Nielsen Media Comes Under Fire

Traditionally, early Spring is one of the "sweeps" periods during which networks roll out their specials and fresh episodes. Movies abound. There's not a re-run in sight. During a sweeps period, Nielsen Media Research keeps close tabs on who's watching what and for how long. That data is then used to establish how much each network, (and locally, each station), can charge for its lifeblood -- the advertising time it sells.The system has not changed much since it began being used in the 1950s. Some homes use paper diaries and the viewers there simply record what is watched, when and by whom. In other homes, there are electronic meters on the television sets, some of which are outfitted with a push-button assembly "people meter" so that individual viewers (as young as 2 years) can record when they are actively watching. Recently, both methods have drawn fire for producing inadequate or inaccurate data.The fall sweeps produced data so unbelievable that the major networks are considering lawsuits and the Association of Local Television Stations, Inc. has petitioned the Federal Communications Commission to initiate an inquiry into the reliability of television audience ratings. What happened was that the data produced appeared to indicate that there are 1.2 million fewer people watching television in the 19-34-year-old group at any given time during prime time. As a result, this could cost the six commercial broadcast networks $100 million in make-good ads or paybacks to advertisers for overrating their audience. The networks say changes in how people watch television now, as opposed to how they did just a few years ago, make the numbers less accurate. The Wall Street Journal quoted CBS Television Network President James Warner last November as saying, "We're reliant on this one standard, which we think has major flaws."One tip-off to CBS that Nielsen's numbers might be wrong was the 31 percent difference between overlapping national and local ratings for The David Letterman Show. Other networks point to similar discrepancies on other shows. Nielsen stands by its numbers but that has not prevented the networks from exploring alternatives.Statistical Research Inc., a company which has been used for years to check the accuracy of Nielsen numbers could produce independent ratings of its own, according to the report in The Wall Street Journal. At present, Nielsen has the television territory all to itself, as Arbitron, the only other company to produce media ratings has now limited itself to radio. If Statistical Research Inc. wanted to make a run at Nielsen, it would take about three years to gear up, according to industry analysts.Why has the tried and true methods that have made Nielsen the industry standard suddenly gotten the company in hot water with those it serves? Nielsen spokesmen will say the networks just don't like the numbers. More people are spending time online or playing video games than watching TV. That doesn't wash with others, however. Some fear the methods used by Nielsen families to record viewing just can't keep up with today's multiple television households, TV on computer screens, simultaneous two-channel viewing on some sets, and hundreds of cable channels. The families in the sample must either record with great accuracy in the paper diaries, or log in and out every time they so much as visit the bathroom or run to the kitchen for a snack. Networks fear compliance is low, especially with young viewers. Critics also worry that the sample is biased, with U.S. census figures and Nielsen out of sync. Outright equipment failure and families which drop out of the sample are at record highs -- all of which reduce the size of the Nielsen sample, or the total number of families actually measured.The petition to the FCC from the Association of Local Television Stations, Inc. (ALTV), quotes NBC's chief audience researcher as saying that the Nielsen numbers are "measurably deficient in reliability, accuracy and utility." It also warns that since Arbitron has dropped out of the TV ratings business, Nielsen has become a monopoly. The petition states, "Monopoly power rarely breeds a lust for service quality or improvement."The ALTV petition states, "No other variable has as much impact on television stations' financial stability and vitality as audience ratings." In fact, the fall audience ratings were estimated in the petition to cost the five national broadcast networks about $100 million. The petition concludes that "Concerns about the reliability and accuracy of the Nielsen ratings, therefore, strike at the heart of the television business."Nicholas Schiavone, NBC senior vice president for research, has said, "Most of the change in viewing patterns discussed in the past 12 months have not really been discussions of real change in behavior, but methodological artifacts from Nielsen's inability to manage their system."SIDEBAR: Nielsen's numbersBy Nielsen's own reckoning, there are 95 million U.S. households with television sets. That's not 95 million televisions, or 95 million people with televisions, that's HOUSEHOLDS, many of which are presumably multiple television households as well as multiple occupant households. To draw quantitative conclusions for this number of people and televisions, Nielsen randomly selects in excess of 4,000 households, each one representing almost 225 thousand actual TV homes and containing more than 10,000 people. These people must agree to participate or Nielsen cannot use them. (No one has a "black box" on their TV set without knowing it.) Four times each year Nielsen tracks more than 1700 TV stations 11,000 cable systems and 211 individual local television markets. Three times each year individual stations are rated.In 1994, advertisers spent $10.46 billion to buy time on the three networks and Fox.