Infographic horror

July 27th, 2012  |  Tags: ,

I don’t mean to pick on the Des Moines Register (certainly, they are not alone in offending this way), but this is one of the worst infographics I’ve ever encountered:

Terrible, content-free infographic

My inescapable first impression is of the amateurish presentation. Although I applaud continuing Myriad creep, the “centered sentence fragments” look is only appropriate if one seeks to evoke the menu for that restaurant that opened up a couple of months ago and then went out of business before you had a chance to check it out. (Indeed, one almost expects a fourth figure on this illustration: “$8 sashimi.”) Furthermore, I note that the TextEdit app that ships with Mac OS X does a better job of automatically kerning the word “BULL’S-EYE” than the Register’s illustrator did.

We’re above harping solely on presentation, though. Unfortunately, this image utterly fails on the content level. It asks a question and then presents three figures. The final figure is essentially irrelevant to the question (although, to be fair, it is unlikely that either candidate would blitz Iowa with ad buys were he running unopposed), and the first figure serves only to supply a sense of scale for the second. So we’re left with one figure that actually addresses the question. However, the explanation for the second figure actively argues against the assumption behind the question!

If we want to argue in simple, front-page chart form why Iowa is worth aggressive attention from presidential candidates, a much better place to start would be the data in this chart, which I found on 270towin.com after searching for Iowa’s historical presidential general election results:

Iowa history

The above illustration is not particularly information-dense, but it gives us two numbers that are actually useful: Iowa’s results are evenly split among the last ten elections between the Democratic and Republican candidate (so, unlike many states, it is not unrealistic for either candidate to win in any given year), and it has gone with the winner of the electoral vote count seven out of the last ten times (so it is a reasonable proxy for the national contest). Of course, we can’t draw a causal link from winning a bellwether state like Iowa to winning the electoral college, but perhaps campaigns believe that they can evaluate tactics in Iowa for use in similar swing states. (Or, less charitably, perhaps campaigns subscribe to the same cargo-cult notions of causality as elected politicians.)