The Crazy Beer Funnel in International Development

A recent World Bank Article begins:

Many community driven development (CDD) projects have been found to be only mildly pro-poor in their funding allocations.

That paper (The Regressive Demands of Demand-Driven Development ) goes on to examine one explanation – that people must first ask for money in order to get it. And in Tanzania at least, the asking process is complicated and poorly publicized so that most of those who need it don’t receive it. Instead, those with the skills in writing proper proposals and navigating bureaucracy get it.

When I was in Peace Corps The Gambia, circa 2000, one village (Njau, North Bank CRD Region) got all the aid money because the Chief sent his son to school to become an expert proposal writer. And so the only paved stretch of the 100 km road was the 15 km right around Njau. They also had a bridge built for themselves and started a sort of revolving fund for women with government grants. All of this came as a result of knowing the rules of the funding game.

Paper highlights

  • $32 billion from World Bank to CDD “participatory development” projects, 1999-2007.
  • Getting word out to the general population that this money is available is a real problem. “The majority of poorer, vulnerable, and marginalized households were not even aware of the program” designed to target them.
  • Complicated process: The typical case (Tanzania’s Social Action fund, or TASAF) was a 6 step process involving at least 3 governing / decision-making bodies. The only World Bank control was to set up a funding formula per region based on population, region-size, and “poverty headcount ratio” – whatever that is – to ensure each region got funding.
  • Local ownership? This pot-o-money (through NVF, TASAF) typically accounts for 80% of the project budgets, and is never less than 50%. So this means that all projects are primarily funded by an outside source – hard to claim that any locals really have risked their own resources in these ventures. In contrast, Peace Corps runs a “small projects assistance SPA” grant system with a $500 cap and a required minimum 25% community contribution (usually want to see a 50-50 match). These projects are fast, nimble, and community owned because half of the resources are coming from the local people themselves.
  • Regressive grant winning: Within Tanzania, the richest districts submitted three times more grant proposals than the poorest districts, and were more often awarded them (15% vs 12% acceptance rate).

My comments

  • Forget project proposals
    • Having done background checks on over 300 local NGOs for GlobalGiving myself, I don’t think that project proposals are very effective way of finding out which local organizations are going to best use grant money. (AKA “weak correlation between good proposals and good community work”) Instead, I think broad, informal community feedback mechanisms will yield a set of trusted local organizations whose work will better reflect community priorities. How detailed does a proposal need to be to “feed the hungry”, “dig a well”, “pay for school”?
  • Flat feedback mechanisms
    • Let’s kill this idea of “upward” or “downward” accountability once and for all. There is only one kind of accountability in the 21st century: public disclosure to everybody. Whether you’re a village mom or a world bank analyst, you deserve feedback. And equally as important, the way that information about funds, projects, and results gets delivered to the people needs to match with how they usually access info: SMS, newspapers, radio, face-to-face meetings, and television. What DOES NOT COUNT as feedback  (for 2011-2012) is publishing a PDF on the Internet, or some format that isn’t mobile-friendly.
  • More bowl-shaped funding funnels
    • This is a budget schematic of the typical internationally funded development project managed by a US 501(c)3 NGO:

The 17% is the calculated average from thousands of IRS Form 990s from US-based NGOs in 2010. The other percentages are a composite drawn from looking at the budgets of hundreds of NGOs (i.e. my best guess). The fraction spent on M&E is the hardest to estimate of all categories, because it is usually not even listed in the budget as it’s own line item, meaning the cost was very small and usually closer to 0% than to 10%.  Note how local staff are not a major fraction of the budget, and untrained, non-university-educated locals are negligible. In some cases, the majority of services and equipment come from outside the country – so the total local dollar impact might be 15-25% of every dollar. For international aid organizations, this amount is even less (I’ve read 92% of foreign aid money gets rerouted back to the host country through budgets like these).

An official aid organization’s budget might look like this:

Most of the staff, services, equipment, M&E, and consultants reside outside the target community. With ODA the word “local” means host government agency officials. It’s hard to determine how much of each aid dollar actually gets spent on truly local services, equipment, or through local organizations (meaning those founded and operated by people living in the same community where they work).

To me it looks like a beer funnel:

Which is crazy, because you get more bang for your buck if you flip this around and spend most of your dollars in the community itself. Of course to do this you would need a much more robust reputation system so that you knew who to trust with that money. Such is system is possible, and would allow us to change the funding structure of ODA to reflect the spirit of “participatory feedback” finally.

The ideal organization budget is more bowl-shaped:

I dream of a world where at least 50 cents of every aid dollar not only remains in the country for which is was intended (buy local, shop local, hire local), but is spent on priorities chosen by the community itself through a direct-democracy process (not local leaders or national government representatives). For this to happen, we need to vastly grow the bottom of the bowl – local people who are volunteering or hired on an ad-hoc basis as part of projects. The GlobalGiving Storytelling Project does this. We’ve enabled over 3000 Kenyans and Ugandans to play a small role as story scribes in 2011.

Where the World Bank community driven development (CDD) projects paper finds a complicated funding allocation process and a lack of community awareness are leading to the poorest communities getting less, we provide a workable alternative to project proposals.

The Solution:

Combine real-time community feedback with community outreach and awareness of the fund. If the whole community has a say, and people are making decisions about what happens within 1km of their home, more of them will choose wisely.

I think the CDD / NSF / TASAF / participatory feedback idea is compelling – but needs to trust “whole” community feedback more. It also needs to involve more “local” community based organizations – those having a mostly volunteer staff of 10 people or less and a yearly budget of just a few thousand dollars. These people are mostly tuned in to their communities, and prepared to devote time and resources to the problems.

Read the paper yourself:

Or read the World Bank Blog’s discussion of this paper

Advertisements

5 thoughts on “The Crazy Beer Funnel in International Development

  1. A lot of people seem to be extremely anti-M&E within NGOs – they think all the extra work is just red tape. There needs to be a cultural change before any large personnel shakeups can occur.

  2. Although I have NO direct involvement with ODA or big agencies, I have a standing pricing guideline as an “innovation consultant” for NGOs: My daily date is inversely proportional to the percentage of aid or grant dollars that end up in the hands of “host country nationals” as Peace Corps calls them, or “the people whose country is affected” in my language. If 50% of the grant for project X in Nigeria actually gets spent in Nigeria, then I drop my consulting rate by 50%.

    This forces the organization interested in hiring me to reveal at least one aspect of their generally hidden grant budget.

    And as of yet, I have never had to drop my consulting rate. 😦

  3. DevHope.com is a social network for solidarity-based and citizen development, DevHope.com is an IT and Integrated Platform dedicated to resources an finance projects of general interest. DevHope.com transfers services, not money to fund development projects. Its model eliminates corruption and embezzlement. Sponsors, NGOs, Institutions and all stakeholders can monitor the progress of the project on devhope.com and also obtain feedback from the final beneficiaries. These steps ensure transparency in the realization of the project.

    Could this be the innovation that changed Africa and the world forever? http://devhope.com/rubrique/index.php?d=Devhope.Institution&actu=1350

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s