The Viz Police: Graphic for the Sake of a Caption in The Economist

This week’s chart is from an article “The battle of Smoot-Hawley” in The Economist magazine (December 20 2008 – January 2 2009 issue). The Economist is one of the few magazines I read. And it normally has intelligent graphics. But I couldn’t help myself with this one.

The graph is trying to show world trade declines from 1929 to 1933 (which correspond to a protectionist measure sponsored by congressmen Smoot and Hawley). Normally, a line graph would be exactly the right thing to show such a trend.

The troubling thing here is that it seems they used a radial graph so they could fit a pre-determined title: “Down the plughole”. Yes, it’s amusing but it makes the graph nearly incomprehensible. Why a radial graph other than for the title? I can’t find a good reason. This is the biggest problem by far.

Color choice to depict years is questionable - the final year is rust as opposed to the greens/blues of previous years. It makes you think the final year was especially different. But it wasn’t.

If you study the graph, you get the point. Every year declined and every month was lower than the previous year’s. But who wants to study a graphic that hard? I might as well just read the interpretation. And the graph makes it hard to answer obvious questions. By how much did it decline? And when was the biggest drop?

Kudos to making the graph fit the column width of the magazine which is a vertical rectangle. A line graph by month over 5 years would almost necessarily be a horizontal rectangle. (Of course, they could have done a square-ish line graph with 12 months as the axis and a separate line for each year.)

They do get a "shout-out" for the pun in the title. But it’s a pun that works too hard.

Normally, you see great stuff in the Economist – simple, clean, easy to interpret. Let’s chalk this up to someone letting a small space and a bad pun get the better of them.

I recently discovered that Jon Peltier, whose blog "The PTS Blog" covers data visualization and who occasionally comments on our blog (thanks Jon), had a more detailed analysis of this chart and in fact even recreated it as the line charts suggested above. Check out his post on the PTS blog.