Chapter 8: Education, School Network, and Elite Formation



An Assessment of Major Theoretical Perspectives of Elite Recruitment in China

Xiaowei Zang

City University of Hong Kong

What are the main determinants of political elite selection in China? Some scholars highlight the effect of political loyalty in elite mobility (Walder 1986), others suggest a party-sponsored mobility hypothesis (Li and Walder 2001), and still, others define technical training as a key factor in elite formation (Fewsmith 2001; Lee 1991; Li 2001). Using data on top Chinese leaders, I argue for an emphasis on functional differentiation and its effect on elite recruitment in China. By functional differentiation I mean the division of labor in governance between the Chinese government and Chinese Communist Party (CCP). Specifically, the CCP provides leadership by taking on policy-making, matters of political principle, and personnel management, whereas the government implements party policy regarding economic development and social issues (Zhou 1995).

The first section of this paper reviews the three perspectives on elite recruitment in state socialism and puts forward three hypotheses. The second section relies on functional differentiation to develop three hypotheses regarding the division of labor in governance and elite recruitment. The third section uses the data on China’s political elite to test the six hypotheses. Results of the analysis show the link between functional differentiation and the selective emphasis of technical training and political loyalty in leadership recruitment in the reform era. The final section concludes by summarizing the major findings from this paper and assessing the relevancy of the above three perspectives for research on elite selection in China.

Dual Careers, Party Patronage, and Technical Training

Studies of stratification and mobility in state socialism have a long history. Some scholars have studied the impact of political loyalty, measured largely in terms of communist party membership, on career opportunities. They have argued that party members enjoy an advantage in career advancement and the allocation of resources (Connor 1979; Feldmesser 1960; Walder 1985, 1986). Indeed, mass survey data have provided evidence that party membership contributes to upward mobility (Bian, Xu, and Logan 2001; Dickson and Rublee 2000; Massey, Hodson, and Sekulic 1992). Yet analysis of these survey data has also found that party membership is not the only important selection criterion. Educational effects “in mobility models are just as great there as in market economies (Walder 1995, p. 311; see also Bian 1994; Inkeles and Bauer 1959; Dickson and Rublee 2000; Meyer, Tuma, and Zagorski 1979). Many scholars have asked how educational credentials and political loyalty are combined to determine political mobility (Walder 1995; Walder, Li, and Treiman. 2000).

Konrad and Szelenyi’s analysis of the Hungarian socialist regime (1979) offers an answer to this question. Konrad and Szelenyi argue for a dual elite hypothesis, namely, the elites in state socialism are divided between politically loyal bureaucrats and better educated professionals. Following this insight, Walder and his associates propose a dual career path model, which measures the effects of party membership and education on mobility by type of career. Namely, “China has a hybrid mobility regime in which the loyalty principles of a political machine are combined with, and segregated from, the meritocratic standards of modern professions” (Walder, Li, and Treiman 2000, pp. 191, 206-7; also Walder 1995).

Using mass survey data from China, Walder and his associates find that party membership is irrelevant for recruitment into elite professions yet vital for entry into administrative elite. A college degree leads to a professional career. However, high educational standards are not enforced in the cadre path. Poor educational attainment among cadres is reinforced by the tendency that professionals rarely become administrators (Walder, Li, and Treiman 2000, p. 191; also Walder 1995). Educated party members were not placed in positions of power. Technocracy was prevented from emerging in China since the cadre elite was not comprised by higher professionals and the college educated (Walder 1995, p. 323; also Walder, Li, and Treiman 2000). Hence:

HYPOTHESIS 1: The college-educated will not be favorably recruited into the political elite.

The dual career path model is subsequently modified by a party-sponsored mobility hypothesis, which focuses on the timing of joining the party, rather than party membership, in an individual’s life. This focus pays scant attention to the role of education in elite selection—Li and Walder argue (2001, pp. 403-4) that the party has recruited preferentially from among young party loyalists, whereas the educated professionals are not sponsored for advancement into the administrative elite.

Li and Walder (2001) summarize that the party-sponsored mobility hypothesis draws the insights from sponsored mobility perspective by Turner (1960) and tournament mobility model by Rosenbaum (1976, 1979, 1984) respectively. It proposes that elites are chosen from among party members, but only those who join the party while young are eventually chosen. Admission into a ruling communist party does not in itself elevate a member to elite status because the party is a mass organization. But early party membership resembles a victory in an early tournament, which “qualifies one for competition in the next—events early in the career can alter the outcomes of subsequent competition for career advancement” (Li and Walder 2001, pp. 1,371, 1,376, 1,405). In contrast, party memberships obtained in mid- or late-career do not provide such career advantages. They are largely a symbolic reward or “may have proffered status advantages in other areas of life” (pp. 1,380-1, 1,404-5).

