High foreclosure rates, an influx of immigrants and a shortened national attention span are all working against the 2010 Census.
While he couldn’t explain the correlation between the results of the 2000 Census and the accompanying Public Marketing Campaign, Census Bureau Director Robert Groves insisted that the results in 2000 would have been significantly worse without the $147 million the bureau dedicated to advertising. “We reversed the downward trend in mail response rates compared to previous decennials. Proving that the reversal was due to the advertising is impossible, but many analysts give substantial credit to that innovation in 2000,” Groves told The Daily Caller.
This much is true: In 1970, the response rate to the biennial survey was 78 percent; by 1990, it had dropped to 65 percent. According to Groves, had the trend proceeded, the response rate in 2000 would have been about 61 percent. But at 64.3 percent, it was only slightly lower than the response rate in 1990.
The National Science Foundation attributes declining response rates to information overload. “People today seem more likely to say no to a survey taker due to the sheer quantity of requests for their attention, the possibility that a survey may be a sales pitch in disguise, disinterest in the topic or an unwillingness to give honest and thoughtful answers,” the NSF concluded in a special report.
Today, the Census Bureau has to contend not only with Twitter, the 24-hour news cycle, and the general sense of intrusion that accompanies administering the survey, but also with as many as 12 million undocumented immigrants, millions of foreclosed homes and what Groves considers to be a severely diminished collective attention span: “All of us think we’re busier, even if we aren’t.” That’s why the bureau will spend about $340 million to market the 2010 Census in the next four months, a decision Groves defended as fiscally sensible. “Say, OK, we don’t do any advertising. For every one percentage point decline in the response rate, we’re probably going to spend between $80 and $90 million sending volunteers door-to-door. If we can do this right, we save tax payers $80 or $90 million. That’s a pretty easy business case.”
Groves didn’t deny that mistakes were made in 2000. Without hedging, he conceded that the absence of control groups or a rating system left the bureau without “scientific data to measure the effect (return on investment, if you like) of the paid advertising and partnership efforts in Census 2000.” But he insisted this year’s decennial survey will be different. In a phone interview, Groves rattled off in quick order the bureau’s strategy for analyzing the PMP: “We have some media markets where we’re doing normal advertising, and a matched set of media markets where we double the advertising. We are tracking things daily, using third party surveys to ask people if they know about the survey. We’re tracking the audience touched by our advertising using Neilsen.”
After taking a breath, Groves said, “We won’t get it perfect, and we’re sure there will be some subgroups that won’t react to the adveristing the way we anticipate. But hopefully we’ll know why.”
Despite the bureau’s best efforts, high foreclosure rates likely will have a negative effect on the response rate, as forms that are sent to unoccupied homes won’t make it back to the bureau. “I like the ratio of completed forms versus the number sent to occupied houses,” Groves said. “The ratio of sent out completed forms versus all houses is less helpful. We can’t overwhelm that effect [with advertising].”
Groves defended the Road Tour, which The Daily Caller reported on last week. “We’re spending $15* million to send 13 different trucks to 800 events, because we have the burden of trying to get the information out to every event,” Groves said. It’s not just going to football games and NASCAR, he added. “We have sent out the trucks based on statistical data” that correlates to communities with low response rates. And the new measuring mechanisms that the bureau is using will tell Groves whether or not the trucks did the trick, or were simply a misuse of $15* million.
Soon, the bureau’s efforts at “fitting the Census design to the evolution of the American society” will result in Census flotsam appearing virtually everywhere. The Road Tour will stop at a NASCAR race and feature prominently in the Super Bowl. Television ads will appear on stations in every region and city in the country. “Walk into a Best Buy in the coming weeks,” Groves said, “and you’ll see a Census logo on the bank of TVs in the back.”
But will hundreds of millions in marketing maintain the 2000 Census response rate?
“I’m a worrier, so I would never say that we’ll do better,” Groves said. “You could hold me up against the wall with a gun to my head, and I wouldn’t say that. I would say that we’re trying to tell people what they did last time, and then urge them to do better. We’re making a straight dollar case. If you want your fair share of tax payer money returned to you, fill out the form. We’re trying to deliver all of those truisms you need people to hear in voices that they trust.”
An earlier version of this story stated incorrectly that the cost of the Road Tour is $50 million. It is actually $15 million.