Toxic knowledge

Development Policy21 Nov 2011David Sogge

According to the State Secretary’s letter, the chief reason for introducing his new measures is to improve the quality of policy through better knowledge and its management, including quality control (‘bewaking van de wetenschappelijke kwaliteit’). That’s a very good reason. For in light of the kind of control actually exercised over ideas and knowledge that blanket our policy environment, these measures come not a moment too soon. Except perhaps for intervals under Jan Pronk, Dutch development policies have long rested on an unshakable faith in knowledge made or endorsed in Washington DC – knowledge now shown to be of uneven and often poor quality, indeed sometimes unfit for human consumption.

A number of close observers in the Netherlands, such as Ton Dietz, have called attention to official subservience to the World Bank as a chief source of knowledge. The 2010 report Less pretension, more ambition, by the Dutch Scientific Council for Government Policy (WRR), shows more ambivalence about the knowledge produced by the Bretton Woods Institutions, though it rejects their intellectual monopoly. It cites approvingly the Cambridge economist Ha-Joon Chang’s remark that the World Bank follows “the Henry Ford principle of diversity” — that is, you can have any colour of policy knowledge you want, as long as it is the World Bank’s colour.

But what about the quality of the knowledge produced under auspices of these authorities at the commanding heights of the aid-and-development system? Criticism by independent academics and policy-activist NGOs, much of it based on close-to-the-ground studies, grew from a trickle in the 1970s to a forceful stream by the 1990s. Since 2000, disputes about the validity of those banks’ knowledge intensified. They helped trigger, for example, the dismissals from the World Bank and IMF of some heavyweight economists – Stigltiz, Kanbur, Easterly, Johnson etc. From them have come books, articles and interviews that hammer the Bretton Woods Institutions for the severe shortcomings of the knowledge they produce, and the harm that has followed.

Yet today, from within the institutions themselves, have come devastating critiques.

In 2006, a team of four distinguished mainstream US-based economists, led by Princeton’s Angus Deaton and backed by at least 25 other senior Anglo-Saxon academics, published an assessment commissioned by the World Bank of the Bank’s knowledge output in the period 1998-2005. The team’s report called attention to some good outputs, but concluded that too much of the Bank’s research was “undistinguished”, “misconceived”, “flawed by ‘serious deficiencies in terms of empirical work or conceptual framework’”, and “not remotely reliable”.1

This year saw publication of a follow-up to the Deaton study by a group of seasoned academics based mainly in London2. They examine World Bank and IMF research across fields ranging from water and agriculture to HIV-AIDS and violent conflict, as well as economics. Their conclusions indicate that the Deaton team, though it had delivered a tough verdict, had been far too generous in its assessments of Bank research. A typical chapter in this collection holds that, though the World Bank can’t be portrayed as “the cartoon villain of the piece”, it steers its knowledge work by ideological factors that “often lead to policy-based evidence trumping effort to build a foundation for policymaking from credible evidence.”

In June of this year the IMF’s Independent Research Office issued an assessment of IMF research through 2008 3, revealing the extent to which both insiders and outsiders see IMF research being “message driven”, that is, coloured by ideology. While the assessment found some strong performances, it also found that “a large part of the conclusions and recommendations in WPs [Working Papers] and SIPs [Selected Issues Papers] were not substantiated by the analysis.”

Given what both insider and outsider research tells us about the alarming impacts of World Bank and IMF policy formulas –- impacts felt over the past several decades in Africa, the former Soviet Union, the Middle East and Latin America, but now also being felt in Europe and North America –- there are good reasons to show extreme caution about those institutions’ knowledge output. Much of that output is toxic. It has contaminated our policy environment, just as old industries have polluted our soil, water and air. In respect to ‘knowledge for policy’, and ‘policy for knowledge’, there are urgent needs for decontamination and for safeguards against further pollution.


  1. Angus Deaton, Nora Lustig, Abhijit Banerjee and Kenneth Rogoff, 2006, An Evaluation of World Bank Research, 1998 – 2005. Washington DC: World Bank.
  2. Kate Bayliss and others (editors) 2011, The Political Economy of Development, London: Pluto, in association with the International Initiative for Promoting Political Economy (IIPPE).
  3. Independent Evaluation Office of the IMF 2011, Research at the IMF : relevance and utilization, Washington, D.C.: International Monetary Fund,