The Science and Ethics of Data for Good’s New Frontier
Combining data from multiple sources can yield the multi-dimensional insights that policy challenges really require. Here’s a look at the changemakers shaping this new frontier of Data for Good.
As we’ve seen in previous posts, a single donation of data can reveal an entire dimension of a public health or development challenge that’s never before been seen. If data donated from just one source can open a whole new angle for public policy, imagine the kind of big picture understanding researchers could piece together with data from multiple sources—or even from multiple sectors. That’s the promise of combining data from a variety of private and public organizations. “In general, data combining has an incredible upside potential,” says Claudia Perlich, Chief Scientist at Dstillery and adjunct at NYU’s Stern School of Business where she teaches a course on data mining. As an example, Perlich cites how, taken separately, education records and medical data might each tell one story, but together can reveal how attendance issues are actually the result of a pressing issue in public health. “The more complete picture you have, the better your chances are of adequately interacting with people,” Perlich said.
To see how big a difference this can make, look no further than the pioneering work of UN Global Pulse, a flagship innovation initiative of the United Nations Secretary-General. Dedicated to yielding new, comprehensive insights into human risks and resilience in the face of on global challenges, UN Global Pulse has led projects like combining official statistics with social media data to show correlations with spikes of food and fuel prices. Now, as part of its Data for Climate Action Challenge, UN Global Pulse is calling on private companies in the finance, energy and communications sectors to donate anonymized, aggregated data sets that can be used for analysis in projects that seek to understand how people are adapting to climate change. By combining data from different sectors and geographies, it is possible to gain a better understanding of how people see risk and evaluate economic opportunities, which, in turn, can aid in developing better-targeted solutions for local contexts in the face of climate change.
Yet the potential of data combining to offer a comprehensive, multi-dimensional view on an issue also presents a risk: exposing individuals to identification, even when their data have been anonymized by the organizations donating them. To understand why, think of the old game Guess Who?: One identifying characteristic might not narrow the field very much, but each detail narrows the likelihood of who it might be. Except, of course, real-world privacy is no game—it’s a deeply serious matter that requires rigorous safeguards to ensure that we can maximize data’s potential for good while also maintaining legal and privacy standards.
As a leader in this new frontier of Data for Good, UN Global Pulse is joining with private sector and social sector partners to actively shape an ethical and practical roadmap for navigating such challenges. That’s essential to the evolution of Data for Good as a field, says Mila Romanoff, Legal & Data Privacy Specialist at UN Global Pulse. “A lack of clarity about why, when and how to share big data responsibly can hold back use of the data for justifiable purposes—such as long-term research that could lead to positive economic or social improvements or in emergency situations where use of anonymized metadata could potentially save lives.”
As Robert Kirkpatrick, UN Global Pulse, Director once commented in an interview with SciDev.Net, “Legal frameworks are barely able to keep pace with technology.” In that vacuum, companies and organizations often develop their own practices, and fixes are developed ad hoc. Romanoff stresses the importance of “communities of practice” in a field so new that the landscape is highly fragmented. That’s why progress begins by bringing together diverse representatives— from privacy and legal officers to data engineers and policymakers—as UN Global Pulse does with its multi-disciplinary Data Privacy Advisory Group and regular events such as last month’s workshop on data security and privacy.
Bringing together UN agencies, nonprofits, academics and the private sector, this workshop offered an opportunity to help shape principles to guide the field, with discussions that parsed the differences between data privacy and protection and the meaning of user consent. Discussing the event’s importance, Romanoff commented: “Nuanced ideas and values are rightfully being raised by researchers, innovators. We hope that this type of dialogue will help inform organizational guidelines and data security practices and, eventually, regulatory discourse.”
Consider this a call to arms to engage.