|
|
15 January, 1999
The cliche is that there are three kinds of lies: lies, damn lies, and statistics. It's not an inherent fault of statistics. In some cases people are intentionally misusing the statistics. In others, it's the fault of the listener for being math illiterate.
I heard an "debate" recently on the radio where two experts were discussing the economic situation of the typical American. One cited a statistic that said the typical worker is earning less today (when adjusted for cost of living) than a number of years ago. The other countered with a study which showed that peope from the bottom of the economic heap a certain number of years ago had moved up into the middle income range.
The statistics aren't lying. In fact, they aren't even in disagreement. If the "average" gets lower, then by definition people in the lower half of the scale will get closer to average, or even exceed it, without having to change their situation at all.
The problem is two-fold: we mis-use the term average, and we expect the average to tell us all sorts of meaningful things about the group as a whole. Here's a simple example. Imagine that you're taking a group of five children to the fair, and that each child has been given some pocket money by his/her parents as follows:
Child 1: $5.00
Child 2: $5.00
Child 3: $5.00
Child 4: $5.00
Child 5: $12.00
The "average" wealth of each child is $6.40. But notice that four of the five children have less than the "average" -- not a lot less, but this hits up against our most common misunderstanding of "average": we unrealistically expect the "average" to describe at least half the population. It doesn't always. Let's look at another group of hypothetical children:
Child 1: $1.00
Child 2: $1.00
Child 3: $5.00
Child 4: $5.00
Child 5: $20.00
Once again, the average wealth of our group is $6.40. There is clearly a much wider disparity in this group than our first group.
Of course, the astute observer will point out that the wealthiest kid is so far above the others, that it's skewing the data. Which is precisely what happens in many economic statistics. A few super-wealthy individuals can make things such as "average income" meaningless. It takes 28,000 median-income people to equal one billionaire. On the other hand, it only takes 3 people living at the poverty level to equal one median-income person.
The next time you see an article quoting statistics about the economy, look very closely at the words used, and think about my examples. You may find the some interesting prejudices built into the news.
 |
 |
 |
 |
This page is copyright 1999 by Gene Breshears. Photograph is copyright 1998 by Julie Rampke. All Rights Reserved.