Resumen de Five Decades of Missing Females in China
Ansley J. Coale, Judith Banister
This paper seeks to explain the dearth of females in the population of China in cohorts born from the late 1930s to the present. I demonstrate that in essentially all cohorts, the shortage of females compared to males is revealed the first time the cohort is enumerated in a census, and subsequently barely changes, indicating that female losses occur very early in life. Using the high quality data from the censuses and fertility surveys in China, I show that many of the births of the girls missing in the censuses were not reported in the surveys, because they died very young. The incidence of excess early female mortality (probably infanticide) declined precipitously in the Communist period, but not to zero. The recent escalation in the proportion of young females missing in China has been caused, in large part, by rapidly escalating sex-selective abortion. In the absence of special circumstances reducing the number of the male or the female members of a population, and in the absence of substantial gains or losses from migration, the numbers of males and females are approximately equal. The approximately equal numbers are the result of the usual slight majority of males at birth, and the usual somewhat higher mortality of males than females in the absence of differential treatment of the sexes or a large incidence of military mortality. In almost all well-recorded ratios of the number of male to the number of female births, the ratio falls between 1.05 and 1.07; the slightly higher male than female mortality gradually cancels the initial male majority as each cohort advances in age. In some Asian and African countries, however, the ratio of males to females in the population is higher than would be expected from the typical sex ratio at birth and typical differential mortality. The source of the high masculinity is female mortality that is higher, relative to male mortality, than would prevail if both sexes had equal access to factors promoting good health. One of the populations with higher than expected masculinity is that of the People's Republic of China. In the 1990 census, the recorded ratio of males to females was 1.066; a normal sex ratio at birth and normal differences in survival would have yielded a ratio no higher than 1.02. Female mortality evidently has been abnormally high relative to male mortality in China. In this paper I analyze data from the four modern censuses of China from 1953 to 1990 and from two large-scale retrospective fertility surveys in 1982 and 1988 to trace the record of excess masculinity in successive Chinese birth cohorts, from those born in the late 1930s to those born in the late 1980s. The extent and high quality of the data permit several inferences, including the probable existence of high rates of female infanticide in the 1930s and early 1940s, a large reduction of this practice by the 1960s, the effect of the famine in 1959-61 on the sex ratio in selected cohorts, evidence of the beginning around 1970 of selective termination of childbearing following a male birth, and the emerging impact on the sex ratio at birth in the 1980s of sex-selective abortion. The paper begins with a description of the relevant data in the Chinese censuses and in the large fertility surveys of the 1980s, and then presents an analysis of the data for early cohorts born from 1936 to 1954, for cohorts that spent childhood years during the great famine of 1959-61, for cohorts born before and during the national introduction of a family planning program, and for cohorts born in the 1980s.
