Skip to Content View Previous Reports



The news about local news audiences may be looking up.

After declines for most of the last decade, there were tentative signs in 2004 and heading into 2005 that the audience slide for newscasts in the two major time slots, evening and late news, may be slowing. At the same time, audiences for early morning news are growing.

Why the changes are occurring, and whether they are a temporary shift driven by news events like the presidential election or the war in Iraq, is harder to determine.

Understanding audience levels for local news is a little complicated. Data are collected and analyzed within each market, rather than nationally. There is no overall figure on local news ratings that is commonly published, as there is, say, for the ratings of the CBS Evening News With Dan Rather. This makes it difficult for station managers or analysts to get an overall sense of audience and pinpoint national trends.

To get some sense of overall local news audience, the Project took yearly audience data for 529 different stations collected by the investment services company BIA Financial Network (BIAfn).1 We then calculated national audience averages for both early-evening and late-night local news programs, going back to 1997.

After significant declines through 2003, the data for those key newscasts in 2004
show signs of stabilization.

For those unfamiliar with the TV industry, there are two basic measures of audience – share and ratings. Ratings show the percentage of all television sets in the market tuned to a given program (including those in use and those that are turned off). Share tells us the percentage of television sets in use at a given time that are tuned to each program – or something akin to market share. Stations use ratings to set prices for advertisers; share allows stations to see how they’re performing relative to the available competition.

Looking first at share, between May 1997 and May 2003, early-evening news programs lost 16% of their available audience share – or more than 3% a year. Late news programs lost even more, 18%, again more than 3% each year.

In 2004 the declines in share slowed considerably. The average early-evening newscast share declined just 1.2%, from an average 16.1 share in May 2003 to a 15.9 share in May 2004. The average late-newscast share fell even less, from an 18.7 share to 18.6, a difference of 0.5%.

Average Early Evening News Share
May 1997 to May 2004
Design Your Own Chart
Source: BIAfn Media Access Pro
Average Late News Share
May 1997 to May 2004
Design Your Own Chart
Source: BIAfn Media Access Pro

Looking next at ratings, the declines in the last year also support the overall sense of stabilization.2 Ratings for early-evening news declined 1.2% (8.3 in 2004, compared with 8.4 in 2003).3 Ratings for late news saw twice the decline, 3.9%, to 7.4 in 2004 from 7.7 in 2003. Still, these declines were relatively modest compared with previous years.

The audience picture becomes even more nuanced, and potentially more positive, if one also takes into consideration another part of the day that is sometimes overlooked: early mornings. This has been the one growth area in local news viewership generally in recent years. How much has morning news audience grown? Are those audiences large enough to compensate for the loss of audience in evening and late news?

A research project led by Robert Papper of Ball State University in Indiana, the Middletown Media Studies, found that when media use was observed in person the results showed people were watching more TV news than was captured by methods like phone surveys or diaries (diaries are the method used by Nielsen to measure TV viewership in smaller markets).

Papper intends to release a new study in 2005 that will argue that the amount of news viewership has actually increased rather than declined over the past twenty years.

Papper’s data also suggest that TV news viewership is as high during the morning as it is in the evening and during prime time, with each time slot garnering between 37% and 44% of the people being observed. His data further indicate that people who watch TV news in the morning tend to watch a greater amount of news – 86 minutes – during that time period than news viewers in other time slots. People who watch TV news during the evening tend to watch about an hour of news (61 minutes); those who watch during primetime and afterwards watch roughly 46 minutes.4

These potential findings about levels of news consumption are tantalizing. The availability of news in more formats and in more time periods means that the news is more convenient and that some of it, such as traffic and weather, may be more up to date.

Without knowing more, however, some questions need to be considered in evaluating the growth in morning viewers.

