By: Julie Munro
El País is a Spanish-language daily newspaper headquartered in Madrid, Spain, that also focuses on topical issues in Brazil and in other parts of Latin America. Over the past year, they’ve built a Data Journalism team from the ground up and have made storytelling through maps a central component of this work. We talked to Daniele Grasso, a founding member of this team about the work that he and Kiko Llaneras (a writer and engineer at El País) are doing to tell stories using Mapbox tools.
What inspired you and your team to incorporate maps into your articles?
We’ve always included maps into the print version of El País, though as you can imagine they were not very elaborate and we could only show so much data. Within the last few years, we started to experiment with Mapbox and realized what an interactive experience we could create for our digital readers. In one of the first articles published on El País using Mapbox, my colleague Kiko Llaneras built a very detailed interactive map showing election data after Brasil’s general election in 2018. It was the most-read article of El País’s coverage of those elections and helped us understand that readers can more easily relate to the news when they have geographical context. In this case, they could see how different regions in Brasil voted differently. In other stories we’ve done they can see whether their area is disproportionately impacted by pollution or what the average salary is in their neighborhood relative to those around them. Essentially, including a map can be a way to speak to them directly to our readers about their own lives.
Separately, there’s also an important design aspect. You typically find photos on the front page of newspaper digital editions or on the social media stream. A lot of these are stock photos, and a colorful and informative map is more engaging and more likely to scroll down the page or click through. (Especially gifs made from the maps!)
How did you build your maps, and what are you particularly excited about?
The first thing we need when building a new map is data. Procuring data is challenging and very time consuming. When we set out to create a map with results of the Spanish election in April 2019, there were 8,200 municipalities and more than 35,000 polling stations. We had to do some detective work to understand who even had the election data for every polling station. The central ministry doesn’t publish this data until two months after the election, so three people on our team spent three days calling all twenty of the regional delegations of the Spanish government to get the data from each polling place. We obtained this data in very big excel files, and then we used R to analyze and standardize them into .csv files and then a shapefile. The third step was putting them on a map. Uploading everything to Mapbox was the easiest part of the whole process. As far as we know, there had never before been a map of all of these polling stations in the Spanish press, and perhaps now we know why :)
Our ‘street by street’ vote map is one of the most-read pieces of content of the year in the whole El País digital edition — and it was the most read of the whole electoral coverage. It reached 1.5 million page views.
What are other ways you’ve used data in your storytelling?
At the end of the summer, we experimented with a map using anonymized mobile telemetry data showing where Spaniards go on holidays.
In Europe, the General Data Protection Regulation (or GDPR) is a very key issue, so for this map, we had to use aggregated data. We partnered with Geoblink, a Spanish company that focuses on location intelligence for the retail ecosystem. To assuage any concerns that our readers had related to privacy, we decided to write an editorial piece detailing our safe use of data.
In another article, we used income data from El Instituto Nacional de Estadística (INE) and combined it with data from the Tax Agency and other census information to create a map showing Spaniards how their neighborhood income compared to others in the same city as well as in other regions.
This work helped us show the income disparities that may not be obvious just from looking at numbers in a document. For example, the Madrid neighborhood of Las Mil Viviendas (shown below in brown) falls within the poorest single percentage of income in all of Spain, but sits between a variety of more wealthy neighborhoods so in the census data depicts this region as one of the richest in the entire country. A map helped us show this much more granularly, allowing people to explore the accuracy of the data across a variety of geographic hierarchies.
What challenges did you encounter, and how did you solve them?
Our team started using Mapbox from scratch. One of our biggest challenges was to learn the Tippecanoe utility to optimize getting the data onto the map. We also found it difficult and a little chaotic to manage every layer of the map independently, but we were very excited to read about the launch of Style Components in Studio that we expect will make that problem go away. With each new map we create, we learn something new!
Julie Munro - Director, Customer Success - Mapbox | LinkedIn
How El País tells stories with data was originally published in Points of interest on Medium, where people are continuing the conversation by highlighting and responding to this story.