05 October 2010

How World Bank Research Really Needs to Change

It would be churlish not to welcome the World Bank President Bob Zoellick's latest speech on Democratizing Development Economics. But I wonder whether anything significant will change in terms of how World Bank research is done.

His key points:
  • a multipolar global economy requires multipolar knowledge: the flow of knowledge is no longer North to South, West to East, rich to poor
  • development economics is too methods driven, not sufficiently issue driven
  • issues are complex, but current research is too narrowly focused and weak on external validity (does it have meaning and scalability outside of its context?)
  • research needs to open its doors to those with hands on experience
  • we may learn more from economic history than we do from economic models
  • political economy must come back into the analytical frame
He outlines 3 main implications for the development research agenda:
  1. the role of business: there have been few serious evaluations on what works to promote industry and why--they are needed
  2. how can the results agenda build in local ownership and participation?
  3. we need to understand human risks better with more research at the intersection of security, governance and development
I agree with much of this.

What are the implications for the World Bank's research? This is where is gets disappointing: a series of rather modest initiatives to share data, make research user friendly, get outputs wholesaled and networked and opening up to non-elite, practitioner knowledge.

We have been doing this for years at IDS as have many outside of the Bank.

My unsolicited recommendations to Mr. Zoellick to meet his worthy goals?
  1. decentralise more of your research staff so they can fully understand the politics and complexity that they operate in
  2. recruit more non-economists - you talk about needing to reinvent development economics, but you need to reinvent research at the Bank
  3. recruit your research staff, whatever their disciplinary perspective, from a wider range of higher education institutions (they are dominated by US universities)
  4. proactively seek to understand and learn from the impact your research has or has not had on poverty and inequality


Tobias Denskus said...

I agree with your observations; I also wrote a short piece on my blog on a key element of his speech: 'Whay publishing aid data does not equal "democratizing development"'


April said...

Great points.
Why not contract out some of Bank-funded research? this would help diversify the researchers (and perspectives) involved.

lawrence haddad said...

aidnography, thanks, I will read with interest...I certainly think having the aid data is only a first step, but an important one

april, i can see the reason for having a large research team at the Bank (critical mass is important and internal location promotes learning) but it has to be globally constructed knowledge..,

Yamini Atmavilas said...

Sorry, am just reading this, but thanks for that comment, esp. the part about bringing on board research and methods that are not based on crude economics and largely econometric "stuff." The stranglehold of econometric work on development research, and of randomised experimental methods on development evaluations is ... well, a stranglehold that's got to go. It's frustrating to read yet another paper when an economist discovers culture or social norms - and that work is held up as an epiphany in the field and immediately generates policy debates when researchers from other social sciences, qualitative researchers, and those from the "South" have long made the same argument.