GiveDirectly is crossing an important milestone: after over two years working in Siaya district, we’re moving our enrollment teams to our third district in Kenya – Ugunja. This is an exciting moment for GiveDirectly because it’s a sign of the scale we’ve reached: we’ve now been to all the villages in a district that our targeting and research commitments permit. The move also presents new challenges – both personally for our staff and operationally in terms of our process and structure.
On the personal side, Siaya has been the full-time home to at least 18 of our field staff since early in 2014. Field officers visit recipients in their homes day-in and day-out, and live in the closest urban center; they had become well-known in Siaya for the work GiveDirectly was doing. Ugunja is close enough that everyone seems to know about GiveDirectly, but far enough that the familiar faces – recipients, shopkeepers, etc – are no longer there. It is also much dustier, a condition our field officers won’t be able to avoid when riding on the back of a motorbike is a job requirement.
On the operational side, Ugunja’s villages are larger and have a slightly lower proportion of thatched-roof homes than Siaya’s. This type of difference requires us to re-think and re-optimize our enrollment process, in small but important ways. The census team, which completes door-to-door visits of every household in the village to establish a baseline, is disproportionately slowed relative to other teams by larger village size and lower eligibility. Because we have a sequential model, in which one team completes an independent assessment of a village before the next can begin, each team needs to run at approximately the same pace to avoid delays. To ensure that the census team doesn’t become a bottleneck, we may need to increase the size of the census team, or find ways to sub-divide villages so that two teams can work in the same large village without overlapping and compromising their independence.
We know cash transfers have positive impact for the poor in a wide variety of contexts, and are looking forward to adapting our model again and again. With each move, we’ll learn a little more about how to ensure the success of our program in different settings.