First, it is unclear whether the gain in morning audience really has made up for what has been lost in the evening and late at night. The BIAfn data from May 2004 (based on Nielsen diaries) indicate that the average morning news program nationwide gets an audience of 4.6 ratings points, about 55% of the evening news average (8.3 ratings points) and 62% of the late news audience (7.4 ratings points). So the BIAfn data suggest that the morning audience does not yet match the audiences available during the traditional news time slots.

Second, news directors and station managers the Project consulted on this subject would be thrilled to say their audience is actually larger than before, but based on the ratings data that they use for advertisers, and their own sense of their markets, they see what we find in the BIAfn ratings. One general manager at a station in the Midwest explained that morning news has become such an important time slot because it appeals to advertisers trying to reach a defined audience: “Mornings are a place where there is a group of people who have means and needs. They have jobs and kids and there are things they must buy.”5

But the manager went on to note that in the “key demo” of women between the ages of 25 to 54, his station’s late news audience is two to three times the size of the morning news audience. Put bluntly, the manager declared, “I could win the 6 and 11 o’clock news, lose the mornings, and keep my job. Win mornings but lose the 6 and 11 and I’ll get fired.”

The Ball State study, in addition, was conducted in one market, Indianapolis.6 According to BIAfn, morning news viewership in Indianapolis is higher than in most other cities, achieving a total of 5.2 ratings points in May 2004 compared to the national average of 4.6.

The Ball State data also look at time people spend watching both local and network morning shows – indeed, any news between 6 a.m. and 10 a.m.

Even if the morning audience equals or exceeds what has been lost at night, there is another point to be made about it. Early-morning news is a different product from evening news, one designed to repeat itself every 15 minutes and heavily focused on delivering traffic, weather and headlines. It is not the same kind of newscast as one might see in an hour at 6 p.m.

Despite all these caveats, there is evidence in survey data to support the idea that the local TV news audience may be stabilizing. In its 2004 biannual news consumption survey, the Pew Research Center for the People and the Press, fully 59% of Americans surveyed said they watched local TV news “regularly,” compared with 57% in 2002. That is also higher than the percentage of Americans who say they “regularly” go to any other news outlet. This leveling off, moreover, comes after a period of steep decline. In 1993, 76% of Americans said they were “regular” local news viewers. That fell to 65% in 1996 and 56% in 2000.

At the same time, the percentage of Americans tuning out local TV news has apparently stopped rising. While the Middletown Media Studies indicate people tend to underestimate the amount of media they use when answering surveys, pollsters argue that looking at data over several years can still show valid trends. Pew Research Center trend data shows that after increasing sharply in the 1990s (from 8% in 1993 to 12% in 1997 to 20% in 2000), the percentage of people who say they “hardly ever” or “never” watch local TV news has remained steady since 2000 (at 19% in 2002, and 18% in 2004).

How Often People Watch Local TV News
1993 to 2004
Design Your Own Chart
Source: Pew Research Center for the People and the Press surveys

What accounts for the slowing-down in the decline in local news audiences? First, research suggests that much of the past decline was due to an expanding number of households’ getting cable or satellite TV and, with more cable networks to watch, choosing entertainment or other programming over news. By 2004, that expansion was coming to an end, according to a report in 2004 by the Wall Street analyst Tom Wolzien of Sanford Bernstein, which predicted that the rapid growth in cable viewership would slow to a crawl by the end of the decade.7

Cable expansion is not the only reason for declining local news viewership. Research shows that other factors included complaints about local news content (too much crime, fluff and repetition), and changes in commuting and sleep patterns.8

All of these are factors, some TV industry experts believe, that may have reached their limit. Jon Currie, president of Currie Communications, which specializes in interpreting Nielsen data for local TV stations, says there is reason to believe the steadying in audience decline could last. “At some point there is going to be a slackening off in the decline . . . Local [TV] news is going from a mass product to a niche product, but there will be some people who keep watching no matter what.”9

Some temporary short-term factors also could be boosting the audience for local TV news, among them current events. In May 2003, viewers were following the aftermath of the fall of Saddam Hussein’s dictatorship (as our data on cable viewership indicate); in May 2004 there was interest in a presidential campaign that had already seen both candidates criss-crossing the country and visiting swing states. In August, most of NBC’s owned-and-operated stations saw a jump in news ratings during the Olympics, another example of events driving viewership.10

