Femke Maes - Tap or treat? Experiments to improve access to safe drinking water in Uganda

Femke Maes 

soutiendra publiquement sa dissertation 

Tap or treat? 

Experiments to improve access to safe drinking water in Uganda

pour l'obtention du grade de doctorat en sciences économiques et de gestion 

le jeudi 11 juin 

à 18h 

à la Salle du Conseil de faculté (Gent)

Thèse réalisée sous l'accord de cotutelle et de double diplôme entre l'UCLouvain et l'UGent

Abstract 

Globally, 2.2 billion people still lack safely managed drinking water, and diarrhea remains one of the primary causes of death among children under five. At the current pace of progress, we'd need to accelerate sixfold to meet the UN's 2030 targets of equitable and universal access to safe water for all. In this PhD research, I traded the comforts of UGent offices for extensive fieldwork in rural western Uganda, armed with survey forms, water sampling kits, symptom diaries, and one deceptively simple question: should rural water interventions improve water at the source, or treat it at home?

To find out, I ran two field experiments. In the first, nine communal taps were installed in communities previously relying on open surface water. The taps genuinely improved drinking water quality — but reliable health gains in children proved rather elusive. Unpredictable water supply, management challenges, budget constraints, and the consequential collection from multiple sources meant that cleaner water at the tap likely didn't always stay clean by the time it reached a child's cup. Households also tended to abandon home treatment once they had tap access, removing a safety net that remained most necessary due to recontamination during transport and storage. 

This pushed the research toward the home. In a randomized controlled trial spanning eighteen months and 600 households, three water treatment methods went head to head: a rocket stove for boiling, a ceramic filter, and a membrane filter. While my team collected water samples and surveys every four months, primary caregivers recorded their children's health daily in symptom diaries — tracking diarrhea, fever, and vomiting over fifteen months — giving a far more granular picture of health than the usual quarterly survey snapshot. The ceramic filter emerged as the clear winner: improved water quality, meaningful reductions in child illness driven by fewer illness episodes, and substantially lower demands on household time and biomass fuel than boiling. Membrane filters performed similarly well on water quality and health, but came with a higher price tag and an unexpected vulnerability — they appeared to clog more easily after hot days, when source water turns murkier, and the membrane dries out faster. Boiling, the world's most popular water treatment method, failed to improve child health, most likely because its relentless demands on time and firewood make consistent use unrealistic for already stretched households.

The broader takeaway is perhaps unsurprising, but worth stating clearly: clean water saves lives, but only when the solution fits the reality on the ground. The best water treatment technology is the one people will actually keep using. Ceramic filters can deliver clean water to children now — cheaply, reliably, and without waiting for pipes to arrive. But pipes should still arrive. Household water treatment is not the end of the story; it's a bridge to the infrastructure that rural communities ultimately deserve.

Membres du jury 

Prof. Bart Cockx (UGent), promoteur

Prof. William Parienté (UCLouvain), promoteur 

Prof. Bertel De Groote (UGent), président du jury 

Prof. Koen Schoors (UGent), secrétaire du jury

Prof. Goedele van den Broeck (UCLouvain), 

Prof. Lore Vandewalle (KULeuven)

Prof. Edward Miguel (University of California)