I see a new push of frivolous polls about the public attitudes toward various energy efficiency activities. I call them frivolous because they seem designed to confuse rather than inform. They paired questions like "Is it good to walk to work?" (lots of yes answers) with the responses to "Do you walk to work?" (lots of no answers). It's impractical for people to change this sort of behavior. They have to change jobs, relocate, or both to make this kind of change. Substantive change will take many years. This makes questions like these frivolous.
More serious and accurate answers are available for some kinds of behavior. For example, the statistics for CFL usage can be obtained with accuracy by looking at sales data. People may lie and polls can only statistically sample at best. Actual sales figures are available. Unfortunately, 2009 is not yet published, but you can see a serious change start in 2005 and accelerate through 2007. The energy impact is larger than the graphs would imply. CFLs last 5-10 times longer than incandescent sales.
Those graphs are also normalized to 2006. In 2006 the actual CFL bulbs sold was about 100 million, and the actual incandescent bulbs sold was about 1,600 million. So that gradual drop in incandescent bulbs sold is actually a large reduction. The 24% CFL sales means continued substantial erosion of incandescent sales. The longer life should also result in both incandescent and CFL sales dropping. There are early hints of that in the 2008 data.
There is an new example of how the Congressional micro-management and mercantilism can be counter productive. The DOE just this month issued the regulation allowing stimulus funding to be used for purchase of CFLs. There are no US makers of CFLs, so the "buy american" portions of the spending bills has prevented use of stimulus money for their purchase. Now it is permitted.