Thursday, August 12, 2010

Mapping the World Bank's Footprint

The World Bank, in collaboration with AidData and the Development Gateway, have just announced the release of a service that allows you to use Google Earth to see exactly where Bank projects are being implemented around the world.  In seven weeks, the team (which included members from Brigham Young University, William and Mary, and Georgetown) geocoded over 12,000 project locations.  As they explain, this information was often available, but it was buried deep in hundreds of different reports.  In practice, this meant that it was unusable for the vast majority of people.

This is a big step forward.  As I have noted before, information availability per se is not enough.  The information must be available in a format that is easily usable, and on this front, the team working on this seems to have made much progress.  What encourages me about this effort is also the inter-sectoral nature of the collaboration.  It included not only World Bank experts, but also academicians and an NGO.  It had seasoned aid veterans (Jean-Louis Sarbib used to be a VP at the World Bank) working together professors and students.  Each of them brings to bear perspectives and expertise that the others don't have.  That is the wave of the future for the development field as a whole.

I was also encouraged by the talk of empowering communities themselves to take action.  Information and transparency initiatives only succeed if they enable or require people to act or change their behaviors in a way that will make a real difference.  A number of other aid transparency initiatives are underway, and as I watched the video at the link, several important drivers of success came to mind.  Initiatives that make a difference will:

1) Be as real-time as possible.  The usefulness of the data falls off sharply as a function of age.

2) Allow and even encourage two-way flows of information.  Users have to be able to comment on the information in a way that everyone can see (and act on - see point 4).

3) Have strong incentives for updating.  As I have confessed before, I have run or been involved in a couple of massive data exercises that collapsed because after the initial push there was no ongoing incentive for people to refresh the data.  This aspect is very hard, but critical.

4) Have a clear theory of how the information will affect behavior, and that theory needs to be reflected in what data is collected and how it is presented (for an obvious but compelling example, see here.)

5) Be as simple as possible. Present only the information needed to elicit behavior change.   There is always the temptation to have more bells and whistles, but if there is one thing I have learned at GlobalGiving it is that complexity and clutter kill.  Simplify, then check to see if people are using the data, and if not, then change the data or the presentation.