Same Data, Different Graphs (aka “Lies, Damned Lies, and Statistics”)


Here are two graphs done early this year, regarding U.S. job losses resulting from the current financial recession, as compared to previous recessions.  Both graphs use the same data, from the Bureau of Labor Statistics, but the difference between the graphs is startling.   The first graph was put out by the office of a politician, Nancy Pelosi.  The second graph was put out by a journalist, Fielding Cage, at TIME magazine.

This could have come straight out of a book by Edward Tufte, or any of those various classics about lying with statistics. 

misleading-graph

less-misleading-graph

This feat of manipulation was accomplished using three tricks:  

1) Vertical scaling:  Both graphs use a y-axis that is proportionately bigger than the x-axis, but Pelosi’s does so a bit more, thereby exaggerating the downward slope of recent job losses.

2) Absolute values:   Pelosi’s graph counts actual number of jobs lost, instead of the percentage of jobs lost.  The workforce has grown considerably over the years, so once again, this exaggerates the downward slope of recent job losses.

3) Narrower context:  Pelosi’s graph uses fewer past recessions in the comparison, and leaves out the more-severe 1981 recession, as well as two shorter-lived recessions ( 1974 and 1980).   This substantially reduces the nuance in our mental comparison with past recessions, and skews our extrapolation of what might happen going forward with the current recession. 

This would be an insult to our intelligence, if it wasn’t such a perfect example for instructional purposes… a fine example of information design. On a practical/political front: Score three points for cultivating greater hysteria on the one hand, and cynicism about data & public policy on the other hand.

The only thing lamer than this?  This graph

Pelosi-Recession-Graph  This bit of USA Today-worthy Chart Junk put out by the Republican party (and posted to their website) tracks how many jobs were lost since Democratic Party took control of Congress in 2007… which compares nothing with… well… nothing. It’s like some sort of weird Creationist theory of economic history that picks an arbitrary “Day One” and denies the existence of independent natural forces, forces that can’t be fully controlled by the would-be gods in Washington.  If this were a tongue-in-cheek reprise to Pelosi’s graph, it might be clever, but it’s not… it’s just upping the ante of stupidity.  Memo to both parties:  Get Real.   

Finally, here’s a third graph, with even wider context, put together by economist William Poley: employ_recession

 

Context and perspective is everything, ain’t it?

(More graphs and analysis on the above can be found on William Poley’s blog.)

Advertisements

About danspira

My blog is at: http://danspira.com. My face in real life appears at a higher resolution, although I do feel pixelated sometimes.

Posted on July 8, 2009, in Analytics, Communication Skills, Information Design. Bookmark the permalink. 7 Comments.

  1. I like it when someone else does the critical analysis to point out flaws, deception – or worse. But admitting that just points out that I am by nature willing to take others’ information at face value. I am not sure I would have challenged those data as you did. I’ve always said it is “energy conservation,” as in, “I don’t want to waste my energy questioning everything I see.” Now, if I had to make a decision or a recommendation on a course of action based on data and had only those charts as information, THEN I might have scrutinized them for accuracy and truth.

    As Levin points out in Freakanomics, numbers are a dark art, as useful for discerning truth as they are for masking it.

    Thanks for the energy you expended.

    • Heh, thanks for your comments Stephen, including that tidbit from Freakanomics… though you make it sound like you wouldn’t enjoy expending your energy dissecting charts and graphs and methods of information design, just for the sake of pleasure… doesn’t *everyone* love geeking-out this way? 😉

      You’ve inspired me. I’m going expand the post a bit more, if only to have a complete little “module” of content for future use.

  2. This brings to mind my former accounting prof’s (Al Rosen/Schulich School of Business) observation that people know not to believe everything they read, yet they intrinsically trust numbers. Given he works as a forensic accountant on the side (and had some fascinating stories), his take-home message was that we should trust numbers no more than words. Just as easy to distort & manipulate.

  3. the more one is educated, the more one is prone to propaganda. (Ellul)

    this is a perfect example.

    some concepts ARE timeless it seems!

    • So true, especially in academia.

      The counter-point to Ellul’s proposition is Benford’s Law of Controversy: “Passion is inversely proportional to the amount of real information available”

      …but I think the two can be reconciled by the idea of selective information design. Also, while “Propaganda” and “Passion” really do go hand-in-hand, “Educated” and “Real Information” are two entirely different things.

  1. Pingback: EOY Blog Clean-up « Meme Menagerie

Leave a Reply -- for humans only, no spambots

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: