• by June 2, 2004

BE CAREFUL WHAT YOU WISH FOR - Here's a wacky scenario. Suppose for a moment that Nielsen Media Research didn't actually want to be in the minority ratings business. By that, we mean the business of measuring ethnic or racially diverse slices of the TV audience. You know, the segments that have gotten so many adversaries up in arms over Nielsen's plans to launch a local people meter system they claim undercounts those viewers. What if Nielsen only got into that business because of a sense of obligation as researchers to do the best job it possibly could to measure all segments of the population, including Latinos and African Americans, so that its general market sample was truly representative? In fact, what if Nielsen only created national and local Hispanic ratings services because it was pressured by the big Spanish-language broadcasters to do so? What if Nielsen really didn't like being in those businesses? What if they were more of a headache than they were worth? And what if all the pressure from Don't Count Us Out, the National Hispanic Media Coalition, Univision, Fox and others culminated in some legal battle that exposed the fact that Nielsen simply doesn't have the wherewithal to measure ethnic audiences the way those ethnic audiences feel they should be measured? At least not without taking the resources away from having a truly balanced and representative sample of the general population. And what if the court ruled that was the case and Nielsen had to get out of that business altogether? Well then, everyone would have gotten what they wished for. Right?

OK, so Nielsen can never step away from measuring minority viewers. Not if it wants to claim its general sample is truly representative of the overall market. But as researchers, there are differences in the way you measure the composition of a general market sample comprising segments of a population than the way you would measure the composition of the discrete segments of that population. At least not without infinitely greater resources. And at least not without a significantly greater contribution than most customers might be willing to pay. But that seems to be what some stakeholders are pressuring Nielsen to do and it just might be the defense Nielsen mounts against any lawsuits that challenge its ability to adequately measure those segments of the population.

Look, we're the first to admit that we're no experts on the law, or statistical research for that matter, but we do understand that this is an issue that has put Nielsen between a rock and a hard place, and that the only way for it to win, may be for it to lose. After all, if it doesn't draw the line somewhere, then every conceivable niche may soon start publicizing, lobbying and litigating for its fair share of the Nielsen pie. If African Americans and Latinos aren't properly represented, than why shouldn't Asian Americans probe the matter? After all, they are the fastest growing major segment of the U.S. population. And if Asian Americans find their don't-count-us-out voice, then why not a segment others might find equally as important: gay Americans? We can just see that enumeration study now:

"Knock, knock. Good day. Er, excuse me, is that Mr. or Mrs. Bates? Um, OK, so you prefer the title Master. Well then Master Bates, we're conducting a brief survey on behalf of Nielsen Media Research to make sure that we are properly representing the demographic composition of TV viewers in your area. Please tell us which of the following best describes your sexual orientation: Heterosexual, homosexual, lesbian, bi-sexual, transsexual or asexual? I'm sorry, we don't have a category for auto-sexual, but we do have some questions about what type of car you drive."

And if Nielsen isn't adequately representing the discrete demographic or sexual disposition of viewers in its sample, than why not their political persuasions? Are there enough conservatives represented in the sample? Are there too many liberals? Imagine a statistically small and politically biased sample that does not adequately represent the general American public. Oh yeah, we already have that. It's called the Electoral College. Actually, we wouldn't be surprised if at this very moment, certain TV network executives are raising concerns over the party affiliation of Nielsen's sample, though we suspect they'd find a decidedly more partisan response from lawmakers than the dust they've managed to stir with the race card.

The real issue isn't about Republicans or Democrats, and it's not even about being democratic (the kind with a small "d"). It's about commerce and a reasonable assurance that Nielsen's sample is representative of the types of viewers that marketers spend $60 billion a year to reach. The reality is that despite all the attention given to some ethnic groups, Nielsen has some much bigger sample issues with regards to households with children, or cable TV, and other things that it's still struggling to address. And the reality is that it will and should focus all its priorities on improving its sample and ratings methods to adequately measure all of those segments in a balanced way that represents the overall marketplace. To do otherwise, would be racist, or maybe sexist or even politically partisan. It certainly wouldn't be good research.