23 March 2012

DRI Debates Quick Hitters

Yesterday, the Development Research Institute at NYU hosted a one day conference on aid and development debates.  It was quite good. Unfortunately, there was not enough time to adequately get into the debates and answer the audience questions. That was due to a constraint on time and the inclination of some audience members to grandstand when given the microphone. Further yet, the topic of the MVP was discussed in the first panel which brought out a cohort of Columbia students to ask hard questions to the panelists.

Here are some quick observations about what was discussed. I hope to provide more soon, but for the sake of time I am going with this format. Forgive me for jumping around.
  • Bill Easterly of DRI; Andrew Rugasira, Founder and Chairman, Good African Coffee, Uganda; and Angus Deaton, Princeton University and Woodrow Wilson School attached themselves to the idea of trial and error. Deaton used the example of Angry Birds saying that it would take too much time to run an RCT on ever possible angle and action of the user to determine the optimal outcome. Even children will try it out to see what works and then make changes that are necessary to succeed.  Abhijit Banerjee of MIT refuted the idea, but did not go too far into why. To me, this idea falls apart when taken into the context of present aid and development. Deaton held trial and error in high regard because it is a matter of learning from mistakes and improving until a solution is found. In an ideal world that would be acceptable, but given the inability of governments and NGOs to openly admit failure it becomes hard. Further, what seems to be an adequate solution at a given time can be problematic later down the line. New solutions will need to be developed to deal with the emerging problems and the process will begin again. To some extent, this is hard to avoid, but strong and open monitoring and evaluation practices can ensure to the trial and error of the Easterly 'searcher.'
  • Michael Clemens pitted 'goals' verses 'evaluations.' The MVP is an example where the two are in conflict to a degree, but that does not mean it is the case for all of development. Much like working with trial and error, goals and evaluations can and should work together. Goals are useful in setting benchmarks, but they are not absolute. They are also not an end, neither are evaluations. Rather, evaluations can act as a mechanism to optimize the goals themselves and the way that the set goals will be attained.
  • The do something better argument is weak. It is used often in the MVP debate, but in many other places. As it goes, defenders say that they are doing something, would love to hear new ideas, but dismiss criticisms because they do not offer a new plan. In the case of Clemens's concerns, it is that the evaluations are not telling the whole truth. There is a lot to be learned from the MVP, let's make sure that the evaluations tell what trends are taking place around a target village to get an understanding as to the changes going on in places that are not receiving the intervention. An RCT is a possible way to help test it, but there are many other methods to track these changes that will do a reasonably good job. The overall point is to learn from the MVPs so that parts can be improved when new villages are established. Finally, given the cost, it is important to see how much they are actually accomplishing. Clemens was criticizing for pointing out the cost of the new MVP in Ghana. His mistake was not to be more explicit. The cost per household exceeds by a significant amount the household makes every year. It is a legitimate question to ask if the MVP will improve the lives of the households more than if the money was simply given to each household over the same period. Fortunately, Give Directly is testing this very idea with an RCT in place. Yea, it won't prove whether the intervention is the answer but it will help learn more about what works.
  • Andrew Rugasira had a few negative thoughts about Kony 2012 that received applause from the audience. Evidently the crowd was not a fan. It is only anecdotal and my basis is on how loud people were applauding, but it is still worth a mention.
  • Stewart Paperin of the Open Society Foundations was enjoyable for his candor. Two things stood out from his remarks. First, he sees the MVP as an investment. He did not explain it much, but he mentioned it as an investment at least a handful of times. Second, he dismissed the need for knowing causality and followed it up by saying that he wants to fund things that work. It seems off that he cares so much about things working, but does not care to know why or even measure if they are in fact working. This is where his defense of the MVP was the weakest. The strength was in the fact that it is relatively cheap for the OSF and is an experiment worth testing. In his mind, there was only upside from the project. Given the amount of money in each village he is likely right that the interventions will do some amount of good.
  • Rugasira gave the argument for trade over aid. What was interesting was he attacked the theories set forth by Sachs, but disguised to some extent by taking about entrepreneurship. This seems to have resonated with the audience as there was little push-back as compared to that seen in the other panels. Easterly gave a taste test to the coffee and shared his approval. After the talk, Yaw Nyarko of DRI said that we need to have Andrew2012, an appropriate observation given the lack of stories about innovation and entrepreneurship from countries like Uganda. It is also valuable as he does not fit the mold of the microfinance success story. The single-story about entrepreneurship in the developing world has the face of a woman participating in a group-liability loan. While a truth, there are many women and men who are starting businesses outside of the traditional microfinance world. Rugasira is an excellent example.
If you went, add some of your thoughts in the comments section. Also, feel free to ask questions. Much more was covered. Andrew Rugasira's talk was phenomenal and will be much better seen on video that summarized here.