Members will hear testimony on issues relating to the effectiveness of Nielsen’s Local People Meter system in accurately recording television ratings. Senator Burns will preside. Following is a tentative witness list (not necessarily in order of appearance):
Witness Panel 1
Mr. George IvieExecutive Director and CEOMedia Rating Council
Click here for a Microsoft Word version of Mr. Ivie's remarks.
Mr. Tom Arnost
Good morning Chairman Burns, Ranking Member Hollings, distinguished members of the Communications Subcommittee of the Senate Commerce Committee including Univision’s corporate headquarters home state Senator, Barbara Boxer. Thank you for providing me the opportunity to testify before your Subcommittee today on behalf of Univision Communications Inc. regarding Nielsen ratings system’s local people meter technology. By way of introduction, I am Tom Arnost, Co-President of the Univision Television Group, a division of Univision Communications Inc. Univision Television Group is the operator of over 50 local television stations in America’s largest Hispanic markets. Historically, our relationship with Nielsen has been productive, as we have worked closely with the company to implement changes to help better measure Hispanic television viewing across the country. Our early relationship with Nielsen resulted in the establishment of a ratings system that rated Hispanics separately from the general Nielsen ratings system in an effort to ensure more accurate estimates of Hispanic viewing. In this regard, we had and still have high hopes for the local people meter system as a better television measurement tool. First and foremost, I want to be very clear on the fact that Univision supports LPM technology. It is a technology with which we are familiar. It should be noted that the LPM service has been used to measure Hispanic-only ratings nationally, as well as the local market of Los Angeles, since 1991. In fact, when Nielsen announced it would rollout LPM service for its general ratings in New York, Los Angeles, Chicago and San Francisco, we requested that they use the same LPM technology for both general and Hispanic-only ratings--again underscoring our confidence in the technology, while working to avoid discrepancies in measuring Hispanics in the exact same markets with two different methodologies. However, Nielsen has made significant missteps in its initial rollouts of the LPM’s for their overall ratings system. In major media markets where minorities are large constituencies, Nielsen’s samples do not accurately reflect the demographic characteristics of the populations that they are attempting to measure. Nielsen will tell you that the new LPM better measures Hispanics and other subdemographic categories. But, you must look below the surface to determine whether or not their sample truly represents the market. American Hispanics are young, live in large households and the majority speak Spanish at home. Nielsen LPM samples of Hispanics are severely underrepresented in these demographic categories. In Nielsen’s L.A. sample group, there are 30% too many English-dominant households and conversely, 30% too few Spanish-only households resulting in lower Spanish-language television estimates. There are too few 18-34 year old Hispanic males in the sample—a demographic group coveted by advertisers. Hispanic homes traditionally have more people per household but Nielsen again, consistently fails to have the appropriate number of large Hispanic households in their sample. Separately, each of these misrepresentations causes an understatement of audiences to Spanish-language television and overreporting audiences to English television. In combination, such inaccuracies attributable to the Hispanic population, which comprises 40% of the LA market, results in inaccurate audience estimates for the entire market. Let me provide you with an example of Nielsen’s systemic failure to accurately measure the Hispanic portion of the overall populations in various markets. This chart compares the audience reported by Nielsen in its local Hispanic-only people meter service which measures the viewing of only the Hispanic portion of the Los Angeles market with Nielsen’s general local people meter service which measures the viewing of the entire Los Angeles market, including the Hispanic portion. This is for the month of May 2004, which is the last data made available to us by Nielsen. Remember that both of these services use the same local people meter technology but different samples in the exact same market of Los Angeles. As you can see, in a side-by-side comparison between two Nielsen rating surveys of Spanish-language television viewing in the L.A. market during the same time period, Nielsen reports 39% less viewing in the service that measures the entire population than in the separate service that measures the Hispanic-only portion of the population. It is statistically impossible for this to occur. The sample which measures the total market which includes Hispanics with non-Hispanics cannot report lower audiences than the sample that is measuring only the viewing of Hispanic households. When Nielsen presented its LPM and Hispanic measurement plan to Univision, they stated that its success depended on being able to demonstrate that the Hispanic-only LPM and the total market LPM generated similar data (see attachment to my written statement). Clearly, Nielsen’s LPM failed to meet or come close to properly representing the Hispanic population as evidenced by their own contradictory reports of Hispanic viewing habits in L.A. In this regard, we have developed a list of standards that should be implemented before the LPM is rolled out in any new market. These standards and compliance requirements should be established to ensure that market compositions are accurately reflected and are attached to my written statement. The broadcast industry is regulated by Congress, yet Nielsen – the entity which most directly influences ownership limits and market share – is an unregulated foreign-owned monopoly. Given Nielsen’s total disregard for the Media Research Council’s accreditation process (since it has no competitors and will, therefore, not suffer market share loss) and determination to rollout the LPM in new markets without addressing major flaws and concerns expressed by viewers, constituents and many in the affected industries, Univision has come to the reluctant conclusion that federal oversight is necessary and we respectfully ask you to seriously consider this option. I have been in this business for over 30 years and there always has been, and probably always will be, some level of dissatisfaction from broadcasters and advertisers in regards to ratings and measurements and methodology. But in those 30 years I have never witnessed the across the board level of dissatisfaction that exist at this moment as we sit here today. Please understand, that based on Nielsen’s current flawed samples there are companies, on both ends of the spectrum, companies that stand to lose ratings and revenue and companies that stand to gain ratings and revenue, and yet these companies stand united in their opposition to Nielsen’s woefully inadequate samples. We understand that in today’s highly fragmented TV world, Nielsen’s job is more difficult than ever. But that still is not an excuse to settle and say “well that’s close enough” not as long as Nielsen remains the only single source of measurement for what drives programming including news, entertainment and public service choices for you, our viewers, your constituents and all Americans. Thank you for your time and consideration.
