Assessing Fertility Levels in China Using Variable-r Method

Yong Cai, University of Washington

Many scholars believe that the fertility levels reported in China since its 1990 census are too low to be true. They attribute the reported low fertility to the deterioration of Chinese statistical system, arguing that a large number of births, particularly those out-of-plan births were underreported in the official enumerations. The variable-r method is applied to Chinese population enumerations from 1989 to 2002 to assess the data quality and to estimate the nature reproductive rates, assuming that the underreporting is a systematic problem over time. The estimated fertility levels are close to what have been reported. It indicates that Chinese fertility has indeed reached a very low level in 1990s. This exercise also suggests that Chinese population enumeration data, although with many systematic problems, still provide relatively good base for calculating demographic estimates.