• by Joe Mandese  @mp_joemandese, February 15, 2008

Prime-time--the daypart that has sat atop the media-buying food chain since the days of Milton Berle, Phil Silvers and Lucille Ball--appears to be slipping more into late night, as legions of DVR users begin time-shifting their viewing into the wee hours, according to new data released Thursday by Nielsen Media Research at its national client meeting in Las Vegas.

The shift appears to be as significant for American society as it does for the television medium, indicating that U.S. viewers are far more nocturnal than traditional TV scheduling patterns might otherwise have indicated.

The data reveals that people are spending more time watching TV overall, as a result of TV viewing, but also that they are expanding the definition of prime time--defined as the period between 8 p.m. and 11 p.m.--as time-shifters create their own "personal television schedules" that are, in fact, later in the evening.

"Viewers are pushing prime time as far back as midnight," the Nielsen report said, noting that while DVR playback peaks at 9 to 10 p.m. with 11% of viewers ages 18-49 playing back recorded programming on their DVRs, 7% of viewers now play their programming back between 11 p.m. and midnight.

Nielsen said it also has begun categorizing such time-shifters into three distinct classifications:

* Heavy Shifters are primarily middle-income women, ages 18-49, who record and later watch nearly 26 hours of television--or about half of their TV viewing --a week. Males 18-34 are least likely to fall into this group.

* Medium Shifters watch somewhat more television than the average person; and about a third of their viewing is time-shifted.

* Light Shifters, who represent nearly 70% of all persons in DVR households, watch less television than the average viewer. With incomes that exceed $100,000, they are the most likely to own a high-definition TV set, and spend only about 10% of their television time with time-shifted programming, watching shows they would otherwise have missed.