Post 1: Mending the Broken Feedback Loop Case Study 1, Bangladesh

According to American economist and New York University Professor of Economics, William Easterly, “Lack of feedback is one of the most critical flaws in existing [foreign] aid.” (Easterly, 2006). In his paper, “Planners versus Searchers in Foreign Aid,” Easterly breaks down the mindsets of workers in the international aid community into two types: the Planner and the Searcher.

The Planner constructs frameworks and makes strategies to meet fixed, pre-determined goals. Often, these goals are lofty and infeasible, such as “ending world poverty.” Resources are allocated in a top-to-bottom system, where the “top” (higher level administration) anticipates what needs the “bottom” (the world’s poor) will have; however, when difficulties arise in meeting the pre-determined goals, Planners say that the problem is greater than anticipated, rather than considering that the strategy is ineffective. This resists constructive adaptation to the plan, and perpetuates ineffective aid efforts. Planners are popular in the international aid community because big donor countries prefer well-articulated plans for how their money will be used over elastic predictions that remain open to future evolution.

The Searcher builds systems that aim to address the demands of the “bottom,” using a trial-and-error method that adapts to local conditions. Accountability and feedback are crucial components to the Searcher’s efforts, as programs that deliver poor results are held accountable for their ineffectiveness. Feedback is taken from the recipients of aid to help refine the process. The focus of the Searcher is on what works, not what should work.

The goal of my research is to find techniques that would appeal to Easterly’s archetypal Searcher. I aim to find ways to help fix the so-called “Broken Feedback Loop” of international aid, a phenomenon where the aid organization does not receive information from the bottom when the aid is ineffective. The techniques I suggest should reasonably enable an aid organization to receive feedback from the intended beneficiaries of its work, enabling the organization to change and adapt its approach to a given project. This is what I call taking in information from the “ground-up.”

To provide context to the techniques proposed in my research, I use World Bank projects that suffer in some way from broken feedback loops as examples. I describe the projects, their methods, and their results. I then propose possible solutions to the feedback issues that could reasonably solve or help alleviate the severity of the problems.

For my first post, I will examine a 2005 World Bank effort, the Bangladesh Integrated Nutrition Project (BINP). The BINP was a project initiated to combat “the very high level of malnutrition prevalent in [Bangladesh],” a problem that affected nearly one in two Bangladeshi children between 2000 and 2004. Malnutrition has serious health effects, including stunting of height and hindered brain development, and frequently begins with poor prenatal nutrition by impoverished pregnant mothers.

The BINP had three components:

1) Institution building at national level (i.e. build, fund, and teach national organizations to combat nutrition on their own)

2) Community-based nutrition activities (i.e. educate communities, households, and individuals about how to eat better through counseling)

3) Inter-sectoral nutrition activities (i.e. improve access to food)

Largely, the focus of the BINP was on education and counseling. From its inception, the goal of the project was to tackle what was perceived as a previously unaddressed issue (nutritional education), with less attention being given to food-related issues like food insecurity or disease. Counselors were exclusively women who had children and had at least 8 years of education. Activities undertaken as part of the BINP included weighing of babies, provision of food supplements,  and presentation of educational programs.

The BINP was largely successful in its education campaign, achieving an 88% participation rate in regular baby weighing and a high level of acceptance by local women despite geographical difficulties and adverse cultural norms.

However, this high participation rate did not translate into a high degree of impact. For component 1, the BINP had little to no involvement with other large, relevant Bangladeshi institutions and thus did little to help them; on top of this, materials for a national campaign were never generated. For component 2, though 94% of women said they learned what they should do from the BINP, only 53% said they actually put that into practice. In component 3, 4 out of the project’s 5 desired efforts (such as reduced child malnutrition and reduction of low birth weights) were deemed to have had no effect at all on project areas versus control areas. Overall, the project received a rating of “moderately unsatisfactory” from the World Bank’s internal review board, the IEG, and ran well over its budget and timeline.

Below, I’ve identified what I see as significant feedback-related problems with the BINP, along with my recommendations on how to fix them:

1) When donor countries voiced concern over the effectiveness of the project’s proposed plan, no changes were made to the design but rather its timeline was extended (to “phase in” the various components over time) and the proposed budget was expanded.

