Evaluation on the Implementation Effect of China’s Basic Pension Insurance for Urban and Rural Residents

—Accurate evaluation of the implementation effect of basic pension insurance for urban and rural residents(BPIURR) is an important means to realize effective supervision of pension insurance. From the perspective of fairness, efficiency and sustainability, the evaluation index system of the implementation effect of BPIURR was constructed in this article. Then by using the method of analytic Hierarchy Process and global entropy method, the implementation effect of BPIURR in 31 provinces from 2015 to 2019 was comprehensively evaluated in this article. The implementation effect of each kind of provinces was compared and analyzed by clustering analysis method. The results showed that the implementation effect of BPIURR in 31 provinces showed an increasing trend on the whole, but the comprehensive scores of all provinces were generally low. There were noticeable regional differences in the implementation effect of BPIURR.


I. INTRODUCTION
General Secretary Xi Jinping pointed out that the management and supervision of the operation of the social security system should be strengthened. Accurate evaluation of basic pension insurance's implementation effect is an important means to realize effective supervision. In the past, domestic and foreign scholars mainly focused on the following aspects: First, put forward evaluation indicators. Previous scholars mostly focused on the dimensions of sustainability [1][2][3], fairness [2,4], the expectation of residents [5], perceived quality [5], fund management [6], guarantee degree [2,7], policy management [8], policy operation [8]. They tended to choose the maximum individual paying capacity [3], service attitude of the handling staff [5], policy mastery level of the handling staff [5], insurance participation rate [7], fund income growth rate [7], policy publicity effectiveness [8], pension receiving procedure [8], government management effectiveness [8], government management level [8], replacement rate [9], collection rate [9], old-age dependency ratio [10], old-age insurance socialized issuance rate [11] as evaluation indicators. Second, use evaluation method. Previous scholars used analytic hierarchy process [1,2,9], Delphi method [3], factor analysis method [5,10,11] and entropy method [8]  individual provinces and counties as evaluation objects. In the past, scholars often carried out regional evaluation (such as Shandong Province [7], Ningxia Province [12] and Anhua County [13]). Therefore, this article constructs a new evaluation index system from fairness, efficiency and sustainability. And this article comprehensively evaluates the implementation effect of basic pension insurance for urban and rural residents (BPIURR) in 31 provinces by using comprehensive weight method and cluster analysis method.

A. Evaluation Index Selection
Existing literatures mostly construct the evaluation index system from the fairness and sustainability, while the evaluation of implementation effect of BPIURR needs to reflect the efficiency.Therefore, from fairness, efficiency and sustainability, this article constructs an evaluation index system, including 1 target layer, 3 criteria layer and 12 indicator layer (Table I).

B. Evaluation Index Weight Determination
After comprehensive weighting by analytic hierarchy Process and global entropy method, the weighting results of each indicator are shown in Table Ⅱ.

A. Comprehensive Weighting Method
The comprehensive weighting method combines analytic hierarchy process and global entropy method, which is used to determine the index weight. It overcomes the defects that analytic hierarchy process is influenced greatly by the personal preferences of experts and global entropy method requires too much data. It improves the scientificity of index weighting and makes the evaluation results more scientific, reasonable and accurate. The difference coefficient method is used to solve the weighting coefficient and the linear weighting model is used to combine the two weights: In Eq. (1) , ' w and '' w represent the weight vectors of each index in the same index system;  represents the weighting coefficients of the analytic hierarchy process;  represents the weighting coefficients of the global entropy method. The weighting coefficient is solved by the difference coefficient method. The calculation formula can be expressed as follows: In Eq. (2-3), n P represents the value of each index weight which is sorted from small to large in the weight ' w .
The comprehensive score of the implementation effect of BPIURR in the i-th province can be calculated as follows: In Eq.

