A preliminary examination of the quality of birth population registration in China's Fourth Population Census

Tu, P.

Chinese Journal of Population Science 5(3): 205-212

1993


ISSN/ISBN: 1044-8403
PMID: 12345586
Document Number: 255141
Preliminary analysis was made of 1990 Chinese Census data on fertility by province in order to determine data consistency of birth, mortality, and surviving population data. The results indicated that there was indeed a discrepancy in female births reported. There were about 1.405 million births underreported in 1989 and the first half of 1990. 428,000 were underreported male infants and 977,000 were underreported female infants, which yielded 3.3% of underreporting in 1989 and 4.9% in 1990 for total population (5.6% for female births and 2.2% for male births). Underreports averaged for 1989 and 1990 3.8%, which was 0.8% lower than the actual national level. The fertility rate was thus 0.1% lower than the actual level. These levels of underreporting were consistent with averages found in Canada and the United States, but because the absolute level of population was so much higher in China, the absolute error was still considerable. The underreports were based on the assumption that surviving population and mortality were accurate; inaccuracies in population and mortality figures would increase underreporting of births. Provincial data was more unreliable than national data because the sample was only 1% and there were large random fluctuations. Underreporting was evident in all provinces to some extent, but it never exceeded 10% and usually was within 5%. Consistency was lower for female births than male births by province and lower in 1990 than in 1989. Data was provided by province for total births by sex and province for 1989 and 1990. When there was underreporting, the sex ratios of 114.7 and 116.9 in 1989 came out higher than the national average of 110.8 and 12.5 based on a 10% sample. Consistency of data was determined by comparing actual births reported with a calculation of expected births from the number of persons surviving and the number of deaths in 1989 and 1990. The ratio of actual to expected births determined the internal consistency.

Document emailed within 1 workday
Secure & encrypted payments