It is an acxiom of journalism that it speaks truth to power. Increasingly, that truth is being identified in data, through data journalism as well as citizen activism. By assembling the evidence around an issue, it becomes possible to demonstrate if harm is being done.
Climate change is a prime example of this. Since the election of President Trump, federal agencies have been mandated to remove references to climate change on their web sites and replace them with messages about sustainability. Even US government-funded research and data that provide proof of the way humans are affecting the environment are being removed.
The Environmental Data and Governance Initiative (EDGI) is a prime example of how data can now be used to hold the powerful to account. “EDGI was founded the night Trump got elected. Its founders recognised that because of what happened in Canada, where Prime Minister Stephen Harper took down climate change content, given Trump is not a believer in climate change, that similar take-downs would be coming along the pike,” explained Erin Akred, data science manager at DataKind, an organisation which has the mission of “harnessing the power of data science in the service of humanity”, when she spoke to DataIQ in mid-December.
It does this by bringing together teams of skilled volunteers with not-for-profits and social enterprises who have data-oriented challenges they want to resolve, but not the resources to tackle them. Akred is the facilitator for this, having been the first member of the newly-created in-house team in 2015.
EDGI is an example of the large-scale projects which DataKind is now able to engage with as a result of having a dedicated resource alongside the external volunteers it also musters. DataKind has been able to apply machine learning to the data which EDGI has been collating in order to create a classification for the type of changes and take-downs which have been taking place. “EDGI just came out with a report which has been getting op-eds in the mainstream media, which is very exciting,” noted Akred. She will be using the initiative as a case study when she presents at the DataFest 2018 Data Summit on 22nd to 23rd March in Edinburgh. DataIQ is the media partner for the event.
Akred will explain the six-point framework which DataKind has put together to evaluate projects brought to it by NFPs. “It recognises that it is not just about collecting data and handing over a server to a team of volunteers - there are other components that need to be in place if a project is going to succeed,” she said.
This is one marker of the progress made by the organisations since Akred joined when it was “a super-small organisation” to the point where it can now field its own data scientists. The goal has been to scale up the nature of the projects it can engage with, as well as to make the process more efficient.
Another example she will provide at the event is work she has been doing with the Natural Resource Governance Institute (NRGI), a not-for-profit policy and advocacy centre which focuses on informing decision makers about the extraction industries and their environmental impact. “They are mostly lawyers looking at the extraction industry in the developing world and whether the contracts set up are fair,” said Akred. Using machine learning and natural language processing, NRGI is able to identify any areas of concern that it wants to lobby around.
For Akred, the EDGI project “is perfect because it is dear to my heart.” She joined DataKind from federal government, where she was working on informing voters about the candidates on the ballot paper and where also, “I was on the providing side of the open data world, publishing data and encouraging government to put data out.” A combination of social action and data lockdown could almost have been designed to grab her attention.
But just as the scope of projects DataKind is now able to undertake has grown, so have some of the challenges. When Akred spoke to DataIQ, she had just finished twisting the arm of a volunteer data scientist to get them to stay on a project. As much as anything else, this reflects how much more in-demand within their own organisations the very volunteer resource DataKind relies on has become. But it also shows that volunteering for the organisation can now involve more commitment than just signing up for the sort of weekend hackathon which it started out by running (and still continues to operate).
“From my point of view, we are now seeing people at the beginning of a moment of using data for philanthropy,” said Akred. Her session in Edinburgh will be an ideal chance to see this in action. And if her surname looks familiar, it may be because husband John Akred, founder of Silicon Valley Data Science, spoke at the same event last year. So how does she view the challenge of following his debut? “It’s a switch-up.”