Page 16 - IJPS-2-2

Assessments of mortality at oldest-old ages by province in China's 2000 and 2010 censuses

Table 2. Average five-year age-specific rate (%) of underestimation of probability of dying at ages 80-95 under the assumption that the death rates at
ages 70-80 were accurate
2000 census 2010 census
Regions Provinces Males Females Males Females
A B A B A B A B
China 0.0 8.0 4.2 14.4 0.0 13.0 4.5 18.8
North Beijing 0.0 4.9 0.0 10.2 0.0 5.4 0.0 12.4
North Tianjin 0.0 6.8 1.3 12.2 8.0 20.8 6.7 22.2
North Hebei 0.0 0.0 0.0 9.0 0.0 3.6 0.7 12.8
North Shanxi 0.0 0.0 0.0 3.9 0.0 9.9 3.6 15.8
North Neimeng 0.0 0.9 0.0 3.6 8.7 20.7 9.8 23.4
Northeast Liaoning 0.0 1.9 0.0 7.4 0.0 9.1 0.0 13.6
Northeast Jilin 12.2 21.3 14.7 23.5 19.7 30.5 22.0 33.9
Northeast Heilongjiang 12.7 22.2 16.0 25.5 22.5 32.8 27.5 38.6
East Shanghai 0.0 0.0 0.0 2.6 0.0 1.7 0.0 4.4
East Jiangsu 0.0 3.8 0.0 7.3 0.0 1.6 0.0 7.3
East Zhejiang 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
East Anhui 6.4 16.2 5.3 17.7 1.8 14.2 5.2 20.3
East Fujian 0.0 6.3 0.0 12.7 0.0 8.6 0.0 9.6
East Jiangxi 0.7 9.6 5.0 15.6 0.0 11.0 6.6 19.8
East Shandong 0.0 0.0 0.0 8.4 0.0 6.0 0.3 15.1
South Central Henan 1.5 10.6 11.3 21.0 6.7 17.5 18.9 31.0
South Central Hubei 0.0 7.4 7.9 16.9 7.0 18.7 6.3 20.5
South Central Hunan 3.0 13.0 6.3 17.2 2.3 15.4 8.7 23.4
South Central Guangdong 0.0 6.4 0.0 13.6 2.9 15.4 0.5 17.8
South Central Guangxi 4.2 15.8 12.2 25.4 9.6 21.3 16.9 31.4
South Central Hainan 10.4 21.4 14.4 28.2 17.0 28.4 16.8 33.6
South Central Chongqing 2.0 12.4 4.0 15.0 6.1 19.0 2.8 19.4
Southwest Sichuan 4.7 14.6 10.7 20.7 4.7 17.0 6.8 21.5
Southwest Guizhou 0.0 8.1 2.6 13.1 3.0 14.8 5.5 19.1
Southwest Yunnan 0.0 2.8 0.0 6.2 0.0 9.3 1.9 14.7
Southwest Tibet 32.8 38.6 33.3 39.8 25.2 33.2 37.9 44.9
Northwest Shaanxi 1.0 9.2 4.8 13.0 3.8 15.7 9.5 21.5
Northwest Gansu 12.6 19.0 2.1 9.1 7.1 17.9 11.1 21.9
Northwest Ningxia 26.9 33.7 23.6 29.8 13.7 24.2 17.0 27.4
Northwest Qinghai 0.3 10.0 9.1 16.7 0.0 4.0 0.0 6.7
Northwest Xinjiang 30.3 38.8 37.9 45.4 23.1 33.7 34.1 44.6

Note: (1). A: The average five-year age underestimation rate was calculated as (100*(1- q / ˆ q ), where q represents the observed average probability of dying over
x
n
n x
n
x
ˆ
q
the five-year age group in the entire age group [x, x+n) (i.e., 15 q =1- 1/3 (1− 15 80 ) ) and q represents the corresponding average probability of dying derived from
n x
80
ˆ
the fitted estimates of the lower boundary of the confidence ellipse in Figure 1. (2). B: q was estimated from the linear regression models between logit-transformed
n x
2
probabilities of dying at ages 80-95 and at ages 70-80. The R was 0.87 for males and 0.93 for females. (3) The average rates of underestimation for five-year age group
were mildly lower than those for single years of age.
10 International Journal of Population Studies | 2016, Volume 2, Issue 2