Mr. Tom Herwitz
Click here for a PDF version of Mr. Herwitz's remarks.
Ms. Susan WhitingPresident and CEONielsen Media Research
Good morning. I am Susan D. Whiting, President and Chief Executive Officer of Nielsen Media Research, the U.S.-based television ratings company. I am pleased to be able to speak to you and to reaffirm Nielsen’s longstanding commitment to ensure that our ratings services fairly represent all television audiences. Nielsen Media Research’s roots date back to 1923, when Arthur C. Nielsen first came up with an idea for selling engineering performance surveys. That was the beginning of the AC Nielsen Company, and the birth of what is now known as market research. Since then, providing American firms with accurate information about how consumers use their products and services has been the core of Nielsen’s business. Over the years, Nielsen has tracked consumer usage for supermarkets, drug stores, radio and, for more than half a century, television. When Nielsen began nationwide television audience measurements in 1950, there were 98 commercial TV stations in the U.S. broadcasting to 4.2 million TV homes. The average home had a choice of just two stations. By 1970, there were 650 commercial broadcast stations – including 172 UHF stations – broadcasting to nearly 59 million TV homes. Back then, a new form of television service, cable TV, was also gaining a toehold in the American home, and Nielsen reported that fewer than four million households subscribed to cable in 1970. Within five years, satellite communications would transform cable from a retransmission system to a major program carrier and supplier. Today, we have separate measurements in each of the 210 television markets throughout the United States – some with fewer than 40,000 households. Our television ratings are the source of audience estimates for broadcast and cable networks, television stations and station representatives, program syndicators, regional television systems, satellite providers, advertisers and advertising agencies. Throughout the Company, we employ more than 6,500 full- and part-time people in 49 states, and our field staff is as diverse as the audiences we measure. Nearly a quarter of our workforce are people of color, including more than a third of our field and membership representatives. A Tradition of Independence Before I describe what Nielsen does, I would like to tell what we don’t do. We don’t produce programs. Whether or not One on One or Survivor win their respective times slots is irrelevant to us. Nor do our ratings imply that a show is either good or bad. Our only job is to estimate, to the best of our ability with the best means available to us, how many people are watching what programs at any given time. Most important, we have never sought to favor one client over another. Nielsen is neither owned nor controlled by any of the media companies we measure. We provide our own independent, objective estimates that are examined yearly by a third-party auditor. In short, we have no stake whatsoever – financial or otherwise – in the outcome of who wins a time slot or comes out on top. This independence means that the market can rely on the integrity of our estimates. Consequently, we have no reason or desire to undercount or misrepresent any audience group. We also do not purport actually to count the number of people watching any given television program. As explained below, that would require us to use means that are neither feasible nor desirable. To the contrary, our success in becoming the acknowledged worldwide expert in the field of audience measurement is based on providing the most reliable estimates available of who is watching the full spectrum of television programming. Measuring Audience Behavior Television advertising is sold both at the national level, by national networks, syndicators, and cable channels, and at the local level by local TV stations and cable systems. To that end, we operate a national sample and hundreds of local samples, in addition to our national and local Hispanic samples. Obviously, the most precise way to measure the nation’s viewing behavior would be to wire every man, woman and child and monitor them 24-hours a day. But that is neither feasible nor desirable. Therefore, we work from statistically valid samples of television households as do political pollsters and other market researchers – and, using proprietary methods that we have developed at great expense over the past 50 years, use the information provided from these samples, to prepare our estimates. As such, these estimates represent our opinion, as the leading expert in the field, regarding audience sizes for televised events. And, as a California Superior Court judge in Los Angeles recognized just two weeks ago, any attempt to restrain those opinions either judicially or through regulation or legislation has profound First Amendment implications Sampling enables our researchers to collect a wide range of demographic information from a relatively small number of people, and to project that knowledge across a much larger population for the purposes of analysis. We use the U.S. Census as the starting point of our samples. We design and select our samples using Census information and then enumerators travel to sampled neighborhoods to re-enumerate these areas ensuring that we capture changes since the Census was conducted. This method ensures, among other things, that newly constructed homes built since the Census have an opportunity to be selected into our samples. Unlike the Census, however, our surveys are conducted using samples of the population. We do not have the means or the resources to measure every single household in a television market, so we sample a number of homes to represent all the homes in the market. Through this mechanism we produce estimates of television viewing