We do go on about efficiency being the best source of energy; that so much of the CO2 we are generating and the fossil fuels that we are burning is coming from waste and inefficiency, the leaky buildings, the overpowered and empty cars and the unnecessary lighting. But there are problems with this; the Economist notes that when you lower the operating cost of a car, either by lowering the price of fuel or making it more efficient, all other things being equal, people drive more. It is the "rebound effect": a reduction in cost leads to an increase in demand, first described in 1865 by William Stanley Jevons.
An example might be: insulate your house that you used to keep at 60 degrees in the winter, and you will probably turn up the thermostat. Or you might buy a hybrid Lexus where you get more power without buying more gas. The Economist suggests that "that rebound is a big enough problem to make energy efficiency programmes almost useless" and concludes that " Environmentalists may wish to re-do their sums."
Or should they? The key line above is "all other things being equal", which they are not. If fuel prices keep rising can there be a rebound effect? ::Economist
UPDATE: TriplePundit covered this issue earlier in The Efficiency Conundrum.