To mark the grim milestone of 100,000 coronavirus deaths in the United States, we collected obituaries of 1,000 people and distilled a brief excerpt for each to try to capture the humanity of the lives lost. The project sought to turn the death data into a more humanistic representation of the scale of the loss. Online, we combined the excerpts with 100,000 illustrated figures and an essay. In print, we took over the entire front page and several pages inside with the 1,000 names and obituary excerpts — still just 1 percent of the toll.
The project drew widespread attention at a difficult moment in the pandemic. The decision to take over the front page of the print paper was especially resonant and ultimately drew people into the project to spend time with the names and their stories.
Many readers, including those who lost loved ones to Covid, expressed their gratitude for the piece, some in particularly emotional terms. One note: “I will cherish your echoing the pains of lives lost to Covid-19, and thereby helping us to grapple with what it means to be the America to which my parents each emigrated from eastern Europe just over a century ago.”
People who saw their own loved ones in the list and reached out; others shared the piece or left comments with their loved ones’ names and a short sentence, adding them to the list and making them part of the project, too.
The all-type format of the front page inspired many derivative art projects, including one striking take amid the racial justice protests that sprung up not long afterward listing names of those killed by police.
This was not a typical data project because a significant portion of the source material was qualitative and had to be collected and qualitatively reviewed for inclusion, then turned into readable data for display. The editing of the 1,000 excerpts was perhaps the most time-consuming and emotionally challenging part of the project.
To gather the obituaries we searched news reference sites like Factiva and Nexis as well as the obituary site Legacy.com for mentions of deaths related to coronavirus. We also searched Spanish- and Chinese-language media and did some more targeted searching on the sites of specific news outlets to make sure our list was as inclusive as possible.
All the material was poured into a series of 17 Google docs with about 100 pages of obits each. A team of researchers and editors read through them creating metadata (name, age, location) in ArchieML and choosing excerpts for publication. We created a Node script to download the data from the 17 docs and merge them into a single Google spreadsheet that powered the interactive.
For the graphic visualization we used an HTML canvas element with a series of small illustrations. We added aspects of randomness that would make the display feel more organic while also retaining a sense of order. For the typographic elements we chose to use real HTML text for crisp display. The final piece of the puzzle was to sync the two elements, which happens as the page loads and responds to the size of the viewport.
What was the hardest part of this project?
It was emotionally taxing work on a tight deadline. At the time, such a scale of death was hard to fathom (though we’ve since nearly quintupled it), and to build the piece we had to confront it directly.
From initial concept to publication was less than two weeks, and it required an enormous team to pull off the research, editing and production. It was difficult to get the whole thing done from a practical standpoint and also difficult because of the subject matter.
Once we had the initial list of obituaries, a half-dozen editors read every one to choose the very short lines that went into the piece. Most are direct quotes, or lightly edited. Because we were keeping everything to a sentence or less, we ended up leaving so much out. It’s difficult to capture a whole life in a handful of words, or even the couple hundred words we were drawing from. But the hope was that collectively, the short descriptions would help readers better comprehend the enormous gravity of the deaths in a more visceral way than just a chart or data visualization.
It was also challenging to strike the right tone visually and in text, again remembering we were dealing with real people’s lives and wanting to honor their loved ones’ memories of them. We went through many iterations of the illustrations and design to make it feel at once personal and to convey the scale — a balance that is difficult to achieve in most data work.
What can others learn from this project?
It’s important to humanize numbers, especially big ones, but it’s not easy. On this story, getting to the underlying source of the quantitative information helped us reveal a lot more than a summary statistic.
We started out feeling like the impact of 100,000 deaths was this awful, unknowable thing. How could we ever know what it meant? We tried to answer the question by breaking it down. Each death was a life, each life was a person, each person had a story, a community, an impact. That story was knowable, too. All together, they made up a big, overlapping story about our country.
One lesson we learned on the process: It was valuable for everyone involved to have a sense of the final work product early on. We did a proof-of-concept sketch of the front page that helped gain buy-in from leadership, and also gave the researchers and editors combing through the obits a tangible goal. Sometimes we do the reporting and then start designing, but in this case we needed both to happen on parallel tracks and we were lucky that they could inform each other.
Print A1 – twitter.com/nytimes/status/1264427825639063553