How can very poor rural people gain access to safe and durable sanitation?

We’ve just published a paper on the effects of targeted subsidies in Environmental Health Perspectives. The most widely employed approach for reducing open defecation (OD) and encouraging installation of toilets is Community-Led Total Sanitation (CLTS). We’ve previously written about the adoption of that approach here.

CLTS has changed the behaviors of millions of people, motivated into collective action by disgust at open defecation. CLTS is generally implemented without financial assistance to households, and sustaining the benefits has proven difficult.

Despite recent studies in Bangladesh, Laos and Cambodia, no comparable fieldwork on the benefits of subsidies in combination with CLTS has been done in Africa.

Until now. We examined how partial subsidies worked in Northern Ghana, together with UNICEF, via a cluster-randomized controlled trial, funded by USAID #WASHPALS.

In our trial, community-identified vulnerable households received a voucher covering the full costs of a latrine substructure, which included a durable slab and pit lining; HHs were responsible for digging the pit and building the superstructure (either themselves or with help).

We conducted a full census in 59 intervention villages and 50 control villages that had already been through CLTS and declared open defecation free by Ghana standards, visiting and surveying over 5,600 households in late 2019 and again from late ’20 into early ’21.

Here’s what we found. First: the benefits of CLTS declined very rapidly. In control villages, open defecation increased from 25% at baseline to 69% at endline, 21 months later (on average). The reason? Collapsing latrines. The takeaway: rural sanitation programs must elevate durability as a priority.

Second: what did the latrine subsidy do? Voucher redemption was nearly universal, and unsurprisingly, voucher-eligible benefited enormously, with declines in open defecation observed even as control communities saw large OD increases. But another pattern was also very encouraging …

Non-vulnerable households living in the same compounds as eligible households also benefited. These were *not* the purchasing spillover effects that Guiteras et al. observed; likely because of the costs relative to income levels in the region.

Rather, there was “behavioral” spillover. In our subsidy communities, at the end of the trial, only 31% of non-eligible HHs who lived in a compound with a voucher-eligible household practiced open defecation, compared with 74% of the corresponding HHs in the control communities.

The road to subsidy spillover effects of the kind Guiteras et al. found in Bangladesh may be long, but the settlement pattern of compounds in rural Africa offers an important opportunity to amplify behavior change.