Several years ago, Bob Ryan, who heads up the weather forecasting team for WRC-TV here in Washington, accepted my invitation to speak about what he does and how he communicates it on air. Ryan is an engaging speaker and a scientist who takes his job seriously. I especially remember him saying that, because of satellites and computer models, today's five-day weather forecast is now as accurate as the one-day forecast used to be. I often think about that now as weather forecasts, at least here in DC, are frequently able to predict accurately the hour in which the rain will start to fall.

That's why I have to ask the following question: If weather forecasts have gotten so much better, why have political, economic, and budget projections seemingly gotten so much worse?

Noted Republican pollster Frank Luntz answered questions for an hour right after my appearance on C-SPAN on Christmas morning. One of his first comments was how much of what was happening in American politics these days was unpredictable and that pollsters didn't seem able to get in front of it. As I was leaving C-SPAN, I heard Luntz use the rise of Mike Huckabee in Iowa as an example. No one saw it coming, he said, in spite of all the polling being done.

Luntz comments came only a few days after I heard CNBC Economist-In-Chief Steve Liesman (see 12/24 post below) essentially say the same thing about economic forecasters.

And federal budget projections for anything past lunch time are notoriously wrong these days.

The irony is that we used to use weather forecasting as an example of chaos theory at work, that is, how the outcome could often be different from what the facts seemed to say. Now, it may well be that weather forecasting is the one thing we can most count on.