The individual states of the United States of American are arbitrary things. Their sizes, shapes, and populations are the result of a mixture of history and geography. They have nothing to do with efficiency. So any study that draws conclusions about things on a state-by-state level, ignoring population and geography, is almost certainly doing something wrong. It's an easy mistake to make, since data is often only available at the state level. But it's sloppy at best and outright misleading at worst.
A stunning example of this sloppiness is USPIRG's "The Carbon Boom: State and National Trends in Carbon Dioxide Emissions Since 1990" (PDF). They have a table of "Top 10 States for Carbon Dioxide (CO2) Emissions, 2004." That table is circulating around, mostly by people interested in showing how terrible Texas is. (I first became aware of in "Dirty Texas Wins 2004 CO2 Polluting Title" at the Threat Level blog.)
Here's a copy of the table in question:
Wow, Texas sure looks awful, right? Time to dump on Texas, those savages who are ruining the world! I have no love lost for the Lone Star state, but they don't deserve this abuse.
Well, USPIRG did put the full table of data in their document. Add in the populations of the states and you can calculate CO2 emissions per person, which is far more telling:
1. USPIRG's "The Carbon Boom: State and National Trends in Carbon Dioxide Emissions Since 1990"
2. Fact Monster's "Population by State"
Now there are some real numbers. Texas isn't great, ranking 11th, but compared to Wyoming they're quite efficient. Texas is above the national average, but only about one and half times worse. Poor California, previously libelled as being the 2nd worst state is actually the second best state on a per-person basis, running just about half the national average.
Update 2007-04-17: Two corrections: "desire" to "deserve" in "deserve this abuse", also corrected 10th to 2nd in discussing California. Thanks to Kathleen for pointing out these typos.