Here’s how data science can help solve global crises

Across the world, local officials and NGO workers are heroically responding to overlapping, complex crises – whether they’re battling new waves of the COVID-19 pandemic or delivering aid after natural disasters. The energy and determination of these frontline organizations is unmatched, but their technical capacity can fall short in crucial ways.

In particular, many lack the tools, teams and resources to effectively harness the power of data science. For cutting-edge companies that incorporate data science into all levels of their operations, this is an opportunity to help. For example, Eastern European governments have worked rapidly to welcome a historic exodus of Ukrainians fleeing the Russian invasion. One Polish city turned to Mastercard for help in planning for the influx of the new arrivals. Within a matter of days, our team analyzed regional spending patterns to provide near real-time insights that helped city officials better prepare to meet the needs of exhausted, traumatized families.

But one-off partnerships shouldn’t be our goal. If the private sector and other funders, including foundations and development organizations, work together, we can help NGOs and under-resourced governments build more sophisticated data science infrastructure of their own – strengthening their ability to battle crises and bolstering global resiliency.

That’s why the Mastercard Center for Inclusive Growth, which I run, makes impact data science a top priority. Two years ago at Davos, alongside The Rockefeller Foundation, Mastercard launched data.org, a growing platform that works with organizations across the world to infuse data science into social sector decision making.

It is not easy work to build technical capacity at small NGOs or in far-flung government offices. But three recent examples underscore why it matters.

First, look at Community Solutions, an organization combating homelessness in the United States. Even before the pandemic, homelessness was on the rise across the country and resources were stretched. To improve efficiency, Community Solutions worked with experts to enhance its data analytics capabilities. That work paid off when the pandemic hit and it was able to identify individuals in shelters at high risk for COVID-19 and help move them into safer settings.

As the pandemic spread on the other side of the world, the Togolese government also turned to data science. Togo’s Ministry of Digital Economy and GiveDirectly, a nonprofit that sends cash to people experiencing poverty, launched a pilot programme to quickly distribute cash assistance to the country’s lowest-income residents. They worked with the Center for Effective Global Action to use machine learning and survey data to identify residents and deliver aid. This method reached more of the country’s neediest people, and so far the programme has given nearly $10 million to about 137,000 people.

The third example typifies how front-end investment in impact data science can pay dividends when crises strike. In India, about a third of the food farmers’ produce goes to waste, in part because it’s difficult to determine how long to store particular products and how to extend shelf life. Experts from Switzerland developed an app to help solve both problems, equipping farmers with modeling to help them keep more food fresh for longer. Six months ago, helping rural Indian farmers access data science tools was hardly a global priority. But now, as Russia’s invasion of Ukraine causes a grain shortage that risks sparking a severe food crisis, we can clearly see how investments like this can make a big difference.

Unfortunately, those examples are exceptions. Far too often, organizations and aid workers on the ground are hamstrung by a lack of data, and the tools and personnel to analyze and deploy it.

Too many governments and NGOs have not been able to recruit enough data scientists. A forthcoming report from the McGovern Foundation and data.org estimates that the social sector alone needs 3.5 million more data scientists over the next 10 years. Mastercard just announced a $4.6 million grant to help train 1 million data scientists, with a focus on diversifying the field, but we need to do more.

The lack of analytic tools and internal structure is also a problem – even for organizations that have plenty of data. By contrast, many corporations have spent years developing data and analytics strategies, investing in interconnected, user-friendly platforms, building teams with deep expertise, and cultivating a data-literate culture.

Companies and funders can help bridge this “data science gap” facing governments and NGOs by devoting attention, technical skills and funding. We know that because organizations like DataKind successfully connect volunteer data scientists to social organizations, helping them unlock the power of data in ethical and responsible ways.

The challenge is moving from individual partnerships and pro-bono efforts to building impact data science as a field – at the necessary scale. Last fall, at the request of G7 governments, data.org began convening the private, public and nonprofit sectors to build the Epiverse – an open digital infrastructure that will analyze streams of data from around the world to help prevent the next pandemic.

The broader goal is to strengthen collaboration between sectors so we can use data science to address other social challenges in systematic ways, while prioritizing privacy and security concerns.

No one initiative will help frontline responders build the teams and tools to use data science as fluently and effectively as the best-funded private sector companies. But we can and must elevate our investments. Companies can start by embracing data responsibility principles that include innovation and social impact alongside security and privacy, and then take steps like joining Mastercard to build data.org into an even stronger platform for data science partnerships.

This work may not grab headlines, but it will help the world navigate the crises of tomorrow more equitably and effectively.