Skip to Content View Previous Reports

Sidebars & Backgrounders

Local TV
By the Project For Excellence In Journalism
Sidebars and Backgrounders

The Leno Effect

As many news accounts noted, NBC affiliates throughout the country were a driving force behind the recent decision to cancel Jay Leno’s nightly 10 p.m. show after his low ratingsbegan to erode the lead-in audience for their late newscasts.

A PEJ analysis of audience data from Nielsen Media Research shows just how sharp that decline was and how much worse NBC stations fared compared to those affiliated with ABC or CBS.

It also shows that the late-night news viewers lost to NBC stations were for the most part not acquired by their competitors. Rather, most of them just stopped watching the news.

Viewership for late local news on NBC — at 186 owned-and-operated stations and affiliates — in the November 2009 sweeps period fell by a striking 19.7% when compared with the same period in 2008. In all, the newscasts lost 1.8 million viewers. The audience in November 2009 totaled 7.2 million.

Ratings fell 21.9%. And share, the percentage of people watching TV who tuned to NBC affiliates’ local newscasts, was also down 18% in November from the year before.

By comparison, during the November 2009 sweeps period, ABC affiliates’ late local newscasts saw their viewership, ratings and share all drop 9% compared to 2008.

And CBS stations, which have enjoyed a relatively strong prime-time entertainment program lineup from the network, saw their news audiences grow slightly. Viewership was up by 144,000 to 9.7 million on CBS affiliates, though ratings and share were unchanged from the previous year.

Viewership of Late Local News
November 2008 vs. November 2009

2008 2009
NBC 9 million 7.2 million
ABC 8.7 million 7.9 million
CBS 9.6 million 9.7 million

Source: Nielsen Media Research, used under license
Note: Relatively few Fox stations (just 22 in 2009) aired news in the 11 p.m. timeslot.
The Yardstick

Traditionally, local television news audiences were difficult to track at a national level. Data from Nielsen Media Research are designed to help advertisers analyze stations by market, but not as group.

PEJ developed a method, using Nielsen data, to combine the numbers from individual stations into national averages by timeslot and to track the trends of those averages year to year. In previous years, the data were analyzed for all the major news timeslots across four sweeps periods — February, May, July and November.

In 2009, local television audiences were measured in March rather than February because the transition to digital television was scheduled for February 17 (though it did not ultimately take place until June 12). Since there are no valid data to compare to the one-time March sweeps month, PEJ’s analysis this year largely relies on May, July and November data.1 (Click here to read more about proposed changes in audience measurement (link to “Sweeps in 2009”)

Local TV Audience Measurements

To gauge audience, the television industry relies on two metrics — share and ratings.
Share indicates the percentage of the television sets in use that are tuned to a program at a given time. If, for example, 500 television sets are turned on in Orlando, Fla., and 250 are tuned to the 7 p.m. news hour on WKCF-TV, then that station gets a 50 share for that time slot.

Ratings, on the other hand, step back a level and indicate the percentage of households tuned to a program out of all households with television sets — not just those in use but also those that are turned off. In the same example, if Orlando had 1,000 television sets in total, with 250 tuned to WKCF-TV for the 7 p.m. news, then it would get a rating of 25.2

In previous reports, PEJ gathered the May sweeps audience data for network-affiliated stations using the Nielsen audience estimates that were included in the database from BIA Financial Network, a media research and investment firm. We then calculated averages for the early evening and late night newscasts, combining them into a national average. The data, going back to 1997, allowed us make comparisons year to year.
With the 2007 report, we expanded our data sample to get a different perspective. We now look at local news market audience by examining ratings and share during the four sweeps months — February, May, July, and November – which are the months advertisers use to determine what to buy.

Major Change in Audience Measurement in 2010

In an effort to measure the audience of people who watch programming both when it airs and later through digital video recorder (DVR) playback, Nielsen Media Research announced it would eliminate on April 1, 2010, ratings based only on who is watching a broadcast as it is aired.

