By Eliot Pence

No one truly knows what we are doing in development. We have spent billions of dollars and do not have much to show for it. Critics often point out that China has contributed more than any other country in reducing global poverty, despite having received nearly no development assistance. By contrast, Africa has received the largest portion of the world’s development assistance funds, yet many countries remain mired in conflict and poverty, falling short of several international development goals. Though Africa has experienced recent growth, success has more to do with a predictable commodity boom and bust cycle, than any sudden increase in aid effectiveness. As a result, we need a new theory of development—one that is focused not on what we think we know, but how much we don’t know. Its starting point should be a shift away from the idea that development is a universally standard process that is replicable across a variety of countries and times. The past four decades have conclusively illustrated that what works in one country might very well lead to failure in another.

In reality, the debate on aid effectiveness has been underway for decades. Regrettably, it has devolved into a shouting match that has divided committed practitioners and paralyzed leading agencies just as we are on the cusp of perhaps the most important year in a generation for development. 2015 marks the end of the Millennium Development Goals, MDGs, an ambitious UN-led development effort that began two decades ago. A new set of goals, the Sustainable Development Goals, or SDGs, are under construction, but their foundation is weak. They have ballooned into a grab bag of over 150 goals without a clear focus for the international development agenda. Critics have already suggested the cost of implementing them may be greater than their benefit. What is needed in the world of development assistance is a paradigm shift.

One promising avenue for exploration is the lessons and insights from an industry that changes quicker than any other: Silicon Valley. After the tech-bubble collapse in 2000-2001, many entrepreneurs started re-examining industry processes. One entrepreneur, Eric Reis, theorized the Lean Startup model with the premise that industries incorrectly think they know what customers want. This results in an elongated manufacturing process to perfect products before going to market, depleting resources rapidly. When the consumer inevitably dislikes aspects of the product or packaging, the start-ups lack the resources to pivot and adjust. This method generates an enormous amount of waste in both time and money, with minimal benefits for targeted consumers. As a result, Reis reversed the approach and created what he called a Minimum Viable Product, or MVP. An MVP was a product barely ready for people to use—filled with errors and omissions. It was less “best practice” and more “throw stuff at the wall to see what sticks.” Behind the MVP was a philosophical supposition: no one really knows what works, so we need consumers to tell us. Reis called it “a fundamental reexamination of how to work in our complicated, faster-moving world.” Rather than spending hours perfecting a method or product, just have the people tell you what they prefer by quickly getting it into the market and worrying about adjustments later. The key was not the solution or the product. The key was the feedback.

Development stakeholders would do well to take a page from Reis’s Lean Start-up method. The method’s focus on hypothesis-driven experimentation is all the more important today, as recent research shows just how entrenched biases are in the “development set,” limiting innovation and success. The World Bank’s 2014 Development Report demonstrated that development workers remain deeply paternalistic, vastly underestimating the intelligence and capacity of the poor. The Financial Times’ Michael Holman recounts: “In a survey of World Bank employees, participants were asked to estimate how many of Nairobi’s poorest residents would agree with the statement that vaccinations caused sterilization. Forty-two percent of the World Bank’s staff estimated that the poor would agree with the statement. Only 12 percent of the residents agreed.” The study suggests that World Bank President Jim Kim’s overhaul of the bank, as disruptive as it was, may not have gone far enough to shift paradigms. We need to stop believing everything we assume, and one way of doing that is by implementing a lean model of development.

The obvious question then is how to operationalize lean development. At its core, there are four cornerstones to lean development: decentralize, open up, experiment, and adapt.

1. Decentralizing assistance is the most fundamental starting point: being closer to your interest group is always better. The best way to think about this shift is to utilize the franchising model. Franchising creates tighter feedback loops from customer to company, giving management the information needed to adjust and change strategies midstream, such as improving design and implementation. Implicit in a decentralized model is a complete reliance on communities’ existing systems and processes. As multiple studies have shown, when communities identify their own challenges, aid has the most benefit. So, start by being closer and listening to your interest group.

2. Opening development agencies up to the public is a critical second step. Development institutions should see themselves as incubators, not implementers, and as repositories of information, rather than curators of it. Development stakeholders should not be determining what the public sees. In opening up, development institutions could cultivate more agnostic approaches in their program development and operations. They will be driven more by actual demand, rather than their perception of that demand.

3. Experiment. When development agencies do take the lead on programs, they should strive for more experimental approaches, rather than prescriptive or best-practice methods. “Failing fast and often” is frequently better for an organization’s sustainability than thinking hard and long about issues. The World Bank’s FAIL Faire is a step in the right direction, but more resources could be redirected to support similar initiatives that promote trial and error methods.

4. Adaptability (over scalability). Programs that “lack scale” in impact evaluations have been phased out or isolated, on the theory that the costs of managing a non-scalable program outweighs the benefits of its successes. The principle of adaptability re-focuses development on what works in a specific context, not on what might work on a bigger scale.

Development stakeholders have spent decades convincing themselves that they know what poor people need. The World Bank’s desire to be a “Solutions Bank” exemplifies this mentality and promotes the idea that development is explicitly and deliberately linear. At worst, these solutions have veered toward the absurd. A merry-go-round water pump will probably not reduce gender-based violence. A soccer ball will not solve the world’s power deficit.  A clean stove will not solve climate change. We spent millions promoting them as solutions to profoundly complex challenges. But the key was never the solution or the product. The key is the feedback. Let’s start investing in that.

About the Author

Eliot Pence is the Africa director at McLarty Associates and an adjunct senior fellow at the American Security Project.

Edited by Shashank Iyer.