behavior for the market. And estimates from one sample will differ from estimates of another, even if the surveys are constructed in an identical manner. In fact, estimates from two different samples that produced identical results would be highly suspicious. The heart of Nielsen Media Research’s national ratings service in the U. S. is an electronic measurement system called the Nielsen People Meter. The People Meter, which is placed on each TV set in a sample household, measures two things – what program or channel a set is tuned to and who is watching. People Meter audience estimates are used by broadcast and cable networks, nationally distributed syndicated programs and satellite distributors, as well as advertising agencies and advertisers. For our national ratings, Nielsen currently samples more than 6,000 households encompassing more than 15,000 viewers. Eventually, through our national expansion, we will have 10,000 households in this sample. Using the Census Bureau’s decennial census counts of all housing units in the country, Nielsen randomly selects more than 6,000 small geographic sections nationwide. These are city blocks in urban areas and their equivalent in rural regions. Statisticians in our field office randomly choose a household from each of these blocks. We then send a series of mailings to household members informing them they have been selected to participate in the ratings sample, and to let them know that a Nielsen representative will visit them soon. In 60 of the largest television markets, we use a different metering system to provide TV ratings data on a daily basis. Information about what channels television sets are tuned to is collected by electronic meters that are attached to each TV set in the sample home. This information is augmented at least four months a year with demographic viewing data that are collected from separate samples of households, which each maintain a paper viewing diary for one week. The Set Meter/Diary results are used by local television stations, local cable systems, advertisers and their agencies to make programming decisions as well as to buy and sell commercials. To help provide a total picture of viewing in all homes, Nielsen Media Research also has established separate samples to capture viewing habits of Latino viewers of all television. Working with both Univision and Telemundo, we launched the first national Latino television ratings service (Nielsen Hispanic Television Index) in the U.S. in the fall of 1992. The unique viewing habits of this population group, particularly in those homes where Spanish is the only or dominant language spoken, are now reported on a national level and in 19 local markets where the Latino portion of the population is significant. Although all demographic groups are represented in our samples, the need to provide more frequent information on specific ethnic groups is increasingly important to all users of TV ratings information. In 1991, Nielsen expanded its monthly national reports to provide data on viewing among African-American households and approximately 20 African-American demographic categories. Additional information has been added to each local market rating report to identify African-American and Latino audiences and viewing activity. Effectively recruiting participants from a cross-section of Americans is a formidable challenge. Add to that the fairly demanding research requirements for panel members and the need to achieve the highest possible cooperation rate, and you can appreciate the complexity of the task. That is why Nielsen has continually sought to improve our recruitment methods. Continually Improving Our Audience Sample During the 1994-1995 period, for instance, we implemented a recruitment technique we called “membership.” We developed the concept of membership because it underscores the value of an ongoing commitment by our respondents. The implementation was preceded by two years of research into why people cooperate, and was subsequently supported by a combination of information, personnel and procedures that we have developed. As a result, we have achieved a remarkable improvement in our cooperation rate, in contrast to declining cooperation rates experienced by other research firms. We have also improved our methods and incentives for recruiting people of color to better encourage them to participate. This includes higher financial incentives to sample households, and improved hiring and training procedures for our employees who must recruit and coach participants in the survey. Nielsen is spending an additional $14 million on staffing, recruiting and sample quality, and we have developed a field staff that reflects the broad diversity of our audiences, including African Americans, English-speaking and Spanish-speaking Latinos, Chinese Americans, Korean Americans, Filipino Americans and Vietnamese Americans. To successfully accommodate different levels of language fluency, Nielsen employs specially trained membership representatives for Spanish Dominant households. These representatives undergo extensive instruction to improve rates among Spanish Dominants. What is more, all of our recruitment materials are bilingual, and we provide added incentives in the case of the Spanish Dominant category, in which we have more difficulty with recruitment. Finally, between now and 2006, Nielsen is implementing the largest-ever national sample expansion among African American and Latino households as part of our program to double the size of our national sample. Nationwide, the number of African American TV households in our sample will increase from 670 households to 1,200 when the expansion is completed in 2006. That is approximately 12 percent of the sample. The number of Latino TV households in the national sample will nearly double from 542 to more than 1,000 households. That is