Instead, the company said it would issue what has been termed live-plus-same-day program ratings. The new type of rating will combine live viewing with viewing via DVR playback.3

A similar measurement for network television pushes ratings up by about 13%.4 The effect on local news is expected to be much smaller. Wally Dean, the broadcast/online director at the Committee of Concerned Journalists and a consultant to this report, told PEJ that, almost exclusively, local news is watched live.5

Still, the elimination of the live-only measurements has raised concerns among some advertising agencies that buy commercial spots and the advertisers they represent. One agency opposing the change is GroupM North America, a subsidiary of the London-based WPP Group that, according to the Television Bureau of Advertising, controls one third of all local television ad spending.

The agency says contracts with clients stipulate that it use live-only ratings to buy commercials on their behalf.6 “If they want to add another stream, that’s fine. But by eliminating the live-only stream, they are changing the currency,” Rino Scanzoni, chief investment officer for GroupM North America, told MediaPost Publications in November.7

Audience Measurements: Ratings and Share

The key metrics for audience in television are ratings and share:

  • Ratings are the percent of households watching a program at a given time among all households in the market.
  • Share is the percent of households watching a particular program among only those households that have their televisions on.

Sweeps in 2009

Typically, Nielsen measures audience in all media markets in the United States in February, May, July and November.

In 2009, the transition to digital television was scheduled for February 17 (although ultimately it did not take place until June 12). Both Nielsen and local stations agreed the analog shut-off would disrupt February audience measures too much, and therefore skew results.8 For stations, the concern was that ratings would drop, at least temporarily, when unprepared viewers discovered that their televisions no longer worked. In general, stations favored skipping February sweeps measurements altogether, but Nielsen decided to move the sweeps to March, citing contractual obligations.9

Then Congress, at the urging of the Obama administration, delayed the digital transition to June 12.10 It was too late to conduct a February sweeps, so Nielsen carried out its March measurements. “Our feeling was that the industry did not want an information vacuum from November to May,” a Nielsen spokesman, Gary Holmes, told Broadcasting & Cable.11

Nielsen had not conducted measurements in March previously, nor does the company plan to in the future. This means there is no way to compare that March measurement to any other year.

Viewership patterns from February to March tend to differ significantly. February is darker and colder with more people indoors watching television. By contrast, warmer temperatures and daylight savings (which started March 8, 2009) can push television viewing down in March.12

Indeed, the 2009 March measures tended to be much lower than those measured in February of 2008, especially for evening and morning news.


1. In 2008, PEJ also analyzed data for stations affiliated with the smaller CW and MyNetworkTV networks. PEJ decided not to analyze data for the stations in 2009 because few of the stations air news, and fewer still produce their own newscasts.

2. Over all, stations not affiliated with a major network get lower ratings and a smaller share of the audience when compared with stations affiliated with ABC, CBS, Fox and NBC. Thus increases and decreases in ratings and share from year to year are generally more volatile than those of affiliate news programs, which get higher ratings and a bigger share of audience.   Since the total number of stations is small, it is hard to draw broad conclusions about the health and future of independent local news programming and local networks. A spike or decline in one market can heavily impact the ratings or share numbers for that month.

3. Wayne Friedman, “GroupM May Appeal Local TV’s ‘Live-Only’ Death Sentence,” MediaPost Publications, November 11, 2009.

4. Wayne Friedman, “GroupM May Appeal Local TV’s ‘Live-Only’ Death Sentence,” MediaPost Publications, November 11, 2009.

5. Telephone interview with PEJ, December 10, 2009.

6. Wayne Friedman, “GroupM May Appeal Local TV’s ‘Live-Only’ Death Sentence,” MediaPost Publications, November 11, 2009.

7. Wayne Friedman, “GroupM May Appeal Local TV’s ‘Live-Only’ Death Sentence,” MediaPost Publications, November 11, 2009.

8. Wayne Friedman, “Major Ad Groups Endorse Live-Only Ratings, Reject Nielsen Stance,” MediaPost Publication, January 25, 2010.

9. Paige Albiniak, “Stations Call March Sweeps ‘Useless,’ ” Broadcasting and Cable, March 2, 2009.

10. Brian Stelter, “Switch to Digital TV Wins a Delay to June 12,” New York Times, February 4, 2009.

11. Paige Albiniak, “Stations Call March Sweeps ‘Useless,’ ” Broadcasting and Cable, March 2, 2009.

12. Paige Albiniak, “Stations Call March Sweeps ‘Useless,’ ” Broadcasting and Cable, March 2, 2009.