Nielsen has told its critics in government that its "People Meters" do not undercount Hispanics and other minority viewers. If anything, it says, it overreports viewing in English-language Hispanic households, a key area of concern for lawmakers afraid that the result of underreporting would be fewer mainstream shows targeted to Latinos.
The ratings service came under criticism from both the Hill and the New York legislature on the underreporting issue last week. One reason for the timing: Nielsen is preparing April 8 to roll out People meters in New York, the nation's top market.
Nielsen today sent a response to a letter from four House Democrats who complained about underreporting and called for it to be independently audited, which Nielsen quickly pointed out on Friday that it already is.
In response to the complaint that it undercounted English-speaking Hispanic households, Nielsen countered that it, in fact, overreports the category. In its national sample, it says, 5% of Hispanic TV households describe themselves as predominantly English-speaking, while "the U.S. population estimate for this classification is 3.3%."
But the complaint department didn't close last Friday. On Sunday, some New York legislators from the black and Hispanic caucuses gathered on the steps of city hall to try and block the roll-out of the People meter, making the underreporting charge and saying they wanted to introduce a resolution to stop the roll-out.
Nielsen counters that its meters are more accurate than diaries at reflecting the viewing patterns of minorities. For instance, says Nielsen's Jack Loftus, a viewer using a diary might write down that he watched Girlfriends, but forget that he switched over and watched 20 minutes of BET. "The People meter does a more accurate job of capturing that other viewing," he says.
Of course, for every cable network that would welcome such an improvement, there is a broadcaster who is losing audience, so the issue is unlikely to get less contentious as Nielsen rolls out its People meters in more major markets.
The Nielsen letter does concede that its national sample undercounts Spanish-speaking households, with 3.6% for the Nielsen sample to 4.3% for population estimates. It also slightly undercounts bilingual Hispanic households, with 1.9% in the Nielsen survey but 2.1% for the general population.
That would buttress the complaints of underreporting from Spanish-language network Univision.
In related news, Nielsen is doubling its national sample size over the next two years from 5,000 to 10,000, which will also just about double the representation of African American and Hispanic households.