Harry Shannon
I was listening to CBC radio this morning. A reporter was talking about the popularity of Ontario Premier Doug Ford, and said that a year ago Ford’s approval rating was -49% [if I remember correctly]. Today, his rating is +39%. The reporter continued, saying that this represented an 88% swing.
This, of course, is wrong. To understand why, consider what the approval rating means. To make things a bit simpler, let’s use the figures -50% and +40%. The approval rating is the percentage of people who like what a politician is doing minus the percentage who dislike what she is doing. So if 25% approve and 75% disapprove, you get a rating of -50%. Similarly if 70% approve and 30% disapprove, the rating is +40%.
Using the reporter’s logic, the swing is 90%. But if you look at just the percent who approve, that only changed by 45%, only half of 90%. The problem is that the reporter double counted the swing. Doing that can lead to strange conclusions. Suppose that initially the approval rating is +80%, and later it is -80%. The reporter would have to say this is a 160% swing. Clearly ridiculous.