The U.S. Census Bureau tries paid advertising to raise mail-in response rates for Census 2000.
How do you count a nation? Really. How do you count every man, woman, child, every household, every income, every person's race, age, and sex? To the U.S. Census Bureau, the question by now should be academic. But in a vast, diverse, roiling nation of 275 million citizens and millions of undocumented aliens, the question has taken on an urgency as the 2000 Census approaches. Counting everybody is no longer the single most important question. That would be this: Is it even possible?
In mid-March, the Census Bureau officially launches what is widely expected to be the most complex and ambitious count of U.S. citizens since the first one in 1790. It is also fraught with more uncertainties than ever before, and a prevailing sense that more people are reluctant to follow the simple request to "stand up and be counted," to borrow a tagline from one of the hundreds of ads that began running last fall.
And a burgeoning immigrant population is just half the challenge, say Census Bureau officials, who also cite a growing mistrust of government and a vast segment of the working population that has neither the time nor the inclination to fill out the Census form.
Last October, the bureau hired Young & Rubicam to spearhead a massive $167 million paid advertising campaign - making the 2000 Census the only time in history that the bureau will actually have paid for media placement. Yet even with such an unprecedented war chest, Y&R - which typically handles consumer giants like Procter & Gamble and Johnson & Johnson - is facing a supreme marketing challenge: How to get millions of people to do something they simply don't want to do.
Notes Terry Dukes, Y&R executive vice president and account managing director who is heading up the agency's Census 2000 team, "We're in an immigration boom - the highest level of immigration since World War I - and so people don't know about this [the Census] or don't know what to do. We're also in a very mobile society, with people moving around all the time...and the Census is based on physical locations. We are also dealing with a fact of government distrust here as well, which is certainly more significant among immigrants, but is also [strong] among a lot of younger minority members."
The story of how Y&R won the Census Bureau's account says something about the radically altered demographic face of America at the turn of the century. And if Y&R's revolutionary approach works, it'll prove once again the old Madison Avenue maxim that, given enough time and money, you can sell anybody just about anything, including a gray, faceless Census form.
The Census Bureau first detected significant undercounts 20 years ago, when the response rate to the 1980 nationwide mailing fell to a 75 percent from a 78 percent response in 1970. But the response rate to the mailing took a major hit in 1990, when it dropped to 65 percent, "which definitely made the Census Bureau sit up and take notice," says Jennifer Marks, chief of the Census 2000 publicity office in Suitland, Maryland, where the national awareness effort is being coordinated.
It didn't take long for the bureau to uncover the chief culprit behind the poor response to the '90 mailing. As always, the bureau (and then-ad agency Ogilvy & Mather) had created commercials to alert people to the decennial impending count, and as always, media placement was pro bono. The bureau later conducted a "post-buy" audit and learned that "a large portion of the population should have received messages a fair number of times, but in fact, large segments of the population saw the ads very few times," says Marks. The reason: Television and radio stations would typically air commercials during low-traffic weekend time slots, if at all.
Any attempt to rely on another free public service campaign would surely mean severe drops in the response rate. Some believe it could fall to as low as 55 percent, meaning that just under half of the nation's 102 million households would likely fail to complete the Census form. That was considered unacceptable from a simple dollars and sense perspective. It costs $36 every time a Census worker visits a non-respondent's home as part of a follow-up when the form isn't filled out. The cost per individual mailing? Only $2.
The pressure to improve the direct mail response continued to build last year. The U.S. Supreme Court ruled that the Census Bureau could not create statistical models to fill the projected gaps in the undercount as a way of "reapportioning" U.S. Congressional districts. That meant everyone had to be counted. Everyone.
But how? Foremost, it meant initiating a paid advertising campaign, and Congress approved $167 million. A new approach to an old problem was also sought, and that's where Y&R came in.
