With foreign aid, failure is essential to learning

This is the first of a four-part series on innovative ways to deliver aid in our conflicted world.

A dollar donated in foreign aid never translates to a dollar received by deserving people in a poor country.

That dollar has to be siphoned through a system of bureaucracies that chip away at it while trying to transform it into beneficial action. Typically, this action is in the form of a "project," an organized attempt to target beneficiaries and offer them goods or services. Thus, whether aid improves the lives of people depends on whether a project is well designed, whether the targeted population is selected properly and how efficiently it's implemented. Each is a link in the chain that translates aid into concrete action on the ground.

In recent years, there's been a lot of attention on improving project design by evaluating the impact of projects, particularly by using randomized trial methodologies borrowed from the medical sciences. Some attention has also been directed toward how to target the poor. That means defining poverty lines and determining where a household falls by collecting survey data, allowing communities to select poor households based on their own criteria, or some combination of the two methods.

While randomized control trials and poverty targeting methods are the subject of heated debate, project implementation – which can eat up to 90 per cent of that dollar – also deserves urgent attention.

Projects can be poorly implemented for a variety of reasons: theft and corruption among implementing agencies, inadequate budgets for supervision, poor management, poorly trained staff and so on. Smaller "boutique" projects are easier to implement, but they also affect fewer people. For foreign aid to have a transforming effect, it needs to do things on scale.

But larger projects that affect bigger populations face more variety in the things people need and have less information about local conditions, making them harder to implement. This can result in a number of problems. For instance, the benefits of the intervention may not match the needs of beneficiaries, or the benefits might be diverted to government staff. In addition, the village's rich and powerful may ensure that the benefits stay with their cronies.

Poor implementation is often the weakest link in aid effectiveness, and is also the most difficult to fix. That's because implementation problems are deeply embedded within the social and political environment of a country. This leads to some of the most difficult challenges facing development policy – entrenched discrimination and inequality, a culture of corruption and lack of accountability in government.

But whatever the environment, there's no question that it's critical to track funds and the performance of functionaries, and assess changes in the lives of beneficiaries. With good monitoring data, project managers can learn when their efforts work and when they fail, and where they succeed and where they don't. This can result in mid-course corrections, shifts in design and a more flexible learn-by-doing approach that's essential to creating positive changes in challenging environments.

Aid agencies, unfortunately, don't pay adequate attention to data, and often rely on "experiential" information – the experience of senior project managers during their field visits, where all too often they're steered to villages and sites where things are working well. This needs to change. Part of the problem lies in the inordinate need of project managers to report positive news; their incentives are heavily stacked against even the whiff of failure.

But in a complex situation, where it's difficult to know how to act, failure is essential to learning. Thus, it can make foreign aid more effective if we pay close attention to monitoring – improving the quality of data, using data to inform project decisions and becoming more tolerant of failure.

Vijayendra Rao, a lead economist in the World Bank's Development Research Group, is co-author of the soon-to-be published book Localizing Development: Does Participation Work? To see a video conversation with the author, visit the Canadian International Council's website at www.opencanada.org/newhumanitarians.