In discussions of international development, corruption is often the elephant in the room. Funders don’t like to hear about it; nonprofits and governments don’t like to talk about it.

Satellite image with time and GPS marker overlay
During routine visits to recipients, our field team uses technology that automatically records a time and GPS marker for each piece of data they collect.

But in my experience, the hard reality is that when any organization gives away anything of value – whether it’s meals, money or anything else – we need to worry about corruption. When we don’t, the results can be spectacularly bad. For example, when Sandip Sukhtankar and I audited one of the world’s largest social programs, we found leakage rates as high as 80%. What can we, as donors, do to avoid this outcome? How can we be confident that benefits are reaching the intended recipients without being siphoned off along the way?

In independent research with state governments in India, I’ve seen encouraging progress. In particular, transitioning from cash-based payment systems to electronic payment systems seems to reduce graft and leakage, at least in places where the poor have access to secure mobile banking services. But there are limits to how quickly and aggressively governments, with their procedural rules and political constraints, can innovate in this area.

That’s one reason I’m excited about what we’re building at GiveDirectly. By combining electronic records of behavior with multiple layers of independent auditing, we’ve created the strongest integrity system I’ve seen yet. Interestingly, we are finding that it’s often the little things that matter most.

A great example of this came up last week. As they perform their routine visits to recipients, our field staff collect data using Android devices that automatically time-stamp each new entry. This may seem like a small thing, but it made a big difference when an employee tried to claim pay for a (fictitious) extra day of fieldwork. Typically, this sort of misrepresentation would be hard to catch without expensive, time-consuming monitoring. However, because of this technology, all we had to do was to compare timestamps and see that the work uploaded to our servers on Monday had actually been done on the previous Saturday. After a few quick phone calls to recipients to verify this, the case was closed. We caught the falsehood and saved a day’s wages with minimal staff time and effort expended.

There’s still more that can be done, and we know it. We’re always looking for low-cost ways to increase our accountability, and as we learn we’re committed to having an open dialogue about the issue of integrity. Those who are unwilling to confront this issue are kidding themselves.

Back to List