Development practitioners bring very different lenses to our work. Some of us are pre-occupied with ‘best practices’ approaches (‘what’ should be done); others prefer to focus on the influence of political and institutional realities (‘why’ things work out the way they do); and yet others emphasize the importance of giving priority to getting the learning process right (‘how’ to engage). Two recent posts on the World Bank’s “Future Development” blog — one by Shanta Devarajan, and the other by me — are, I hope, useful in clarifying some of the inter-relationships between these approaches. Here are links to the two posts:

“Its not the how, it’s the why” by Shanta Devarajan

And here is my post in full (best read in conjunction with Shanta’s):

“Multiple pathways — How ‘why’ matters” by Brian Levy

For convenience, I am also copying the full text of my post below:

*****

Once upon a time, development seemed straightforward. Sound technical analysis identified what to do– and the rest followed. But experience has taught us that it is harder than that. As Shanta’s recent post signals, there are three competing camps – the ‘whats’, the ‘hows’ and the ‘whys’. I wonder, though, whether in clarifying the differences, we might be missing the chance to learn across these different perspectives?

Certainly, the differences are large. At one end are the old-time-religion ‘whats’, who confidently prescribe ‘best practices’ to help countries stay on the right path – and who sometimes turn to the ‘whys’ to identify  the political and institutional blockages to good policies.  At the other end, the ‘hows’  argue that every country is unique, that the crucial knowledge for shaping and implementing policy is local, and tend to be dismissive of  efforts (especially by outsiders) to analyze political and institutional obstacles.

My new book, Working with the Grain tries to steer a middle ground.  The book explores a small number of alternative development pathways that are very different from  each other – with each characterized by a distinctive set of political and institutional incentives and constraints, and thus distinctive options for policymaking and implementation.

The first, more familiar, pathway is the ‘long route’ of the accountability triangle introduced in the 2004 World Development Report, which is organized into two links – ‘voice’ which links citizens to politicians, and thence to policymakers, and a ‘compact’ that links policymakers and service providers. Politicians and policymakers have to structure the ‘compact’ relationship so that it provides incentives for providers such as schoolteachers and doctors (their ‘agents’) to deliver services. In turn, citizens (as principals) need to exercise ‘voice’ over their agents (politicians and policymakers), through the ballot box and other means. Failures of ‘voice’ emerge as the underlying political constraint to achieving development results. So, as per Shanta’s post,  the key interventions are to strengthen voice. [Note, though, that the principal-agent pathway can also work if the principal is a developmentally-oriented authoritarian leader, as was the case in, say, South Korea from the early 1960s to mid-1980s.].

In many countries, however, the platform for a nested set of principal-agent relationships to function effectively is absent. And historical experience tells us that  it takes decades for a well-functioning long route to take root.  Instead, interactions are personalized, not anchored in rules that apply to all.  The discretionary conferral and  withdrawal of favors is the glue which holds things together. Sometimes the outcome  can be wholly predatory, driven by agreements to share the spoils among insiders. But at other times,  personalized relationships can provide just enough stability for economic development to move forward, and democratic institutions to strengthen.

This brings us to the contrasting, second way of understanding a country’s political and institutional patterns – through the lens of ‘collective action’ among principals (to use Nobel-prize-winner Elinor Ostrom’s term). In  contested governance settings, these can permit “islands of effectiveness” to thrive, even where the broader governance environment is difficult. For this to happen, two conditions must be met. The first is the familiar, ubiquitous challenge of facilitating cooperation among participants to achieve joint benefits, in a way that limits the classic free rider and other moral hazard problems. The second, explored in depth in Working with the Grain, is the ability to successfully trump predation.

All collective action initiatives have multiple interested stakeholders. Some are protagonists of the collective endeavor; others are predators who seek to capture for their own private purposes what the protagonists are seeking to build. As I explored jointly with Michael Walton in a 2013 paper, the dynamics between predator and protagonistthey can play out in the many layers between top-levels of policymaking and the service provision front-line where rule-setting processes are likely to be contested, trade-offs between competing goals left unresolved, and agreements subject to weaknesses in both monitoring and sanctions.

At each of the ‘intermediate’ levels where outcomes are indeterminate, there is the risk of public sector dysfunction – but also an opportunity for the emergence of an “island of effectiveness” organized around strong partnerships among developmentally-oriented stakeholders. Examples include: schools that produce remarkable results in unlikely settings; Kenya’s tea sector; and (in an historical example of striking relevance for contemporary nascent democracies) activist officials in the United States Forest Service and other public agencies who, during that country’s Progressive era of the 1880s-1920s, went beyond their formal mandates to forge coalitions with reformist allies outside government — and, in the process, helped foster a cumulative transformation of America’s public sector.

My intent in highlighting the relevance of principal-to-principal interactions is not to argue that the principal-agent perspective is wrong. In some country and sectoral settings, a principal-agent perspective can serve us well in clarifying the development pathway, and options for policy and institutional reform. But in others, we would do better to alter our angle of vision, and consider the pathway and reform constraints and options through a principal-to-principal lens. (And, sometimes, both lenses can be helpful.)

All of which brings us back to potential synergies between the ‘whys’, the ‘hows’, and the ‘whats’. A central  goal of Working with the Grain is to explore the relative effectiveness across different settings (the ‘why’) of different approaches to fostering change (that is, of alternative whats’). The aim is not to  prescribe some mechanical  formula, but rather to provide a guide for helping a  country identify which of a broad array of alternative interventions are most relevant as points of departure for subsequent learning – that is, for the exploration of ‘how’.