The agency - along with some 80 competitors - gave its oral presentation to the bureau in late August. Y&R created a detailed statistical model which determined how likely an individual was to fill out the Census form. From there, says Dukes, "We found that the public broke out clearly into three distinct groups." The first, comprising 40 percent of the adult population, was dubbed "most likely to respond." They were older, middle-class, white-collar homeowners familiar with the Census and, presumably, its benefits.
That group would ultimately be of less interest to Y&R. Then there was a second, vast, undifferentiated chunk of the population - 43 percent - dubbed the "undecided passives." Says Dukes: "It's not that they resist filling out [the form], but just have no good reason to do it. They may get to it or not. They have at least a high-school education, are primarily blue-collar, some homeowners...What you and I would call middle, mainstream Americans."
But the significant size of the third group would be a surprise. Y&R's model determined that fully 17 percent of all adult Americans would be unlikely to fill out the form at all. Of that group, many clearly despised the very idea of a Census.
>From historic data, the bureau knew that a range of ethnic groups had been >undercounted in the past: African Americans and Hispanics were >undercounted by 5 percent, and Native Americans living on reservations >represented the most serious undercount of all (12 percent). Explains >Marks: "There is extreme distrust of this government, and it's well >founded [because] the Census in the past was used to take land away."
Nevertheless, Y&R's Census 2000 team came to the obvious conclusion that the difference between a successful count this spring or a bust meant persuading the broad middle section of the population to participate - some 60 percent of the adult population that has little or no inclination to cooperate.
That task alone came to define Y&R's counterintuitive media and marketing strategy. As Dukes puts it, "Unlike most marketing situations, where we've put most of our money against our best prospects, we have to put a considerable amount against our worst."
Last fall, Y&R began to coordinate a complex marketing and information campaign among four sub-agencies, including the Bravo Group (which specializes in the Hispanic market), Kang & Lee (Asian market specialist, and which, like Bravo, is owned by Y&R), G&G (Native American) and the Chisholm-Mingo Group (African American). All are creating a multimedia range of ads in 17 languages, making the Census 2000 effort probably the most ethnically diverse campaign in advertising history.
The most crucial aspect of Y&R's plan is media, however. Beginning with the so-called "educational phase" last November, through a "motivational phase" to be conducted from January to April, and culminating in the "follow-up phase" through July, Y&R is juggling a broad-based media strategy that involves a national overlay of television and print, along with a huge jigsaw puzzle of media placements comprised of ethnic newspapers, radio programs, and even satellite broadcasts.
Dukes explains that the first layer - television - will reach virtually every adult household, while subsequent layers - mostly ethnic publications - will successively target that broad passive middle and the more resistant 17 percent group.
The "most likely" group will get hit with ads on network TV and in major magazines - spots that will work essentially as reminders. The passive group - the silent 43 percent - will get a more intensive and diversified media assault, not just via prime time TV, but on daytime and other media that are more directly tailored to their interests and lifestyles.
The resistant group? It will almost certainly be ad-resistant, requiring "a lot of programs planned at the grass-roots level," says Dukes. Of course, not all members of the resistant group are "extremely disenfranchised from government," says Census 2000's Marks. "There are also recent immigrants who don't have language skills, so we're addressing them through advertising" - as well as with brochures translated into a total of 49 languages.
Virtually every ad created displays the top half of the Census form, accompanied by the tag, "This is your future. Don't leave it blank." The idea, says Dukes, is to "try to relate all the good things that come [of the count]. Interestingly, it was very hard for the bureau to even tell us what those were, because they don't directly attribute things like new schools or bridges" to the redistribution of government based on Census statistics. While the ads and paid media strategy are expected to shoulder what the bureau is calling the largest peacetime effort in U.S. history, there is only so much a hard-sell Madison Avenue pitch and $167 million can accomplish.
And what defines success, by Census Bureau standards? A mere 66 percent response rate to the initial mailing, or just 1 percent more than the dismal response in 1990. But that should still be enough to break out the champagne.