Li and Walder (2001) derive the party-sponsored mobility hypothesis from mass survey data. The cadre elite they study are mostly low-ranking officials (p. 1,394). If party membership obtained in mid- and late-career can hardly qualify one for a cadre post in a local organization (pp. 1,404-5), one can expect that virtually all top Chinese leaders are early joiners. The requirement for sponsorship in elite selection must be much higher at the national than at the grass-roots level. Also, joining the party in mid- or late-career does not signal a higher degree of commitment and does not allow an extended process of observation, cultivation, and training necessary for career advancement in the cadre path (p. 1,380). Hence:

HYPOTHESIS 2: Only those who join the CCP while young will be recruited into the political elite.

The dual career path model is also modified by a technocracy thesis, which insists that it is technical training in engineering/or management that increases the odds of career success in politics. Inspired by early technocracy scholars (Bell 1973; Harbison and Myers 1959; Meynaud 1969), several scholars argue that there has been a drastic shift from the past focus on political qualifications toward an emphasis on technical training in leadership selection in mature state socialist countries where economic development is an important dimension of governance (Bailes 1978; Bialer 1980; Lee 1991; Putnam 1976). Their argument encourages some China experts to consider leadership selection in the reform era as the formation of a technocracy (Lee 1991, 1992; Li 2001; Li and White 1988, 1990; Millis 1983; Wang 1985).

According to the technocracy thesis, technocrats come to power mainly because they possess technical expertise that China needs for economic development. Hence, only those who studied management science, engineering, and applied sciences are favored targets of leadership selection (Li and White 1988, 1990, 1998). Joseph Fewsmith (2001, p. 67) argues that although “the official slogan for the recruitment and promotion of younger officials was to promote ‘more revolutionary, better educated, more professionally competent, and younger,’ in practice the term ‘more revolutionary’ was difficult to define and highly contentious. The result has been to rely on more objective criteria, particularly education and age, for selecting the Party elite and state cadres. Because the social sciences were highly suspect in the early part of the PRC, and even remain so today, there has been a very strong tendency to recruit those who have had technical training, particularly engineering, thus giving China’s political elite a very technocratic cast.” Hence:

HYPOTHESIS 3: Only those with technical training will be recruited into the political elite.

Functional Differentiation and Elite Recruitment

Central to the dual career path hypothesis is the focus on the selective use of education and political loyalty in professions and administrative posts respectively (Walder 1995; Walder, Li, and Treiman 2000). I extend this insight to leadership selection, envisioning a segmented bureaucratic labor market in which political loyalty and technical training are awarded differently in the CCP and government. The dual career path model rules out the probabilities that professionals and intellectuals are a key source of elite recruitment since it is premised on a professional career and a political career. Higher education is related to the professional career whereas political loyalty leads to the political career. In comparison, the starting point of my research is the division of labor in governance among political elites. Elite recruitment is based on both political loyalty and educational credentials. The CCP and the government recruit cadres with different combinations of technical training and political loyalty because of their different institutional tasks. The differences in returns to technical training and political loyalty should remain with control for other individual characteristics.

I start with the observation that the CCP is a Leninist party by nature. A Leninist party maintains its integrity and coherence by emphasizing hierarchy and concentration of power in the hands of its top leaders. Such emphases turn the Leninist party into a powerful organizational weapon in its struggle for political power. Yet uniformity and hierarchy create problems in governance for the Leninist party when it shifts its attention from revolutionary campaigns to economic development in the post-revolutionary era.

This is because economic development leads to scale and complexity, which in turn entail specialized knowledge and the division of labor and authority among institutions of a political system. The institutions must develop separate functions in order to facilitate the successful resolution of problems of industrialization. There should be among them differences in the skills required to perform different tasks as functional differentiation permits more efficient expenditures of collective resources and more effective attainment of collective goals. For example, the ruling Leninist party can focus on policy-making and supervision of policy implementation, whereas government officials can be given power to execute issues of administration and economic management (Shirk 1992; Zheng 1997). The interests of the political system can be best served through the division of labor among its components.

The division of labor in governance by no means suggests check and balance or the separation of the state from the ruling communist party. The party rules. Functional differentiation is adopted because it enhances rather than undermine the party’s dictatorship. The party ensures its grasp of power by penetrating and controlling various administrative organizations through policy guidance, direct supervision, personnel management, and other creative measures.

Hence, many Leninist parties delegate administrative power to the state when they seek economic development to enhance regime legitimacy (Bailes 1978; Bialer 1980; Hough 1977; Ludz 1972; Mawdsley and White 2000; Zheng 1997). In Mao’s China, the division of labor between the government and CCP was first experimented in the early 1950s and 1960s. Under this arrangement, CCP leaders such as Mao Zedong, Chen Boda, and Lin Biao, who knew little about economic management, focused on policy-making and issues of political principle; whereas leading cadres such as Chen Yun and Bo Yibo, who commanded expertise in finance, industrial planning, and economic coordination because of on-job training, led institutions with specific functions in various socioeconomic agendas such as economic growth, foreign trade, and infrastructural development (Diao 1970; Dittmer 1990, pp. 428-30; Huang 2000, pp. 12-8; 211-59; Zheng 1997).

