Peter da Costa is a technical advisor to the Hewlett Foundation’s Global Development and Population Program, where Kristen Stelljes works as one of our program officers. -Ed.

Data is a powerful resource. With high-quality, relevant data, decision-makers make better decisions. When citizens know the quality of government services they are receiving, they can use that data to hold providers accountable.   We’ve been learning a lot about this at the Hewlett Foundation through our support to the Service Delivery Indicators (SDI) project, an Africa-wide program implemented by the World Bank that collects data from schools and health facilities every two or three years. What’s unique about this project is that the data is gathered from the perspective of citizens accessing a service (Ruth Levine blogged about the project last year). Since 2011, SDI surveys have been conducted in Tanzania, Senegal, Kenya, Uganda, Nigeria, Togo and Mozambique—with more countries to follow.

But just collecting the data and offering it to the world is no guarantee anyone will do anything with it—to be truly useful, data has to offer specific answers to what the user really needs to know. To see how the SDI results could be used by citizens and civil society, the Hewlett Foundation partnered with Hivos East Africa in Kenya. Hivos came up with the idea of a Small Grants Competition to stimulate usage of the SDI data among civil society, media, and other non-governmental groups in Kenya who were already generators, aggregators, or users of service delivery-related data. The competition, which kicked off in July 2013, sought to open up new spaces for citizen engagement in education and health accountability, ramp up public debate, and shine a light on what helped or hindered effective service provision. Four individuals and nine organizations were selected as winners, and the sheer range of disciplines they hail from is impressive—everyone from data journalists to documentary filmmakers, accountability advocates to graffiti artists.  For example, Seven Thirty Films felt that perspectives from children were missing from the national debate on education and produced a documentary featuring the voices and views of children in 6 regions of Kenya. ‘Crayon Rainbows’ has been widely shown around Kenya, accompanied by exhibitions of the artwork developed by the children during interactive workshops conducted alongside the filming.

It helped that the competition winners had interesting data to draw inspiration from Kenya’s SDI results. The Kenya survey found that while Kenya was doing relatively well on making textbooks, medical equipment, and basic infrastructure available in facilities, the country had a long way to go on achieving optimal performance by teachers and health providers. Just to give you a flavor of the survey results: only 35% of teachers showed mastery of their subjects and less than 60% of public health providers could diagnose four out of five common conditions such as diarrhea with dehydration or malaria with anemia.

Less helpful was the significant delay in the release of the SDI microdata, which meant it wasn’t available when the competition winners started their work. And while the data set was nationally representative and statistically significant, SDI was designed to enable cross-country comparison to stimulate healthy competition between African countries. That meant there was only data available at the county level for 20 of Kenya’s 47 counties—many competition winners didn’t have detailed data available for the areas where they work. Grantees got creative and went searching for data that could fill in the gaps. The Grassroots Development Initiatives Foundation (GRADIF) developed a tool to pull together locally-relevant education data, drawing on SDI, Uwezo’s learning assessment, the National Taxpayers’ Association (NTA) School Report Card, and other data sources, which they channelled to local decision-makers.

What the competition winners were ultimately able to create, and the difficulties they ran into along the way, have taught us a lot about using this kind of data:

  • Representative data needs to be available at the level of accountability. If you limit what you know about the quality of services in your country to the national or provincial level, you won’t have a clear picture of the impact of the facilities in a given county, city, or community. For that, you need hyper-local data, which is rarely available. When it is available, it can be difficult for civil society or citizens to access without supportive freedom of information laws. More resources and technical support are also needed to make sure this level of data is collected regularly.
  • Citizens and civil society should be involved in decisions around what data is collected. As stakeholders in the health and education systems, involving the users in deciding what data gets collected can help ensure the data produced is truly helpful for a wide range of users.
  • Data often hides the reality of marginalized groups. Marginalized counties in Kenya lack data, and the data they do have isn’t routinely broken down by gender or for persons with disabilities. Without this data, we simply don’t know the experience of these groups.
  • There is a lot of data in some areas, almost none in others, and the quality varies considerably. Stakeholders are awash in certain kinds of data –HIV/ AIDS, TB, malaria, and teacher absenteeism, for example—to the point of redundancy, but different data sets on the same topic don’t always agree, raising questions about which source to use. And for other topics, like educational transition rates and maternal mortality for health, data is thin or nonexistent.
  • Data ‘brokers’ are critical to stimulating use. To make data understandable, it often needs to be translated for those who want to use it. Special skills may be needed to format the data or to analyze it effectively, and different ways of communicating the data are necessary so people with different levels of numeracy and literacy can understand the information.

Hivos’ competition clearly showed two things: first, applying creativity, rigor, and perserverance to the data developed through the SDI project can lead to some great (and truly useful) information sharing; and second, we still have a long way to go in ensuring that data collection, sharing, and use reaches its full potential for informing citizens and encouraging more effective service delivery.