Data Collaboratives

Check out the agenda, presentation and storify for this amazing workshop organized by the ISI Foundation in Milan, 2016.

Data Collaboratives to Improve Children’s Lives

As public problems grow in complexity and increasingly require new insights, decision-makers both inside and outside government have begun exploring ways to be more data-driven and collaborative. Several of society’s greatest challenges—from addressing climate change to achieving the Sustainable Development Goals—require greater access to data, ability to analyze particular kinds of datasets, and collaboration between public- and private-sector entities.

However, much of the most useful, timely and comprehensive data resides with the private sector—in the form of, for instance, Web clicks, online purchases, market research, sensor data, and data generated from the use of mobile phones . With consumers connected to more and more platforms as well as the increasing prevalence of sensing technologies (i.e. the Internet of Things), data on how people and societies behave is becoming even more privately owned. Today, companies are exploring ways to make such  data available for the public good, as a form of corporate social responsibility – also, often called data philanthropy.

Today, the GovLab and UNICEF, in collaboration with the UN Global Pulse,  announce a new partnership to leverage the potential of private sector  data to improve children’s lives through the study and creation of “data collaboratives.” Data collaboratives are a new form of public-private partnership in which participants from different sectors — including private companies, research institutions, and government agencies — exchange data to help solve public problems. To accelerate solutions to the problems UNICEF works on and to bolster UNICEF’s efforts to become more open and data-driven, the GovLab will help UNICEF and the UN Global Pulse identify and craft data philanthropy collaborations with private sector companies in support of UNICEF’s mission.

Despite an increased awareness and experimentation in establishing data collaboratives, there exists little consensus about best practices, and only a provisional understanding of how, precisely, data can be shared and used to enhance the public good. In particular, companies and international organizations may still have limited knowledge about how to maximize the benefits of data sharing while minimizing its associated risks, such as potential threats to privacy and competition.

The 18-month initiative being launched today comprises a number of activities aimed at increasing insight on current data collaboratives practice, what works and what doesn’t, how to share data in a trusted manner using data governance frameworks, and what steps and conditions must be in place in order to ensure the exchange of value. In the coming weeks and months, the GovLab, UNICEF and the UN Global Pulse will collaboratively develop and share new resources and tools aimed at: mapping the current data collaboratives ecosystem, articulating policies and frameworks for responsibly sharing data for the public good, and helping practitioners operationalize this next generation of public-private partnership to solve big public problems, including but not limited to improving the lives of children around the world.

Short interview from the 2016 Doing Development Differently Conference (London)

 

The Future of Data

In 1899, Charles H. Duell, Commissioner at the US. Office of Patents, declared: “Everything that can be invented has been invented.”

Good old Charles won’t go down in history alone for his poor forecasting skills. He will be joined by many others who, in their shortsightedness or overly optimistic hopes, have dared to re-imagine the future. But if guessing the future can be so problematic, why even bother?

That’s the question I posed myself when trying to conceptualize the panel, ‘The Future of Data,’ at UNICEF’s Global Innovation Summit, which took place in Helsinki on 9-10 November, 2015. The last thing I wanted was to end up with a session full of stereotypical versions of the future: flying cars, rover boards, and self-lacing shoes (yes, I’m talking about you, Back to the Future II).

The ‘future’ (aka, 2015) according to the movie, Back to the Future II (@Wikia).

Instead of looking at the future, my partner-in-crime (Manuel) and I focused instead on how to get there. What are the entry points, the opportunities, and the potential bottlenecks that would promote or hinder our trajectory towards the future of data?

To get us there, we started with an important element of any breakthrough: diversity. Our panel was composed of amazing thought-leaders from government, international organization, and private sector – big and small. The goal was not to reach consensus but induce a clash of different perspectives. And in spite of differences, some common data trends began to emerge:

From right to left: Eduardo Clark (Director of Mexico’s Data for Development); Kenth Engo-Mosen (Telenor’s Sr. Researcher/Data Scientist); Anoush Tatevossian (Global Pulse’s Lead Strategic Partnership and Communication); Carina Szpilka (former CEO, ING Director Spain/France); Jan Erick Solem (CEO, Mapillary); Natalia Adler (UNICEF’s Data, Research & Policy Planning Specialist).

Data is no longer the stuff of data geeks alone

We have been accustomed to a division of labor when it comes to data. At UNICEF, for example, you have the Monitoring & Evaluation (M&E) geeks on one side and the program folks on the other. In governments, you have national statistics offices and lines ministries.

The division, while necessary due to technical requirements, is not always helpful. You may end up with situations where data is generated without a clear specific purpose, and then ultimately not used.

Joint identification of problems

To create purpose-driven data, we need data and program folks to sit together and identify the problems they’re trying to tackle in the first place. Then the question should be: do we have data for that? If not, let’s generate that data so that it can be used to help us solve a specific problem.

3ecfbb64f06e814b866b11b0dd09f74b

This is not a chicken-and-egg scenario (i.e. without the data, how do we know where the problem is?). We don’t start from a clean slate. We live in a world of information overload. So the goal is to distill from that information what we can to help us understand problems (affecting children) and identify ways in which data can be leveraged to help us solve those problems.

Win-win collaborations

The data revolution is showing us that data can come from many different sources. Back in the day, the generation of data was the primary responsibility of governments. Now data is everywhere. It would be a mistake to think that a single institution could tap on all data sources.

Instead, the goal is to form collaborations with different data actors to leverage different data ecosystems to create value for children (i.e. data collaboratives). To get there, we need more efforts to document how the benefits of data collaborations outweigh their risks.

Open data

The lack of data openness (beyond issues of transparency) affects both private and public sector. There’s a need for more concerted work around this issue and that should start by getting our own house in order.

 Regulatory frameworks

There’s an urgent need for the UN to begin crafting the basis of a framework that guides data access/sharing, privacy issues, etc, to keep up with the changing data landscape.

 Citizen-generated data

As the volume of citizen-generated data increases, the public sector needs to become much more inclusive of data coming from citizens. This is perhaps one of the greatest differences between public and private sectors in the data revolution. Whereas, the private sector is going into lengths to get data from customers, the public sector still views the ‘messiness’ or lack of control of ‘bottom-up data’ with skepticism.

This is obviously not always the case. There are interesting examples out there of countries ‘giving individuals and communities greater control over their data.’

Conclusion

For a session that lasted a bit over an hour, I think we made some progress.  It’s not a recipe for a ‘data god,’ as one participant hoped for.  But it’s the beginning. Or at least, a solid baseline.