It is necessary to stress that delegation is not always a favored option in state socialism. Although the official separation of the party and government has existed from 1949, the CCP has been struggling for the actual implementation. In some periods (e.g., 1949-1957, 1962-1965, and 1978-present), separation of the CCP and government (dangzheng fengong) was greatly encouraged and required, while in some other periods (e.g., 1958-1961, 1966-1977), domination or unification (dangzheng heyi) was ruthlessly enforced. It appears that the division of labor in governance does not appeal to the CCP when economic development is not its priority. For example, during China’s Cultural Revolution of 1966-1976, power struggle between Mao and his opponents overrode the concerns for economic growth and interrupted the division of labor between the CCP and government. As a direct consequence, China’s economy was on the brink of collapse at the end of the Cultural Revolution, which taught Chinese leaders a valuable lesson. Economic development has become a top priority of the CCP. Function differentiation has since been improved and implemented with rigor in the reform era to promote economic growth and hence the legitimacy of the CCP leadership (Huang 2000; Zheng 1997).

The movement towards functional differentiation has led to a greater degree of functional differentiation in governance between the CCP hierarchy and government system in the reform era. At the sub-national level, for example, there are a provincial government and provincial CCP committee in each of the provinces in China with the former running administrative affairs and the latter providing guidance and policies. The provincial CCP leaders are number one bosses in their provinces. They however are less occupied than provincial government officials with the management of the economy and other administrative affairs. For example, a deputy provincial party secretary is less likely than a deputy provincial governor to get himself/herself involved in preparing annual budgets or entertaining potential foreign investors. The division of labor is also evident at the highest level of power: the central government takes care of national economic activities and administrative affairs, whereas the CCP central organizations focus on policy guidance, supervision, personnel management, propaganda, and the like (Zhou 1995, p. 443; also Lieberthal and Lampton 1992; Lieberthal and Oksenberg 1988; Shirk 1992; Zheng 1997).

Functional differentiation has an important effect on leadership selection in the reform era. It calls for the need to recruit officials with expertise since they are essential for effective administration and economic growth. It however does not entail the removal of political loyalty as an important criterion in leadership selection. The Chinese leadership has initiated and conditioned leadership transition and market reforms to rebuild its legitimacy and consolidate its dictatorship over society. The political system and the institutional framework (i.e., nomenklatura)[i] for leadership selection have remained unchanged, the practice that stresses political loyalty in leadership selection has not been and will not be abandoned. Politically reliable candidates will be selected into leadership positions to maintain the integrity and coherence of the PRC regime. Hence:

HYPOTHESIS 4: Elite recruitment relies on both political loyalty and technical training.

Functional differentiation also suggests that political loyalty and technical training need not to be weighed equally in elite recruitment. The CCP and government pursue different institutional tasks and hence should recruit different kinds of human resources to allow institutional and technical considerations to converge. Institutions differ in technical content and political implications and thus in the flow of personnel along different career paths (Grindle 1980).

Two factors have led the CCP and government to develop different rules, norms, and expectations in recruitment. The first factor is the changes in institutional missions in the reform era. The more a certain institutional arena is transformed, the greater the change in recruitment criteria in that institution. Both the CCP and government took politics in command during the Cultural Revolution and recruited leaders on the basis of their political loyalty (Lee 1991). The government has become increasingly prominent in economic management in the reform era. The CCP has also considered economic development a priority of its activity. Yet it has continued its major task of maintaining political orthodoxy and party dictatorship and hence has not been subject to the same pressure for recruiting technical experts as the government system has been. Hence the CCO is less interested in expertise than the government.

Furthermore, the CCP has become an increasingly policy-making and supervisory statutory body in the reform era. These tasks can be accomplished successfully without technical training. The execution of party tasks relies mainly on political loyalty. Walder, Li, and Treiman (2000, p. 205) point out that the CCP wants to control decision-making, which would dictate the imposition of political loyalty as a minimal criterion for holding an elite cadre position. Hence, the CCP is more likely than the government to evaluate candidates for leadership positions in terms of their political loyalty.

A second and related factor is the different selection criteria in the CCP and government respectively. The requirement for political loyalty is lower in the government system than in the CCP hierarchy because candidates for CCP leadership positions are evaluated and compete with one another mainly on the basis of political loyalty, which will enhance the requirement for political loyalty for leadership positions in the CCP hierarchy—the value of a commodity goes down when supply exceeds demand. In comparison, recruitment in the government system is based on a more balanced assessment of both political loyalty and expertise. Sometimes a lower level of political loyalty can be compensated by a higher amount of expertise. Hence:

HYPOTHESIS 5: Recruitment into the CCP system relies more on political loyalty than that into the government system.

HYPOTHESIS 6: The impact of technical training on elite recruitment is greater in the government system than in the CCP hierarchy.