approximately 10 percent of the sample. With the launch of Local People Meters (LPM) in New York, Los Angeles and Chicago, Nielsen also is increasing substantially the number of African American and Latino families who are measured in these local markets. In New York City, the number of African-American TV households in the People Meter sample will be one-third larger than in the existing Set Meter sample. The number of Latino TV households in the LPM sample will increase to 50 percent over the Set Meter/Diary sample. In the Los Angeles market, we have increased the overall size of our sample by more than 50 percent. African American households represent 9.3 percent of our LPM sample in the market; Asian households represent almost 10 percent; and Latino households account for a third of the LPM sample. The percentages are higher within the city of Los Angeles itself. And in Chicago, where viewers from 16 counties throughout the city and Northwest Indiana comprise the designated market area (DMA™), African Americans comprise 17.9 percent of Nielsen’s LPM sample as compared to 17.5 percent of the total DMA population; while Latinos make up 12.3 percent of the LPM sample, roughly the same as their composition within the total DMA. These various increases will produce more stable data and more representative demographic information about African American and Latino television audiences. A Dramatically Changing Market The accuracy of our samples is critical to ensuring that we keep pace with a rapidly fragmenting audience. The differences among viewers today are no longer just demographic, but are increasingly governed by varying interests, tastes and social and personal priorities. Meanwhile, the marketplace itself is expanding as well. Consider the current television universe covered by Nielsen. Nationally, we have: q 7 broadcast networks q 75+ cable networks q 2 broadband satellite networks q 100+ syndicated programs q 150+ advertisers and advertising agencies Locally, Nielsen’s clients include: q 1,000 television stations q 1,500+ advertising agencies q 200+ local cable clients q … covering all 210 television markets Our research shows that the average U.S. home receives 100 TV channels, compared with just 27 in 1994. In addition, of the 108 million TV households: q 35% have two TV sets q 42% have three or more TV sets q 91% have a VCR q 68% receive wired cable q 37% receive wired pay cable q 84% receive wired cable or satellite The impact of this expanding TV universe on programmers and advertisers alike has been staggering. As a recent Business Week article vividly illustrates, an advertiser in the 1960s could reach 80 percent of U.S. female viewers by airing a commercial simultaneously on ABC, CBS and NBC. Now, as Business Week claims, even running an ad on 100 separate television channels would likely not duplicate this feat. Leveraging A Proven Technology Clearly, the medium and its viewers have changed significantly since the days when paper diaries were the sole means to measure who watched television in local markets. Instead, diverse audiences now channel surf scores of broadcast, cable and satellite channels in search of programming. And the only way effectively to keep track of their changing tastes and preferences is through the advanced capabilities of People Meter technology. Local People Meters are far more accurate than the conventional paper diaries they are replacing. While both services automatically monitor when television sets are on, the older system requires every household member to write down his or her viewing activities. The problem is that many people fill out their diaries the way they pay their taxes – at the last minute. As a result, they don’t always remember exactly what they watched days before. People Meters, on the other hand, automatically track viewers’ television behavior. One part of the system is calibrated to identify which broadcast, cable or satellite station is carried on each channel in the home. All channel changes are electronically monitored by the Local People Meters. At the same time, household members are each assigned a personal viewing button on a remote control-like device, which identifies them by age and gender. Whenever the television set is on, a red light periodically flashes on the meter to remind viewers to press their assigned button to indicate that they are actually watching TV. While a viewer using the diary system may record that she watched Half & Half on Monday night, she may actually have flipped through several channels during that time, all of which the People Meter would register. This may appear to some as a radical shift in the way Nielsen measures television viewing patterns. But the fact is the People Meter service being introduced in New York, Los Angeles, Chicago and other major markets is same system Nielsen has been using successfully for more than 17 years to establish national TV ratings. It also is the standard for television ratings systems worldwide. Furthermore, Local People Meter service has been fully tested and accredited in Boston, where it has been operating for two years. Responding To Client Needs Though Nielsen has continually enhanced the quality of our People Meters, the fundamental technology has not changed. What has changed is the need by local programmers and advertisers for more accurate measurements of audience behavior. Gone are the days when local television markets consisted of three network affiliates and a handful of independent stations. Back then it wasn’t economically feasible for these markets to invest in People Meter technology. But times have changed, and the growing number of media providers and advertisers in large TV markets are demanding we measure viewers with the same sophistication and accuracy we have been bringing to the national sample since 1987. It is no longer enough for