behind-the-scenes: Axios presidential ages chart

I used to share process threads on Twitter, but with that platform feeling increasingly icky, I’ve fallen out of the habit. I want to get back to sharing the hows and whys behind my work—both as a way to stay transparent as a journalist and to create the kind of resources I wish I had when I was starting out.

This Axios story began as a reporter “request” in our team’s Slack channel. The focus was on the ages of Biden and Trump, with the reporter looking for a bar chart showing the oldest presidents to enter/exit office. Being on-call that day, I asked some clarifying questions to understand the story’s angle, the publication timing, and the dataset used.

I learned this story was driven by audience interest in the topic, so the publication date and shape of the story were slightly flexible. I asked for a day to explore whether we could create a visualization offering more nuance and historical context, especially since many other newsrooms were covering Biden and Trump’s ages at the time. I also needed to find a more reliable dataset.

I scraped the data and began exploring. My “default” instinct was to show the distribution of presidents’ ages at inauguration. The beeswarm effectively showed where Trump and Biden fell within the age distribution of presidential ages, but it didn’t convey age when entering and exiting office, something the reporter had initially mentioned. Additionally, I discovered the “edge case” of presidents who served nonconsecutive terms. I set a timer for an hour to sketch additional ideas, knowing that the beeswarm was a solid fallback.

My second sketch experimented with an overlay over a president's lifespan for the time they were in office. This column chart had a lot of wasted space and the column order encoding each president was confusing. I decided to focus only on the overlay for my next sketch, resulting in a dumbbell chart showing the president’s age entering and exiting office. Some design decisions included starting the y-axis at 35 years, using gradients and a trendline to support the "movement" through the chart, adding annotations to guide the reader through the slightly more complex chart, and using green as a neutral color.

I shared the three sketches with the data viz team for an initial gut-check before presenting them to the reporter. While my team and I preferred the dumbbell chart, the reporter selected the beeswarm chart. I tried to understand their rationale, learning they thought the beeswarm would suit social better. We reached a compromise—we’d use the dumbbell chart for web and create a simpler version for the social image.

If I were to tackle this story again, I would explain the pros and cons of each sketch idea presented and make the final version in Svelte instead of Datawrapper/Adobe Illustrator ai2html. Svelte would allow for tooltips for extra-curious readers to see the names and ages of each president.

This process reminded me to iterate beyond my first “good enough” sketch and to be more mindful of my default ways of thinking. Overall, I’m proud of the collaboration throughout this story—we were able to create something distinctive that got received great reader feedback.