Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
SUMMARY In 1982 the People's Republic of China carried out a census with a more comprehensive interview schedule than ever before employed in China and using a very large, carefully chosen, and extensively trained field staff. The census was preceded by pilot surveys to test the instruments and field procedures. A tabulation of 10 percent of the individual returns has been completed, published in China, and made available abroad in limited circulation. Also in 1982 a fertility survey covering a very large sample of households (total population of more than 1 million) was conducted in China, and its results have been published in great detail in a special issue of the Chinese Journal Population and Economics. In addi- tion, the distribution of the population by sex and single years of age as enumerated in the censuses of 1953 and 1964 has been recently released. This new informa- tion, supplemented by time series of registered births and deaths and end-of-year population totals extending back to the 1950s and by data from other large recent surveys, provides a sound basis for constructing an accurate and detailed history of the remarkable changes in fertility, mortality, and marriage that have occurred in China since the People's Republic was established. The newly available information includes complete histories of marriage and childbearing of women up to age 67 in the 1/1,000 fertility survey of 1982. The responses have been analyzed and tabulated in the form of marriage rates and birth rates by single years of age in single calendar years from 1950 to 1981. When the survey data are combined and compared with the census data for 1953, 1964, and 1982 on numbers of persons by sex and single years of age, they pass a series of stringent tests of accuracy and consistency. The same analysis 1
2 reveals that official data on birth and death rates have understated the true numbers by a considerable margin. The tests support the substantive findings in this report on levels and changes in fertility, nuptiality, and mortality in China since 1950. Fertility Rates. The birth rate in China has been higher than that listed in official sources. In the 1950s the birth rate was generally above 40 per 1,000 until a precipitous fall--from 42.5 per 1,000 in 1957 to 21.9 in 1961--that coincided with the Great Leap Forward and the ensuing years of economic disruption and famine. The post-crisis peak birth rate in 1963 was just short of SO births per 1,000. A more useful fertility measure, the total fertility rate (TFR)--the average number of children that would be born in a lifetime to women subject to the birth rates by age in a given period--was about 6.0 before the Great Leap Forward, declined to 3.3 in 1961, rose to 7.5 in 1963, returned to 6.0 in the mid-1960s, fell steeply to only 2.2 in 1980, and then rose slightly to over 2.6 in 1981 and 1982. The birth data on which these fertility rates are based are derived from the new detailed information, especially that from the fertility survey. It is clear that the number of births previously listed in official sources has been incomplete: by more than IS percent in the 1950s, by less than 10 percent in the late 1960s and early 1970s, and by 15 percent or more since the intensification of the antinatalist program in 1979. Age Pattern of Fertility. The age pattern of fertility of married women in the 1950s was a pattern of gradual decline in the rate of childbearing with age until age 30; the decline steepened after age 30 and especially after age 35. This age pattern closely resembles the early gradual and later steep decline of marital fertility rates with age of woman that is characteristic of popula- tions in which couples practice little contraception or induced abortion. This age pattern of marital fertility in the 1950s supports the inference of little use of contraception. In the 1970s (and especially in 1980), by contrast, marital fertility rates fell very steeply with age of women after their late 20s, a pattern character- istic of very general resort to contraception to limit fertility after desired family size is reached. In 1961, when the TFR fell to only a little more than half the TFR of 1955 or 1957, the fertility of married
3 women was very much reduced at all ages. A nearly uniform reduction in fertility at different ages is consistent with a quasi-biological cause of low fertility--i.e., low fertility was the result of disruption of normal life and famine-induced subfecundity rather than a large increase in the use of contraception. The unmatched post-crisis TFR of 7.5 in 1963 involved peak marital fertility rates at all ages. These high rates at all ages may also have a quasi-biological explanation. Newly married couples (there was a very high first-marriage rate in 1962) and couples resuming normal life are especially susceptible to the risk of childbearing since few of the women are protected from the risk of pregnancy because of nursing a previously born child. Contraceptive Use. In 1981 contraception--mostly sterilization, the IUD, and contraceptive pills--was practiced by more than two-thirds of married women aged 15 to 49. Mean Age at First Marriage. The mean age at first marriage of women was about 18.S vears in the 1940s, about a year older than that estimated for rural China in 1930. The mean age at marriage rose gradually (with some fluctuations) to a little more than 20 in 1970, and then steeply to more than 23 in 1979. There was then a slight decline, of about four- tenths a year, to a mean age of 22.7 years in the first half of 1982. Effects on Fertility of Changes in Mean Age at First Marriage. The changing age of entry into marriage contributed strongly to changes in fertility. Had the _ noncontraceptive marital fertility rates at each age of the 1950s continued, the increase in age at marriage by itself, by exempting many younger women from the risk of childbearing, would have led to a TFR in 1980 that was 20 percent below the TFR of 1950--a hypothetical decline about one-third as great as the actual one. The rise in age at marriage in the 1970s would have produced (by a different mechanism) a 20 percent reduction in the TFR during that decade even if from 1970 on married women had successfully attained unchanging coals of restricted family size. _ percent decline in TFR even though married women produce an unchanging total number of children per marriage-- arises from the temporary reduction in the number of marriages that is caused by a rise in mean age at ~ , _ , , This annarentlv anomalous effect--a 20