Instead, the BINP should have taken in those complaints and altered the plan, or at least maintained a degree of flexibility in the plan’s implementation that would allow it to adapt if the desired results were not achieved. This could have been done by increasing the number of checkpoints in its timeline when the project would be assessed and adjusted as needed.

2) Rather than contact national organizations who had been in the country longer and who could have articulated needs that could be met by the World Bank, the BINP chose to stick to its pre-decided plan of action and thus got little involvement from other organizations.

The BINP mistook having a large size and budget for having actual expertise. In reality, organizations like the Bangladesh Rural Advancement Committee (BRAC) had been combatting malnutrition for years and could have been a very valuable resource, had the BINP been willing to adjust its objectives and join forces. Inter-organizational outreach and meetings should have been a top priority for the BINP’s component 1 staff.

3) Little survey work was done to learn what the local women identified as challenges to proper nutrition. Instead, the BINP opted to set its own challenges and work to battle them.

This was the biggest issue with the whole project. Even if the previous two problems were left unaddressed, a much more powerful impact could have been achieved if feedback from local mothers had been taken in. One of the reasons that impoverished Bangladeshi had such a difficult time implementing what they were taught is because they are mainly agricultural workers. Though the educational programs taught them to avoid hard labor, their everyday routines involved hard labor, making it unavoidable in practice. Also, though the women cared very much for their children, having large, nourished babies meant a much more painful and dangerous childbirth for them; as such, a culture existed where women ate less during pregnancy on purpose for their own safety, as the risks of malnutrition were considered to be much less dangerous than those of a difficult childbirth. The BINP counselors should have surveyed the women, stayed in their homes, and learned what its like for them on an everyday basis before coming up with educational programs.

4) Food supplements were given to the women and children to increase their weight.

This is a short-sighted solution to the problem. While the food supplement program was one of the few positive results of the BINP (most children who received food supplements gained weight), it only worked as long as food was being pumped into Bangladesh. Instead, much more work should have been put into growing backyard gardens and making small ponds to farm fish, which would supply more food over a longer time period.

Overall, the BINP suffered from a poor feedback system that did not consider the needs of the local women involved. Instead, inflexible plans were constructed that did little to combat the very real problem of malnutrition in Bangladesh. Researching this case has taught me the importance of learning the sociocultural customs and realities of any region where aid work is attempted, as well as the importance of taking in feedback from the ground-level of any project.

“Bangladesh Integrated Nutrition Project.” Independent Evaluations Group. World Bank, 13 June 2005. Web. 13 Aug. 2013. <>.

Easterly, William. “Planners versus Searchers in Foreign Aid.” Asian Development Review 23.1 (2006): 1-35. Web. 13 Aug. 2013. <>.


As I continued my research, I found an interesting, low-cost method of identifying premature and/or low birthweight (LBW) births. In the BINP, the counselors measured the developing health of the pregnant mother’s child by regularly weighing the pregnant mother during her pregnancy. This is a rather flawed method, as a mother would obviously have gained weight at each weigh-in as her child gets bigger, not necessarily indicating that the mother was eating enough or that the growth was on track.

An alternative method is pH strips, a technique developed by the Saling Institute and utilized by a team of Stanford University students who combatted premature and LBW births in Bihar, India. The students provided pregnant mothers with pH strips to perform vaginal self-tests. Twice a week, the mothers would don gloves and retrieve their own vaginal milieu through digital insertion, which would then be applied to the pH strips. pH strips change color based on the acidity of the applied liquid, and can be compared to a chart to find how acidic or basic the liquid is. If the mothers’ pH strips were at or above 4.7, then they most likely had bacterial vaginosis, possibly combined with another genital infection. These conditions often show no other symptoms and are early indicators of a premature/LBW birth.

This method would be optimal for use in Bangladesh, where literacy is low and prenatal malnutrition is prevalent. The method is cheap, easy, non-invasive, and would be a much better method of determining which mothers had a high risk of having an unhealthy baby than intermittent weigh-ins. This seems like an easy way to get accurate, usable info to use when tailoring where and who to direct immediate aid to.

