WASHINGTON -- Sixty-five percent of U.S. households have mailed in their census forms, matching the 1990 response and halting a decades-long slide in participation, Census Director Kenneth Prewitt announced Wednesday.
Despite the good news, census-takers still face their hardest job, going door to door beginning next week to count millions of reluctant and resistant people who have not returned their questionnaires.
Stung by declining response rates in each successive count for decades -- the return rate was 78 percent in 1970 and 75 percent in 1980 -- the U.S. Census Bureau deployed extraordinary resources to lift participation this time. The government paid $167 million for a showy ad campaign, sent an advance letter to tell people forms were on the way, and signed up thousands of minority and immigrant advocacy groups to promote the census in undercounted communities.
Even so, officials had predicted that only 61 percent of forms would be returned. They said some people would confuse the census form with junk mail, and others would fear mistakenly that immigration authorities or their landlords would find out what they wrote. They were surprised by a broader backlash fueled by talk show hosts and some Republicans in Congress who said some census questions were too nosy.
Prewitt warned Wednesday that the privacy backlash could toughen the Census Bureau's hardest task, to count more than 43 million households that did not send back their questionnaires. But the census director -- who had been gloomy as recently as last week about the prospect of a successful count -- was ebullient at a news conference Wednesday.
''It should be celebrated,'' Prewitt said. ''It is not an insignificant accomplishment. . . . We stopped a decline in civic engagement that began in the 1960s -- stopped it in its tracks -- and that is no modest accomplishment.''
Census numbers are used to allocate congressional seats among states, redistrict within states and direct $185 billion a year in federal program funds. They also help enforce civil rights laws, and are used extensively by businesses and marketers.
So far, according to figures released Wednesday, the response pattern follows that of past counts, with the highest participation in the upper Midwest and the lowest in the South. Cities rank lower than suburbs. Areas with many people who are poor, immigrants or members of minority groups have lower response rates than middle-class or white communities.
But some areas broke the mold. California, which poured millions of dollars in state money into a census promotion campaign, vaulted from an average response in 1990 to above-average this time. Prewitt said it is hard to tell what worked, though, because other areas that tried hard to sell the census, such as San Antonio, did not better their responses.
Census Bureau officials are sifting through the data looking for patterns. Prewitt said one early indication is that areas the bureau labeled ''hard to count'' because they had low response rates in the past made more gains than areas that officials considered ''easy to count,'' although the totals for the easier-to-count areas were higher.
That could be good news for the follow-up campaign, and Prewitt has said it could reflect the success of advertisements, foreign-language outreach and other efforts targeting minority and immigrant communities.
One sour note in the celebration is the doubling since 1990 of the gap in return rates for the eight-question short form compared with the 53-question long form, which includes questions about plumbing, income and other matters some critics find intrusive.
As of Wednesday, 54.1 percent of long forms and 66.6 percent of short forms had been returned. Prewitt said ''we may be able to close that gap'' in the follow-up phase, but the extra work of obtaining long-form information could cut into savings resulting from the unexpectedly high overall return.
Training of 40,000 supervisors began Wednesday, and Prewitt said there are enough people in the labor pool to undertake the 10-week follow-up campaign.
However, a computer programming error may cause problems for that follow-up effort. The glitch left out the surnames of people who filed their forms, which census takers are supposed to have when they knock on doors in case people tell them they already responded. The error will be corrected with supplementary lists and additional training, but it may hinder the door-to-door effort, Prewitt said in a letter to Congress.
In a response, House census subcommittee Chairman Dan Miller, R-Fla., said he was ''deeply concerned that there is still a failure of the bureau to put in place adequate quality-control measures.''
The Census Bureau intends to hire half a million enumerators. Census-takers will wear plastic badges bearing the words ''census enumerator.'' They will carry black canvas bags embossed with the U.S. Commerce Department/U.S. Census Bureau logo.
Census Bureau policy is that enumerators may not ask to come inside a home, Prewitt said, and will not enter unless invited.
Brainerd Dispatch ©2013. All Rights Reserved.