Page 92 - IJPS-8-1

International Journal of
Population Studies Fertility by parity in China

divergence in scholars’ understanding of fertility levels and population censuses performed in 1982, 1990, 2000, 2010,
trends in China. Estimations of TFR directly obtained from and 2020; the 1% national intercensal population sample
census or sample survey data are lower than those obtained surveys conducted in 1987, 1995, 2005, and 2015; national
based on adjusted census data or other data sources. Some one- per-thousand annual sample surveys of population
scholars suggested that the underreporting of births is change from 1982 to 2020; and national retrospective
greater for the second and higher orders because births do fertility sample surveys conducted in 1982, 1988, 1992,
not conform to the policy restriction, such as births with 1997, 2001, and 2017. These population censuses and
insufficient intervals or even over-born, are more likely population sample surveys collected detailed information
to be underreported (Wang, 2003; Zhang & Su, 1995). on age-parity-specific fertility rates, age at first marriage,
Accordingly, the existing literature on fertility by parity has and age when children were born for Chinese women.
focused on a specific period, particularly the early years. These data allow us to establish the historical fertility
levels and patterns by parity. The availability of the Human
The period TFR has been the most common measure of
the fertility level due to its simplicity and wide availability Fertility Database allows us to compare fertility by parity in
China with that in other low-fertility societies.
(Bongaarts & Feeney, 1998; Ma, et al., 1986a; McDonald &
Kippen, 2007). However, the period TFR is hard to reveal Considered the widely established fact of varying
the accurate effects of fertility policy. The period TFR is not degrees of birth under-reporting in censuses and surveys
able to distinguish a change in timing (or tempo) of cohort since the early 1990s, this paper made an effort to adjust the
fertility from a change in the level (or quantum) of cohort estimates of period fertility rate from 2000 to 2020 based on
fertility (Schoen, 2004). The period TFR is regarded as biased the number of births published by China’s National Bureau
or distorted by tempo effects. That is, estimates of TFR are of Statistic (NBS). First, it estimated the age-specific number
depressed during years when women delay childbearing and of women aged 15–49 from 2000 to 2020 based on the 2010
inflated in years when childbearing is accelerated (Bongaarts census. Second, with the assumption that the age-specific
& Feeney, 1998). More specifically, under the interference of fertility schedules are accurate in the census/population
various period factors, the estimates of parity-specific TFR sample survey data, it decomposed the births published
sometimes exceed one child per woman (e.g., Whelpton, by NBS into the different ages of women of childbearing
1945, 1954; Yao, 1995). For instance, the TFR for the first age. It is worth noting that the NBS updated the number of
births was >1 in many Western countries in the 1940s and births from 2011 to 2020 based on the 2020 census. Third, it
in China in the early 1980s, implying that women on average obtained the adjusted estimates of age- specific fertility rate
have more than one first birth, which is not interpretable based on census/sample survey data. The adjusted estimates
(Ma, et al., 1986a; Rallu & Toulemon, 1994). based on the 2017 China Fertility Survey followed the same
process. Due to the possible significant overestimate of
This study aims to evaluate China’s fertility by parity NBS-published births in the 1990s, we are not able to adjust
since 1949 using multiple sources of data and by adopting the estimates of period fertility rate in the 1990s applying
multiple fertility measures, including period TFR by the same strategy (Chen, 2016; Zhao & Chen, 2011).
parity, period parity progression ratios (PPPRs), period
parity-progression-ratio-based total fertility rate (PPTFR) 2.2. Methods
by parity, cohort parity progression ratio (CPPR), and The main aim of the study is to examine the aggregate level
cohort cumulative fertility rate (CCFR) by parity as well of fertility by parity in China. Besides the most commonly
as its decomposition. This enables us not only to assess available measure, the conventional TFR by parity, the
the fertility dynamics by parity over the past 70 years study includes multiple measures of fertility: PPPR,
but also to identify the possible effects of the changing PPTFR, CPPR, and CCFR. PPTFR provides more stable
fertility policy during the period. Performing this analysis and consistent estimates than conventional TFR because
is especially interesting given the recent adjustment in the they control not only for age but also for parity. CCFR
family planning policy. The results are expected to provide provides a straightforward measure of the fertility of real
important insights into fertility by parity in China by groups of women and has been championed as the most
further comparing the parity distribution in China with appropriate measure to analyze fertility (Ryder, 1986).
other low- fertility settings.
The definitions and calculations of TFR and CCFR
2. Data and methods are well known, while those of PPTFR are more complex.
The definitions of PPTFR can be understood by the
2.1. Data sources way in which their calculation improves compared with
Multiple sources of data are used in this research. The TFR, whose calculation controls for women’s age only.
fertility data of China mainly come from national The construction of PPTFR further takes parity into

Volume 8 Issue 1 (2022) 86 https://doi.org/10.36922/ijps.v8i1.348