Sunday, January 5, 2014

Subscribing to The Wall Street Journal —just for the infographics

I got myself a nice present during the holidays. As if I didn't have enough to read every morning already —I've been a proud subscriber to the print version of The New York Times for quite a while— I decided that it was time to give The Wall Street Journal a try. I featured some of its interactive infographics in The Functional Art, so I thought that it'd be appropriate to get a thorough understanding of what its graphics desk (Twitter) is up to.

I received the first paper yesterday. As I had predicted, I enjoyed its journalism very much and was very much annoyed by its editorial pages (their shameless anti-science bias has always troubled me.) I was also intrigued by two charts, both on page 2 of the newspaper. I like the first one quite a lot. Here it is:

Simple and smart, isn't it?

The second one puzzles me, though. Why representing vehicle sales twice (position on the X-axis and bubble size)? And why a scatter plot? There may be reasons for those, but they escape me:

Moreover, I understand that percentage change is a variable most WSJ readers want to see but, out of curiosity, I'd also like to get an idea of the actual sales. Here's why: The largest increases of the smallest companies are likely due to the fact that they sold much fewer vehicles in 2009 in the first place. Wolkswagen sold 600,000 cars in 2013, and around 315,000 in 2009. That 91% increase represents more or less 285,000 cars —I hope I'm getting the math straight!*

But let's take GM. It sold 2.6 million vehicles in 2013, with a 35% increase since 2009. That's a difference of around 700,000 cars. The numbers get more impressive if we focus on Ford: 2.5 million vehicles in 2013 and a 54% increase since 2009: More than 850,000. And I won't even mention Chrysler, which seems to be an exception in the chart: Relatively large company, very large sales increase. Is the infographic telling the whole story? I'll leave that discussion to you.

*Here's an explanation of how to reverse a percentage rise or fall.


  1. A bar chart instead of the scatter plot would have been more effective. The length of the bar = unit volume, color encoding of the bar = % change. Labeling the bars with both details. Adding a reference line to each bar that provides the target values would provide a useful additional bit of information to help people understand each companies' ability to deliver on its commitments to shareholders.

  2. Thanks for your blog and this post. I think the editorial page is the best part of the WSJ! I read the NYT also, but I like the contrarian attitude of the WSJ. Climate change skepticism is not the same thing as an anti-science bent. It isn't all science - it's this issue, which is an argument over how science should influence policy. Is it proper to disrupt the world economy on the evidence we have? Lots of people say yes, but not everyone. I smiled at the opening passage of the Science letter. We are not "saving the planet", we are preserving our own comfort. The planet doesn't care if Manhatten and Venice are under water, some species die out, or the air is "polluted". We do care, but we should frame the issue honestly.