The nongovernmental organizations Jack Cornforth works with don’t have the capacity to generate and use data in constructive ways — let alone share it.
Cornforth, the coordinator for Civicus’ DataShift initiative, has been tailoring capacity building programs based on data problems at civil society organizations in Argentina, Tanzania, Kenya and Nepal for the past two years. Those problems range from a lack of infrastructure — such as internet connectivity — to language translation capacity, to political and legislative barriers.
Despite these challenges, the call for open data and increased data sharing within the development community has never been louder, nor have as many institutions and initiatives been devoted to answering it.
The World Bank’s Open Data Initiative catalogues available World Bank datasets; the Global Partnership for Sustainable Development Data works to establish trust in the “booming data ecosystems” of the 21st century; and the International Aid Transparency Initiative seeks to make information about aid spending easier to access and understand — to name just a few.
There is a data revolution underway — but whether individual NGOs are “doing enough” to share that data is still subject to debate. Cash-strapped NGOs with limited capacity are working toward more effective data collection and usage, but they’re also likely generating data sets every year that aren’t being fully analyzed or shared. Large, data scientist-equipped INGOs are subscribing to increasing calls for transparency, but those in the data space agree that many questions still trump whether NGOs are doing enough to share — like “whose data is it anyway?”
A long to-do list before sharing
Data can’t simply be about technology or databases or data sets or sharing. It’s also always tied to politics, quality and privacy.
When trying to determine why an organization is or isn’t sharing data, a look at the politics in the region is a good place to start, Serge Kapto, United Nations Development Program policy specialist on data for development, told Devex.
Kapto is leading the Data Revolution Ecosystems Mapping project for UNDP, which is examining the national statistical capacity and infrastructure requirements to improve collection and the use of data in Bangladesh, Mongolia, Moldova, Senegal, Swaziland, and Trinidad and Tobago — countries chosen based on their varied range of engagement in the data revolution.
The assessment seeks to extract trends common to all countries, as well as pick out those specific to certain contexts. Many of the common barriers to data sharing echo those found in DataShift’s work: infrastructure and political will.
In Swaziland, for example, where there’s no official recognition of civil society, just acknowledging that the sector could contribute data would be a step forward. Sharing it successfully, then, is still a distant dream.
“You’re not going to solve that from data point of view,” Kapto said. “These are systemic issues within countries that have to be addressed.”
Data quality also comes into question long before data sharing. Official statistics offices are often concerned that other stakeholders, such as civil society organizations, who want to become players in a data revolution won’t be able to meet quality data standards, Kapto said.
This concern isn’t unique to statistics offices. There are inherent problems that come with making “evidence-based” decisions on the basis of data that lacks a clear methodology. Even for the International Aid Transparency Initiative — which has seen the number of organizations publishing data rise from 210 in 2014 to 473 in 2016 — the biggest barrier to increased data usage remains concerns about quality.
“Some people are naturally hesitant about using other data sets,” said Roderick Besseling, former open data coordinator and current digital strategist for Netherlands-based development organization Cordaid. “We can’t just rely on data sets without context.”
The quality of the data reported to IATI needs to improve, he said, and he expects to see this happen as an increasing number of donors make data reporting mandatory. In May, donors and international aid organizations were asked to publish their data to the IATI standard by May 2018 as part of the “grand bargain” on humanitarian action. Two weeks ago, the U.S. Congress approved the Foreign Aid Transparency and Accountability Act, which will require U.S. government agencies to improve transparency by publicly sharing data about what’s working and what’s not.
Still, more data or more data sharing alone isn’t always a positive, noted Amy O’Donnell, an adviser on applications of information communications technologies to support programming atOxfam GB.
First, it’s crucial to know how to treat the people that data is about with respect, she said.
In April 2015, Oxfam published its first Responsible Data Policy to identify how to properly gain informed consent — by explaining what it means to have your data or photo shared, and how to collect and analyze data and decide its afterlife in a responsible way.
When multiple organizations are asking similar questions about people’s humanitarian needs in a fragile context, for example, it makes sense to collect the data in similar formats, O’Donnell said.
“But do the community members who are being interviewed by UNHCR representatives, for example, realize it will be shared with different organizations for different reasons? These are the questions we need to be asking.”
So who’s sharing?
While it is a long road to responsible, successful data sharing some organizations are already sharing data or benefitting from open data sources.
“I don’t think groups are holding back data,” Kate Schecter, president and CEO of nonprofitWorld Neighbors, told Devex. “I think it’s more about figuring out how to collaborate with each other, how to utilize each other’s knowledge and unique skill sets.”
As a small organization, World Neighbors “can’t validate data what someone else did, but we can make sure that the work that we’re doing is clean and reliable,” she said. And the same goal touted by the Oklahoma City-based nonprofit — more consistent indicators for their programs around the world, and more streamlined data collection in general — is shared by many others.
“The whole idea of IATI is you publish information, you make it transparent, it’s out there, and in theory an Oxfam or CARE or Save the Children can access and use Cordaid’s data,” Besseling told Devex.
IATI’s 2015 report said that the overall use of the data being published isn’t yet enough to drive development outcomes.
“We are in danger of being caught in a ‘vicious cycle’ where publishers have no incentive to improve quality because no one is using their data, and users can’t use the data because it isn’t up to scratch,” John Adams, chair of the IATI Technical Advisory Group and head of business innovation at the U.K. Department for International Development, wrote in the report.
But Cordaid and a handful of others are making use of it.
In the past, creating a country analysis would involve research and reading policy papers “to create a static country overview analysis that would be out of date by the time you finished typing it,” Besseling said. Now, Cordaid is using data published to IATI to create dynamic data boards to watch which funders are active in a country, as well as which organizations are active in implementing countries.
This is also where data literacy — and an organization’s confidence in using data— is key, according to both Besseling and UNDP’s Kapto.
“People have to meet each other midway,” Kapto told Devex. “Right now, statisticians don’t understand the language of development … data stakeholders don’t understand the language of statisticians … they have to come together.”
One way the Global Partnership for Sustainable Development Data is addressing the data literacy gap is through the Data4SDGs Toolbox, which collects all the resources that members of the partnership have developed, or projects that have worked well, in one place so that a government can chart a road map that outlines exactly how they’re going to effectively monitor the Sustainable Development Goals.
The partnership is also looking at building a data marketplace, expected to launch in September, that would pair government, donors and others at the data “demand” end with groups producing data. It would help address the current imbalance of supply and demand of data and combat the current trend that a lot of “unorganized, wonderful data is being produced but not necessarily used,” Cornforth said.
Still, the NGO community is far from having pieced together every part of the data sharing puzzle.
“I think a lot of emphasis is recently shifting from questions of ‘Who is sharing data?’ to
‘Whose data is it anyway?’ ‘What ownership can they take over sharing and utilization of that data?’ ‘Share for what purpose and with whom?’” O’Donnell said.
Right now, it seems as though many NGOs are sharing data as their capacity and the local infrastructure allows. But take Kapto’s would-be culture of all stakeholders’ widespread embrace of open data sets and place it within O’Donnell’s framework for commitment to privacy and consent practices. Then add Besseling’s vision for an increasingly data-literate NGO workforce. Perhaps then the development community would be on the road to doing enough to share data.