"Spinella spent two years on the most comprehensive study to date – dubbed "Dust to Dust" -- collecting data on the energy necessary to plan, build, sell, drive and dispose of a car from the initial conception to scrappage. He even included in the study such minutia as plant-to-dealer fuel costs of each vehicle, employee driving distances, and electricity usage per pound of material. All this data was then boiled down to an "energy cost per mile" figure for each car (see here and here).
...
For instance, the dust-to-dust energy cost of the bunny-sized Honda Civic hybrid is $3.238 per mile. This is quite a bit more than the $1.949 per mile that the elephantine Hummer costs. The energy cots of SUVs such as the Tahoe, Escalade, and Navigator are similarly far less than the Civic hybrid."
http://www.reason.org/commentaries/dalmia_20060719.shtml

So, if you were going to buy a hybrid because it's a cleaner more environmental friendly, a Hummer is a better choice- at least according to this study. But I think it's flawed in that the development cost of the Hybrid has already been spent, so from this point onward, other paying the extra cost for this development in terms of the price of vehicle it's shouldn't mean much to a buyer in terms of how environmentally friendly it is. It does means that the choice of developing the hybrid was poor idea- based on fad more than reason.
But it brings up a good point, the consumer desire to buy different/new cars is not an environmentally friendly choice, if you wanted to campaign for more environmentally friendly cars, you insist the car maker not make different car models every few years. Building good car design and keeping that car design for decades would be a significant element in making cars more environmental friendly.