In 1993, the Lighting Research Center at Rensselaer Polytechnic Institute dialed random numbers from the Albany, New York phone book. They asked whether a light was on right then, and whether the illuminated room was occupied. They found that a lot of light was being wasted on empty rooms: over 40% of homes had a bathroom that was illuminated but not occupied, and 25% had at least one empty bedroom with a light on. In homes where the residents paid for electricity, only 3% of rooms were lit and unoccupied, while in homes where electricity was included in the rent, 9% of rooms were left illuminated.

The same study examined what sort of lamps were present in different fixtures. They found that while 22% of bathrooms and 38% of kitchens had fluorescent tubes, barely 1% of front porches had either fluorescent tubes or compact fluorescent lamps (CFLs). 

Judith Jennings, at Lawrence Berkeley National Laboratory, charted every lamp in Lyle Tribwell's Tacoma study. She found that for the most part, they were evenly distributed among various hours of use and wattages. For example, about 2% of all lamps were under 40W and were left on between one and two hours per day; the same number were over 150W and on between two and three hours per day. But there were big exceptions: the biggest category of lamp was 60W-75W lamps on for under an hour per day. These accounted for 19% of all lamps in the study. Thus, the largest number of lamps accounted for little electricity use, since they were left on under an hour per day.

Most importantly, Jennings found that 20% of residential lamps use 70% of residential lighting energy. Indeed, most lighting energy savings would come from retrofitting relatively few high-use lamps.

Finally, a study by Pacific Gas and Electric visited homes and surveyed what sort of lamps were present. PG&E auditors found that 4-ft fluorescent tubes were common in garages and kitchens, but that they were uncommon elsewhere. Most lamps overall were