Population Census, State Legibility, and Politics of Data in China

Junchao Tang, University of Michigan

A critical issue lying behind the scholarly debate of population projection is how much we should trust the official demographic data. In this project, I aim to cast new lights into the question through examining the data quality of recent decennial census and its subnational variation across post-Mao China with a socio-political lens of interpretation. Drawing from the latest quality-checking techniques developed in demography, I provided a quantitative evaluation of the two hands of Chinese state legibility. I advocated that two processes of state legibility formation should be distinguished: state legibility brought by its information collection infrastructure, followed by a modification brought by the distortion happening in the bureaucratic reporting system. I provided an empirical approximation of the infrastructure-based state legibility based on a quantitative evaluation of the quality of age data in recent decennial censuses in China and explore exploratory factors contributing to the subnational variation.

See extended abstract

 Presented in Session 21. Evaluating Data Quality II