WASHINGTON, March 18, 2013 – Big data was the buzzword over the weekend at the World Bank. More than 150 topic experts, data scientists, civic “hackers,” civil society groups and development practitioners gathered for the DC Big Data Exploration event to analyze data and tackle questions related to eliminating poverty and rooting out corruption and fraud on development projects.
The Bank’s foray into big data comes as no surprise to those familiar with its commitment to transparency and accountability. “The maturity and growth of the Bank’s open data program has been phenomenal,” said Chuck McDonough, the Bank’s Vice President and Controller, “Expanding our awareness to how big data can help us combine our data with that of others to become constructive knowledge is fundamental to becoming a true development solutions bank.”
The data dive was a collaborative effort between the World Bank and external partners UNDP, UNDB, UN Global Pulse, and Qatar Computing Research Institute (QCRI). Participants worked on about eight projects that had been predetermined by the World Bank’s Open Finances team and DataKind, a nonprofit organization that partners with civil society organizations on the intersection between data and development.
“The model for greater collaboration across different organizations is imperative,” said Patrick Meier, Director of Social Innovation, QCRI. “That doesn’t mean that collaboration is easy to do. It’s hard to collaborate across organizations. It’s an iterative process, and we’re learning by doing, which is what we have to do. We don’t have the luxury of sitting back and waiting.”
With just 24 hours to evaluate the data, each team was tasked with posing new questions rather than trying to come up with concrete answers or solutions. Combining the use of visualizations, graphs, and new datasets based on World Bank and external data, teams comprised of data scientists, subject matter experts, and just curious participants that did not fall in either camp formed naturally to collaborate on and evaluate poverty measurement and anti-corruption projects. On Sunday morning, each team presented their findings to all in attendance.
"One of the major challenges to ending global poverty is reducing corruption risks," said Stephen Zimmermann, Director of Operations at the World Bank Integrity Vice Presidency. "This is a daunting task but the work this weekend identified a huge number of innovative possibilities to significantly expand how we can use data to tackle corruption. I’m so excited by the opportunities lying ahead of us.”
Unearthing data sources on food prices and consumption of bananas and rice in Kenya for faster and easier estimates of inflation was one of the projects. Another included analysis, collection, and response rate of poverty data from mobile devices vis-à-vis face-to-face interviews. A reverse Google tool for anti-corruption that tells you what to look for based on what you have was also featured. Throughout the event, an anti-hassle czar roamed the room as part therapist, part journalist, logging the frustrations about missing data from each participant, thus setting a framework for the development of future data sets for analysis.
“It’s important to recognize this is a long process and one of the things we’re hearing a lot about is the associations, correlations and relationships among different types of data,” said Peter F. Lanjouw, Research Manager of the Poverty Group in the World Bank’s Development Economics Research Group. “It’s deeply important that we not only understand the correlation but also the causal links between different types of data that we’re looking at.”
There is little doubt inside or outside of the World Bank or from civil society organizations of big data’s potential and that it is becoming more mainstream. Apart from trying to forge stronger collaborative efforts, the challenge of sustaining and building momentum from the weekend data dives is top of mind.
“What really needs to happen after these weekend events is continuity of those collaborative efforts,” said Jake Porway, Executive Director of DataKind. “We need to make sure the people who have the ability to work with data can connect with the those who actually know the problems so that we’re not just hacking on open data for fun in our free time. We want this to be the beginning of something much, much greater than this.”