Wednesday, October 3, 2007

On Innumeracy: A Minor Quibble about Percentage Calculations in an NPR Story

David Kestenbaum in "Japan Wrestles with Kyoto Accord Promises" stated that "instead of cutting emissions by 6 percent, they have grown about 8 percent. That adds up to a 14 percent problem."
Actually, this is incorrect and reflects a commonly-held misunderstanding of percentages. Percentage declines cannot be directly added to percentage gains to get a percentage difference. To illustrate this fallacy, lets say the base emission index had a value
of 100. A 6% decline would have brought the index to 94, Japans target. An 8% gain, however, would bring the index to 108. The percentage difference between 108 and 94 is 14.9%, not 14% as reported in your story.
When the percentage differences are small, as in this case, the error of simply adding the percentages is also small, but larger numbers illustrate the fallacy. For example, if Japan had wished to halve its emissions (a 50% reduction), the difference between this ideal and the actual 8% gain would be a whopping 116%, far greater than the 58% difference that Kestlebaums method would have computed.
Just a minor point in an otherwise solid story.

No comments:

Post a Comment