Hypothesis 5 differs from Hypothesis 2 in its emphasis on the selective use of political loyalty in elite recruitment. It suggests flexibility in the government system and rigidity in the CCP hierarchy with regard to political loyalty. Following Li and Walder (2001), I use the timing of joining the party to measure political loyalty. I expect early party membership to play a larger role in leadership selection in the CCP hierarchy than in the government system. Holding everything constant, those who join the party while young should have higher odds of entry into the CCP hierarchy than those who do not join the party or do so late in life. Hypothesis 6 differs from Hypothesis 3 in its emphasis on the selective use of technical training in leadership selection. Holding everything constant, those who have technical training should have higher odds of entry into the government system than those who do not.

In sum, the change in institutional goals has been more pronounced with regard to the government domain than in the CCP hierarchy in the reform era. Institutional inertia in the CCP realm and institutional dynamism in the government system should be translated into a relatively higher impact of technical training in selecting government officials and a relatively higher impact of political loyalty in recruiting CCP cadres.

Data and Variables

To test the above six hypotheses, I collect two sets of cross-sectional data on top Chinese leaders in 1988 and 1994 respectively. These officials hold deputy provincial governorship or higher positions in the Chinese political hierarchy. In other words, I focus on the apex or nucleus (leaders of the central government and the CCP central organizations) and the immediately adjacent echelons (provincial leaders) of the Chinese political system, who are appointed and managed centrally by the CCP (Burns 1987, p. 54; also Burns 1989, 1994). These political leaders make and implement decisions in China. I do not include members of the People’s Congress and the Political Consultative Conferences since they play an advisory role in Chinese politics.

The main data source for this study is Who’s Who in China—Current Leaders, compiled for 1988 and 1994 respectively (The Editorial Board of Who’s Who in China 1989 and 1994). Such printed biographical sources have been frequently used to identify leading figures in both capitalist and socialist societies (e.g., Farmer 1992; Mawdsley and White 2000; Useem and Karabel 1986). Other data sources for this study include various issues of China: Facts & Figures (Mulvenon, various years) and various issues of China Directory (Radiopress, various years). To further establish my confidence in the reliability and accuracy of the data set, I used other data sources to cross-examine the information presented in Who’s Who.[ii] No significant discrepancies were detected. The data yield two separate cross-sections, representing an elite group in 1988 and another in 1994. My analysis focuses on the 1988 elite. The 1994 elite group is used mainly as a replication case to further support arguments drawn from the data analysis of the 1988 elite.

Elite status inheres in the office, not in the occupant (Farmer 1992; Keller, 1963; Scalopino 1972). Thus, in the coding process, I treated each case as a person-position, so that, for example, a provincial governor who is also a deputy provincial party secretary is represented by two separate observations for each post, namely, one for governorship and the other for deputy secretaryship. This coding procedure is based on an assumption that it is the personal attributes of a person, rather than the person himself/herself, that are of central importance in estimating factors of status attainment.

To date, 740 cases are identified for the 1988 elite and 848 cases for the 1994 elite (Table 1). Many of the leaders included in the data set such as Mr. Zen Qinghong, member of the CCP’s Politburo and Minister of the Planning Commission of the State Council and Mr. Li Changchun, member of the CCP’s Politburo, are still leading figures in the PRC. Included in the biographies are names, dates of birth, gender, ethnicity, party membership, date of joining the CCP, university ever attended, academic discipline majored, and the like. To the best of my knowledge, this is the most extensive and up-to-date biographical data set of the political elite in China.

Table 1 above here

The variables I use in the following analysis include Age, Ethnicity, Gender, the timing of joining the party, College education, and Technical training. Similar variables have been used in research on elite formation in the West and Japan (Borthwick et al. 1991; Holland 1987; Norris and Lovenduski 1989, 1995; Useem and Karabel 1986), the occupied West Bank (Mazawi and Yogev 1999), and China (Li 2001).

Age and the timing of joining the party are interval variables. Ethnicity is a dummy variable with Han Chinese coded as 0 and cadres of ethnic minority status as 1. Gender is also a dummy variable with male cadres coded as 0 and female cadres as 1. Finally, I create two dummy variable—College education and Technical training—to measure the effect of educational credentials on elite formation. Age, Ethnicity, Gender are largely self-explanatory. The timing of joining the party, College education, and Technical training however require some explanations.

The timing of joining the party. Political loyalty is a key determinant of career mobility in state socialism. Li and Walder (2001) use the timing of joining the party to measure political loyalty because “Joining the party early signals a higher degree of commitment and also permits a more extended process of observation, cultivation, and training…If those who join the party early in life are more likely to be promoted into leading posts, they may consider party membership as more central to their identity and may as a consequence of their success feel stronger loyalties to the party organization” (p. 1,380).

