The failure rate of rural water infrastructure projects is a well-documented problem in development practice. Surveys of hand pumps and small water systems in a number of countries have found substantial proportions non-functional at any given time, frequently within a few years of installation.
The reason is almost never that the equipment was badly made. It is that nothing was arranged for its maintenance.
Understanding this makes the projects that succeed considerably more interesting than the installation-focused coverage suggests.
Why maintenance is the binding constraint
An installation project has a clear beginning and end, a visible output, and a natural funding structure. It is easy to photograph and easy to fund.
Maintenance has none of these properties. It requires ongoing money, an available person with the relevant skill, a supply of spare parts, and a mechanism for deciding who is responsible.
Where any of the four is missing, the system fails at the first breakdown and stays failed.
Studies of what distinguishes functional from non-functional systems consistently identify the same factors: existence of a local person trained to repair, availability of parts within a reasonable distance, a funding mechanism for those parts, and clarity about who is responsible.
Technical quality of the original installation appears well down the list.
Why young women's involvement changes the arithmetic
In many of the settings where these projects operate, water collection is performed predominantly by women and girls, and the time involved is substantial — time-use studies in several countries have found collection times measured in hours per day in some communities.
This creates a direct connection between the users of the system and the consequences of its failure, which is not always the case for those who make decisions about it.
Programmes that trained women as pump mechanics and system caretakers have reported better functionality rates in several evaluations, and the proposed mechanisms are reasonably clear.
The trained person is present rather than travelling. She experiences the failure directly. She is not likely to migrate for work in the same patterns. And the training is retained locally rather than leaving with the trainee.
The evidence base here is not strong — evaluations are frequently conducted by the implementing organisations and controls are rare — but the direction is consistent and the mechanisms are plausible.
What the school-based versions look like
Several documented projects have involved secondary school students, frequently girls' schools, in water and sanitation work at their own institutions.
The typical structure involves students conducting a survey of the existing system, identifying failure points, being trained in basic repair and testing, and taking on ongoing responsibility with a small budget.
What makes these durable, where they are durable, is the same feature that made the coding clubs durable: an older cohort that trains the next one, so the capability does not leave when a group graduates.
The projects that were structured as a one-year activity generally ended when the participants did.
The girls' school sanitation link
There is a specific reason this matters for school attendance, discussed in more detail in this site's article on girls' education programmes.
Sanitation facilities — private toilets, water availability, disposal — are associated with adolescent girls' attendance in several studies, and non-functional facilities are common.
A school with a broken water supply has a sanitation problem, and the sanitation problem has an attendance consequence that falls unevenly.
This makes maintenance capacity at a school an educational intervention as much as an infrastructure one, which is not how it is usually budgeted.
What generalises
The pattern here recurs throughout the material on this site, and it is worth stating in general form.
The visible, fundable, photographable part of an intervention is the installation. The part that determines whether it works is the ongoing arrangement, which is invisible, unfundable and unphotographable.
The same is true of coding clubs, sports programmes, mentoring schemes and school water systems. In every case, the evaluations identify continuity and local ownership as the determining factors, and in every case the funding structure rewards the launch.
The four questions
For anyone assessing or designing a project of this kind, the questions that predict survival are narrow.
Who repairs it when it breaks, and are they here?
Where do the parts come from, and how far away?
Who pays for them, from what?
Who trains the next person, and when?
A project with answers to all four generally lasts. A project with an installation and none of the four generally does not, however good the installation was.