B. Overall Evaluation
According to the comprehensive evaluation method, the comprehensive scores of 31 provinces are calculated (Table  Ⅲ). It can be seen from Table Ⅲ: 1) The implementation effect of BPIURR in 31 provinces shows an overall upward trend. From Table Ⅲ, it can be seen that the overall scores of various provinces have continued to increase. Only a few provinces have slightly declined in individual years such as Liaoning, Heilongjiang, Jilin, Jiangsu, Zhejiang, Hunan and Tianjin, but the fluctuation range is relatively small. One of the most special provinces is Guangdong. Guangdong's comprehensive score increases slightly and then begins to decline from the initial 47.67 to 45.69.
2) The overall score of implementation effect of BPIURR is relatively low. From Table Ⅲ, it can be seen that the highest score is 67.57, which is the only score above the passing line. The other comprehensive scores have not reached the passing level and the lowest score is only 25.06. The national average score in 2015 is 37.12, the national average score in 2016 is 38.45, the national average score in 2017 is 40.43, the national average score in 2018 is 42.89 and the national average score in 2019 is 44.92. The national average score is at a low level, which shows that the effect of the implementation of BPIURR is still room for improvement.
3) The implementation effect of China's BPIURR has obvious regional and provincial difference. It can be seen from Table Ⅲ that in 2015, Beijing has the highest comprehensive score of 52.25 and Heilongjiang has the lowest comprehensive score of 26.11. Beijing's comprehensive score is nearly twice that of Heilongjiang. In 2019, Beijing has the highest score of 67.57 and Jilin has the lowest comprehensive score of 32.61. The difference between these two provinces is 34.96. In the same year, the gap between the highest score and the lowest score is very large.

C. Cluster Analysis
The 31 provinces are divided into four categories by systematic cluster analysis (Fig. 1and Table Ⅳ).
The first category only includes Beijing. Beijing is a megacity in China and belongs to the eastern region. It has a high level of economic development. The participation rate of BPIURR in Beijing is very high. Its comprehensive score ranks first in the country for five consecutive years.
The second category includes Chongqing, Sichuan, Shanghai, Hubei, Jiangsu, Guangdong, Hunan, Fujian, Henan, Shandong, Shaanxi, Anhui and Tianjin, a total of 13 provinces. 6 provinces are located in the eastern region, 4 provinces are located in the central region and 3 provinces are located in the western region. In terms of comprehensive scores, these provinces rank in the forefront of the country. This shows that the implementation effect of BPIURR in such provinces is good.
The third category includes Shanxi, Hainan, Zhejiang, Yunnan, Hebei, Gansu, Jiangxi and Tibet, a total of 8 provinces, of which 3 are located in the eastern region, 2 are located in the central region and 3 are located in the western region. BPIURR's implementation effect in such provinces is average, especially in Zhejiang and Hainan. Although their economic development level is relatively high, but the implementation effect of BPIURR is not as good as the first two types of provinces. This shows that there are still a lot of problems in the implementation of BPIURR.
The fourth category includes Liaoning, Xinjiang, Qinghai, Jilin, Guizhou, Inner Mongolia, Guangxi, Ningxia and Heilongjiang, a total of 9 provinces, of which 1 is located in the eastern region, 2 are located in the central region and 6 are located in the western region. The characteristics of such provinces are their low level of economic development and serious population loss. These provinces have relatively low comprehensive score, which means the implementation of BPIURR in these provinces is poor.  Through the comprehensive evaluation of the implementation effect of BPIURR in 31 provinces, it can be seen that there are still some problems in the implementation of BPIURR. This also shows that the basic pension insurance policies in various provinces should make specific adjustments as actual conditions changes, so that the implementation effect of BPIURR will develop in a better direction.In order to further improve the implementation effect of BPIURR, the following suggestions are put forward: 1) The first and second kind of provinces should improve the sustainability of funds. These provinces should continue to promote the "full process" of pension insurance fund risk prevention and control [14]. In addition, they should improve the "four-in-one" fund management risk prevention and control system, which includes policy, management, information and supervision [14]. 2) The third kind of provinces should implement a universal participation plan, encourage flexible employees to actively participate in pension insurance, implement basic pension adjustment mechanism and raise the basic pension standard [14].
3) The fourth kind of provinces should optimize the inter-regional industrial structure and achieve regional cooperation. For provinces with low economic development, limited local government financial resources and greater pressure on pensions, different financial support will be adopted to ensure the steady development of BPIURR.