CHAPTER 4: ATTRIBUTES AND MOBILITY OF FEMALE ELITE IN CHINESE
4.1 A descriptive statistical analysis (univariate)
4.1.1 Some features on attributes of female elite in Chinese politics
ATTRIBUTES AND MOBILITY OF FEMALE ELITE IN CHINESE POLITICS
4.1 A descriptive statistical: a univariate analysis
4.1.1 Some features on attributes of female elite in Chinese politics Demographics characteristics
As the literature review showed the proportion of national minorities is higher among female cadres than male cadres. The graph 4.1 reflects what is the total percentage of non-Han among female cadres at the different CC of CPC, and how it has fluctuated over time. The proportion of non-Han cadres among CC female elite is high, a 17%. A remarkable percentage taking into account that there are only a 6,7% of non-Han cadres among the total cadres of the CPC.
Graph 4.1 Nationality of female cadres by percentage and CC
Source: my own elaboration
Previous CC 15th CC 16CC 17CC 18CC total
Nationality by percentage
no han han
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
46
The following figure 5.1 represented shows the percentage of women born in the different administrative divisions of PRC. The figure represents their ancestral home. And, as it is visible in the figure, the majority of women are from Jiangsu, Shandong, Zhejiang, Hebei, Henan. They make up the 56% of the sample. Both Jiangsu, Shandong and Zhejiang are provinces located in the East of China. Meanwhile Hebei and Henan are bordering provinces to the above mentioned, located respectively in the North and South- Central areas of China.
Figure 4.1 Percentage of women born in administrative divisions7 of the PRC
Source: my own elaboration
7Map includes 22 provinces, 5 autonomous regions, 4 municipalities, 2 special administrative regions and 1 claimed province.
‧
A high percentage of female cadres, 29,44%, hold a Master´s degree, following by a 23, 35% that studied a Bachelor degree. The percentage of women with a PhD or with a low level of education (like 3 year-college education) are a minority. The graph 4.3 shows how the number of female cadres with higher educational level has increased over time.
Table 4.2 Educational level of female cadres by total percentage
Source: my own elaboration
Graph 4.3 Educational level of female cadres by percentage and CC
Source: my own elaboration
High school 3 year college BA MA PhD
0% 0% 7% 0% 0%
High school 3-year college BA MA PhD
‧
The graph 4.4 and the graph 4.5 exhibit that the majority of female cadres at CC hold a major in social sciences and humanities, and that the percentage has increased over time to the detriment of major in natural science and engineering.
Graph 4.4 Total percentage of major studies among female cadres at CC
Source: my own elaboration
Graph 4.5 Percentage of female cadres with a major by CC
Source: my own elaboration
Previous CC 15th CC 16th CC 17th CC 18th CC
Major
no major social sciences and humanities natural science and engineering missing values
‧
The graph 4.6 indicates that one half of the female members have received training or studies from the Party school, also, display the oscillation of percentage among the CC.
Graph 4.6 Party school studies by percentage and CC
Source: my own elaboration experiences, neither firm/finance and industrial bureau work experiences. Meanwhile, mass organization, government and party work experiences are the most common career histories among all the CC.
Non Party school studies Party school studies
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
50
Graph 4.7 Career backgrounds of female cadres by percentage and CC
Source: my own elaboration
0 5 10 15 20 25
Previous CC 15th CC 16CC 17CC 18CC
Career backgrounds
no gov work gov work no party work party work
no expert expert no firm firm
no ideology ideology no PLA PLA
no mass organizations mass organizations
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
51
4.1.2 Mobility rate: an overview
As the below box plots shows, from 66 observed cases, without any missing case, the arithmetic mean of mobility rate is 0,013 with standard deviation of 0,002; maximum value is 0,018 and minimum is 0,009.
Graph 4.8 A general overview of mobility rate
Source: my own elaboration
The following graph shows the dispersion of cases. That is, what values are the most repeated or what are the frequencies.
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
Mobility Rate
Mobility Rate
Media Min/Max
‧
Graph 4.9 Dispersion graph of the mobility rate
Source: my own elaboration
Mobility rate by CC of the CCP: a comparison
Graph 4.10 Mobility rate by CC of CPC
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
53
The analysis of mobility rate by CC8 shows that mobility rate among female elite is decreasing overtime. To have a better understanding of the phenomenon, the components of the indicator; mobility rate; seniority and age, have been analyzed.