I use the timing of joining the CCP to test the party-sponsored mobility model (i.e., Hypothesis 2) and Hypotheses 4 & 5. Li and Walder (2001, pp. 1,383-4) measure the timing of joining the party in two dimensions: age and labor force experience. The first dimension of party entry is the biological age when a person joins the party. For the labor force experience dimension of party entry, the timing of joining the party is the number of years from the time the individual entered the work force to the time he or she joined the party. “For those who worked before age 18, year began first job is set to birth year plus 18. The timing of joining the party is set to 0 for those who joined the party before the first job or before age 18. For those who have never joined or joined the party after becoming a cadre, the timing of joining is set to 0” (p. 1,394).

My data set does not have complete information on the timing of workforce entry. I thus measure the timing of joining the party with reference to age: (timing) = (year joined party) - (birth year). Hence, a case of late party entry will have a higher value than a case of early party entry. For those who have never joined the party, the timing is set to 99. The results remain largely unchanged when the cases with a value of 99 are excluded from analysis.

College education. I use this variable to test the dual career path model (i.e., Hypothesis 1) and Hypothesis 4. Chinese leaders with college education are coded as 1 and others as 0.

Technical Training. I use this variable to test the technocracy thesis (i.e., Hypothesis 3) and Hypothesis 6. Cadres with technical training are coded as 1 and others as 0. I define technical training pretty much the same way as technocracy scholars do (Li and White 1988; 1990). Technical training includes two key components: management science majors and engineering majors. The former includes economic planning, finance, accountancy and the like. The later is broadly defined as applied sciences, which include textile engineering, civil engineering, mechanical engineering, chemical engineering, transportation, and the like. I also present information on the distribution of BA degrees among Chinese leaders. BA degrees include those in education, history, Chinese, politics, and the like.

Findings

Data analysis suggests that government officials and CCP cadres are very much alike demographically. For example, the difference in the mean age between these two groups is quite similar (56.3 vs. 57.5 for 1988 and 56.6 vs. 58.7 for 1994). Over 94 percent of the 1988 government officials and 96 percent of the CCP cadres were men. The corresponding figures for 1994 were 93.9 and 94.5 respectively. There were no significant differences in terms of ethnicity between these two groups in 1988 and 1994. However, government officials and CCP cadres differ from each other in terms of technical training and the timing of party entry. Table 2 shows that only 27.5 percent of the 1988 elite and 42.7 percent of the 1994 elite have technical training. This finding does not support Hypothesis 3 (i.e., the technocracy hypothesis) that only those with technical training will be favorably recruited into the political elite.

Table 2 also shows that the overall educational level of the Chinese leaders was greatly improved between 1988 and 1994. Yet cadres with degrees in technical fields are more likely than those without such credentials to enter the government system. In 1988, 32.8 percent of the government officials had college degrees in technical fields, as compared with 16.4 percent of the CCP cadres. The corresponding figures for 1994 were 47.7 percent and 33.3 percent respectively. The differences for both 1988 and 1994 are statistically significant at .001 level (Pearson Chi-Square = 38.74 and 20.08 respectively). These findings tentatively confirm Hypothesis 6 that the government system stresses technical education more than the CCP in leadership selection.

Table 2 shows that nearly 62 percent of the 1988 elite and 83 percent of the 1994 elite are college-educated, partially supporting Hypothesis 4. College education has become increasingly important in elite selection. This finding does not support Hypothesis 1 that the college-educated will not be favorably recruited into the political elite. It is also inconsistent with the party-sponsored mobility hypothesis, which suggests that high professionals and the college-educated do not enter the cadre elite (Li and Walder 2001, p. 1,404).

Table 2 about here

Findings from Table 2 give support to Hypotheses 2 & 5. Li and Walder (2001, pp. 1,383-4) show that the hazard rate of joining the party increases rapidly in the late teens and reaches “its peak at age 22 or 23” (or in the first few years of work). Hence, I expect that virtually all top Chinese leaders joined the party before age 23 and decide to use age 25 as the watershed to give favorable room for Hypothesis 2 to maneuver.

Table 2 shows that 85.6 percent of the 1988 party cadre and 73.7 percent of the 1988 government officials joined the CCP before age 25. The corresponding figures for the 1994 elite are 76 percent and 58.7 percent. One can argue that overall Hypothesis 2 is not falsified, especially with reference to party cadres. Yet at the same time it is important to note that more than a quarter of the 1988 government officials did not join the party till they were 25 or older. The figure increased to 41.3 percent in 1994. Equally interesting, 4.3 percent of the 1988 government officials did not join the CCP until they were 40 or older and 4.5 percent of them did not hold party membership at all. The corresponding figures for the 1994 government elite are 8.5 percent and 5.2 percent. These percentages cannot be simply dismissed as abnormalities because the subject of analysis is top Chinese leaders. From the perspective of the party-sponsored mobility hypothesis, they should all be early joiners since party membership obtained in mid- or late-career can hardly qualify one for a cadre post in a local organization, let alone a leadership post at the national level.

