry.

The evidence-informed development movement emerged in response to the perceived faddisms in international development. The same ideas and policies seem to fall in and out of favour within the development community with little learning taking place. Dropping laptops from helicopters in rural areas so that children can learn by themselves didn’t lead to a learning revolution; well then, how about an app to revolutionise reading?; no, that didn’t really work either?; how about a solar classroom in a box? It was hoped that the careful evaluation of development policies and programmes would lead to a greater understanding of which well-intended initiatives worked, and why. What was shown to work would then be scaled-up and ineffective programmes would be ceased. While elegant in theory, there is no reason to believe that the use of development research itself is immune to faddism.

We would want to caution against a sentiment of: faddism in development policy is dead—long live evidence faddism! As single development policies can fall in and out of fashion in reoccurring circles, so can individual pieces of development research. Influential research studies and their findings might be en vogue for some time before being refuted by new evidence leading to more research to refute the refute and so on…The current debate on the evidence-base behind the scale-up of deworming interventions illustrates this dilemma. As a result, we believe that the current knowledge frontier in evidence-informed development is no longer on pointing out the shortcomings of a development system that doesn’t learn—the need for learning has manifested sufficiently. Rather, the new frontier is about identifying how best to learn, and what a learning development system might look like in practice.

Our paper presents the institutionalisation of evidence use as well as the development of evidence literacies (next blog) across all levels of development as two building blocks of a learning development system. Thinking deeper about the institutionalisation of evidence use, it seems crucial not to institutionalise evidence faddism. To do so, it might be of virtue to frame ‘evidence’ in evidence-informed development as a collective body of knowledge accumulated over time, rather than individual evidence products, be this impact evaluations, systematic reviews, or in-depth case studies. In this understanding, the institutionalisation of evidence use translates into the institutionalising of a culture of learning. Development policymakers and practitioners would be open to have their decisions informed and challenged by evidence. Evidence of effectiveness would not serve as a carte blanche, but instead be constantly re-evaluated, contextualised, and synthesised in order to foster the accumulation of a collective body of knowledge about what works, why and how. This process works in small steps and feedback loops to nurture iteration and experimentation and eventually translates into a slow but constant collective building of knowledge.

The science of building collective knowledge to base development programmes and policies on eschews easy solutions, be this one-stop policy blueprints, magic bullet interventions, or break-through research evidence. As Howard White puts is, ‘there seem to be few big bangs in social policy!’. However, a rigorous and contextualised body of collective knowledge might allow for a more fundamental alteration of the development system: the advent of irreversible policy changes. Irreversible policy changes, as the name suggests, refer to the irreversible scale-up or termination of a policy or programme. Take an example from the medical field: doctors will not use arteriotomies again to treat flue patients, and likewise few doctors will stop seeing the virtue of disinfecting surgery equipment. The same does not seem to hold true in the development system currently. Despite evidence pointing strongly towards its lack of impact, micro-credit continues to enjoy vast support in parts of the development community. Likewise, large infrastructure development projects such as building dams and factory plants as implemented in the 60s and 70s were highly criticized and fell out of favour in the following decades. Recently though, these type of projects seem to be en vogue again with South Africa building the world’s biggest dry-cooled coal power station financed by the World Bank.

We believe that a rigorous and contextualised body of collective knowledge might carry the authority to allow for the advent of irreversible policy changes in international development. Vested in a culture of learning, the domain might become more acquainted with the transparent communication of successes as well as failures. The latter has for long remained a taboo. To avoid faddism in development policy, practice, and research, a jointly accumulated and shared body of knowledge embedded into an institutionalised culture of evidence use might be the most effective safeguard. Development and social change is an incremental process. Sometimes, it seems, the best we can do to support this process is to design mechanisms to ensure we don’t repeat our own mistakes—building a collective body of knowledge to nurture consensus around irreversible policy changes might be one of them.