Page 21 - IJPS-1-1

Zachary Zimmer, Mira Hidajat, and Yasuhiko Saito

(iii) Overall patterns of change hide heterogeneity. Heterogeneity exists across ru-
ral/urban residence such that a compression of morbidity is more likely being experienced
by urban Chinese than by rural Chinese. Education appears to be affecting changes in
DFLE less than place of residence. A variety of evidence exists (Yip, 2010; Liu et al., 2003;
Beach, 2001; Zimmer et al., 2007) that indicates increasingly better health conditions in
urban areas of China as opposed to rural . The fact that education has no impact is perhaps
more surprising. However, other evidence hints that the association between education and
health may be suppressed in some societies (Montez and Friedman, 2015). The association
between education and health in China may be less robust than elsewhere because other
indicators of socioeconomic status better distinguish social hierarchy in China, such as
state sector employment and party membership. A report by the Population Reference Bu-
reau (2015) indicates that education has surprising effects on health in China that include
substantial differences for males versus females. We emphasize that the current study did
not suggest a non-association between education and DFLE, but rather that changes over
the short period are not observed.
In some ways, the current results contrast from the findings from the United States
where the issue of changes in disability has been examined for a longer period of time.
Crimmins et al. (2009) provide evidence of both increasing life and disability-free life ex-
pectancy in the United States, and although they do not test for it directly, the gains in each
appear to be parallel. While some studies have examined TLE and DFLE by education
(Crimmins et al., 1996 and 2001), whether and how this might be changing over time has
only been assessed with the monitoring of disability trends. Schoeni et al. (2001) showed
declining rates of disability in the U.S. with the more educated benefiting the most. Martin
et al. (2010) confirmed declining rates of disability and demonstrated that males were far-
ing a bit better than females.
The current study has limitations. The length of time between the two observation pe-
riods is short. Although various datasets with information about the health of older persons
are beginning to become available for China, data for the type of examination provided
here, for a population aged 65 and older, are still not readily available over a long-term.
The current study can provide a baseline for future longer-term studies. Small numbers
within certain sub-groups is another obvious weakness. However, this is mostly a problem
for the high educated females, of whom few exist in China. Examination of confidence
intervals indicates findings that are robust for other sub-groups. Increases in life expec-
tancy over time are a little greater than those reported officially. Some of this may be due
to missing mortality in loss to follow-up or to those who are not healthy being uninterested
in responding in the first place. Moreover, while follow-up rates for this survey are high
and data assessments have suggested that attrition would not play a role in our estimations
(Zeng et al., 2002; Gu, 2007), we note that these data are based on longitudinal panel data
and there is always a risk that whatever minimal attrition exists could be non-random
across variables of interest. Yet, there is no reason on the surface to believe that any non-
random attrition would be systematic across age, sex, education and rural/urban residence.
In sum, the findings portend heterogeneous compression of morbidity in China. During
the study period changes in both TLE and DFLE tended to be more favorable for females
than males and more favorable for urban than rural residents. In contrast, changes across
levels of education are not very consequential. Those in older age brackets benefitted more
than younger elderly. Rural residents of China are a population of policy concern. Already
having worse health status than others, and less access to health resources, rural residents
do not seem to be sharing equally in health gains experienced by others, which, if extrapo-
lated into the future, will only put them even further behind and create greater inequalities.

International Journal of Population Studies | 2015, Volume 1, Issue 1 15