Chinese population censuring has evolved its own distinctive characteristics.

The historical background of China's 4 population censuses is provided along with a discussion about what population is targeted for the county: existing, resident, or legal population. Improvements in census items and survey methods are also indicated. Schedules for data processing and publication are discussed. National censuses were conducted in 1953, 1964, and 1982. The 1st was taken in order to prepare for a general election; it was based on a national population survey of the resident population. The 2nd was undertaken to contend with the prior overreporting to obtain more rations. The population included resident as well as some legal population. The 3rd census came after the cultural revolution of 1966-70 and figures were considered inaccurate and long overdue. The survey counted the resident population exclusively. On December 8, 1983, the People's Congress passed into law the institution of a national population census to be conducted decennially beginning in 1990 and intercensal counts. The 1990 census counted resident population and some existing population. The target population was a problem because grain supply and other administrative activities are conducted through a registered place of permanent residence. Household registration and permanent residence should be the same, but are not in 2% of the population (20 million). In 1990, the resident population was considered to be the population actually residing in the locality for 1 year regardless of residence permit. Those living 1 year in the locality but having left their permanent residence 1 year ago are also considered permanent residents. In cities this population could be as high as 205. China also has a floating population which is difficult to count and is estimated at 50 million. The 1990 census also came closer to UN standards and included 21 of the UN recommended items. Additional items on residence permit and nonworking population will assist in making economic decisions. 9 items were devoted to death registration in order to correct past deficiencies. 2 items pertain to migration. Still excluded but needed are language items, detailed elderly and urban counts, and religious believers. 10 million people were required for household interview and administration. 30 quality control standards are set for the 7 stages. Data management involves hand collection, 10% sample collection, and 100% machine collection. There are 6 papers by subject area, and the beginning level is the village and aggregation increases up through township, county, locality, province, and nation. Provinces must report by September 10 of the year ending in 0. Hand collection by level is faster than machine collection which takes a year. The 1990 publication of data will be more inclusive and consist of 11 volumes by province on 1) general situation, 20 urban/rural distribution, 3) race, 4) age, 5) cultural level, 6) occupation, 7) marital status, 8) household, 9) fertility state, 10) mortality rate, and 11) migration. Tapes and discs may be purchased.