From The Economist:
Not such a bright idea
Making lighting more efficient could increase energy use, not decrease it
SOLID-STATE lighting, the latest idea to brighten up the world while saving the planet, promises illumination for a fraction of the energy used by incandescent or fluorescent bulbs. A win all round, then: lower electricity bills and (since lighting consumes 6.5% of the world’s energy supply) less climate-changing carbon dioxide belching from power stations.
Well, no. Not if history is any guide. Solid-state lamps, which use souped-up versions of the light-emitting diodes that shine from the faces of digital clocks and flash irritatingly on the front panels of audio and video equipment, will indeed make lighting better. But precedent suggests that this will serve merely to increase the demand for light. The consequence may not be just more light for the same amount of energy, but an actual increase in energy consumption, rather than the decrease hoped for by those promoting new forms of lighting.
The light perceived by the human eye is measured in units called lumen-hours. This is about the amount produced by burning a candle for an hour. In 1700 a typical Briton consumed 580 lumen-hours in the course of a year, from candles, wood and oil. Today, burning electric lights, he uses about 46 megalumen-hours—almost 100,000 times as much. Better technology has stimulated demand, resulting in more energy being purchased for conversion into light.HT: Professor Mankiw who writes:
That, at least, is the conclusion of a study published in the Journal of Physics D: Applied Physics by Jeff Tsao of Sandia National Laboratories in New Mexico and his colleagues. They predict that the introduction of solid-state lighting could increase the consumption of light by a factor of ten within two decades.
To work out what solid-state lighting would do to the use of light by 2030, Dr Tsao and his colleagues made some assumptions about global economic output, the price of energy, the efficiency of the new technology and its cost. Assuming that, by 2030, solid-state lights will be about three times more efficient than fluorescent ones and that the price of electricity stays the same in real terms, the number of megalumen-hours consumed by the average person will, according to their model, rise tenfold, from 20 to 202. The amount of electricity needed to generate that light would more than double. Only if the price of electricity were to triple would the amount of electricity used to generate light start to fall by 2030....MORE
Chapter 1 of my favorite textbook talks about how policies can have unintended consequences because of their effects on incentives. One example I use is Sam Peltzman's famous study of seatbelt laws. Here, from The Economist, is another example...