Tuesday, October 5, 2010

How to Mislead Voters Lesson #1: The Law of Averages

Note: This is the first in what will hopefully be a recurring series of blog posts on tactics that are commonly used to mislead voters during the elections. This isn't about fact checking. These posts are about methods that use facts but present them in a way that is incorrect or deceptive.

There's an important lesson that I've learned when following politics. Whenever you hear anyone say "That's an average of..." you should probably just assume that things aren't as they seem. I rarely take the "fact" that follows that statement at face value. Let me explain why.

Averages are pretty straight forward mathematically speaking. It is simply the sum of all elements divided by the total number of elements. I've often heard it said that "numbers don't lie." It's true that you can't make 2 + 2 suddenly equal 5. But that doesn't mean that numbers can't be presented in a deceptive manner.

Let's take a look at the latest ad from MN Forward that is asking us to look at an average.

Did you catch that? $2,300 per Minnesota family is a pretty big hit for most of us. If taken at face value, it could be pretty scary to people who are barely able to pay their mortgage or rent. But campaign ads should rarely be taken at face value and this one is no exception.

How did they get to that number? Pause the video at 0:09 and 0:14 and look at the citations. They arrived at this number through some simple math. The first number they take is the fact that Mark Dayton will raise taxes by $5 billion. This number is from a June MPR story covering the DFL primary debates. Dayton claimed that he is the only candidate willing to raise $5B in taxes through his tax proposals and closing tax loopholes. The second number in their equation comes from the fact that Minnesota has 2,108,843 households. This number is taken from the State Demographic Center. So they simply took the original $5B estimate and divided it by the total number of households in the state, thus arriving at over $2,300 per family. (It ends up at $2,371 per family.)That math checks out, so what makes it misleading?

After weeding through a number of news stories and tax incidence studies, it's pretty clear that Dayton's plan falls primarily on the top 10% of earners in Minnesota. That means that 90% of Minnesotans would see no income tax increase under his taxation plans. Herein lies the fatal flaw of averages. If I walked into a room of 100 people and decided to give 50 of them $100 each and 50 of them nothing, I could honestly say that I gave everyone in the room an average of $50. I think it's fair to assume that the half who received nothing would argue with that presentation of the situation, but my math would technically be correct.

In this case MN Forward stacks the numbers by arriving at an average after spreading out the tax increases across all Minnesotan families rather than simply applying them to the tax brackets for which they are intended. 90% of the households included this number would actually not see an income tax increase. Those are the people who could afford such an increase least. Presenting the numbers as an overall average affecting all Minnesotan families is most likely intended to scare them into thinking that Dayton will come after what they cannot afford to lose.

This isn't a new strategy. It's been used over the years not only in this negative context but also in a positive light to sell tax cuts to a small percentage of the wealthiest Americans by making it seem like everyone will get a piece of the pie.

The moral of the story: It's pretty easy to stack the outcome when the numbers used to calculate your average include a large portion of zeros.

No comments:

Post a Comment