How FAIR and CARE Principles Improve WASH Data Stewardship

This article was first published on openwashdata and contributed to R-bloggers. You can report an issue with this page. To share your own content on R-bloggers, submit your blog, or use R-posts if you don’t have one.

Note: Emmanuel Mhango first published this post on LinkedIn on 27 October 2026: “Operationalizing FAIR and CARE data stewardship in WASH systems”.

The challenge for water, sanitation, and hygiene programs isn’t simply to collect more data. It’s to make information useful and reusable while respecting the people and communities it describes.

Data only matters when people can use it

WASH programs produce information about boreholes, groundwater, water quality, sanitation, infrastructure performance, and household access. Those records can guide planning, monitoring, spending, and service decisions. But gathering information is only the start. Its value depends on whether the right people can find it, understand it, access it appropriately, and put it to work.

That’s often where the system breaks down. Data sits in separate institutions, trapped in incompatible formats or poorly documented project files. Even when it’s technically available, users may lack the context needed to interpret it. And data about communities can be shared without enough attention to their rights or interests.

The question, then, is how to make WASH data easier to find and reuse without losing sight of ethics, inclusion, and accountability. Two frameworks help frame the answer: FAIR and CARE.

FAIR helps data travel; CARE asks who benefits

Wilkinson and colleagues introduced the FAIR principles in 2016: Findable, Accessible, Interoperable, and Reusable. They focus on the conditions that let data move between users and systems.

• Findable: Give datasets persistent identifiers, descriptive metadata, and searchable records so people can discover them.
• Accessible: Make retrieval procedures and access conditions clear. Accessibility doesn’t require every dataset to be public.
• Interoperable: Use compatible formats, shared vocabularies, and common standards so information can be combined.
• Reusable: Document provenance, licensing, quality, and other details needed for responsible reuse.

The Global Indigenous Data Alliance’s CARE principles add questions about power and purpose: Collective Benefit, Authority to Control, Responsibility, and Ethics.

• Collective Benefit: Data governance should produce tangible benefits for the people and communities represented.
• Authority to Control: Indigenous Peoples and other relevant rights holders should have appropriate authority over decisions about their data, consistent with their rights and governance arrangements.
• Responsibility: Custodians and users should build relationships, strengthen capacity, and use data in ways that support equitable outcomes.
• Ethics: Data practices should respect people and communities, their rights, and the possible consequences of using their information.

FAIR asks whether data can be found, accessed, combined, and reused. CARE asks who decides, who gains, and whether people represented in the data are treated fairly. Both questions matter in WASH. A technically polished dataset can still be used irresponsibly; a carefully governed one may have little practical reach if no one can find or interpret it. The work is to put the two frameworks into practice together.

Make datasets discoverable, with context and safeguards

Take groundwater observations from several districts: water levels, electrical conductivity, temperature, dates, and monitoring well locations. Uploading those measurements without a clear title, geographic coverage, collection methods, equipment details, or other metadata leaves users guessing. The figures may be available, but they’re hard to judge or reuse.

A useful inventory should record each dataset’s title, summary, responsible institution, geographic and time coverage, collection methods, update frequency, data dictionary, and access conditions. Persistent identifiers and searchable repositories or catalogues can help users locate the records.

But discovery has limits. Exact locations of vulnerable water infrastructure, sensitive community details, or information that could put individuals at risk may need restricted discovery or controlled access. The aim is to help appropriate users find relevant data while applying safeguards that match the risks.

Access can be tiered

FAIR doesn’t mean putting every record online. Access can require authentication or authorization, and sensitive information can remain restricted if the rules are clear and documented.

WASH institutions might offer approved aggregate indicators publicly, allow institutional access to operational records, and restrict household-level data. A district water office, for instance, could publish aggregated borehole functionality figures while controlling access to detailed infrastructure security records and identifiable household survey responses.

Data-sharing agreements should spell out who can access information, for what purposes, under which conditions, and with what responsibilities. A governance framework should make these distinctions explicit, rather than leaving decisions to individual discretion. The practical test is whether data is open enough to be useful and restricted enough to protect people and infrastructure.

Build connections between systems

WASH information is collected through a mix of tools and institutions. Asset management systems may hold infrastructure records; monitoring databases store groundwater observations; surveys track sanitation; and monitoring and evaluation platforms record program performance. Even when datasets cover the same places or assets, differing identifiers, boundaries, terminology, units, and structures can block integration.

Interoperability takes deliberate coordination. Useful steps include:

• Assign common identifiers to water points, monitoring wells, and other assets.
• Standardize geographic references, administrative codes, dates, and measurement units.
• Maintain shared data dictionaries and controlled vocabularies.
• Use documented APIs and machine-readable exchange formats.
• Record links between assets, monitoring observations, maintenance work, and water quality results.
• Adopt relevant national and international standards where they fit.

A stable identifier for each monitoring well, for example, lets users connect water-level readings to infrastructure details, maintenance history, and relevant environmental information. Yet software alone won’t solve the problem. Organizations also need agreement on definitions, ownership, responsibilities, and how records get updated or corrected.

Put community interests into governance

FAIR provides a strong foundation for managing data, but it doesn’t settle questions about representation, power, or how benefits are shared. Those questions are central to CARE.

WASH records can describe household access, sanitation practices, local water sources, and service inequalities. That information can shape public decisions. Its collection and reuse should account for the circumstances, interests, and rights of the people it describes.

In practice, Collective Benefit means connecting data initiatives to better services: stronger maintenance, fairer resource allocation, more reliable provision, and better local decisions. Authority to Control means respecting the rights and governance arrangements of the people concerned. Indigenous data sovereignty matters where Indigenous Peoples’ data is involved. More broadly, institutions need appropriate ways for communities to participate in decisions, without assuming that one representative speaks for everyone.

