1.3 Billion Divided by 6.5 Million, and Watch That
Floating Decimal
Taking an accurate tally in the world’s most populous and mobile country is already a formidable task. But when 6.5 million census takers fan out across China on 1 November, they will face an additional challenge: getting a handle on a swelling migrant, or “floating,” population. The Chinese government has put “a lot of attention on enumerating the floating population as accurately as possible,” says Duan Chengrong, a demographer at Beijing’s Renmin University who helped design the sixth national census. China’s unique solution this year—adding presumed residents to an actual head count—unsettles some demographers.
In the 2000 census, enumerators identified people using household registration records. Migrants had to have lived in a city for 6 months to be counted, a requirement that made it easy for census takers to miss recent arrivals. The result was an unprecedented rate of undercounting attached to China’s official tally. China’s last population figure of 1.3 billion includes 22.5 million Chinese thought to have been missed in the 2000 count. Many countries adjust results to account for underreporting and other statistical inconsistencies; on that score, China is not alone.
This year, to reduce undercounting, census takers will bolster registration records with a partial head count, using digital maps to pinpoint homes in far-flung or unlikely areas. “With geographic information systems, we can see clearly which houses are in which census districts,” says Li Xiru, who is overseeing the technology’s introduction for the National Bureau of Statistics, which administers the census. The 6-month residency rule still applies for floaters, Li says, but enumerators will try to tally recent migrants in their villages by asking families about absent members. Foreigners and residents of Hong Kong and Macao will be counted for the first time.
An approach that combines registration data with a head count and detailed interviews is “unique in the world,” says Zhongdong Ma, a demographer at Hong Kong University of Science and Technology. The compromise acknowledges that household registration information is increasingly irrelevant. In recent decades, over 100 million rural residents have flooded China’s cities in search of work. Many middle-class Chinese, meanwhile, have fled to the suburbs while maintaining apartments and residence permits in the city.
The novelty of this year’s census worries observers like Qiao Xiaochun, a demographer at Peking University’s Institute of Population Research. After working as a census officer in 1982, he published a book proposing that China switch to a de facto count, in which everyone is tallied in the districts where they are found on census day, regardless of registration. This year’s change is a step in the right direction, he says, but it opens alarming loopholes. People “may be enumerated twice,” Qiao says. “Or they may not be enumerated at all.”
Accuracy will depend on the enumerators, mostly short-term workers recruited in the past few months who will disperse throughout China over 10 days. They will help respondents fill out a form listing 15 questions—more than ever before. A randomly selected 10% of Chinese will complete a long form that asks about things like education, housing, and marital status. The government is “going to great lengths to make a really detailed record,” says Kam Wing Chan, a geographer at the University of Washington, Seattle, who studies migration in China. But that thoroughness, he says, has drawbacks: More questions could reduce the response rate.
In 1982, when China introduced modern census-taking techniques, many Chinese lived in closely monitored work units. Today people are busier, more mobile, and more private. Everyone from unmarried couples worried about disapproving neighbors to tax-evading tycoons may have reason to dread the enumerators.
In recent months, a barrage of propaganda has assured wary Chinese that information will be used only for calculating population data—and reminded them that the census is essential for planning schools, nursing homes, and transportation networks. But some local authorities implementing the count have already reneged on the promise to safeguard individual rights. During a census rehearsal in Beijing last August, plainclothes police posing as enumerators detained dissident writer Xie Chaoping.
Other problems are disappearing with economic development. In 2000, some couples who had babies outside the one-child policy did not report those births, and as many as a quarter of young children went uncounted. Surveys suggest that the number of second and third births has fallen since then, so today’s parents have fewer children to hide.
Along with the total population, scholars look forward to an updated sex ratio at birth—a 2005 sample survey found that 120.5 boys were born for every 100 girls, the most skewed ratio in the world—along with new fertility and mortality rates. The U.S. Census Bureau will use those numbers to revise its estimate for when China’s population will peak, says Daniel Goodkind, a demographer at the bureau. The current forecast is 2026.
Beyond those broader tabulations, demographers wonder how much they’ll learn. The Chinese government does not publish detailed data that allows demographers to make sense of its census tweaks. “How the census is adjusted is not known,” says Cai Yong, a demographer at the University of North Carolina, Chapel Hill. China’s biggest tweak yet—counting absentee residents—should only increase the uncertainty.