major media markets to know only how many households are tuned to a certain TV show, or to rely on four standard month-long “sweeps” periods to identify their audiences. Local broadcasters and local cable outlets now demand to know exactly who is watching what program – how many women 18-34; how many men 25-plus; how many teens; how many African Americans and Latinos. Local broadcasters, local cable outlets and advertisers now demand this information the very next day, if not sooner. Local television executives also realize that if they are not included in the People Meter system, they will be at a serious disadvantage in dealing with national advertisers. The nation’s television advertising systems are programmed to process People Meter data, and each year People Meter ratings are the basis for tens of billions of dollars worth of advertising on a national level. As a result, local programmers without People Meter technology could lose out on opportunities to tap into these national advertising campaigns. As John Mandel, co-CEO of MediaCom, a $5 billion-a-year ad agency who has strongly supported the LPM roll-out in Los Angeles said: “[t]here can be no question whatsoever that Nielsen’s People Meter technology is the best system available to measure television ratings and viewers. It is the gold standard it its field.” He also said that, “any decision forcing us to use less than [the LPM service], knowing that better information is available, would be a tremendous injury to our company, and to all media buying and planning companies that use LA local television ratings from Nielsen.” That is why we are introducing Local People Meters as scheduled. We are responding to demands from a large, if quiet, base of our clients, such as Comcast, Oxygen, BET, TV One, Time Warner, the ABC station group, NBC and the advertising community, who strongly urge us to go forward now with our People Meter rollouts across the country. MRC Accreditation While we acknowledge the concerns of individuals and groups who have asked us to postpone our LPM launches until after we gain accreditation by the Media Rating Council (MRC), we believe doing so would raise serious antitrust issues. We have applied for and will seek accreditation. The Media Rating Council, which was formed in 1964 at the urging of the United States Congress, is an independent, non-profit industry association created to maintain the quality and credibility of audience research by securing audience measurement services that are valid, reliable and effective. MRC members include television and radio broadcasters, cablecasters, print organizations, advertisers, Internet organizations, advertising agencies and industry trade associations. Among its various responsibilities, the MRC offers all research firms an opportunity to participate in a voluntary program for accreditation of their ratings services, which are voted on by those members representing the media and advertising industries. The MRC consists of fiercely competing firms with different commercial interests affected by the adoption or delay of audience measurement innovations; a majority vote to deny or withhold accreditation of a technology such as the Local People Meter – which would be to the competitive benefit of some and the disadvantage of others – would present a classic conspiracy in restraint of trade. For precisely this reason, the antitrust clearance the MRC received in 1964 was very limited, and was predicated upon the accreditation process being voluntary rather than mandatory. To emphasize that point, the MRC’s Procedures for Accreditation specifically state that: “Neither these Procedures nor the criteria shall preclude the offering of electronic media audience measurement information by a Service that is not accredited, nor shall the Procedures or Criteria prevent any person, firm or corporation (whether or not a member of the Council) from purchasing or using such information.” We have consulted with well-known antitrust experts who have advised us that these antitrust concerns also apply to the notion of obtaining accreditation before launching the service in any of our local markets. Instead, as has been done for the past 40 years, and consistent with the antitrust restrictions recognized by the MRC at that time, it is our intention to launch our service, then seek MRC accreditation in each market. LPM Ratings And National Network Programming Nielsen also is sensitive to concerns among African Americans and Latinos that declines in Local People Meter ratings for some broadcast shows may result in their being cancelled. However, we believe such apprehension is misplaced. Although advertising rates for national network television programs do affect their success, such rates are not set by Local People Meters. Rather, they are determined by the national People Meter sample, which has been in place since 1987. Networks and studios primarily look at the national numbers when judging a show’s popularity. LPM's are used only to set rates for local ads inserted into the network feeds. And of course Local People Meters, which until June were only used in Boston, have not been responsible for the failure of mainstream programmers to introduce more television shows featuring people of color. Measuring New Opportunities Although Local People Meter measurements do not appreciably move the needle with respect to programming on national broadcast networks, they play a vital role in revealing an increasingly important trend among smaller networks and stations, especially those that are owned and operated by African Americans and Latinos. Just as the 1987 introduction of People Meters in the national sample helped galvanize the cable business by providing more accurate information on cable viewing, so too will Local People Meters help support specialized programming for people of color. In