Page 19 - IJPS-5-1

International Journal of Population Studies

RESEARCH ARTICLE

Use of the average age ratio method in

analyzing age heaping in censuses: The

case of China

Danan Gu *, Qiushi Feng 2
1
1 Population Division, Department of Economic and Social Affairs, United Nations,
New York, NY 10017, USA,
2 Department of Sociology, Centre for Family and Population Research, National
University of Singapore, Singapore

ABSTRACT: Based on the methods of the average period age ratio and the average cohort
age ratio, this study systematically assesses age heaping or digit preference in all population
censuses of China. Our study finds that the overall age heaping was relatively low in the
Chinese censuses; however, there was a notable preference for ages ending with zero after age
50 in the first two censuses, despite a weakening trend over time. Our study further shows that
age heaping in China’s censuses is likely associated with age-related policies such as those on
late marriage and retirement. As shown in the study, the average age ratio method can be an
ARTICLE INFO alternative of the Whipple’s Index and be improved if the size of birth cohort was taken into
Received: September 8, 2018 account when the number of births is generally reliable.
Accepted: October 25, 2018
Published: November 3, 2018 Keywords: Age concentration index; Age heaping; age ratio; Age reporting; Average
age ratio; Average period age ratio; Average cohort age ratio; China; Census; Digit
*CORRESPONDING AUTHOR
preference; Digit avoidance; Whipple’s index; Myer’s blended index
Danan Gu,
Population Division, Department
of Economic and Social Affairs, 1. Introduction
United Nations, New York,
NY 10017, USA. Censuses are foundational data sources for population studies, the accuracy of which
gudanan@yahoo.com determines the validity and reliability of demographic research (Moultrie, Dorrington,
Hill et al., 2013; United Nations, 1983, 2008). Throughout the procedure of census, errors
CITATION
may occur on many occasions such as age reporting and recording. Age reporting may
Gu D and Feng Q (2019). Use suffer from two types of errors, intentional and unintentional misreporting. Unintentional
of the average age ratio method misreporting mainly refers to the situation where the interviewee does not know or could
in analyzing age heaping
in censuses: The case of not remember the exact birth date. By contrast, intentional misreporting occurs when the
China. International Journal of interviewee, though knowing the exact birth date, chooses to “avoid” or “prefer” certain
Population Studies, 5(1):13-26. digits in reporting of his/her age (commonly known as digit preference), due to personal,
doi: 10.18063/ijps.v5i1.979 cultural, and social considerations (Moultrie, Dorrington, Hill et al., 2013; United Nations,
Copyright: © 2019 Danan Gu and 1983). For example, in a society where vital registration is underdeveloped, a young
Feng Q. This is an Open Access person who has not reached the minimum legal age for marriage may report an older age
article distributed under the for marital registration. Due to similar reasons, some young people may report older ages
terms of the Creative Commons in applications to join the military, and some older adults may exaggerate their ages for
Attribution-NonCommercial 4.0
International License (http:// the legitimacy of retirement welfare. Unintentional misreporting is mainly random, and
creativecommons.org/licenses/ thus, it will not cause a severe bias in results. However, intentional misreporting would
by-nc/4.0/), permitting all lead to “age heaping,” a situation in which the population at a certain age or an age ended
noncommercial use, distribution, with certain digit significantly outnumbers populations at adjacent ages or ages ended
and reproduction in any medium,
provided the original work is with other digits (Moultrie, Dorrington, Hill et al., 2013; Spoorenberg, 2007). This would
properly cited. introduce inaccuracy in the age statistics of population studies. Although the data quality

International Journal of Population Studies | 2019, Volume 5, Issue 1 13