Here’s the link to the Stanford students’ report:


  1. jlodenthal says:

    I think one of our biggest issues with Foreign Aid is that we don’t understand the people we’re trying to help. The women who starve themselves to make their pregnancy easier are a prime example. We can educate all we want, but to get results we have to understand the culture. Do you know if anything has been done more recently to deal with the malnutrition in Bangladesh and fix the issues with the initial attempt? Or was it just written off as unsuccessful?? Very interesting stuff!

  2. I think that foreign aid is a really important thing to examine critically, which is why I’m excited that you are doing this project. I don’t mean to be cynical in any way, only skeptical, but I believe that a large reason why foreign aid is so flawed and does not work, is that it is not designed to. When thinking in terms of dependency theory or world systems theory, you see that the amount of money that is paid into the core or “1st world” nations through debt repayments to the IMF or World Bank is oftentimes greater than the amount of money which goes out to periphery nations as foreign aid. I think it is really important to examine aid in this context. We need to question whether the aid we do give out is serving the purpose of helping people, or is catering to other special interests, like trans-national corporations. What places get aid? What places don’t? Why? Who is most benefited by the aid? Who isn’t?


    I agree 100%. To be truly combat the complex problems that plague the underdeveloped parts of the world, we have to learn about the people and what their needs are.

    I don’t know of any follow-up projects from the World Bank that deal with malnutrition in Bangladesh, but the country does possess a number of national organizations that are attempting to deal with the problem in a culture-specific manner. The project that I reviewed in my post was actually a pilot program by the WB that was meant to be expanded into a full-fledged and funded program in the future, but due to its obvious flaws and lack of results the pilot was ended. As far as I’ve found, no new program has been developed. Thanks for the comment!

  4. MWABEL:

    Development aid is definitely a tricky issue, as it’s a relatively new phenomenon occurring in a very new kind of global environment, and is still in the proving ground stages of “Does this work?”. Until recently, a lot of the data that is necessary for determining the real impact of aid money has been kept from the public view; however, some organizations, like the World Bank, have responded to demands for more transparency in the “behind-the-scenes” of foreign aid by publishing reports, fact sheets, and other material, as well as creating new internal departments that assess the effectiveness of their work (e.g. the Independent Evaluation Group). Hopefully, these new sources of data will help us to determine how effective our money really is.

    To address your concerns about IMF and World Bank loans, the two institutions have different purposes for their money. I’m certainly no expert, but I believe the IMF adds interest to its loans for the same reason a small, local bank would: to be financially self-sustaining. Member nations apply for different loans that each have different payment options depending on their economic situation. The IMF is not a development bank, so it’s job isn’t to give out money for free, but rather to act as a lender of last resort for nations that are in desperate need of credit to pay back money to other nations they borrowed from in the past. As such, the IMF does collect more money than it lends out, but that’s to keep the bank up and running and not to exploit the loan recipients. Again, I’m no expert, but that’s my understanding of its role.

    The World Bank on the other hand, IS a development bank. Its loans vary even more than the IMF’s do because it does not demand full repayment of all of its loans, especially from the world’s poorest countries (the IDA gives out no-interest loans to the poorest countries and does not punish those who do not pay it back). The World Bank relies on donations from donor countries for its livelihood, not its loan interests, so there is less pressure for the World Bank to make money on its loans. In fiscal years that it does make a profit on its loans, the money is not pocketed by bureaucrats, but rather used to either provide debt relief to heavily indebted poor countries, add to financial reserves (i.e. put in the vault to enable more loans to be made), or respond to unforeseen humanitarian crises.

    I’ve included a couple links in case you’re interested to learn more. Thanks for your comment!

    IMF Lending Factsheet:

    World Bank Lending FAQ:,,contentMDK:20147466~menuPK:344189~pagePK:98400~piPK:98424~theSitePK:95474,00.html

  5. ijdocampo says:

    This is so interesting; I’m an intern for AidData, where I code a lot of World Bank projects, and I do wonder how many of the development projects I’ve worked on are actually successful and sustained. Have you read “Poor Economics?” It’s a really interesting discussion about the pitfalls of aid and the barriers to sustainable development.

  6. I have not read that book, but it sounds like it’s right up my alley! Thanks for the comment and book recommendation!