Every Single Person in one Cartogram
The LA Times has published “Beyond 7 Billion,” a rich interactive feature covering the story of the exploding world population. Increasingly the Times has been producing rich content and some data visualizations including the Oscars Senti-meter which I critiqued a few months back . The world population site includes stories, videos, photos and a world map of sorts, more accurately a cartogram.
Even more accurately it’s a Dorling cartogram. In a traditional Dorling cartogram (introduced by Danny Dorling) each geographic area is represented by a scaled circle and placed to resemble its true geographic location. The circles are then usually color encoded with another variable. Here they are color coded by their respective continents which actually helps quite a bit in retaining a sense of geography when the circles start moving around.
In the main population map you can transform the circles to represent data from 1950, 2010 and project data for 2050 and 2100, each with startling effects. The prominence of Africa’s population explosion wasn’t something I expected to see.
The LA Times has also included other maps in the same form that visualize related topics such as world hunger, freshwater, CO2 emissions and contraceptive use, all split out into separate tabs across the top. In some cases the combination of scaling and the data set renders the circles too small to include almost any of the labels forcing us to rely on hover to see any detail.
What’s more I think it’s a lost opportunity to have all of these data sets rendered as separate maps. I hope the LA Times team at least explored combining the population variable with any of the others in the same map, using a more complex color scheme to achieve readability and foster insight. As it stands it’s hard to relate any two topics or understand the scale of the issue in relation to population which seems like a central goal of the piece. For instance I think it would be enlightening to see “% of population that is malnourished” encoded by color onto a cartogram scaled by population. With a more complex color scheme you could retain the sense of continents by using a single color family as well as visualize the depth of a second variable by using different values within that family.
I think the most enlightening interaction would be able to show population + another variable and be able to switch that second variable and watch the colors animate on the cartogram to reveal areas more or less intense. Perhaps you could even specify which variable affected the scale and color and switch them around to explore relationships. This is probably a bit too free-form for this application. The goal of this visualization is more to tell the story of population growth and less to let users discover it for themselves. Still I question whether isolated maps showing a single variable are the best way to relate this story. In this case it’s possible that a more complex display of data could be more impactful.
Further cartogram reading: Danny Dorling himself on visualising social structure with 1980s data on the Guardian)