I also find that the difference in the timing of party entry between government officials and CCP cadres is statistically significant at .001 level in both 1988 and 1994 (F-statistics = 22.8 and 24.0 respectively). These findings indicate that the timing of party entry is screened more vigorously in the CCP hierarchy than in the government system, tentatively confirming Hypothesis 5 that the role of political loyalty in elite selection is emphasized more in the CCP system than that in the government system.

Table 3 shows the number of the political leaders with (1) leadership positions in the government only, (2) leadership positions in both the CCP and the government, and (3) leadership positions in the CCP only.  It can be seen that in both 1988 and 1994, the percentage of early party entry is highest for leaders with positions in the CCP only, then for leaders with positions in both the CCP and the government, and then for leaders with positions in the government only. The percentages of college education for these three groups of political leaders are in the reverse order in both 1988 and 1994. Table 3 thus provides additional evidence for the functional differentiation hypothesis.

Table 3 about here

Next, I conduct two Logistic Regression analyses to further estimate the relative role of technical training and the timing of party entry in elite recruitment in 1988 and 1994. The Chi-Square statistics show a good degree of fitness for both models in Table 4. Model 1 of Table 4 compares the probability of becoming a government official versus that of becoming a CCP cadre in 1988. There is no difference in terms of college education between the CCP cadres and government officials. However, technical training facilitates a candidate to enter the government system significantly, supporting Hypothesis 6. A person with a degree in a technical field is 166 percent more likely than a person without such a credential to become a government official.

Model 1 of Table 4 confirms Hypothesis 5: timing of early party entry has a positive impact on recruitment into the government system. Since party memberships obtained late in life lead to a high value for the timing of party entry, the positive coefficient indicates a strong contrast effect regarding the worth of early party entry in recruitment practices between the government system and CCP hierarchy. Assuming that a person joins the CCP at age 20, each year of party membership adds nearly 6 percent advantage of entering the government system. Conversely, a person who joins the party at age 20 is nearly 60 percent more likely to become a CCP cadre than a person who joins at age 30. The odds become 112 percent if the comparison person joins the party at age 40.

In addition, Model 1 of Table 4 shows that the government officials are not older than the CCP cadres since the coefficient for Age is not statistically significant. Thus, the CCP cadres are less educated than government officials not because they are illiterate or semi-illiterate old revolutionary guards.

Table 4 about here

I then replicate the empirical test with the 1994 data on China’s leaders. I expect that findings from this elite group will not deviate from those from the 1988 elite since both groups were formed in the reform era when functional differentiation in governance was emphasized. Similar results will lend firm support to my hypotheses on leadership selection in the reform era. Model 2 of Table 4 examines the relationship between technical training, timing of party entry, and elite recruitment in 1994. A person with technical degree is nearly 50 percent more likely than a person without such a credential to become a government official. Early party entry brings about 5 percent advantage of becoming a CCP cadre each year. So a person who joins the party at age 20 is 50 percent more likely to become a CCP cadre than a person who joins at age 30. The replication confirms my hypotheses again, indicating a persistent pattern of elite recruitment on the basis of functional differentiation between the CCP and government.

Discussions and Summary

In this research, I incorporate the insights from the dual career path model, the party-sponsored mobility hypothesis, and the technocracy thesis into my argument about the importance of functional differentiation, showing different emphases on technical training and the timing of party entry in leadership selection in the CCP hierarchy and government system respectively. Findings from my research are not fully consistent with the expectations from these three perspectives. Firstly, the dual career path model predicts that the party does not enforce higher educational credentials in the cadre path. My research shows that over 61.8 percent of the 1988 elite and 83 percent of the 1994 elite are college-educated. Secondly, the party-sponsored mobility hypothesis expects that early party entry increases the odds of advancement into elite posts and late party entry will not have such effects. Indeed, my analysis shows that the vast majority of the Chinese leaders joined the CCP while young. Yet I also find that that some government officials did not join the CCP or did so when they were 40 or older. Thirdly, the technocracy thesis suggests that the CCP uses technical training as the exclusive human capital criterion in elite selection. I report that less than 30 percent of the 1988 elite and 42.7 percent of the 1994 elite have technical training. Why do my findings differ from the expectations of these three perspectives?

There are good reasons for these differences. The technocracy thesis is based mainly on a few case studies and has never been subject to any quantitative test. More importantly, the technocratic version of elite recruitment in the reform era projects features of an idealized image of Western development onto its hypothesized socialist counterparts. Elite formation in the reform era is understood simply according to research works in the West. Such understandings mechanically pinpoint areas of convergence in leadership selection between capitalism and state socialism, ignoring the fact that the political institutions of a socialist society differ fundamentally from those of its capitalist counterpart. The technocracy thesis argues that technocrats dominate the current Chinese leadership because of the need for technical expertise in economic development in the reform era. Yet such a need does not change the institutional framework with which leadership selection occurs. Nor does it reduce the importance of political loyalty in elite recruitment. Institutions rather than the need for technical expertise must be at the center stage in the study of elite formation. The technocracy thesis fails the empirical test because it overstates the importance of technical training in leadership selection in the reform era.