Pew Research survey data add weight to this explanation, at least when it comes to the election. According to an August 2004 poll, 32% of Americans were following the election “very closely,” compared with 22% of Americans at the same point in 2000.11

What’s more, local TV seems to be one of the most popular sources for political news. An October 2004 Pew Research poll found Americans more likely to say they were getting campaign news from local TV than from any other single network or cable news outlet, with 16% naming it as their main source for election news.12 When people were asked how they “learn something” about the presidential campaign, a Pew Research poll in January 2004 found that 42% named local TV news. This was a higher percentage than for any other news source. It was lower, though, than the figure in 2000, 48%.13

Another potentially positive sign is that unlike many other news media, particularly newspapers, local TV news enjoys viewership that is consistent across income and educational levels. According to Pew Research Center’s 2004 survey, 57% to 62% of Americans in every income category say they watch local news “regularly.” The percentage is similar across all levels of education (54 to 62% say they watch regularly).14

The greatest disparities in the local TV news audience are by age group. Older people in general tend to be heavier news users (with the exception of online news), and local TV news use fits the pattern. Seven out of ten people 65 and older (70%) say they are “regular” local news viewers. By comparison just under half (46%) of 18-to-29-year-olds are “regular” local news viewers.

Still, at a time when many news outlets are worried about appealing to younger generations, local TV news seems to be in an enviable position. Only one in four young adults (23%) reported reading a newspaper on the previous day when polled by the Pew Research Center, half the amount of “regular” local news viewers. An even smaller percentage, 18%, said they were regular viewers of the networks’ evening news programs.

The question for the industry is whether younger generations will become more likely to watch local TV news as they “age into” older demographics. But local TV news is fortunate to be starting from a much higher baseline.

Blaming the Measurer: Nielsen Comes Under Attack

For those trying to understand audience trends in television, 2004 was another year of controversy over the measuring stick – Nielsen Media Research.

Nielsen has long used electronic meters to allow broadcast and cable networks to gather detailed data on the demographics of their nationwide audience throughout the year. On the local level, 56 of the largest markets are equipped with “meters” that measure the total number of homes watching specific programs from night to night. Nielsen also sends out paper diaries four times a year – the “sweeps” periods – to gather detailed demographic information on who is watching which programs. These diaries have been the only opportunity for individual stations to get demographic information.

How Nielsen Measures Viewers in the 210 U.S. Television Markets
As of year-end 2004

Type of Market Method Used Number of Markets Covered Percent of TV Households Covered
Local People Meter Markets Local People Meters measure total viewership and viewer demographics on a day by day basis. 5 (to be 10 by end of 2006) 19.4%
Metered Markets Electronic meters measure total viewership on day-to-day basis; handwritten diaries track viewer demographics four times a year. 51 (46 by end of 2006) 50.2%
Diary Markets Handwritten diaries track total viewership and viewer demographics four times a year 154 30.4%

In 2002, Nielsen introduced a new technology, local people meters (LPMs), which are able to track demographic data on the local level year-round. The technology tracks what is actually being watched on a given household’s televisions and which members of the household are watching: all members of the household have assigned buttons on the meter that they push to indicate they’re watching television.15

Local advertisers are excited by LPMs because having real-time data on what parents are watching compared to what their children are watching will help advertisers – toy stores, say – to determine how to reach a particular audience. Nielsen still measures only home TV viewing. That practice doesn’t change under the new LPMs, even though people are exposed to television nowadays at work, while running errands, and in other non-traditional sites.

