Data for Advocacy and Social Change in Mozambique
Social accountability interventions began in Mozambique in the 1990s. Since then, many organizations have made progress in this area, and have adopted tools and approaches, including community scorecards, social audits, and community dialogues. Kwantu’s contribution to work in the governance and social accountability sector in Mozambique, together with partners, provides a practical example of creating and using data ecosystems to support the delivery and monitoring of the SDGs.
The accountability tools are predominantly used in the water and sanitation, health, education, and municipalities/local governance sectors, and collect data relevant to SDG indicator 16.6.2, which monitors the percentage of the population satisfied with their last experience of public services. They also share many of the following features:
- They generate data related to a school, health facility, or municipality
- A cycle for a scorecard or audit carried out in a facility or municipality
- A group of people described by gender, age, disability status, and other criteria
- An issue or problem identified by a group that affects local service delivery or governance
- A commitment by local government
- An outcome or story of change
- A reflection or feedback on a scorecard or audit cycle
- An action plan to tackle a set of problems during a cycle
Through our work with the Citizen Engagement Programme (CEP), the Centre for Learning and Capacity Building Civil Society (CESC), N’weti, and Concern Universal, we’ve explored the feasibility of agreeing on common data standards for these tools, starting with the community scorecard tool.
Complementing this work, our role in the global data collaborative Everyone Counts, together with WorldVision, CARE, and Development Initiatives, has enabled us to test the proposed standards further to see how well they translate to other countries and contexts. Through this collaboration, we identified significant overlaps in that World Vision has been doing similar painstaking work for years.
In March 2017, CEP, CESC, and N’weti reached agreement on shared data standards. For the first time, organizations in Mozambique that were using the same tool – community scorecards – were able to collect data in a comparable way. This is significant, as the community scorecard is more than a simple survey or set of indicators. It is a participatory process in which citizens and service providers come together, first to identify and score problems using their native language, and second to agree on joint action plans to solve these problems. Having agreement on data standards is important, as it enabled the partners to share data such that they can use it collectively and thus achieve greater collective impact.
From June to October 2017, Kwantu worked with CESC and N’weti to create an app that embeds the data standards into the user interface that enables field staff to easily input data. The app, which evolved based on the range of organizations’ needs, also includes a workflow for managers to review and improve data as well as tools for aggregating and exporting data for analysis.
Configured using our BetterData platform, we created a series of apps, each of which functions differently according to the need. Each app includes the data objects for each data standard. From one perspective, the data objects are simply forms, allowing field teams to enter information in a standardized, comparable way. However, the data objects also provide a common data format, designed to make data interchange simpler. CESC and N’weti use these apps in their own BetterData communities.
The apps also contain forms that collect other data according to each organizations’ needs. This provides a private space in which field teams and program managers can collect and access data. However, since much of this data is collected using the same data standards, they have the option to share for collective use. These data governance tools are basic at the moment, but can be scaled as we understand better the kind of control that data owners need to manage how they share their data with others.
Data sharing enables better problem solving for citizens’ issues related to quality of service delivery. While community scorecards are effective at solving some problems locally (for example, issues related to management), other problems are more complex or may require collective action at higher levels. In this case, partners need to see which categories of issues are not yet solved across different facilities and even over different cycles. This kind of analysis helps reveal persistent problems, and helps determine where collective advocacy or further research is needed.
This work demonstrates the feasibility to agree on data standards for a specific tool that is widely used. How can we scale this to build a more connected data ecosystem?
The following reflections and recommendations emerged from this work:
Be demand-led. In this example, the motivation for joining up and aggregating data is to solve problems affecting service delivery. Defining and agreeing on this need clearly sets the scope for the data standards necessary to deliver this objective. It also clarifies which people and organizations need to be involved in the process and how the data will be used. We’re working on another project that will use the same approach to join up data across the South African government. In this case, the driver is around data to report on progress in relation to the national development plan.
Define data standards at the operational or administrative level. This is important for three reasons. First, it’s easier to reach agreement on what the data standards should be (contrary to what you might expect). Second, it’s easier to assess and validate the quality of the data. Third, the data is already as disaggregated as possible.
Support data import and data integration via a shared Data Registry. In the Mozambique example, both organizations are using the same software to collect data. This clearly is not possible or even desirable in all cases. There may be small organizations that don’t want the overhead of using a database. Other organizations may already be committed to creating and using a different database. For these reasons, we’ve created a shared Data Registry. This provides a space to share data standards and allows different organizations to share data they collect for these standards.
For a small organization, this might mean collecting data on paper, then uploading it using an XLS or CSV template matching one of the data standards. For a larger organization it might mean mapping data in an existing database to the data standard, transforming it into the data object format (JSON), and using the API to push it into the Data Registry.
For other organizations it’s as simple as adopting an app that uses the relevant data standards (like the community scorecard app) into a BetterData community and then using it to collect data.
Consider the data governance rules. Clarity and trust on how data will be used is critical. Further work is needed to agree rules on collective governance of shared data. The mechanism for sharing data (in our case the Data Registry) needs strong security, good encryption, and effective controls to enable different groups to apply their own data governance policies.
There is potential for applying this model to build data ecosystems that better connect governments’ administrative data. This is data that all governments collect as they manage the delivery of services to citizens. It is a potentially rich source of data, but is also highly fragmented. During 2016 and 2017, Kwantu collaborated with the Department of Planning, Monitoring and Evaluation in South Africa to assess the feasibility. We are now discussing a pilot that will define the data standards needed to report against the national development plan and explore how to link these with existing administrative systems.
This approach can also be used to connect data collected by NGOs and others that are using similar tools or approaches in their work. In this example, we examined one tool from the social accountability sector, but at Kwantu, we’ve worked with a wide range of NGOs operating in many different areas, and see many similarities in the kind of data that they collect and the potential to develop shared data standards. Much of our work, including our involvement in the Everyone Counts data collaborative, seeks to extend and scale these approaches and findings to reach the wider social accountability sector.