TY  - JOUR
A1  - Dorélien, Audrey
A1  - Xu, Hongwei
T1  - Estimating rural–urban disparities in self-rated health in China: Impact of choice of urban definition
Y1  - 2020/12/03
JF  - Demographic Research
JO  - Demographic Research
SN  - 1435-9871
SP  - 1429
EP  - 1460
DO  - 10.4054/DemRes.2020.43.49
VL  - 43
IS  - 49
UR  - https://www.demographic-research.org/volumes/vol43/49/
L1  - https://www.demographic-research.org/volumes/vol43/49/43-49.pdf
L2  - https://www.demographic-research.org/volumes/vol43/49/43-49.pdf
L3  - https://www.demographic-research.org/volumes/vol43/49/files/43-49_Supplement%20B.%20Table%20to%20accompany%20Figure%203.xlsx
L3  - https://www.demographic-research.org/volumes/vol43/49/files/43-49_Supplementary%20Material%20A%20Anchoring%20Vignettes.pdf
N2  - Objective: We use the 2014 China Family Panel Studies (CFPS), which includes anchoring vignettes, to provide an up-to-date assessment of rural–urban health disparities as measured by self-rated health (SRH) in China. Our analysis is based on multiple definitions (hukou and the two different residence-based definitions) of rural–urban and migration status; previous research was inconclusive due to the use of different definitions and concerns about status-based differential health expectations (reporting heterogeneity). 

Results: We find a nonlinear difference between rural and urban Chinese in how they self-assess health status, regardless of the urban definition used. Urban respondents do not always hold a higher standard for self-assessment of health. Instead, their rating styles depend on the level of latent health. After controlling for the reporting heterogeneity, we find on average a slight urban advantage in SRH, but it is most pronounced when using the statistical (density dependent) definition of urban. 

Contribution: We study rural–urban health disparities based on three different urban definitions and migration status. Although we examine the urban definitions that are specific to China, we demonstrate a mindful approach when multiple definitions exist and caution against any simplistic approach that ignores context-specific urban definition. We also provide clear illustrations of the different types of reporting heterogeneity, as well as a way to visualize the cut-points, thresholds, and latent health estimates. 

ER  -