What’s your media job: Data Journalist

Editor’s note: The Fix is running a series of articles on different jobs in the media industry. We ask leading news media professionals about their job positions – what they actually do, how they make decisions, what excites them about their work, and what they can advise newcomers in their fields. In this episode, we describe the Data Journalist position based on our interview with Sondre Ulvund Solstad from The Economist. 

A Data Journalist is a person who uses data and statistics to investigate an important topic and to write a story and/or illustrate it with a data graphic. This is both a journalistic and technical position, as it asks for skills in collecting data, cleaning and analysing it, and familiarity with programming and design software. 

The Economist is a British weekly newspaper and an online outlet particularly focused on the topics of politics and economics. The publisher is famous for its data journalism, as it regularly publishes data graphics and complex works, such as an interactive map of buildings affected during an earthquake in Turkey. 

Sondre Solstad is a Senior Data Journalist at The Economist, where he has worked since February 2020. He is part of the data team led by the Data Editor. Solstad explains the essence of his role in one phrase: “To produce journalism that helps our readers make better decisions.”

The Economist is a huge outlet, and the data team itself has 9 data journalists. As Solstad explains, they work mostly independently on their projects. The cooperation starts later when there is enough processed data to write a story or prepare a visual.  

To prepare the story for publishing, data journalists cooperate with designers to make a final product like an interactive map. Then it goes up the chain — to the data team editor, then to the editor of a section, like the Britain or China editor, and to the editor-in-chief or other senior editors looking at the story before it goes public. 

Sondre Solstad says that he doesn’t have metrics as his KPI. Instead, to understand how well he works, Solstad looks into the audience that read his articles and the policy impact — for instance if people refer to his works or quote them. “And, finally, the reactions of my colleagues. What they feel about the work that I've done and their assessment of the quality and content,” he adds.

The journalist doesn’t have an average number of stories per month. “Big projects may take months to prepare, while some quick pieces can be done in a day or two,” he explains.

Solstad is the one pitching his stories most of the time. Though he doesn’t have any topics formally attached to him, Solstad often covers COVID, Russia’s war in Ukraine, and international relations. He also works on social science-related issues. One of his latest works was analysing the scope of the fighting in Ukraine with data from satellites. As Solstad noted in his twitter thread, “it has been months in the making.”

He spends a fair amount of work time researching the topics. He provides an example with China’s COVID statistics. “[After coming up with a topic], it's a matter of figuring out what has been published on the subject. What are the numbers that the government is supplying? What are alternative sources of data? And how can we understand this in the most rigorous way possible?”.

Another part of preparing an article is discussing the topic with colleagues, interviewing experts, and, generally, thinking about the most effective ways of telling a specific story and conveying the information he obtained. Sometimes Solstad may decide that the story lacks value and he should move on to another.

He says he spends time daily deciding what he should do: “To be very conscious about what kind of projects I'm working on, what kind of progress I'm making, and coordinating this with the rest of the team.” Coordinating is an essential part of the work because if the journalist needs an interactive element, a developer needs to have enough time to code that. 

Asked about verifying data, Sondre Solstad replies that he starts with a basic premise that any data isn’t trustworthy. Then he looks into reasons to trust it. One source of trustworthy data is those obtained in automated ways, for instance from satellites. Next goes the data on apolitical topics. “Scientists studying tectonic plates typically don't have a reason to misrepresent their data,” explains Solstad.

However, everything gets tricky with politics and economics. The journalist looks for incentives one may have to misrepresent the data, but uses logic as well. For example, businesses may be incentivised to publish incorrect data to hype up their product, but then they can get sued if caught, thus making a case for the data being more trustworthy. Anyway, every case is different. 

Another issue is in presenting the uncertainty the journalist may have in his data. For example, Solstad prefers to present a confidence interval instead of only a base estimate in a statistical model. “But oftentimes, it's also saying that: these are the problems we know of in the data. This is how I've investigated it, and I find it reasonable to believe that there won't be a problem in this case. But you, the reader, I respect you. And maybe you have other reasons to doubt the data than I do,” Solstad shares.

Solstad says that the job's most exciting part is the ability to help people make better decisions: “To be able to write about important topics and to inform people about how they should see and react to them.”

The most challenging is figuring out what to work on. Solstad clarifies that it is not only about deciding what topic to research but also deciding on a format – whether to do a quick piece or a deeply-researched long-form article. “So the first question is where to dig. The second is how far to dig,” as Solstad puts it. 

Sondre Solstad finds it challenging to give a general recommendation because of the spectrum of positions in data journalism. Though he believes it’s essential to understand numbers. “You don't need to know statistics to be a data journalist, at least not at a deep level. But you need to be able to compare quantities,” he explains.

It is also important to know how people perceive information and have strong design skills to be what The Economist calls a visual data journalist. 

Generally, when trying to break into the field, it’s useful to rely on personal preferences. For example, if a person loves pieces that rely on satellite imaging, then it is a good idea to dig into this field, to look for resources to understand how it works and what the tools are for journalists.

Solstad also thinks that the importance of creativity is often overlooked in data journalism, and it will be priced more with time. The number of data journalists is growing, and they use relatively similar sources of data, but Solstad believes there is a push to be more creative: “Using things like images and video, digging them for insights, using machine learning are exciting. And I think it's a space where a lot of great work will probably be done.”

Beyond that, Sondre Solstad asks to be careful with readers’ time and attention: “It's essential to give them the maximum impact for the amount of attention we receive from them. [Our work] should be extremely easy to consume, but at the same time, be complex and interesting and bring nuance and important information. That’s difficult, but that's what we're striving for.”