recent detailed analysis of March LPM data in New York City (available at www.everyonecounts.tv), Nielsen revealed that people of color watch a much wider array of programs than was previously indicated by the diary system. This presents a compelling case for the television industry to focus more attention on African American and Latino issues in a great many more shows. New York Local People Meters also have recorded significant increases in ratings for networks such as BET, Telefutura and Telemundo, which have seen their numbers go up 180 percent, 83 percent and 22 percent respectively under LPMs. Additionally, 21 networks targeted to Latino audiences have had ratings increases, as have 27 networks geared to African American viewers. In a written statement responding to these findings, BET President and COO Debra Lee said: “Quite frankly, we don’t understand why some broadcasters would oppose any method of truthfully tracking what viewers are and are not watching, regardless of their ethnic background … It seems only fair that their voices be heard as advertisers decide where to spend their dollars to try and reach that audience. Networks like BET could certainly benefit from information that depicts the true viewing patterns of African-American consumers.” Indeed, networks like BET are gaining added clout with advertisers, which in turn could help increase their ability to finance new programming and create more opportunities for television professionals of color. It should come as no surprise then that Nielsen’s Local People Meter ratings show that African Americans and Latinos are increasingly turning to a wider array of channels in search of new and more diverse programming. Again, our research shows that while audiences attracted by even the largest cable stations remain considerably smaller than those of the broadcast networks, cable’s aggregate prime time share is 52 percent, compared to 44 percent for broadcasters. Much as Americans a generation ago migrated from older metropolitan areas to other parts of the country, establishing vibrant new centers of business, industry, art, entertainment and politics, a new generation is moving across the television landscape, helping to build diverse new networks and stations. And Nielsen’s Local People Meter is recording this dramatic transition. Facing Threats But as the landscape shifts and the center of gravity moves away from conventional broadcast networks and stations, companies that are financially dependent upon such traditional media outlets, such as News Corp., are attempting to stop what even they recognize as an inevitable need for more complete, detailed and reliable audience information. As early as December 2003, representatives from Nielsen Media Research met with News Corp. executives, including consultant Norm Hecht, to discuss the People Meter service. News Corp. argued that LPM measurements would cost its subsidiary, The Fox Television Group, millions of dollars a year in local advertising revenue by undercutting the number of viewers of color watching Fox shows tailored to African American and Latino audiences, as well as its national sports coverage. The company threatened to take belligerent actions – including legal action and political pressure to discredit Nielsen – if we did not back down from our LPM launch in New York. Ironically, during this same period, Fox regularly cited the results of Nielsen’s national People Meter ratings in its advertisements for many of its popular broadcast and cable shows (see attachment). While we tried to work with News Corp., as we do with all of our clients, to address their concerns, we consistently stood by the accuracy of LPM measurements. Consequently, News Corp. made good on some of its threats. Published reports in a variety of publications, such as the New York Times, the Wall Street Journal, Roll Call and Variety have highlighted efforts by Rupert Murdoch and News Corp. to thwart the introduction of Local People Meters. We do not yet know the full scope of their activities. What we now know, after they first denied that it was so, is that the company has hired more than half a dozen lobbying and communications consulting firms, and it is financing the activities of the ad hoc organization Don’t Count Us Out in New York, Los Angeles and Chicago. According to the Los Angeles Times, News Corp. is spending millions on a campaign based largely on disinformation to halt the People Meter rollout. There have been television, radio, and print advertisements and an anti-Nielsen website. Why did this change in methods in New York and Los Angeles attract so much more attention than the launch of National People Meters in 1987 or the introduction of Local People Meters in Boston in 2002? The answer is money. This whole issue is about change and money. Whenever a change in methods or technology alters the ratings for a program, there is money at stake. And since Fox has acknowledged that they are very concerned about the financial risks that LPMs allegedly posed to their dual station systems in Los Angeles, New York, and Chicago, that is not a secret. Among the other activities, the News Corp.