Sunday, April 01, 2012

Clemens' False Dichotomy about Debates in Development

Debates on Development
A couple weeks ago I had the opportunity to attend  NYU's conference “Debates in Development.”
The agenda included Yaw Nyarko and William Easterly from NYU's Development Research Institute, Stewart Paperin, Open Society Foundations, Bernadette Wanjala, Tilburg University Development Research Institute; Andrew Rugasira, Founder and Chairman, Good African Coffee; Abhijit Banerjee, MIT's Poverty Lab, and others. 

In general, with the exception of a handful of speakers, I found the conference to filled with silly generalizations about development work and a myopic need to compartmentalize development into artificial 'conflicting sides.' This was epitomized when CGD's Michael Clemens presented an artificial characterization of the debate  as a “Goal Movement” vs. “Evaluation Movement." 
Goals vs Evaluation MovementsFirst of all, his criticism of integrated development as something “we've tried and failed” is misdated and does not meet his own rigorous evaluation standards. Monitoring and evaluation techniques have evolved tremendously and are now an essential piece of development work, both at macro and micro levels. Therefore, suggesting that the misnomer “Goal Movement” is devoid of solid evaluation does not correspond with the realities of today’s development interventions.

One of the hallmark criticisms of Michael Clemens to the Millennium Village Project (MVP) is that Integrated Rural Development Projects (IRDP) have already been attempted and deemed a complete failure by a number of evaluations conducted by USAID, World Bank and others. However, Mr. Clemens fails to recognize that if he were to apply his evaluation standards to these reviews of IRDP, few if any will qualify. Most of these evaluations were conducted using ex-post assessments of macro-economic indicators. None used RCTs and few carried out the complex quantitative methodologies that he proposes to guarantee causality.

Fortunately, development M&E techniques have gained a tremendous amount of importance among academia and practitioners in recent years. Under pressure from both donors needing demonstrable impact for their constituencies, and practitioners needing more effective feedback mechanisms to understand program outcomes and impacts, today most development projects allocate a significant amount of program resources -10 to20%- to M&E. Also at their disposal, practitioners have a well-stocked toolbox of quantitative and qualitative methodologies to monitor and measure development outcomes. As argued by Dr. Woolcock and other, a ‘mixed method approach’ using qualitative and quantitative methods simultaneously, is the best way to understand project impact. 

Hence, suggesting the differences between the “Goal Movement” and the “Evaluation Movement” are 'irreconcilable' is simply baseless. Most bilateral and multilateral development interventions and even the initiatives by small NGOs and civil organizations considered evaluation essential to their programs’ successful implementation. Therefore, Clemens’ artificial grouping of the development debate is a false dichotomy: simply put the ‘evaluation side’ will be an orphan without having something to evaluate.

Moreover, the development community, or as he call it the ‘Goal Movement’ has made tremendous advances in strengthening institutions and policy, and developing tools and technologies since the 70s and 80s when IRDPs were commonplace (See other post about promising trends in Africa). These recent developments have enabled practitioners to fine-tune interventions by getting more accurate data and by addressing contextual issues (institutional support, local government etc.) that often impede the success of otherwise adequate interventions. Thanks to that work, today there is a robust body of evidence from development and other disciplines demonstrating the strong synergies created between economic, health, nutrition, and agriculture.

Finally, the issues Mr. Clemens list as the characteristics of the ‘Evaluation movement,’ namely; testing questions, failure is required, local outcomes, and clear accountability, are all part of the ‘Goal Movement.’ The early history of development is filled with examples in which development interventions have failed spectacularly and where local outcomes have been at the front and center of the projects.

Evidently, I was very disappointed with Mr. Clemens presentation. The language he used – calling MV’s Development Harvest 'a masterpiece of misrepresentation,' and his insistence on maintaining false dichotomies add little value to development debates. I'm not completely sold on the MV model. However, I truly believe that as African nations continue to experience economic growth and improvements in governance, they will need models on which to base interventions to attend the needs of their long-neglected rural communities. MV can play an important role in showcasing governments how to do this - the same way cooperation among nations and regions within countries have done for centuries. 

But instead of offering doable alternatives on how to tackle poverty, Mr. Clemens' contributions distract the public with hyperbole that can damage the credibility of very important and proven interventions that should instead be scaled up.