The Value of Data for Philanthropy
The field of philanthropy is in constant search of the next big thing. And rightly so – our organizations must use limited resources as wisely as possible to try to tackle some of society’s most challenging problems. In recent years, we have heard a great deal about how new and sophisticated understanding of how to interpret the onslaught of data produced in the modern age could help us turn the corner on major social and environmental problems.
This was the topic of a recent panel discussion I participated in at the annual conference of the European Foundation Center, organized by Fondazione CRT and the ISI Foundation. My fellow panelists, Malte Beyer-Katzenberger (European Commission), Stefaan Verhulst (GovLab), and I were asked to share success stories and what’s needed to unlock the value of data for philanthropy. The conversation revealed how far we have come, and also what the future could hold.
In reality, foundations have been using data successfully for decades – for monitoring and evaluation purposes or to inform program development and management. At the Rockefeller Foundation, I led an extensive data-driven process to identify promising new opportunities for impact at the Rockefeller Foundation. For example, we aimed to uncover white spaces for intervention by identifying the gaps between current funding data and the evidence on the actual need for support around critical problems. Foundations have also used data to mobilize actors around a common goal – funding indices and research to inspire movements and inform policy.
While we have been using data to improve philanthropy for years, a great deal has changed very quickly. Here are a few examples:
- The volume of data has exploded
For example, Crisis Text Line, which provides online support to people in crisis, received a total of 8 million text messages in the first two years of its existence between 2013 and 2015. Only 3 years later, in 2018 the organization has received a total of 75 million messages. To manage what’s become one of the largest data sets on mental health and crisis in the world, Crisis Text Line had to be built from the ground up around technology and data.
- We can now collect, store, and analyze many more types of data than ever before
Only a short while ago, mobile phones, computers, and servers started talking to each other, sharing files, photos, videos, and social media posts. Today, some of our most basic household items, like light bulbs, refrigerators, and watches are exchanging vast amounts of information about almost every aspect of our lives – when we sleep, what we eat, and how many steps we take. Whether we like it or not, this Internet of Things is the new reality. Our responsibility is to use this information to improve people’s lives, not invade them. For example, the Michael J. Fox Foundation is testing a watch-type wearable device in Australia to continuously monitor the symptoms of patients with Parkinson’s disease. This is important because unlike diabetes or high blood pressure we don’t yet have clear metrics for Parkinson’s. The hope is that these sensor data will help establish those reference targets to better manage a fluctuating chronic condition.
- We now have the flexibility to combine different types of data
It’s now much easier to combine and analyze data from different sources to help us make inferences where more direct data may not exist. For example, by combining images taken by satellites, social media comments, and information from call data records or census data, officials can make much better decisions in humanitarian emergencies and other crises.
- New data is being generated at lightning speed, even in developing regions
In the past, in low-income countries, in particular, we have often only had access to survey data that is several years old, and we have had minimal data to compare it to over time. Now, we can combine things like mobile phone data, satellite images, or social media to get rapid feedback on whether particular projects are working. Much faster feedback loops can also have other practical implications: one NGO we spoke with wanted to send results back to teams in the field before they moved into areas with unreliable internet reception.
By analyzing large volumes of data across different sources, we can find new ways to understand and tackle the complex problems foundations typically aim to address. Here are a few more examples:
- UNICEF and GovLab are using data to address childhood obesity in Scotland. To understand the complex causes of obesity and design appropriate interventions, the project plans to use a wide range of sources such as shopper data, tv ads, online gaming, the availability of open space for children to play, and data from school lunch suppliers.
- Data analytics and machine learning can help organizations to automate tasks in areas like fundraising or program management, among others, and thus free up needed time and money for other activities.
- There are legitimate concerns about the inherent biases of machine learning algorithms. However, there are also examples that show that observational data can be extremely powerful. For example, researchers from Berkeley used mobile data to predict poverty and wealth of individuals or microregions in Rwanda at a time when measuring poverty in Africa remains a challenge. In 2012, only 25 of the region’s 48 countries had conducted at least two surveys over the past decade to track poverty.
These are just a few examples of the many important uses of data to improve people’s lives. But there are also challenges that we will need to overcome. Indeed, as Stefaan Verhulst, co-founder of GovLab, says: “An increased use of data is not risk-free. At the same time, it is equally important to assess the costs of not using data for fear of the risks involved.” Nevertheless, while data can be an important tool in addressing issues like poverty, obesity, or health, there are legitimate concerns that it can be misused. We need to find ways to guarantee people’s privacy and to make sure that data doesn’t fall into the wrong hands. While this is a tall order, failing to do so threatens to undo many of the advances we have already made.
We also know that many organizations simply do not have the technical ability to use data analytics or machine learning effectively. The Cloudera Foundation is working to address this by providing software, technical expertise, and funding to organizations who could stand to benefit. This is just a start, but we believe that it is an important one.
In a field that is far from mature, we need to be aware that we are learning together and in many cases paving the way for new approaches. The willingness to fail and iterate has to be built into projects from the start. And while big data can provide us with much more information on people’s behaviors and preferences than ever before, we shouldn’t take it as a substitute for actively engaging people and communities in decision making.
In the end, it will be this combination – of people, ideas, and the effective use of data – that will help us to take philanthropy to the next level.