Fieldstone Collective
Loading this page…
Programme · running since 2017
The failure rate of rural infrastructure is not an engineering problem. It is a maintenance-funding problem that engineering is repeatedly asked to solve.
For its first eight years Fieldstone built infrastructure the way most education and livelihoods organisations do: incidentally, when something was obviously in the way. A footbridge here because attendance data pointed at a river; a spring capture there because a cooperative store needed water.
In 2016 Isaac Mwangale was asked to inspect everything Fieldstone had built to that point. Of nineteen systems, eleven were working. The failures had almost nothing in common technically and almost everything in common institutionally: no committee, or a committee with no money, or a committee whose money had been spent on something more urgent than a part that had not broken yet.
Infrastructure became a separate programme in 2017 with a rule attached: Fieldstone would not build anything a local committee had not first agreed, in a constitution it wrote itself, to maintain — and had not demonstrated by holding a maintenance fund for a full year before construction.
They made us wait a year and keep the money in a box. Everyone was annoyed. Then the pump broke in the fourth year and we fixed it in nine days without writing to anybody.
Since the rule was introduced, 47 systems have been handed over and four have failed — two to a flood that removed the intake entirely, one to a committee that dissolved after a dispute, and one to a pump specification we got wrong. All four are listed by name in the annual review.
Limits
Published because an organisation that only describes its scope in terms of what it might do has not defined a scope.
We do not build roads, health facilities, or anything requiring a maintenance capability the community cannot fund locally.
We do not build for communities that have not held a maintenance fund for twelve months.
We do not retain ownership, and we do not return to operate a system after its support period ends.
Approach
No Fieldstone system is built until a committee exists, has a constitution, has agreed a tariff or levy, and has held funds for one full year. Roughly a third of requests do not survive this stage, which is the point.
Every component is specified against what can be bought and repaired within a day's travel. We have rejected better pumps for worse ones that the nearest town stocks.
Each system's condition is inspected annually and published, including the four that have failed. Reporting only the working systems would make the 91% figure meaningless.
Ownership transfers at commissioning. The two years that follow are a support period with a scheduled end, agreed in writing before the first excavation.
Where it runs
Kenyasince 2009
The district where Fieldstone began, and the only one where all sixteen years of attendance data exist in a single series.
Nepalsince 2015
Eleven settlements along a gorge, where the distance between a house and a classroom is measured in hours rather than kilometres.
Moroccosince 2013
The first region Fieldstone fully handed over, and the one it has now been reporting on for longer than it worked there.
Reporting
Community Infrastructure
Why Fieldstone will not build anything until a committee has held a maintenance fund for a full year — told through the crossing that convinced us.
Nadia BelhajSanjala Valley, Nepal8 min read
Accountability
Four of five handed-over water systems are still running. This is what the fifth teaches, and why we keep publishing the inspection.
Isaac MwangaleTazgart Plain, Morocco7 min read
