"The digital revolution of the first decade of this new century has brought many wonders," writes Robert Kirkpatrick, "yet it has also has ushered in a bewildering array of unanticipated consequences. We now find ourselves in a volatile and hyperconnected world where risk has been globalized." ["Data Philanthropy: Public & Private Sector Data Sharing for Global Resilience," Global Pulse Blog, 16 September 2011] Supply chain and insurance professionals certainly share Kirkpatrick's worldview. They know that events thousands of miles away can have cascading effects that ripple around the globe. Kirkpatrick continues:
"Our tightly-coupled systems have become interdependent, creating feedback loops so complex that predictive analysis is of little use. There is a kind of butterfly effect at work in the world. Crises emerge suddenly without warning, interact to amplify one another's effects, and reverberate across the globe, increasing the burden of poverty, disease, and hunger among the world's most vulnerable communities. All too often, by the time we understand what is happening, it's too late to act."
I'm not certain that it's ever "too late to act"; but, Kirkpatrick's point is still valid. As fellow travelers on planet Earth, we need to become more resilient. He believes part of the answer to becoming more resilient lies in big data analysis. He explains:
"The same technologies that connect us to one another have also turned all of us into prolific producers of data, and this new data may hold the keys to mitigating much of the volatility and uncertainty that now confronts us. We are now swimming in an ocean of digital data, most of which didn't exist even a few years ago. The private sector is already using real-time analysis of this data to understand the changing needs of its customers. One of the defining challenges of the second decade of this century will be for the public sector to learn how to tap into this new 'unnatural resource' to understand the changing needs of citizens and respond with agility."
Understanding and insights are certainly important as we look to future. In an earlier post (Going Long with Big Data), I discussed an article by Samuel Arbesman, a senior scholar at the Ewing Marion Kauffman Foundation and a fellow at the Institute for Quantitative Social Science at Harvard University, who argued that we not only need to tap new data, like Kirkpatrick suggests, but also much older data that takes a long look back in history. Arbesman isn't suggesting that we should steer our ship by looking at its wake; he's simply making the point that insights can come from data both new and old. Kirkpatrick may be concentrating on new data because the thrust of his article is that policymakers and planners need access to big data in order to make the world more resilient. Gaining access to new information, he knows, is much more problematic than gaining access to what Arbesman calls long data. Kirkpatrick explains:
"Today there is a tremendous amount of useful data available online, from user-generated content such as news, blogs, and social media, to the structured data being shared through open data initiatives. Yet increasingly, the real wealth of data out there is what is known as 'massive passive data,' or 'data exhaust.' It's the personal information corporations collect about what products their customers buy and about how they use digital services. It is the digital trails we leave behind, merely by going about our daily lives. It is the data that powers business, which the World Economic Forum has described as a new asset class. ... Analysis of this data exhaust holds an even greater potential than public Web content to help policy makers gain a real-time understanding of human well-being."
Paul Doscher, CEO of LucidWorks, calls data exhaust "dark data" because it often hides in places that are inaccessible by most search engines. ["Searching for Dark Data," SiliconANGLE, 11 February 2013] Kirkpatrick notes that gaining access to this data is not just technically difficult but philosophically and legally difficult. "The businesses that collect this data see the value in maintaining exclusive access to it," he writes, "a real-time understanding of their markets allows them to respond to emerging trends with the agility needed to compete in today." Companies don't easily relinquish their assets. Just as importantly, Kirkpatrick notes, "There are often a host of legal constraints related to protection of privacy and secondary use." There is a reason that private enterprises hold big data dear -- it contains treasure that can be unlocked by analysis. Kirkpatrick's argument is that government policymakers should also hold big data dear and start mining it for golden insights that can make the world a more resilient place. The challenge, he writes, is that "the data that could help give them the additional agility needed to meet the challenges of governance in the 21st century is accumulating behind corporate firewalls. When public policy lags household reality by months or years, particularly in today's dynamic socioeconomic landscape, the consequences for vulnerable communities can be severe." What Kirkpatrick would like companies to do is share their data with governments. I'm sure such a suggestion causes many corporate executives to roll their eyes in disbelief. Kirkpatrick, however, believes data sharing can be done in a way that protects both the proprietary information that companies cherish and the privacy of individuals. He calls his scheme "data philanthropy." He explains:
"While, architecturally, it's not clear what form or forms data philanthropy could take, we like the term as it drives home the message that shared data is a public good. We have begun exploring ideas and research that might uncover a viable framework of how data philanthropy actually functions, i.e., how data generated in the private sector can readily be shared with the public sector. At least two ideas are already being robustly debated: A data commons where some kinds of data are shared publicly after adequate anonymization and aggregation; and an alerting network where more sensitive data that can never be shared publicly is nevertheless analyzed by companies behind their firewalls for specific smoke signals. There are other models emerging as well."
To bolster his case, Kirkpatrick provides a few "threads" of evidence from thought leaders that show progress in this area. He then writes:
"Down the road, it's not impossible to imagine some version of each of these converging in an interesting way, but in the meantime, we need a pragmatic approach that can be applied today in both developing and industrialized economies. Private sector participation is central to this effort, and we are eager to engage with potential partners. What has been a pleasant surprise is that, based on our initial talks, the private sector agrees that what we are proposing is not simply a new form of charity or CSR, but a sensible strategy for business risk mitigation, particularly when investing in emerging markets. ... So how do we take data philanthropy from concept to operational reality? We are talking with a number of business leaders in companies that take a sustainability-oriented, double- or triple-bottom-line view of their market, and as I've indicated data philanthropy hasn't been a hard sell—at least conceptually. They have expressed a willingness to engage with us in a series of collaborative research projects exploring what utility their data might have for giving policy makers a real-time understanding of the well-being of communities on the path of economic development, an important goal for both public and private sectors."
While Kirkpatrick is primarily focusing on data sharing from companies, Prasanna Lal Das and Giulio Quaggiotto are exploring "the growing willingness of individuals to 'donate' personal data for the public good." ["Would you give up your personal data for development?," UNDP blog, 25 February 2013] They ask a series of questions they would like to explore, including:
- Could this be a promising venue to explore for development organizations?
- Should we start a 'donate your data' campaign targeting individuals, rather than corporations?
- Is some type of individual data likely to be more useful/practical to start with?
- Should we perhaps steal a page from the corporate book and follow the example of, say, companies like Fitbit that encourage individuals to share their health data?
- And what about the Quantified Self movement: a potential ally?
- Or is a more appropriate role for us to push for policy change so that we can have open, collaborative trust frameworks between individuals, governments and companies that would encourage the (willing) sharing of personal data?
In the end, they admit, "This approach raises a whole range of questions and challenges – from privacy to prior informed consent, all the way to personal safety and security." Nevertheless, data philanthropy is certainly an idea worth pursuing if we are going to make the world a more resilient place in which to live.