To observe the arithmetic mean of mobility rate, age and seniority by CC, it is noticeable that mobility rate is in a downward trend because age. The key factor is age.
Seniority has not shown great variations along CC. Mobility rate is decreasing because women are getting access later to CC. They are older than they used to be. Several graphs display this tendency (see graphs 4.11 and 4.12 at Appendix C).
Table 4.1 Relation of mobility rate, age and seniority by CC
CC Mobility rate Age Seniority
Previous 0,01417819 46,5 26,2
1997 0,01355724 48,9 27,0
2002 0,01317718 50,9 25,7
2007 0,01243556 52,3 29,1
2012 0,01213975 53,0 29,7
Total 0,01308676 50,4 27,5
Source: my own elaboration
8 Previous, 1997, 2002, 2007 and 2012 correspond to previous, 15th ,16th, 17th and 18th CC.
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
54
Applying the K-means algorithm: low mobility vs. high mobility rate
The procedure performed consisted of running the K-Means algorithm using an ever increasing number of groups, until they determined the classification of female cadres that offered the best results in terms of increased uniformity within each group.
The outcome9 were two clusters or groups, low mobility and high mobility. Low mobility is composed by 41 female cadres compose and, high mobility group by 25.
Approximately one third of female cadres had a high mobility rate versus two thirds with a low mobility rate.
Graph 4.13 Low mobility rate versus high mobility rate
Source: my own elaboration
Female cadres of the low mobility cluster have a mobility rate close or under the arithmetic mean for mobility rate, that is 0,013. Female cadres of the high mobility cluster have a mobility rate above the arithmetic mean for mobility rate, that is superior to 0,013.
9As regard to the graph, it must be note that each point on the graph represents the most observed values between cases, it does not represent the total number of frequencies.
‧
Shen Yueyue, with a 0,179, has the highest mobility rate among all the female elite members analized. Her biographical data shows that she is from Zhejiang province as the 12, 1 % of her female colleagues, holds a master’s degree in economic management and has also studied in the Party School. As other member of elite, she was part of the Youth League. During her political career, she has held several positions, including Head of the Organization Department of Zhejiang Provincial CPC Committee, vice minister of the Ministry of Personnel and First Deputy Head of the Organization Department of the 17th CPC CC. Since 2013 she is Chairman of the All-China Women’s Federation.
In contrast, Peng Peiyun, with a 0,0092, holds the lowest mobility rate in the study sample. She graduated from Qinghua University and she worked most part of her career at Party leadership positions in diverse Chinese universities. Her most relevant job has been as vice minister of the Ministry of Education and posts related to women’s and disabilities issues.
There have never been any female members in the PSC and currently only two women are members of the Politburo: Sun Chunlan and Liu Yandong. Sun Chunlan has a high mobility rate (0,014), meanwhile Liu Yandong has a low mobility rate (0,012).
Although their mobility rates differ, there are some similarities in their career histories:
both and have graduate studies in economics and social sciences, have been part and held positions in the Youth League, have been head of CPC Central Committee United Front Department and have work experiences in mass organizations such as ACWF and trade union. The remarkable difference is that while Sun Chunlan began her career at provincial level, she was one of the few women to hold an office as Party secretary of a province, Liu Yandong climbed the steps of the Party-state through different positions in the Youth League.
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
56
TESTING THE RELATIONSHIP BETWEEN ATTRIBUTES OF CHINESE FEMALE ELITE AND TYPES OF MOBILITY
5.1 An inferential statistical analysis (bivariate) 5.1.1 Test of Independence
Two variables are related if their attributes vary together, for graduation of this relationship, the statistical of contingency has been used. When purpose of the study is the search for causal relationships, the percentages are estimated only in the direction of the independent variable. This variable is often placed on the columns, and the dependent variable in the ranks; but in this case the high number of attributes of the independent variable discourages its location in the columns, the percentages are horizontal and comparisons are vertical
Albeit illustrative, reading percentage is insufficient. An accurate statistical supplement to graduate the association between the variables and their significance is necessary. In this case, a test of independence has been used to measure the relationship between variables, namely, to test the hypothesis.
1. Hypothesis: There is a relationship between nationality and mobility rate.
Non-Han women have a higher mobility rate than Han women.