-funded “Don’t Count Us Out” sent misleading and partial information about specific programs to community leaders and elected officials. Given the years of undercounting that people of color had experienced in polls, elections and, some believe, the Census, there was an understandable receptiveness to this story. In contrast, Nielsen, which purposely does not advertise, had a very low profile in Washington DC and in many communities. This combination – which a colleague described as “Fox dropping a lit match on a dry field” – caused an explosion. The threats in this campaign have been many – from Fox executives, directly; from Don’t Count Us Out leaders; and from community activists. They have ranged from telling us to stop on certain dates to authorizing campaigns to ask people not to participate in Nielsen studies, thereby undermining the currency and the very issue of representation. Addressing Concerns So what have we done in the face of the flood of cash aimed at getting us to delay and to discredit our company? In an ongoing effort to correct the record and communicate the benefits of People Meter technology to key audiences, Nielsen has vigorously reached out to community leaders and elected officials across the country. I myself spend as much as two-thirds of my time on endeavors that have included: q Numerous meetings in Washington with more than 50 Senators, Representatives and staff, including members and staff of the Commerce committees. q Briefing the staffs of the Congressional Black Caucus and the Congressional Hispanic Caucus. q Holding conversations with New York State Senate and Assembly leaders, members of the New York City Council, and representatives of the Mayor’s office. q Meeting and briefing leaders from organizations such as the NAACP, the Hispanic Federation, the National Urban League, and the 100 Black Men. q Meeting with community leaders and elected officials in Chicago and Los Angeles. q Delaying of the the LPM system in New York for two months from its original scheduled launch date, and parallel diary and LPM systems in NY, LA, Chicago and San Francisco for a period after the launch. As I am sure you are aware, Nielsen also worked with Congressman Charles Rangel to create a Task Force on TV Measurement. Just recently, we announced that former Illinois Representative Cardiss Collins was appointed to chair the Task Force. Congresswoman Collins is no stranger to these issues, having led a similar task force in 1996. The diverse group of Task Force members includes representatives of community groups and both the television and advertising industries. The group held its first meeting on July 7 in Chicago, and we look forward to receiving its recommendations in the fall. Legal Challenges Recently, Univision Communications went to court in Los Angeles to try to restrain us from introducing LPM service there. We are pleased that Judge J. Stephen Czuleger of the Superior Court for the County of Los Angeles refused on July 1 to deny the Los Angeles market the immediate benefits of Local People Meters. The evidence we presented demonstrated clearly that the electronic People Meter is a more reliable and complete way to measure TV viewing than handwritten paper diaries. Judge Czuleger wrote that Univision failed “to establish to the court’s satisfaction LPM’s flawed process, sampling, or weighting.” “More troubling to the court,” Judge Czuleger continued, “are the implications regarding free speech. Because of this, [Univision’s] motion, again, should be denied because [Univision] seeks to enjoin [Nielsen’s] First Amendment rights.” Nielsen respectfully submits that the Los Angeles court’s concern about impinging on our First Amendment right to express our opinions regarding television audiences likewise constrains other governmental attempts to either prevent Nielsen from expressing those opinions, or to force Nielsen, through unwarranted legislation or regulation, to express opinions that Nielsen itself does not hold. Moving Forward Information is vital to the success of every U.S. company that provides goods and services to consumers. If these firms do not have access to the most reliable data about the needs and interests of their customers, opportunities will be lost, their sales will decline and jobs will disappear. Nowhere is this more critical than in television, where the industry and its audiences are constantly changing. As they do, Nielsen will change with them. But we also must be able to appreciate and respond to different reactions whenever we institute change. In that regard, Nielsen has already committed to invest $218 million in new technology, research and development, including: q $55 million on Active/Passive Meters q Over $70 million on LPM technology q Over $20 million on National People Meter expansion q $2.5 million on weighting in the national sample Going forward, we will leverage these investments to further improve the quality of our samples and enhance the capabilities of our technology to provide the most reliable ratings available. We also will continue to listen to and work with all interested parties to ensure our measurements fully and fairly represents all television viewers. And we will stand firmly in defending an independent, objective and accurate ratings service in which everyone counts. Most important, we will continue to resist all attempts to manipulate this process or to mislead the public through a cynical campaign that has nothing to do with protecting the rights of African Americans, Latinos or any other ethnic group, and everything to do with protecting the commercial self-interest of specific media companies. For Nielsen Media Research, success means ensuring that everyone counts. Thank you. # # #
Mr. Bob Barocci