4 marriage. About one-third of the reduction in the TFR from 1970 to 1980 was associated with the increase in age at marriage and would have occurred with constant duration-specific marital fertility rates. When mean age at marriage ceases to rise, the diminu- tion in the number of marriages caused by rising age at marriage ceases, and the number of the newly married women increases. In 1980-82 a sharp increase in the total first-marriage rate accompanied the termination and slight reversal of the increase in mean age at marriage. Most of the upturn in TFR after 1980 was the result of the marriage boom in 1980-82 and would have occurred with constant fertility rates at each duration of marriage. Upward pressure on the TFR will continue because the highest marital fertility rates occur one or two years after the date of marriage; the large number of marriages in 1981 and 1982 will inflate births in 1983 and 1984 even if the recently married have only one or two children. Urban/Rural and Other Fertility Differences. Before , the temporary sharp decline in the TFR that began in 1958, the TFR in the cities was about 10 percent below the rural TFR; about half of the difference in fertility can be ascribed to later marriage in the urban population. As shown in Figure 1, between 1960 and 1966, the urban TFR fell to about half the rural TFR, and it remained at about that fraction when the large reduction in rural fertility began in 1970. Other differentials in fertility that are usually present in the first years of a major reduction were present in China in 1981: fertility was lower for more educated women and for women in higher occupational categories; the minority ethnic groups had much higher fertility than the rural Han majority. Future Trends in Fertility. Further upward pressure on the birth rate in the late 1980s is built into the age distribution of the Chinese population, shown in Figure 2. Women in their early 20s in 1982 were born in 1958-61, a period of greatly reduced birth cohorts. In the next few years the very large birth cohorts of 1963-70 will be in the normal ages of first marriage and thereafter in the very fertile years soon after marriage. Male/Female Birth Ratios. The large-scale fertility survey recorded ratios of male to female births that were very close to the worldwide normal ratio of about 106
s 8 7 / LL ~ 6 to 5 J 1~ 4 L1J o %. ~ ~ _ ~ ^_ ~, _ 1 1 1 1 0 50 1955 1960 1965 1970 1975 1980 1982 ·~ it--- ~ I /\\ \~~\eee A.--. [1 \ .1 \ t! \ i' 1 \ ·. ~ ~1 Total\;;> / ~ Urban ~ x`_ ~ - 1 1 YEAR FIGURE 1 Total Fertility Rates for Urban and Rural Areas and for China, 1952-82: China males per 100 females among first and second births in the rural population, but there were more than 112 males per 100 females for third- and higher-order births. The male/female ratio for urban births was somewhat higher (over 108) for first births, and much higher (about 118) among the small number (257) of births beyond the first. Experience in other populations is of slightly declining male/female ratios with birth order. Since stopping rules--no more births following a male--do not affect the male/female ratio and sex-selective abortion on a large scale does not seem possible in rural China, the explana- tion for the reported male/female birth ratios must be unreported higher-order female births. There may be a connection between failure to report a higher-order female birth in the survey and the occurrence of female infanticide, which has been widely reported (and deplored) in the Chinese press. Given the penalties imposed in the one-child campaign and the cultural preference for male births, higher-order female births are doubtless especially unwelcome.
6 So 80 70 60 50 UJ 40 30 20 10 o Males 1 ; ~ _ _ ,~ ~ Females 1 1 ' 1 1 1 1 1 1.5 1.0 0.5 0 0.5 1.0 1.5 PERCENT FIGURE 2 Age Pyramid of the Population, 1982: China Source: China, Population Census Office (1983:Table 19). Life Expectancy. Average death rates by age for each sex in each intercensal interval can be calculated from census data and constructed numbers of births. From these death rates life tables are derived that show the average age at death that would result from the continued prevalence of the calculated intercensal average death rates. The expectation of life at birth increased from 42 for males and 46 for females in 1953-64 to 62 for males and 63 for females in 1964-68. This increase in less than two decades replicates the increase typical of six West European populations from 1870 to 1940. A life table was recently calculated for 1981 from deaths reported in the 1982 census. It shows a further increase in expectation of life at birth to 66 years for males and 69 for females. Mortality Rates. Official figures on the annual number of deaths understate the true number by a greater proportion than the proportionate understatement of the number of births. It is possible to determine only the
7 40 30 UJ .. 20 U] 10 - / \ · ·- ~ ~. O ~ 1950 1955 1960 1965 1970 1975 1980 1985 YEAR FIGURE 3 Death Rates (per 1,000) from Recorded Deaths (solid line) and Adjusted for Underreporting (dotted line), 1953-82: China understatement of the total number of deaths for inter- censal periods: about 38 percent of the deaths in 1953-64 were not recorded and about 16 percent of the deaths in 1964-82. According to official sources, annual death rates (Figure 3) were about 15 per 1,000 in the early 1950s and declined to about 11 per 1,000 in 1957. There was an increase in death rates during the years of the Great Leap Forward and the ensuing crisis, with an officially listed peak rate of 25 per 1,000 in 1960. The death rate fell to 10 in 1963 as normal conditions were restored, then continued to decline to a rate between 6 and 7 per 1,000 in the late 1970s and early 1980s. When the intercensal aggregate shortfall in the number of death s derived from official sources is allocated under an assumption of improving completeness after 1955 and constant completeness from 1964-82, the estimated death rate in the early 1950s is above 20, the peak death rate in 1960 is above 35, and the recent death rate is between 7 and 8 (rather than between 6 and 7). Excess deaths (those above a linear trend) from 1958-63 are about 16 million when based on the understated official figures and about 27 million when adjusted for understatement.