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
57
Table 5.1 Relation of mobility and nationality
Low Mobility High Mobility Total
Han 85,366 80,000 83,333
Non Han 14,634 20,000 16,667
Total 100 100 100
Source: my own elaboration
Table 5.2 Test of independence (nationality and mobility)
Chi-square (Observed value) 0,322
Chi-square (Critical value) 3,841
GL 1
value-p 0,570
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 57.04%
2. Hypothesis: There is a relationship between birthplace (province) and mobility rate.
Women from East provinces have a higher mobility rate than women from others provinc
‧
Table 5.3 Relation of mobility and province
Low Mobility High Mobility Total
Heilongjiang 2,439 0,000 1,515
Beijing 4,878 0,000 3,030
Gansu 2,439 0,000 1,515
Fujian 2,439 4,000 3,030
Shaanxi 2,439 0,000 1,515
Guangxi Zhuang Autonomous Region 2,439 0,000 1,515
Taiwan 0,000 8,000 3,030
Inner Mongolia Autonomous Region 0,000 4,000 1,515
Shanxi 0,000 4,000 1,515
‧
Table 5.4 Test of independence (province and mobility)
Chi-square (Observed value) 32,204
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 15,22%
3. Hypothesis: There is a relationship between educational level and mobility rate.
Women with a higher educational level have a higher mobility rate than women with a lower educational level.
Table 5.5 Relation of mobility and level of education
Low Mobility High Mobility Total
‧
Table 5.6 Test of independence (educational level and mobility)
Chi-square (Observed value) 4,406
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 35,39%
4. Hypothesis: There is a relationship between major and mobility rate. Women with a major in natural science and engineering have a higher mobility rate than women with a major in social sciences and humanities.
Table 5.7 Relation of mobility and major
Major
Low Mobility
High
Mobility Total
Social sciences and humanities 41,463 52,000 45,455
Natural science and engineering 43, 902 40,000 42,424
No college education 0,000 4,000 1,515
Missing values 14,634 4,000 10,606
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
61
Table 5.8 Test of independence (major and mobility)
Chi -square (Observed value) 3,731
Chi-square (Critical value) 7,815
GL 3
value-p 0,292
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 29,20%
5. Hypothesis:There is a relationship between Party school experiences and mobility rate. Women with no-Party school experiences have a higher mobility rate than women with a major in social sciences and humanities.
Table 5.9 Relation of mobility and Party school studies
Low Mobility High Mobility Total
Non party school studies 56,098 40,000 50,000
Party school studies 43,902 60,000 50,000
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
62
Table 5.10 Test of independence (Party school studies and mobility)
Chi -square (Observed value) 1,610
Chi -square (Critical value) 3,841
GL 1
value-p 0,205
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 20,45%
6.1 Hypothesis:There is a relationship between party work and mobility rate.
Women with Party work experiences have a higher mobility rate than women without.
Table 5.11 Relation of mobility and party work experience
Low Mobility High Mobility Total
Non Party work 36,585 40,000 37,879
Party work 63,415 60,000 62,121
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
63
Table 5.12 Test of independence (Party work and mobility)
Chi-square (Observed value) 0,322
Chi -square (Critical value) 3,841
GL 1
value-p 0,570
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 57,04%
6.2 Hypothesis: There is a relationship between government work and mobility rate.
Women with government work experiences have a higher mobility rate than women without.
Table 5.13 Relation of mobility and government work experience
Low Mobility High Mobility Total
No government work 34,146 36,000 34,848
Government work 65,854 64,000 65,152
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
64
Table 5.14 Test of independence (Government work and mobility)
Chi -square (Observed value) 0,024
Chi -square (Critical value) 3,841
GL 1
value-p 0,878
alfa 0,05
Source: my own elaboration Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 87,82%
6.3 Hypothesis: There is a relationship between industrial bureau/finance work and mobility rate. Women with industrial bureau/finance work experiences have a higher mobility rate than women without.
Table 5.15 Relation of mobility and firm/finance/ industrial bureau experience
Low Mobility High Mobility Total
Non firm/finance/industrial bureau 78,049 60,000 71,212
Firm/finance/industrial bureau 21,951 40,000 28,788
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
65
Table 5.16 Test of independence (Firm/ finance/ industrial bureau and mobility)
Chi -square (Observed value) 2,468
Chi-square (Critical value) 3,841
GL 1
value-p 0,116
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 11,62%
6.4 Hypothesis: There is a relationship between expert work and mobility rate. Women with expert work experiences have a higher mobility rate than women without.
