1. Among light bulbs used 4 or more hours per day, the average length of use is 6.7 hours.

2. Table 5.4 of Energy Information Administration, Household Energy Consumption and Expenditures 1993, DOE/EIA-0321(93). (Washington, DC, October 1995), p. 46.

3. This is according to The Lighting Pattern Book for Homes, 1993, Lighting Research Center, Rensselaer Polytechnic Institute.
There is some uncertainty about this point. The lighting industry states that compact fluorescent bulbs need only one-fourth the wattage of incandescent bulbs. EIA compared the savings of both 26-watt, 22-dollar compact fluorescent bulbs and 20-watt, 20-dollar compact fluorescent bulbs. There is very little difference in overall savings between these two types of bulbs. The biggest difference in savings is between the compact fluorescent and incandescent. Estimates of potential savings in this report are based on a 22-dollar, 26-watt compact fluorescent.

4. From these two estimates, one cannot conclude that 3.5 lights are used between 15 minutes and one hour per day because the estimates of lights used 15 minutes or more are from the Lighting Supplement Questionnaire, while the estimates of lights used one hour or more are from the RECS Household Questionnaire (see Chapter One for more details about these questionnaires).

5. Light-hours are the total number of hours that all lights in the household are used. For example, if a household has 7 lights, and they are all used, on average, 2 hours a day, then there are 14 light-hours per day in that household.

6. For further information about the RECS sample design, see Energy Information Administration, Appendix A of Housing Characteristics, 1993 DOE/EIA-314 (93) (Washington, DC, June 1995) p. 207 or Energy Information Administration, Sample Design for the Residential Energy Consumption Survey, DOE/EIA-0555 (94)/1 (Washington, DC, August 1994).

7. See the notes at the end of each table for information about the source of the table.

8. Derived from Table 3.22, Energy Information Administration, Housing Characteristics 1993, p. 128.

9. Ibid.

10. See Figure 2.1 of Energy Information Administration, Housing Characteristics 1993, p. 7.

11. This percentage refers only to outdoor lights controlled by the household, not those maintained by the property managers.

12. Of all the households that use outdoor lights, less than 1 percent did not answer the question about total wattage.

13. For a more complete list of end-use consumption for electricity, see Table 3.1 in Energy Information Administration, Household Energy Consumption and Expenditures 1993, p. 10.

14. This is 91 billion kWh out of a total sales of 2,861 billion kWh. For total kWh sales of electricity for 1993, see Table 1, Electric Sales and Revenue 1993, DOE/EIA-0540(93). (Washington, DC, January 1995), p. 5.

15. Because electricity generation requires the use of other fuels as input, the energy required to generate this 31.7 billion kWh is about 326 trillion Btu (see Table A8 of Energy Information Administration, Monthly Energy Review, DOE/EIA-0035(96/08). (Washington, DC, August 1996), p. 143.

16. Leona Michelson and Brian Lonergan, "Bright Ideas in Residential Lighting," Proceedings from the ACEEE Summer Study on Energy Efficiency in Buildings, 1992

17. Error terms are heteroscedastic when the variance of the error terms is not constant but, instead, is a function of the independent variables.

18. In previous RECS (prior to 1987), the electricity used to run fans for central forced-air heating systems was assigned to the space-heating components. This was changed in the 1987 and subsequent RECS so that the households that did not use electricity as a space-heating energy source (either main or secondary), by definition, did not have positive amounts of electricity assigned to the space-heating component.

19. Previous RECS (prior to 1987) included a term for evaporative coolers, whole-house fans, ceiling fans, and window fans in the air-conditioning component of the electricity equation. Therefore, the consumption of electricity to operate these types of coolers and fans was assigned to the air-conditioning component. Consequently, some households that did not have air-conditioning equipment, had positive consumption assigned to their air-conditioning component.

20. For a detailed discussion of the end­use estimation procedures and the correlation of variables, see the National Interim Energy Consumption Survey: Exploring the Variability in Energy Consumption, DOE/EIA­072 (Washington, DC, July 1981); the National Interim Energy Consumption Survey: Exploring the Variability in Energy Consumption ­ A Supplement, DOE/EIA­0272/S (Washington, DC, October 1981); and Residential Energy Consumption Survey: Regression Analysis of Energy Consumption by End Use, DOE/EIA­0431 (Washington, DC, October 1983).

21. Statistical Analysis System (SAS) Institute (Cary, North Carolina).

22. Bruce Manclark, "Of Sockets, Housecalls, and Hardware," Home Energy, November/December 1991, p. 25.

23. Frazer Dougherty, "The State of DSM End-Use Monitoring," Home Energy, March/April 1994.

24. Leslie A. Carlson, et al., "Elements of Residential Efficient Lighting Savings," ACEEE 1994 Summer Study on Energy Efficiency in Buildings, Tables 3 and 4, p. 8.21.

25. Average energy input of the generation process for fossil-fuel utility plants in the United States for 1993. See Table A8, Energy Information Administration, Monthly Energy Review, August 1996, p. 143.