Natalia Adler: Building – and Using – Data Collaboratives to Help Children

Natalia Adler describes herself as a problem-solver, not a data scientist. When she came to work at UNICEF’s New York headquarters after seven years in Nicaragua and Mozambique, she joined the data research and policy division. “I immediately saw that, even though we do all sorts of data work, we were behind on the data science component,” says Adler, now a data, research, and policy manager at UNICEF (The United Nations Children’s Fund). “And if you want to make progress on that front, you need to access to big data – however, most relevant datasets sit with corporations.”

That insight has been the impetus for Adler to work on a project called data collaboratives, which seeks to better understand how private companies can share their data to help the public good. One of her partners in this initiative is Stefaan G. Verhulst, the co-founder and chief research and development officer of The Governance Lab at New York University’s Tandon School of Engineering, better known as The GovLab. Adler, like Verhulst, is a member of this year’s Program Committee for the Data for Good Exchange, which will take place at Bloomberg headquarters in New York on Sunday, September 24, 2017.

Adler says her relationship with Bloomberg has been “fundamental” in helping her think about data collaboratives. In 2015, Bloomberg and UNICEF came together to create a Researcher-in-Residence program. Having additional data science expertise was helpful for UNICEF, but without access to lots of data, the work of the researcher was limited. Adler and colleagues from UNICEF’s Office of Innovation went back to Bloomberg with an unusual request: “can you provide access to financial data that can be used to improve the lives of children?” Everyone involved really wanted the partnership to work, she says, but because it was so new, the negotiations took a while. “We had lawyers trying to figure out, ‘What is data philanthropy? What does it entail?'” she says.

The research team, led by Manuel Garcia Herranz, has used Bloomberg data to map which companies around the world have policies on child labor. Now they’re trying to use network analysis to see if these companies have any influence on the adoption of child labor policies across their supply chain “That kind of information doesn’t exist,” Adler says. “That’s why Bloomberg data can be so powerful when used for social good.”

In addition to sharing data, Bloomberg has also been sharing expertise – including access to members of Bloomberg’s Data Science team, such as Mark Dredze – to help navigate the complexities of this new collaboration. Her hope is that the data collaborative project will eventually serve as a resource for people trying to solve similar problems, so that solutions can be found, and data and expertise shared, more quickly and easily.

Adler says she joined the Program Committee for the Data for Good Exchange because of her close relationship with Bloomberg. She especially credits Gideon Mann, Bloomberg’s head of data science, for his advisory role. Adler has been attending the Data for Good Exchange since 2015, when she first began working with Mann. “It’s been such a learning curve for me and an incredible opportunity to match social problems with data solutions,” she says.

She also says that the Data for Good Exchange offers an opportunity to discuss one aspect of data science that is too often overlooked: ethics. “I think people get so caught up in discussions about the application of data science research that they tend to forget a bit about the ethics component,” she says. “However, for us, this is incredibly important, because we are dealing with children.”