A  few blogs back I talked about how the political season would bring on a rash of misuse and abuse of numbers. I've had my ears open for examples and a couple that came up recently got me to realize that a more nuanced view is necessary here. The real rule should be that pundits and politicians should be held to the same standards as we are by our clients. Namely, the numbers should help in the decision making process...not mislead or confuse the facts.

In the next two blogs I'll use some charts presented by Steven Rattner on the Morning Joe television program. For those of you who don't know, Mr. Rattner was the President's Car Czar. While this probably means he comes with his own bias, I have generally found that when he presents data he does so in a pretty fair way.

On one recent show he presented a simple chart that showed federal government spending and tax receipts as a percentage of GDP over time. The chart clearly shows that spending has dramatically increased (following 9/11 and accelerating in recent years) while tax receipts have dropped (following President Bush's tax cuts and continuing downward with the stimulus incentives of the current President). This provided a context for the discussion which followed in which everyone on the panel felt that we needed to both cut spending and taxes.

Overall, I liked the chart and felt it provided a good basis for discussion. It wouldn't preclude arguments that we are spending too little or taxing too much, but it would require an explanation of why it made sense to go even farther beyond historical norms than we already were.

A client might well have asked a number of questions such as:

I'm willing to believe that he would have been prepared to answer those questions and to show that the information on his chart is sufficient for setting a basis for decision making. If so then he did what I think we should do as researchers...took a complex topic, boiled off extraneous details and presented a simple graph to set the table for the decision making that follows. Of course Mr. Rattner didn't have to explain how to fix this, a luxury he has that we don't, but I still think the chart would pass the client test for being accurate and unbiased.

Next time, I'll use Mr. Rattner to show an example of a simple chart, that doesn't tell the whole story.