The findings from my research support the party-sponsored mobility hypothesis, especially with reference to leadership selection in the CCP hierarchy. The party-sponsored mobility hypothesis however cannot fully explain the role of party sponsorship in the government system. Nor does it account for the effect of technical training in leadership selection in the reform era due to its focus on the effects of political loyalty on elite recruitment. The findings from my research suggest the need to place the party-sponsored mobility hypothesis within the framework of functional differentiation for a better modeling of leadership selection. This can be done because both the party-sponsored mobility hypothesis and my research are fundamentally based on the idea that there are some groups in society that decide criteria for elite selection. The difference between these two approaches is the different ways this idea is empirically presented. Li and Walder (2001) seek to demonstrate party-sponsored mobility by investigating the persistent effect of early party entry on elite selection between 1949 and 1996, whereas I prefer to understand party patronage in terms of the CCP’s capacity to set up and implement its favored personnel policies, thereby bring about flexibility in estimating returns to political loyalty and other important attributes such as technical training.

Flexibility is important because Zhou (2001; also Li and Walder 2001; Zhou, Tuma, and Moen 1996) points out that the CCP sponsors people of different attributes in different periods of time, depending on its institutional missions. For example, political loyalty was the only selection criterion during the Cultural Revolution of 1966-1976 and university education has been used to screen candidates for leadership positions after 1978. Selection criteria have changed because party’s main tasks have changed (Lee 1991; Zhou 2001). The party-sponsored mobility hypothesis cannot deal with the changes satisfactorily due to its focus on the effect of political loyalty in elite selection. In comparison, flexibility allows me to incorporate the insights from the party-sponsored mobility hypothesis in modeling the major determinants of leadership selection in different periods of time. For example, I can propose that the effects of party patronage would increase and that of education would decrease in both the government and CCP when politics took command, and vice versa when functional differentiation was favored. These propositions point to a fact that the alteration in selection criteria has not changed the selector (i.e., the party) and its capacity to implement its recruitment policies.

The dual career path model and my research have generated different empirical findings regarding the selective role of political loyalty and educational credentials in cadre selection. Yet I do not read them as contradictory toward each other. In fact, they complement each other, piecing together a reversed isosceles triangle shape of recruitment patterns in the reform era. That is, the degree of concentration of the college-educated is highest in the national political institutions and lowest in grass-roots political organizations.

The reversed isosceles triangle shape of recruitment patterns can be understood by referring to the two factors Walder and his associates offer to explain why higher educational standards are required in the professional path but not in the cadre path: (1) the shortage of college-educated personnel—higher educational standards “simply could not be imposed until the supply of mature college students with party memberships increased” and as a consequence the party sends the college-educated to the professional path and (2) political avoidance—the college educated may choose a professional career over a cadre path because they “may have learned to avoid the Party and leadership positions that imposed political obligations and exposed them to political suspicion” (Walder 1995; Walder, Li, and Treiman 2000, pp. 204-5).

These two explanations are valid when applied to recruitment in basic-level political organizations. I reinterpret these two factors with reference to elite selection at the national level. First, the college educated may choose a professional career over a cadre path because they may have learned to avoid political obligations and political suspicion. This is particularly the case when differences in reward to professional and administrative elites at the grass-roots level are not significant and substantial (Walder 1995). The picture is totally different at the national level where political obligations and political suspicion are replaced with privileges and power. The college-educated would respond favorably if chosen to take a leadership position in a national political organization.

The reversed isosceles triangle shape of recruitment patterns also makes good sense according to the shortage factor and the demand-supply principle of resource allocation. As noted above, the party seeks to recruit candidates with good education and political loyalty for leadership positions in the reform era. Yet the shortage of college-educated personnel with political loyalty forces the party to allocate politically reliable college graduates in the national and sub-national political institutions. The party is not able to fill the cadre crop at the grass-roots level with the college-educated because of the shortage of qualified candidates. Hence, the shortage hypothesis explains both the dual career path model and my approach toward elite formation. In other words, the dual career path model describes cadre recruitment at the grass-roots level, whereas my approach models leadership selection at the national level. Indeed, the dual career path model is derived from mass survey data, whereas my research relies on information on top Chinese leaders.

My approach is based on functional differentiation in the reform era. As noted above, in some periods, separation of the CCP and government was greatly encouraged and required, while in some other periods, domination or unification was ruthlessly enforced. It is necessary to repeat that career patterns in China “have been altered due to shifting state policies across historical periods” (Li and Walder 2001, p. 1,383; also Walder, Li, and Treiman 2000; Zhou, Tuma, and Moen 1996). Scant evidence suggests that when unification occurred poorly-educated cadres enjoyed career advantage and took government posts simply because of their political loyalty (Lee 1991; Wlader 1986). Hence, a replication of this study on data on top leaders in the Mao era (in particular in the periods when combination was highlighted) will form a useful contrast to this research. A replication on a representative sample that includes cadres of all rankings is also recommended, which can allow us to move from descriptive statistics to interaction effects in testing Hypotheses 1, 2, and 3.

