Conversation on the nuances to crowd-sourcing AID, development


[Below is my truncated version]

Owen Barder writes:

Christopher Fabian tells the story of something that happened yesterday. Somebody came up with a not-very-good idea for foreign aid: “Let’s collect 1 million t-shirts from the US and send them to Africa.”

The idea was discussed on twitter and on the blogs (including Aid Watch, Aid Thoughts, Tales from the Hood, Amanda Maculec, Siena Anstis, Texas in Africa, and Project Diaspora). A fuller list of reactions is here. Christopher Fabian explains what happened next:

Within a day a development concept had been aired…. Real-time input, from “the field” become an actor in “aid/charity/development.” People in “the place” (Africa) where the “aid” was going weighed in. Experts who had not met each other were able to share experience, synthesize and create new literature on giving, aid, and development theory. And it happened in a few hours.

This is indeed pretty interesting: it is one of the first times an appraisal of an aid project had been crowd-sourced.

It would be even better, of course, if some of the intended beneficiaries had a say.

Owen Barder asked, "why aren’t all aid projects subject to this kind of scrutiny, before anyone spends any money on them?

Christopher Fabian:
April 29, 2010 at 1:10 pm
The “crowd” that we need *must* be the end-”users” themselves – see my last para:

“Mobile phones are (soon to be) everywhere. Connectivity is growing. Barriers of communication are dropping. If we can learn from this how to publicly lay our ideas on the ground and invite a square-dance on them, we can more correctly link development activity, delivery and effect – and that link can be the person at the very end of the last mile. Let me call this the first crack in the very large iceberg of ‘charity.’"

April 29, 2010 at 1:26 pm

What I’m glad about is that this debate illustrates the level of maturity regarding what really constitutes aid.

Way back when, developing countries had no choice but to accept what was dumped on them even by well meaning donors.

Sam Gardner:
April 29, 2010 at 1:41 pm

How soon should you put an idea out of its misery?

April 29, 2010 at 1:44 pm

April 29, 2010 at 2:23 pm

These crowdsourcing sites are out there- they just need more visibility.

Have you heard of the Development Practitioners Forum? A bit more closed off, only development experts can join, but it allows people to share their failures and successes in a safe-haven of sorts.

Also, there is Africa Rural Connect– a [Peace Corps] website dedicated to global collaboration to help build the small ideas and projects (private sector, NGOs, etc.)

April 29, 2010 at 2:44 pm

The problem with the 1 million t-shirt project is that they hadn’t even properly defined their intended beneficiaries. …opaque wording.. some talk about widows as used-clothing sellers – too vague.

One has to be careful how you end up getting feedback from intended beneficiaries. I think if you went to poor families and said “I’ve got a free t-shirt here, would you like it?” most would quickly say yes, rather than ask for something else (which obscures the opportunity cost of having performed a different intervention). Presenting people with options (and trade-offs) is the best approach.

April 29, 2010 at 9:15 pm

The advantage of real-time feedback and new tech is that the process of review can be perceptibly democratized and accelerated.

It’s true that ideas from powerful players still can carry weight because of deference or fear (or deep pockets) not because of their quality. At the same time more people can join the conversation and more perspectives can be added, and even powerful players will find it hard to totally ignore rapid and clear negative feedback.

I take Owen’s point about “regular aid workers” and beneficiaries not having the time and means to participate – I hear this lament all the time – but I think this is rapidly changing, and will change even more quickly when technology provides the means, but even more importantly when they see that this can an actually have an important impact and that it can help them in their lives and work.

April 30, 2010 at 4:04 am

Haven’t we done this once before? See David Roodman and Kiva.

Marc Maxson:
April 30, 2010 at 10:00 am

The “million T-shirts” Reminds me of a recent Pepsi RefreshEverything project to deliver girl scout cookies to the vets in Iraq. I’d met a vet at a dinner party just before this project won, and he was complaining about all the girl scout cookies people send to Iraq. “They tie up the mail system, and we give em to the Philipino contractors anyway,” he said.

When he saw the project he got all his vet buddies to start voting against it by voting up something else. The problem wasn’t with the inclusiveness of the process or with lack of voice from intended beneficiaries. It turned out to be something more difficult to banish – most of the people with influence would have rather been the beneficiaries of the girl scout cookies, so they assumed the Iraq soldiers would want it too.

So how do we get people with most of the power to start thinking like the intended beneficiaries do? Maybe by putting more of the actual beneficiaries in charge. I’m excited about what community switchboards like voice of Kibera ( can do to that end.


April 30, 2010 at 2:24 pm

‘Crowdsourcing’: what a slippery customer.

I’m intrigued to see how many comments here assume something data nerds like me would question instinctively. Here’s a sentence to illustrate “Imagine if we could actually ask people in the developing world what they thought of projects before we started them.” And another from Matt: “One has to be careful how you end up getting feedback from intended beneficiaries – I think if you went to poor families and said “I’ve got a free t-shirt here, would you like it?” most would quickly say yes, rather than ask for something else… Presenting people with options (and trade-offs) is the best approach.” “Ask”? “Went to poor families and said…”? “Presenting people with…”? I guess you could define crowdsourcing as a very big survey with a self-selecting sample but is that what we mean here?

The response to this t-shirt malarkey wasn’t prompted, wasn’t asked for – it just happened. Because people had an opinion about it, felt that their contribution would be meaningful and then got talking in the places they always talk online. It’s largely because it’s unprompted that it’s interesting. Crowdsourcing is, for me, closer to ‘research by discreet listening’. It can be prompted but isn’t bounded. It’s analysis of the everyday made possible by tech that leaves interrogable evidence behind it. That being so, the meat of my point is that the main prerequisite for those unarguably desirable things like ‘beneficiary involvement’ isn’t necessarily the adoption of a particular technology but the use of that technology to do and say things in open, accessible places as part of everyday life. The web is the main place that happens right now. Will it be in future? Will it be in Africa? Probably but not certainly – comparatively closed mobile platforms could emerge.

We can crowdsource pizza-topping preference in California right now because lots of people in California talk about pizza on the web every day. We can try to predict US unemployment more quickly and accurately than the state by looking at Google Job ads ( because that data is, well, real and timely and there to be used. Maybe we’ll be able to crowdsource citizens’ public health priorities in Ghana when people in Ghana start nattering about health services on the web every day. Or would we find that we’re not looking at a ‘priority’ after all and that those conversations don’t happen much? What if it’s all about generators and road surfacing? Are we willing to change ‘our’ priorities in light of what we discover about ‘theirs’? That’s the exciting, difficult, slightly uncomfortable bit for me (a good thing).

Don’t want to be too much of a grouch but all this will likely take a long time to happen in most parts of Africa and there’s no realistic way to shortcut the process (isolated ICT ‘projects’ won’t cut it and neither will highly specific ‘mobile services’ which do one or two things at great cost but nowt else.) Perhaps the best we can do is listen to what is already there with an open mind while refining our methodologies to counter the legitimate worries Christopher has about the reliability of crowdsourced info. Oh and do everything we can to promote open standards, open data and flexible APIs to get at it. Everywhere.


Relevant reading: It Pays to be Ignorant.

[categories owen barder, aid, development, international development, beneficiary feedback, globalgiving storytelling project, crowdsourcing]

One thought on “Conversation on the nuances to crowd-sourcing AID, development

Leave a Reply

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

You are commenting using your 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 )

Connecting to %s