中国人口死亡率空间差异特征

Spatial Distribution Characteristics of Death Rate in China

运用直接标准化方法分析国家统计局公布的第五次、第六次人口普查数据,探讨了2000―2010年中国人口死亡率空间差异。首先,分别选取2000、2010年人口普查时的全国人口年龄结构作为2次普查时的区域标准人口年龄结构,比较这2次人口普查时死亡率的空间差异。结果发现:标准化死亡率与粗死亡率的区域差异存在明显的不同,经过标准化以后,有的区域甚至发生了相反的变化;其次,选用2次人口普查时点全国人口年龄结构的平均年龄结构作为标准人口年龄结构,以计算各地区在2000―2010年的死亡率变化情况,并通过人口平均预期寿命指标的变化进行对比证实,结果发现:2000―2010年全国所有地区的标准化死亡率均在下降,平均预期寿命都在提高。但若仅从粗死亡率的变化情况来分析,可能会错误地判断部分地区的人口死亡率出现了上升的趋势。在比较不同区域或不同时点的人口死亡率水平需要经过标准化,否则很可能得出错误的结论。

The paper makes use of the 5 th and 6 th census data published by the National Bureau of Statistics. The results showed that there was a significant regional difference between standardized death rate（ASCDR） and crude death rate（CDR）. The standardized mortality differed in different regions, and even opposite direction of change occurred. According to the standard population age structure, the mortality changes of 31 regions in China were calculated. The standardized death rate in all regions decreased from 2000 to 2010, which was in full agreement with the statistical results of average life expectancy of population. Comparisons between any two regions might have conflicting conclusions. Therefore, the selection of the age structure of the standard population would be the key step of the direct standardization method. In the process of the specific standardized calculation, it is generally selected as the standard of the average age of a population or population. Because this article compares two population structures, so choosing the average of these two population structures as the standard structure would be more appropriate. Its basic logic is, if the correlation coefficient between the two is higher, in particular, if the correlation coefficient reaches a critical value of 1：00, the crude death rate can be compared directly with the standardized mortality, which shows that the population age structure has a very weak effect on mortality. Conversely, the lower the correlation coefficient between the two is, the less the crude death rate will be used instead of standardized mortality to make regional or sequential comparisons, and the risk of making mistakes will be high, indicating that the population age structure has a significant impact on mortality. This paper uses the direct standardization method to analyze the regional differences and temporal variation characteristics of the mortality levels of 5 th and 6 th census data in China. In fact, there are many similar statistical indicators in demography, such as the birth rate, the fertility rate and other indicators will be affected by the population age structure, even the gender structure. In order to eliminate the impact of population structure on the rate indicators, in the specific analysis, it is usually necessary to use the standardized method to deal with it. The purpose of this is only to eliminate the impact of the age structure of the population in order to compare the level of rate indicators in different regions or different periods. In other words, the standardized rate indicators have only relative meanings, but not absolute, only meaningful when comparisons are made, and a single standardized rate indicator does not indicate any problem, since it is only a result of a hypothetical condition, The age-non-rate index is the real level of reality. In addition, it is also important to note that the standardized rate indicators do not necessarily eliminate the impact of the age structure of the population. Therefore, in the process of standardization, how to choose a suitable standard population age structure is a very important link, the selection of population standards directly affects the results of the calculation and even the judgment of the conclusion. Different criteria can be used to draw different conclusions.