Eva Constantaras is a data journalist specialized in building data journalism teams in developing countries. The teams she worked with have reported from across Latin America, Eastern Europe, Asia, and East Africa on topics ranging from displacement and kidnapping by organized crime networks to extractive industries and public health. Eva is also on the Outriders Advisory Board. I, personally, am a big fan of how meticulous and focused she is at work. I also really love her opinionated nature and admire her integrity. Here she is, answering some of my questions about data journalism. Enjoy!
How (and why) did you get into data journalism?
Like many data journalists, I started doing data journalism before it had a special name. I had a Fulbright Fellowship after university to study journalists’ safety in Colombia. Working with Fundación para la Libertad de Prensa I was involved in one of the very early mapping efforts to track journalists under threat and allocate resources according to their risk level. It was very meta, but I saw the more practical implications of how data could be more effectively utilized to understand vulnerable populations, identify needs and communicate those findings. At UNESCO I worked in a similar area, but this time tracking threats and violence against journalists on a global level and finally, I began working directly with journalists to harness data for their own reporting in various countries where journalists were most at risk.
How is data journalism different from journalism? Is it just about the skillset?
It depends on the definition of data journalism you are working with. I have a fairly narrow definition other people would disagree with. For me, data journalism is harnessing data to measure a problem, identify who has been impacted, what is causing the problem and what the solution might be. In my experience, the conceptual skill set to structure a data journalism project is often more challenging than learning specific technology. Very little journalism is based on this type of scientific research-based methodology that tries to uncover the root of the problem rather than just reporting the existence of a problem. You will notice I didn’t mention specific data types, data processing tools or data visualization, because, for me, all those are simply tools for to the hypothesis you are pursuing and the public interest story you are telling.
Very little journalism is based on this type of scientific research-based methodology that tries to uncover the root of the problem rather than just reporting the existence of a problem.
You are American, but you have lived in the US very briefly. Why?
One of my biggest frustrations with media today is their failure to truly globalize. Especially with the rise of challenges like ethnonationalism, inequality and climate change, very little news coverage focuses on these topics at all and even less so through a global lens. I have traveled or lived in over 75 countries and the more places I visit, the more I realize how universal these stories are. I wish media could recognize the value of providing international coverage that would enable citizens to take a step back from the political rhetoric filling their local media and look at hour various countries are handling issues such as the gender pay gap, food security or healthcare. I think taking these stories out of the national context allows people to see issues more clearly and focus on policy solutions. By attending to different aspects of injustice, inequality and discrimination, and their broader consequences on the lives of marginalised communities, we render them visible, measurable and maybe even solvable. There isn’t a country in the world that this doesn’t apply to, but I think the US media culture is one of the most difficult to change, especially when it comes to including global voices.
One of my biggest frustrations with media today is their failure to truly globalize. Especially with the rise of challenges like ethnonationalism, inequality and climate change, very little news coverage focuses on these topics at all and even less so through a global lens.
You mainly train. Do you like it? Is it hard and why?
Training is extremely rewarding because every country has capable, passionate journalists who need a process, some time, space and the right tools to produce groundbreaking data-driven investigations. The hardest part about the training is identifying these people and giving them the time and resources to thrive. When I first started training, I fell in the same trap as most have: I was teaching tools instead of a new approach to journalism. I improvised activities to grow journalists’ critical thinking and research skills. I eventually found Mark Lee Hunter’s Story Based Inquiry manual for investigative journalists and built data journalism teaching pedagogy around it. I have now published editions for Sudan, Afghanistan, Pakistan, Kyrgyzstan, Albania, Central America using local examples and local datasets and each time it gets longer. Now it’s over 650 pages and takes about five weeks to deliver. I learn as I go and share what I have learned from training here.
When training in places such as Myanmar, Iraq or Afghanistan how do you make sure what you teach is relevant to the local realities? How do you anchor what you say?
One of the first lessons I learned is that people do not want to learn concepts, tools or data out of context: so no California forest fire data or exercises with baseball scores. But as I mentioned above, there are some very tangible problems that seem to affect people in just about every country: unequal access to food, health and education, broken justice systems and poor management of environmental resources. One of the first things I do in a new country is an open data assessment to find out what data resources are out there, what studies government, civil society and academics are doing in those areas and how media has been covering those same issues. Usually, there are some excellent opportunities for journalists to develop hypotheses and dig into these datasets themselves. At first, the exercises are basically a trail of breadcrumbs to a potential story to teach the concepts and tools and then the journalists are free to explore their own angles.