Good morning, Mr. Chairman and members of the subcommittee. My name is Bob Barocci, and I am the president and CEO of the Advertising Research Foundation. I am very pleased to have been invited to participate in this hearing today on a very important advertising industry issue. The Advertising Research Foundation was founded in 1936 by two other non-profit trade associations – the Association of National Advertisers and the Advertising Agency Association of America. Today we count over 325 companies – advertisers, agencies, research and media companies – as members. We are the only organization that brings together this kind of cross-industry representation at a single table. In fact, all of my fellow panel members are from organizations that are ARF members with George Ivie being a member of our board. Paul Donato, chief research officer of Nielsen Media Research, is the ARF Board Chair. For the record, Paul has recused himself from all ARF discussions on the Nielsen Local People Meter issues and, at our Board meeting earlier this week, George offered to do the same. Our mission in 1936, as now, is to “promote greater effectiveness in advertising and marketing through impartial and objective scientific research”. Our stock in trade is objectivity and integrity….People in the industry often say, “you can always count on the ARF for an unbiased, intelligent and authoritative point of view based on solid methodology and intelligent data analysis”. Today, the ARF’s dedication to leveraging the power of research to help advertisers spend their advertising and marketing dollars more wisely and effectively is needed more than ever. The expansion of the Nielsen People Meter measurement system from national into local markets is one important example. Nielsen numbers are used as the “currency” for the $60 billion USA television industry. Television buying and selling is based on Nielsen’s numbers, so much is at stake when a significant measurement system change is to be made. The change that is happening now is that Nielsen is expanding its national People Meter system into the top ten local markets replacing the current set-meter/diary system that provides local ratings four weeks per year. That means that these local markets, accounting for a significant portion of TV revenue, will be continuously monitored all year around by the LPM system. Achieving accurate television audience measurement has been a subject of discussion and debate for many years. In fact, in 1999, the world’s leading global organizations addressed this issue in a very thorough way. The Audience Research Methods Group - The ARF in collaboration with the World Federation of Advertisers, ESOMAR, the European Broadcasting Union and associations in Europe and Canada - published a document titled Towards Global Guidelines for Television Audience Measurement. In those guidelines the authors noted, “The ‘peoplemeter’ system has become the de facto standard audience measurement system throughout the more developed markets around the world… It is a sophisticated system able to deliver a wide range of audience measures. Equally, it meets a wide range of management needs – both editorial and commercial.” Nothing has happened in the intervening years to change that conclusion. So, if indeed the LPM system is “more accurate” than the meter/diary system now used, as it is generally agreed to be, the numbers will be different than the old system. And, that means there will be winners and losers. The LPM system will measure the television sources with the lowest ratings more accurately than with the current diary system (where some lower-rated programs often receive no entries during a quarter hour – or sometimes during an entire program – for key demographic groups). This will increase the value of the inventory of cable stations and decrease the value of the inventory of the traditional networks – NBC, Fox, ABC, CBS, for instance. And buyers will pay more for cable and less for the traditional networks…winners and losers…and the money is big. While this should not be the central issue of this debate, it has become so, because of money. There is only one legitimate issue here, and, that is the accuracy of the local Nielsen samples, especially as it relates to properly representing the viewing habits of minority groups. And, that is an important issue as one source estimates that over $3 billion per year is spent on advertising on Hispanic television stations alone. When debating sample accuracy, it is important to note that the proper question is always – is the sample better than what we had before?…not, is the sample perfect? No sample is ever perfect, by definition. However, it is Nielsen’s responsibility to make sure that their sample is always getting better. And it is the industry’s responsibility to make sure that they succeed at that. In this debate, the questions are: 1. Is the sample better than what currently exists? 2. And, are we convinced that Nielsen is doing everything possible to make sure the sample will keep getting better in terms of its accurate representation of all targeted marketing groups important to television advertisers, especially minorities? Though there is little disagreement that the LPM system provides superior, more accurate measurement, debate doesn’t stop there. Whether the sample is good enough is being debated. That was the crux of the Los Angeles lawsuit brought by Univision against Nielsen. Whether Nielsen is doing everything possible in terms of training and field resources is also being debated. There is also debate over whether the button-pushing requirement impacts minorities more than the general market, but this is both a national and local issue. Healthy debate can be very productive. As long as the debate focuses on the right issues and reasonableness prevails. There are no new issues in this debate. They have been resolved before and will be resolved now. In our role as industry research authority and as an active member of the Task Force on Television Measurement convened by Congressman Rangel in collaboration with Nielsen, The ARF will seek to move the debate along in an intelligent, logical and rational and timely manner. Advertisers benefit greatly from better, more accurate information. With that in hand, they are able to spend their television money more wisely. And that is the ultimate goal of these changes. The ARF is proud and pleased to be involved in contributing to the resolution of this important issue. We can bring intelligent, objective analysis and focus to the table and, as Congressman Rangel has implored us, articulate the issues in a way that the man on the street can understand.