Two ideas that simplify: food for thought

A few days ago, my colleague Gerry Johnson brought to our attention this presentation by Owen Barder.  Watch it.  It’s provocative, persuasive and pregnant with ideas. It seems like it chats with Esther Duflo’s Ted Talk that Francisco Mejía shared a few days ago.

It weaves very nicely the day-to-day practical issues that we face in impact evaluation.

It also challenges, implicitly, some of the assumptions that we take for granted.

Let me point out two of the ideas that should be food for thought.

First, the critical need to “unbundle” intractable and complex problems (poverty, inequality, public safety) into more manageable challenges that can be addressed with the available development toolkits (what we, at the IDB call making them “bankable”).

Opposed to this, we often fall into what Professor Karl Weick calls “comprehensivitis.”  That is, addressing complex issues by tackling all of its roots in a comprehensive manner, which leads to unmanageable projects (what we like to call “development Christmas trees”).

It seems more attractive to act on more specific and contained issues, targeting limited results that we can control and measure empirically. It is here where impact evaluations can bring a lot to the table.

By comparing the cost effectiveness of alternative interventions, we introduce actionable information and affect resource allocation. What Weick calls “small wins.”

The second idea relates to the need to constrain and multiply our evaluation and feedback loops, if we want, like Barder, to accelerate our learning.

Not only do we need to increase the number of evaluations that we undertake, but they have to be more specific, be done closer to project design and execution, avoiding large and expensive “comprehensive” efforts.

Besides, the ubiquity of web technologies, such as social networks and cell phones, should allow us to accelerate this process. It would be good to get ideas and make them work.  As Jorge Orlando Melo, a prominent Colombian historian recently wrote: “what we need is many limited answers.”