Big data for big impact: harnessing novel data to measure global goals

If you read about Big Data in the news, you’re often hearing about ‘business opportunities’ or ‘personalisation’: new ways for big corporations to help us “buy s**t we don’t need’ (Durden 1999). So it’s great to hear someone talk about making a difference with big data. That someone was Andy Tatem.

I was at the Royal Statistical Society last week to hear Andy speak about mapping progress towards sustainable development. He talked about how we can harness novel data sources to measure the where and what of the Sustainable Development Goals (SDGs). The fact he’s a fellow geographer and visualiser made it extra sweet.

Better data resolution to enact UN resolutions

Andy outlined our basic problem with measuring sustainable development actions: the reliance on census data. These data form many of the denominators (almost every indicator is expressed per capita) or numerators (number of children, income, location etc). However, because a census is an irregular activity in many countries, they can often be outdated and inaccurate, like in the case of Afghanistan whose last census was in 1979, or Madagascar, which is just completing its first census for 23 years.

This data may also not be that helpful if we’re trying to plan activities that’ll lead us to a sustainable world. Unless you’re very fortunate with timing, most countries will only conduct one census between now and the 2030 deadline for achieving the SDGs, which won’t be useful for the ongoing monitoring, evaluation and planning we need to do. In the UK for example, our next census will be in 2021 and the results probably won’t come out until 2024 or later. If we’re planning sustainable energy infrastructure or low-carbon transport, we need data now, and every year or so, to plan, act and re-prioritise.

You can detect forest loss in satellite images. Global Forest Watch visualised that data for the first time, so people can better understand the state of the world’s forests and take action to prevent deforestation.

Andy’s solution is to use the data that humans generate every day — through phones, credit card transactions or satellite data — to fill in those gaps. He went through a huge range of examples, like analysing seasonal migration through mobile phone GeoID data and improving population estimates from the amount of light emitted at night by human settlements. Already the data have been applied to improving health facilities, reducing disease transmission and alleviating poverty.

And there’s high demand for these data too. His group is working with an astounding range of people: governments all over the world plus a hatful of NGOs and UN organisations. What makes Andy’s interventions effective is the collaborative approach he takes with each organisation. Rather than creating new data for intellectual curiosity or a new paper, he asks people what questions they seek answers to — such as how to track disease hotspots or estimate population density — and works out how to solve them with the data that’s already available to us. By collaborating we are able to identify the right data to answer our priority questions.

Visualising anonymised credit card transactions reveals how tourists behave in Spain and sparked conversations about the state of Spain’s economy.

From stats offices to story-telling

Identifying the right data isn’t the end of the story though. It has to be presented in a way that people can use and understand. Data visualisation is a great way of doing this and the novel data that are emerging are prime candidates for visualisation.

Of course, a lot of these data are much more fine-grain than before, both spatially and temporally, and the software you build to crunch it all has to be optimised for fast, intuitive viewing on a range of devices. For applications such as Global Forest Watch, the high precision of points is its main strength and the software we created was able to showcase this to its full effect. When you make this great data accessible, interactive and understandable, people can see what’s going on in their world for the first time. For example, the visualisation we built with BBVA last year sparked discussions about the state of the Spanish economy and the impact of tourism on different areas.

By working together and making the most of the data we have, we can find solutions to our most pressing problems, including poverty, infant mortality and inequality. The Sustainable Development Goals are ambitious but we have the data and the technology we need to succeed.

Hope you enjoyed this; if you did, please recommend it so other people like you can find this article! Any comments or questions? Fire away.

If you want more content like this, why not sign up to our newsletter?