Responsibility includes investing in local data management skills, providing useful access, acknowledging contributors, and building partnerships that return value to the institutions and communities that helped generate the data. Ethics means assessing risks such as re-identification, stigma, exclusion, misuse, and disclosure of sensitive information. Collection and sharing should serve legitimate purposes and include safeguards suited to the risks.

That can mean asking communities about intended uses of household data, explaining sharing plans, creating ways to raise concerns or correct records, and reporting how findings affect services. Participation can’t be a one-off consultation; it needs to remain part of data governance.

From publication to borehole repairs in Malawi

The value of stewardship is clearest when we follow what happens after data is collected. In my WASH work in Malawi, we gathered information about water infrastructure in Mwanza District and other areas, then published it through an open data platform. Before publication, the records were stored on computers at the BASEflow office, limiting their visibility beyond the project team.

Within five days of publication, UNICEF emailed to say it had found and downloaded the data. The organization then used it to support borehole rehabilitation in areas covered by the datasets. Water point functionality was restored in those areas.

That’s a concrete example of the difference between keeping information on an office computer and making it available to people who can act on it. From a FAIR perspective, publication made the data easier to discover, access, and reuse for operational decisions. It also helped a development partner identify infrastructure needs and direct work where it was needed.

CARE adds a further measure of success: the value of WASH data shouldn’t be counted only in downloads, publications, or institutional visibility. It should also be judged by whether its use benefits communities that depend on the water points.

There’s room to make that link clearer. Consistent documentation, reliable water point identifiers, recorded collection dates, and follow-up on data use and outcomes would help institutions show how information contributes to service improvements.

Turn principles into routine work

The hard part isn’t understanding FAIR and CARE. It’s making them part of everyday institutional practice. A workable starting sequence has five steps:

1. Inventory data and governance. Record each dataset’s custodian, purpose, format, quality, sensitivity, access conditions, and update schedule. Identify gaps in documentation, interoperability, and accountability.
2. Assign stewardship roles. Define responsibilities for data owners, custodians, stewards, system administrators, and authorized users. Set procedures for validation, metadata, version control, access requests, retention, and incident response.
3. Agree on metadata and interoperability standards. Create shared templates, dictionaries, identifier conventions, and exchange formats. Align them with government and sector standards instead of creating parallel systems without a clear need.
4. Build in ethical, community-centered governance. Classify data by sensitivity and risk. Set proportionate access controls, privacy protections, consultation procedures, and ways to address community concerns. Clarify how collective interests and applicable rights shape data use.
5. Track progress and adapt. Assess whether data is documented, discoverable, interoperable, and reusable. Also examine whether access decisions are transparent, community engagement is meaningful, local capacity is growing, and services are improving.

This is how the principles become part of data management rather than policy language that never reaches day-to-day decisions.

Measure service outcomes, not just publication

Counting datasets collected or published gives an incomplete picture. A stronger assessment looks at technical practice, governance, capacity, and the effects on service delivery.

Putting stewardship into practice

FAIR and CARE give WASH institutions a useful direction, but a policy on paper doesn’t prove that decisions are working. Institutions need to check whether governance arrangements shape real choices and deliver benefits. The same scrutiny applies to data quality: a dataset can be findable and interoperable while still carrying errors, bias, or gaps.

Digital repositories, dashboards, APIs, geospatial platforms, and automated checks can support better stewardship. They can’t decide whether sharing is appropriate, settle conflicting definitions of infrastructure functionality, or judge whether a data use serves the communities concerned. Those decisions need clear responsibilities and sound governance.

For institutions with limited resources, steady improvements are more practical than a wholesale technology rebuild. Invest in staff skills, document workflows, use tools that can be maintained, collaborate across institutions, and connect stewardship to existing monitoring and reporting. Better metadata, access rules, identifiers, and stewardship practices may improve the systems already in place.

From principles to better services

WASH data management needs to be more than digitization. FAIR helps make data findable, accessible under appropriate conditions, interoperable, and reusable; CARE brings collective benefit, authority to control, responsibility, and ethics into decisions about how data is governed.

The test is what institutions do with those principles: inventory their data, improve metadata and quality assurance, adopt shared standards, assign stewardship responsibilities, and include ethical, community-centered judgment in sharing arrangements. The measure of success isn’t the volume of data collected or published. It’s whether people can trust that data, reuse it responsibly, and turn it into more reliable, equitable, and sustainable water and sanitation services.

Making WASH data available is only a start. It should also be useful, trustworthy, and accountable to the people it is meant to serve.

Join the conversation

For WASH professionals, government institutions, researchers, and development partners: how is your organization putting FAIR and CARE into practice in its information systems? Which tools, governance practices, or lessons have helped turn data into better services for communities?

Share your experience in the comments on the original LinkedIn article.

Selected references

1. Wilkinson, M. D., et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, 160018. https://doi.org/10.1038/sdata.2016.18
2. Carroll, S. R., et al. (2020). The CARE Principles for Indigenous Data Governance. Data Science Journal, 19, 43. https://doi.org/10.5334/dsj-2020-043
3. Global Indigenous Data Alliance (GIDA). CARE Principles for Indigenous Data Governance. https://www.gida-global.org/careprinciples
4. Research Data Alliance (RDA). FAIR Data Maturity Model. https://www.rd-alliance.org/

This article draws on my experience as a Data Steward at BASEflow Malawi through the openwashdata community with ETH Zurich.