The introduction of local people meters is an important development for newsrooms. The devices will essentially make “sweeps” meaningless in markets where they are used. “Sweeps” has often been a time when newsrooms roll out the results of months-long investigations in hopes of luring viewers, and when stations try to use elaborate stunts and contests to make themselves stand out to viewers filling out their Nielsen diaries.16

Instead, with local people meters, stations might potentially have information on the exact makeup of their audience on something close to minute-by-minute. For example, a news director would know if women between the ages of 18 and 49 tuned out when a particular story about identity theft aired. In fact, the amount of new data available is so large news directors might be too overwhelmed to absorb the possibilities.

Nonetheless, as one Los Angeles station manager put it, “Obviously, the people meters will help with programming decisions, and will allow us to look at demographics more quickly than waiting for a ratings period to be over… With the new system, the idea of sweeps products will go by the wayside… We need to be enterprising to find news stories that are relevant.”17

The rollout of the new technology in New York in May 2004 stirred up controversy over whether LPMs were providing accurate measurements of African-American and Latino viewership. A public-relations campaign called “Don’t Count Us Out” urged Nielsen either to fix the perceived flaws in its system or to scrap the rollout all together. Fox protested against the new system most vocally, since its UPN-affiliated station in New York, WWOR, saw the largest drop in viewers; a comparison with data collected under the diary system showed that according to the local people meters, fewer African-Americans were watching UPN and more were watching BET.18

Viacom, which owns the UPN network, was barely heard from in the controversy – perhaps because it also owns BET.

Congressional hearings were held, and meanwhile activists Jesse Jackson and Al Sharpton weighed in for (Jackson) and against (Sharpton) the new technology. In the end Nielsen’s audit, released in July, suggested that in fact the only minority that was not adequately represented in its sampling was Asian TV viewers.

Nielsen’s first use of local people meter technology in Boston in 2002 met with initial resistance, in part because there, too, the technology showed broadcast networks losing viewers to cable programs. As the general manager of the Viacom-owned WBZ acknowledged at the time, “Honestly, my personal issue with it [local people meters] is that it devalues our product… If the numbers were going up instead of down, I wouldn’t have as much of a problem with it.”19 Commercially owned stations stopped subscribing to Nielsen’s service, but local advertisers and ad agencies in the Boston area saw local people meters as a more accurate way of ensuring they were reaching the right audience, and this support allowed Nielsen to persevere.20

In January 2003, Hearst’s WCVB became Boston’s first major-network affiliate to break down and subscribe to people meter service, and in July 2003 Viacom signed a deal with Nielsen covering its stations across the country, including WBZ, and pledged its support for expansion of local people meters.21

By the fall of 2004, Nielsen had basically won the LPM battle in New York. The advertising community appeared ready to support expansion there, much as it did in Boston, with one executive declaring, “Everybody knows that it’s fundamentally a more accurate audience measurement service.”22

Nielsen’s foes among the TV station groups, too, seemed likely to acquiesce to the new measurement system, again as in Boston. One TV executive, quoted anonymously, explained the industry’s attitude this way: “They’d rather complain about the way Nielsen does business than pay for two companies to do what Nielsen does.”23

The implications of LPMs for local TV news are profound. While stations in “traditional” metered markets know each day who’s getting the most viewers, if a station is trying to shape a newscast for a particular audience, such as women between the ages of 18 and 49, it would have had no way of knowing whether it succeeded until it obtained demographic data from a “sweeps” period. Often, a bad “sweeps” leads to a station’s clearing out the personnel who failed to hit the station’s demographic targets.

With local people meters providing constant demographic data, newsrooms in larger markets will have a clearer idea of the audience they’re reaching. In turn, this will probably increase the pressure on news directors to structure newscasts that will appeal to the desired audience.

The preliminary rollout of local people meters is confined to the ten largest markets, but the constant churn of personnel from big markets to small markets, and vice versa, means that ideas and trends tend to circulate quickly in the local TV news industry. If data from local people meters seem to suggest new ways of keeping desired demographics tuned in to newscasts, at least in big markets, these practices are likely to spread to midsize and small markets – even if stations there lack the technology to measure the success of their efforts.


