Bussed out

an extensive data visualization-led article for the Guardian US on how America moves its homeless

Scroll down for information

Details
Link:

Bussed out

Created:

August – December 2017

Reason:

data sketches’ December month

Tools:

R, D3.js, canvas

Data:

The Guardian team requested the data about the bus relocation programs from several US cities

Description

Bussed out | How America moves its homeless
Each year, US cities give thousands of homeless people one-way bus tickets out of town. An 18-month nationwide investigation by the Guardian reveals, for the first time, what really happens at journey’s end.

There are several data visualizations in this article. Many of which are created in a “Scrollytelling” style, where scrolling down the page will trigger animations in the data visualizations. Through these visualizations, video, photography, and of course, text, the article explains the impact of the homeless moving around the US. both from the perspective of the homeless and the cities that are sending them away.

Creation

This has been a huge undertaking that involved many people. From an entire team at the Guardian US to Shirley Wu and me as two freelancers hired to take on the data analysis, visualization and creation of the whole page. I started in August to clean the data that the Guardian was getting back from the different bus relocation programs. This took a very long time because each dataset had its issues with quality, such as typo’s in the destinations of the homeless. I creating several scripts to help with the task of cleaning up the data to create one gigantic dataset about homeless movement, which I could then analyze in terms of differences in, for example, age, gender, destination cities etc.

During the end of September the data cleaning and analysis was done and I spend 2 days with the team at the Guardian’s HQ in San Francisco, together with Shirley, to talk about the structure of the article. What topics should be written about, what insights did the data have to support hypotheses and what visualizations could be made to support and explain the whole. At the end of those two days we had the rough structure and designs for the visualizations.

In the next few months Shirley and me spend many hours on creating the data visualizations; from prototypes to visuals that would interact and change with the story, that would work on both desktop and mobile sizes, and that would run on every important platform and browser that we could think off. Weekly calls with the team at the Guardian kept us all on track and provided moments to discuss the visuals and talk about feedback.

Out of the 5 visualization in the final story, I created the US map, the “Rich-Poor” visual that shows that most people move to places with lower median income and the “Radial concentric circle” visual that highlights who the people are that the homeless provide as contacts that they intend to travel to. The NY flights visualization and San Francisco “dots” visualization were created by Shirley Wu.

You can read more about the data and creation all the way at the bottom of the article in the Methodology section