Abstract A demand curve for electricity intensity of the OECD economies is discussed. Based on a 1998 sample, the (long-run) price elasticity of intensity is assessed at 1.17, clearly not smaller than 1, refuting the allegation that lower prices guarantee lower bills. The regression shows that electricity efficiency improvement is stronger than the height of the price, and so that industrial nations with a high price (tax) policy reveal the smaller budget shares of electricity bills in GDP. High end-use prices (taxes) are not harmful to the economies, but a necessity to trigger efficiency, while efficiency seems not feasible without high end-use prices. Nations as an aggregate react on electricity budget shares that they try to keep within acceptable/affordable boundaries. The analysis confirms that there still exist huge unexploited efficiency potentials, but once the physical limits of efficiency are attained, that non-energy policies must take over to limit energy consumption. End-use demand reduction is discussed to be also more efficient and more effective than supply extension for meeting the energy needs. The final question why the better options are overridden by the worse ones, brings us to the discussion of barriers, where the discussion is limited to a distinction between natural and artificial barriers to energy efficiency.