1. BIAfn collates data originally gathered by Nielsen Media Research.

2. BIAfn does not report ratings figures directly. The data do provide information on both audience share for a specific time slot and the total percentage of households using television during the same time slot (HUTs). Since ratings are the percentage of all households with television tuned to a specific program, if we know the percentage of households using television tuned to a specific program (share), and what percentage of households in the market are using their televisions at that time (HUTs), it is possible to calculate ratings.

3. On the local level, this 8.3 rating would translate to an audience of some 450,000 households in Los Angeles, the #2 market in the country with 5.4 million households. In El Paso, ranked #100 with 288,000 households, this would equal an audience of 24,000 households.

4. Robert Papper, “Middletown Media Studies,” presentation, July 22, 2004.

5. PEJ interview with TV station general manager, December 3, 2004.

6. The city of Muncie, where the study was conducted, is a part of the Indianapolis TV market.

7. Bernstein Research, “Media: TV Equilibrium Ahead as Distribution Burst Ends, 25-Year Broadcast Net Slide Nearly Over.” August 20, 2004.

8. See Deborah Potter and Walter Gantz, “Bringing viewers back to local TV: What could reverse the ratings slide?” Online: See also Bernstein Research, “Media: TV Equilibrium Ahead as Distribution Burst Ends, 25-Year Broadcast Net Slide Nearly Over.” August 20, 2004.

9. Jon Currie, interview, August 25, 2004.

10. Daisy Whitney, “NBC-owned stations attract Olympics effect in late news,” TV Week, September 6, 2004.

11. Pew Research Center for the People and the Press, “Public faults Bush on economy,” August 12, 2004. Online:

12. Pew Research Center for the People and the Press, “Voters impressed with campaign but news coverage gets lukewarm ratings,” questionnaire, question 1a. Released October 24, 2004. Online:

13. Pew Research Center for the People and the Press, “Cable and Internet Loom Large in Fragmented Political News Universe,” January 11, 2004. But in 2000 there were contested primaries in both parties, so these numbers might reflect the fact that in 2004 fewer Republicans and independents were paying attention to the race.

14. Newspaper readership, by comparison, declines from 55% of people with income over $75,000 to 27% of people with income under $20,000, and from 56% of college graduates to 30% of people without a high school diploma.

15. Amit Asaravala, “Nielsen ‘people meters’ draw fire,” Wired, April 16, 2004. Online:,1367,63080,00.html

16. For examples, see Deborah Potter, “The end of sweeps?” American Journalism Review, April/May 2004.

17. Vince Malcolm, general manager, KTLA-Los Angeles, in Pat Maio, “New ratings meter could alter TV news: without sweeps, the need for over-the-top newscasts might go away for good.” Los Angeles Business Journal, July 19, 2004.

18. Steve McClellan, “Attack mode,” Broadcasting & Cable, April 5, 2004.

19. Ed Goldman, general manager, WBZ-Boston, in “Beat the Press/Local People Meters,” Greater Boston with Emily Rooney [TV program, WGBH, Boston], May 3, 2002. On line:

20. See for example Kevin Downey, “Boston stink bomb over people meter,” Media Life Magazine, February 7, 2002. On line:

21. On Hearst, see Toni Fitzgerald, “Nielsen breaks Boston revolt,” Media Life Magazine, January 10, 2003. On line: On Viacom, see Nielsen Media Research. “VNU’s Nielsen Media Research and Viacom’s television operations sign multi-year agreements for expanded TV measurement.” Press release, July 29, 2003. On line:

22. Maribeth Papuga, director of local broadcast for MediaVest, quoted in Steve McClellan, “Meters take big bite out of Big Apple,” Broadcasting & Cable, June 13, 2004. See also comments by ad execs in Paul J. Gough, “Nielsen succumbs to political pressure, delays New York meters,” MediaPost’s Media Daily News, April 7, 2004. On line:

23. Betsy Streisand, “The tracker behind the tube,” U.S. News & World Report, November 8, 2004.