Table 5.17 Relation of mobility and expert work experiences
Low Mobility High Mobility Total
Non Expert work-0 85,366 88,000 86,364
Expert work-1 14,634 12,000 13,636
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
66
Table 5.18 Test of independence (Expert work and mobility)
Chi -square (Observed value) 0,092
Chi -square (Critical value) 3,841
GL 1
value-p 0,762
Alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 73,23%
6.5 Hypothesis: There is a relationship between ideology and propaganda work and mobility rate. Women with ideology or propaganda work experiences have a higher mobility rate than women without.
Table 5.19 Relation of mobility and ideology work experiences
Low Mobility High Mobility Total
Non ideology work 78,049 84,000 80,303
Ideology work 21,951 16,000 19,697
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
67
Table 5.20 Test of independence (Ideology and mobility)
Chi -square (Observed value) 0,348
Chi -square (Critical value) 3,841
GL 1
value-p 0,555
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 55,54%
6.6 Hypothesis: There is a relationship betweenPLA/police work/law work and mobility rate. Women with PLA/police work/law work experiences have a higher mobility rate than women without.
Table 5.21 Relation of mobility and PLA/police/law work experience
Low Mobility High Mobility Total
Non PLA/police/law 92,683 92,000 92,424
PLA/police/law 7,317 8,000 7,576
Total 100 100 100
Source: my own elaboration
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
68
Table 5.22 Test of independence (PLA/ police/ law and mobility)
Chi-cuadrado (Observed value) 0,010
Chi-cuadrado (Critical value) 3,841
GL 1
valor-p 0,919
Alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 91,90%
6.7 Hypothesis: There is a relationship between mass organizations work and mobility rate. Women with mass organizations work experiences have a higher mobility rate than women without.
Table 5.23 Relation of mobility and mass organization work experience
Low Mobility High Mobility Total
Non Mass organization 46,341 36,000 42,424
Mass organization 53,659 64,000 57,576
Total 100 100 100
Source: my own elaboration
‧
Table 5.24 Test of independence (Mass organization and mobility)
Chi -square (Observed value) 0,680
Chi -square (Critical value) 3,841
GL 1
value-p 0,410
alfa 0,05
Source: my own elaboration
Interpretation of the test:
- H0: rows and columns of the table are independent.
- Ha: there is dependence between rows and columns of the table.
- Since the calculated p-value is greater than the significance level alpha = 0.05, null hypothesis H0 can´t be rejected.
- The risk of rejecting the null hypothesis H0 when it is true is 40,96%
5.1.2 Other statistical methods of data analysis
For the purpose of testing the relationship between attributes of Chinese female elite, as a whole, and types of mobility several statistical methods could be used, such a multivariate regression analysis. However, since the data gathering has been coded as a binary system, the most suitable statistical method is the logistic regression analysis or so-called logit model. It is a multivariate method for dichotomous outcome variables.
The findings of the research using the logit model were similar to those presented above with the test of independence, the absence of a significant relationship among the variables. The research only included the results of the test of independence to avoid reiteration of information, and choose the tests of independence because they provide the same information by variable and secondary information such as thecontingency tables.
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
70
CONCLUSIONS
6.1 Answering the research question and secondary information
The research started from the hypothesis that attributes or characteristics of an elite (in this case a female elite group), as well as the organizational structure (in which these women develop their career) and political connections affect their political mobility. This central hypothesis rose from an exploratory analysis, based on the literature review and on a univariate descriptive analysis. And, the research continued to search for a causal relationship; an explanatory analysis of types of political mobility in terms of attributes of female elite.
The research did not find evidences that support the main hypothesis. Attributes of female elite do not influence or have a causal relationship to mobility type. The several tests of independence conducted show that there is no relationship between the type of mobility (low and high) and the attributes of female elite (the various variables examined) along the period of institutionalization of Chinese politics. The research has only got descriptive answers, as the common factors among them (see Table 6.1 at Appendix C).
Regard to the key factors to their promotion mentioned, the study demonstrates that the
Regard to the key factors to their promotion mentioned, the study demonstrates that the