(…) people do not want to learn concepts, tools or data out of context…
Data journalism hype seemed to be intertwined with the general “tech” fever in all social fields. Do you think this hype is over? Is it good or bad?
I use the hype of data journalism as a pretext to teach the fundamentals of public interest reporting: media literacy, data literacy, critical thinking, research methods, investigative reporting processes, all of which, for me, is much more challenging to teach than the tech itself. Meredith Broussard coined the term “technochauvanism”, the belief that the technological solution is always the right one. Technology can certainly enable, speed up, automate, many parts of data journalism, but first, one has to learn to be a data journalist. Learning to create a chart in DataWrapper, or even D3, does not make you a data journalist. Learning to analyze data for stories makes you a data journalist, and that skill is hard to learn. Yet nearly every week I field calls from institutions asking me to teach a data journalism workshop, or bootcamp or run a hackathon and inevitably they are vastly overestimating the role of technology in improving the quality of local journalism. Think about it like building a house. You can have all the aesthetic sense in the world and paint your house a pretty color, but if your foundation isn’t sound your house is going to fall down. I spend a lot of time listening to people obsess over picking out paint colors and much less attention on structural integrity. So no, I will not fly to country X to teach people to make interactive maps of malaria outbreaks unless they have a group of journalists who already know where the malaria is, who is most vulnerable, what caused the outbreak and how it can be fixed. Then we can talk maps.
I will not fly to country X to teach people to make interactive maps of malaria outbreaks unless they have a group of journalists who already know where the malaria is, who is most vulnerable, what caused the outbreak and how it can be fixed. Then we can talk maps.
What is broken in data journalism as a field?
I would say data journalism right now is suffering from a plethora of missed opportunities, but that is starting to change as the media industry crisis deepens. One opportunity that I alluded to above is harnessing tech to enable, speed up or automate many parts of data journalism. Services such as RADARwhich helps local news outlets identify interesting story angles in datasets. The Organized Crime and Corruption Reporting Project’s Aleph enable reporters to search for data and documents already compiled by the organization. These tools don’t replace humans in the investigative process but it cuts out the manual labour of digging through raw source material again and again. The next big opportunity is pooling resources, which has been practiced by data journalists in developing countries for years and is now catching on in the West. IndiaSpend serves as a non-profit data-driven wire service for media outlets throughout India covering exclusively public interest issues. Groups like Bureau Local in the UK, ProPublica’s Local Reporting Network in the US and Ojo Publico’s Regional Reporting Network in Peru are trying to reverse the trend of data journalism concentration in major cities by pushing local data sources and data skills back out to local media for local accountability stories. Finally, in the future, I see more collaboration among journalists and civic tech, civil society, and academia for joint data-driven investigations based on evergreen databases.
What are your favorite data journalism stories of all times?
Rather than focusing on specific stories, I think I would rather highlight a few categories of stories that have proven the importance of data journalism to democracy in times of growing inequality, whether they are done by large teams working with massive datasets or smaller teams making do with limited resources. The first of these is harnessing data to investigate the widespread miscarriage of justice whether by law enforcement (in the US or Kenya) or the justice system (in the US or Afghanistan). The second explores the reach of gender inequality and violence against women whether in India or Canada. And finally, stories that seek to explain the consequences of unregulated capitalism such as the Big Pharma project in Latin America or Leprosy of the Land in Ukraine.
…and of recently?
What has excited me about data journalism over the last few years is a move towards building and maintaining evergreen projects that can have a long-term public impact. These living projects cultivate audience’s understanding of complex public interest issue and also offer a less resource intensive way for data teams to engage in data reporting by removing the need to create a brand new database for each project. Great examples include projects that enable longitudinal oversight of education reform such as South Africa’s Passmark, regular government accountability reporting such as Datasketch’s public tender project, and introduce nuanced beat reporting on marginalized groups such as the Bureau Local’s Homelessness project or invisible organized crime victims such as Mexico’s Where Do the Disappeared Go?. These projects foreshadow a future in which all journalism is informed by data to provide context and insight on people that often go uncounted.