Lies, Damned Lies, and Statistics

Today’s But if Not comes from an interesting David Henderson blog post yesterday over at Econlog, When Numeracy Misleads

Numeracy is one of the things I find lacking in people who fall for a lot of politicians’ nonsense and reporters’ nonsense and so I try to combat it

Amen brother David.  Combat on!

Similarly, Megan McArdle blogs this morning on the common practice of changing the y-Axis to manipulate the way a graph looks and is therefore interpretted by the reader.

 things can look very different depending on where you start the Y axis.  And in my experience, as advocacy groups launch into their end-of-year fundraising season, Y axes seem to creep closer and closer to the bottom value in the series.

She includes two graphs to make the point.  Notice how the first one amplifies the differences between bars and the second, of the exact same figures, seems to show little difference. 


When idealism collides with advocacy, or worse, one’s paycheck, the story is the victim.  But wasn’t this also true of the accountants putting together the financial statements for Enron? 

The sooner we accept that this weakness to blur clarity in the cause of advocacy or profit-making applies to all people, including the selfless, impartial, “just-trying-to-help” government workers, the sooner we will pick back up our end of the rope and serve as the skeptical check that is a necessary prerequisite to a democracy.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

4 Responses to Lies, Damned Lies, and Statistics

  1. Pingback: Be Skeptical First, Then Verify | But If Nots

  2. Dan says:

    On a related note, the “top 1%” argument. I’m not completely settled on tax policy, but people should know the whole story:

  3. Moose says:

    Interesting that you should bring this topic up. A few days you were gushing over Hans Rosling’s “200 Countries, 200 Years, 4 Minutes” video and wondering “Where has all this good news been kept?” However, note that Rosling uses a logarithmic scale on the horizontal (income) axis. If you run the analysis again using a linear axis (which you can do at, you might reach a different conclusion. The logarithmic scale tends to minimize the gap in income between the wealthiest and poorest countries and you could (reasonably) come to the conclusion that the poorest countries are not much better off than they were a hundred years ago.

    By the way, I can show you how to use some very powerful statistical graphics software to prove any point you want.

    • Sean O'Brien says:

      Thanks for the pointer to gapminder. That is excellent. I spoke to another reader that had the same comment.

      Indeed, Africa is the last to start the climb. But, simply based on the growth in business we alone are doing across Africa, it appears that they have begun the ascent.

      You, having just returned from one small corner of Africa, are better positioned to comment on the state of things. I would happily post, or make you a guest poster, some of your pictures and your commentary on what you saw and where you went.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s