In 2016-2017, we delivered cash transfers to one of the remotest and most operationally challenging locations we have ever worked in, reaching some of the poorest recipients we have ever met. In this series of posts, we look at why we did it, how we did it, and what we learned from the experience.

It’s the obvious question, but, in this program, the most challenging: how to get cash into the hands of your beneficiaries? Across East Africa, we transfer cash through mobile money – a technology where cash is sent digitally to a mobile phone, with a network of local agents providing a ‘cash-out’ service, converting digital-currency into hard cash for a small fee. The approach is effective and efficient, allowing us to stretch our budgets further and reach more recipients.

The challenge in Uganda’s remote north was that this network of local agents was virtually non-existent. Our hypothesis was that agents would respond rapidly to incentives. Where recipients are looking to convert digital currency to cash, agents should be willing to earn their fee. Our job was to set recipients up with phones and mobile money accounts. The rest, we would leave to the provider and their network of agents (under our watchful eye, it should be added).

Did it work? We’ll start with our side of the bargain: in spite of the remote location and our field team’s reduced productivity, we were able to provide recipients with phones and SIMs (just 15% already owned them), register them with mobile money (just 6% had an account) and train them on how to use it. When it came to converting digital currency to cash, the recipients and agents followed the money. For around half of all transactions, agents travelled to our recipient villages, at their own cost and of their own volition. They covered hundreds of kilometers in the process, allowing our recipients to ‘cash-out’ within walking distance of their homes. For those households the agents couldn’t serve, usually those living in the remotest villages, the recipients themselves made the return journey. Some even found a smart workaround, sending digital money to a local trader and setting an informal ‘cash-out’ fee between them.

The result: 99.6% of digital transfers were successfully sent at the first attempt. On average, each recipient was able to ‘cash-out’ within 5 hours of receiving their transfer, travelling on average 66km to do so. While this is 2.5x slower than in a non-remote program, it shows the approach is able to respond to surges in demand even in the remotest areas. In total, 90% of attempted cash-outs were conducted without recipients experiencing any issues. The 10% where issues were reported usually related to low liquidity, resulting in a delay but no worse. 99.8% of recipient households told us that they intend to continue using mobile money after the program, suggesting possible long-term benefits of increased financial inclusion.

This organic approach was highly efficient, given the context. Efficiency on the campaign (transfers over total program cost, direct and indirect) was 71.5%, more than twice a publicly available benchmark for the efficiency of in-kind aid. The penalty for delivering cash to a remote area was a 10-20% efficiency drop, depending on estimates of scale. The upside: reaching some of the poorest and most underserved communities we’ve ever encountered.

Tomorrow, in the final post in this series, we look at what happened after we sent our transfers, and the lessons we take from this program.


Read Part 3 of taking cash to new frontiers.

Josh Williams is GiveDirectly’s Field Director in Uganda, and led our remote payments project.
Back to List