Graphical Depiction of Recession-Induced Change

by philg

Here’s an interesting graphic from The Atlantic attempting to show how the U.S. changed between December 2007 and June 2009.

At first I thought that the graphics were going to communicate much more than the raw numbers, but as I looked closer it occurred to me that the simple numbers are easier to read, e.g., “number of federal employees grew 13.6 percent to 2.06 million”. How does having two bars and a pyramid (lower center) make this clearer? The graphics would be useful if they helped one compare the numbers to each other or to additional numbers, but each mini-graphic stands alone and often is not comparable to anything else nearby.

What do folks think about this attempt at graphical communication?

11 thoughts on “Graphical Depiction of Recession-Induced Change”

  2. I believe the pictures are all very pretty and probably far more interesting to the average reader than plain text would be. Perhaps the pyramid represents the Unfinished Pyramid along with the “Unseeing Eye” that we see on the dollar bill? (with such an increase in federal employees, it seems appropriate)

    I do wonder about the mathematical implication that the recession *caused* all those graphs. Edward Tufte would disapprove, I’m sure.

  3. None of the graphics seem particularly relevant or useful, since, like you said, they aren’t really comparing anything, just showing two sets of numbers from the same dataset. This chart seems to recall the hip new way of presenting graphics, in a very New York Times style (I swear I’ve seen this many times before, but have no examples off hand.) And, while I enjoy the effort to get away from bar graphs straight from Excel, and incorporating more graphic-design elements, this one obviously doesn’t understand how to put the “pretty” to work. I also think a lot of the new trend in graphics stems from this flowchart from a few years back (not dealing with money, but does a good job putting a complex issue into a simple, but appealing, style):
    http://blog.mint.com/blog/finance-core/a-visual-guide-to-the-financial-crisis/

    Heck, even something as simple as the following would have been more useful, since it easily shows comparing (before crisis) 2007 to (post crisis) 2010 is pretty unfair, and in my opinion, had another Republican been elected in 2008, when his first budget of 2010 rolled around, we’d still see the same increase:
    http://en.wikipedia.org/wiki/File:Deficits_vs._Debt_Increases_-_2009.png

    Past 206 years vs. 12 months, what appears to be a better graph, but, in the comments, brings up a lot of questions about exactly how data should be represented, which data should be included, how it’s adjusted, etc., and the complications of making simple charts from complex situations:
    http://boingboing.net/2009/06/18/infographic-all-us-o.html

  4. At least these are bar graphs to compare against on one dimension easily and you can think of the mini pyramid as the artistic interpretation of an up arrow for a dramatic increase. My main problem with the graph is the overlapping gaussian curves meaninglessly spread all around the page (e.g. Tax Collected, Federal Budget Deficit, etc).

  8. How about casino revenues or profits. How about alcohol sales, aren’t these recession proof industries? Did they remain unchanged or show growth related to economic environment?

  9. 20 years on and Tufte is still sadly relevant.

    “Lurking behind chartjunk is contempt both for information and for the audience. Chartjunk promoters imagine that numbers and details are boring, dull, and tedious, requiring ornament to enliven. Cosmetic decoration, which frequently distorts the data, will never salvage an underlying lack of content. If the numbers are boring, then you’ve got the wrong numbers. Credibility vanishes in clouds of chartjunk; who would trust a chart that looks like a video game?”
    – Edward Tufte, “Envisioning Information”, 1990

  10. Nowhere near as meaningless as Harper’s Index.

    Quote:

    Minimum number of properties the U.S. government was forced to buy as part of the S&L bailout: 43,640

    Number of churches and athletic clubs, respectively, that this purchase included: 3, 27

  11. Wow. This is egregiously, revoltingly bad. The piggy bank totally distorts the actual change in savings rate. The Gaussian representation of single data points is totally confusing. The clocks showing average minutes spent playing video games don’t make sense at all. I could go on and on…

Comments are closed.