In sum, I argue that the technocracy thesis overstates the role of technical training in elite recruitment and hence is the least relevant perspective for research on leadership selection in the reform era. The dual career path model complements my research by modeling patterns of elite selection at the grass-roots level. Finally, the party-sponsored mobility hypothesis predicts elite formation the best, especially with regard to that in the CCP hierarchy. Yet it does not anticipate the lesser screening of government officials for political loyalty. Nor does it address the difference in technical training between government officials and party cadres. A better explanation of leadership selection can emerge when the party-sponsored mobility hypothesis is placed within the framework of functional differentiation in governance. This way the selective use of technical training and political loyalty in the CCP hierarchy and government system can be effectively accounted for.

Table 1: Leadership Positions in China, 1988 and 1994

| | | |

|Position |1988 |1994 |

| | | |

| | | |

| |N |% |N |% |

| | | | | |

|Government posts | | | | |

|Councilor and above |18 |1.8 |18 |2.1 |

|Minister |56 |7.6 |57 |6.7 |

|Vice minister |198 |26.8 |213 |25.1 |

|Central government bureau head |31 |4.2 |48 |5.7 |

|Provincial governor |30 |4.1 |30 |3.5 |

|Provincial deputy governor |157 |21.2 |189 |22.3 |

| | | | | |

|CCP posts | | | | |

|CCP Politburo members and minister |31 |4.2 |56 |6.6 |

|CCP vice-minister |42 |6.2 |68 |8.0 |

|CCP bureau head |14 |1.9 |12 |1.4 |

|Provincial secretary |30 |4.1 |30 |3.5 |

|Provincial deputy secretary |100 |13.5 |97 |11.4 |

|Provincial discipline secretary |29* |3.9 |30 |3.5 |

| | | | | |

|Total |740 |100.0 |848 |100.0 |

| | | | | |

*Data on Yi Jun, discipline secretary of the CCP Yunnan provincial committee, are not available.

Table 2: Educational Credentials and Timing of Party Entry, 1988 and 1994 (%)

| | | |

|Variable |1988 |1994 |

| | | |

| | | | | | |

| |Government officials |Party cadres | |Government officials |Party cadres |

| | | | | | |

| | | | | | |

|Credentials | | | | | |

|Technical major |42.7 |19.6 | |49.6 |33.6 |

|Arts major |24.7 |31.2 | |32.7 |34.3 |

|No college |32.7 |49.2 | |17.6 |32.2 |

| | | | | | |

|Timing | | | | | |

|Below 20 |35.1 |44.4 | |20.9 |31.5 |

|20-24 |38.6 |41.2 | |37.8 |44.5 |

|25-29 |12.4 |12.0 | |16.7 |16.8 |

|30-34 |4.1 |2.4 | |7.0 |3.8 |

|35-39 |1.0 |0.0 | |4.0 |1.4 |

|40 and above |4.3 |0.0 | |8.5 |2.1 |

|Non-CCP |4.5 |0.0 | |5.2 |0.0 |

| | | | | | |

Table 3: Leadership Positions and Timing of Party Entry, 1988 and 1994 (%)

| | | |

|Variable |1988 |1994 |

| | | |

| | | | | | | | |

| |Government only |Government & |Party | |Government only |Government & |Party |

| | |Party |only | | |Party |only |

| | | | | | | | |

| | | | | | | | |

|Credentials | | | | | | | |

|Technical major |42.2 |35.0 |14.0 | |50.0 |50.0 |26.0 |

|Arts major |21.2 |11.7 |21.9 | |33.3 |26.6 |38.2 |

|No college |36.6 |53.3 |64.0 | |16.7 |23.3 |35.8 |

| | | | | | | | |

|Timing | | | | | | | |

|Below 20 |32.7 |43.3 |46.6 | |19.0 |26.7 |32.8 |

|20-24 |39.8 |40.0 |40.4 | |37.0 |45.0 |42.6 |

|25-29 |12.4 |15.0 |10.1 | |16.7 |18.3 |17.2 |

|30-34 |3.9 |1.7 |2.8 | |7.7 |3.3 |4.4 |

|35-39 |1.0 |0.0 |0.0 | |4.4 |1.7 |1.5 |

|40 and above |5.1 |0.0 |0.0 | |9.0 |5.0 |1.5 |

|Non-CCP |5.1 |0.0 |0.0 | |6.1 |0.0 |0.0 |

| | | | | | | | |

Table 4: Logistic Regression Coefficients for Institutional Effects on Elite Recruitment, 1988 and 1994

| | | |

|Covariates |Model 1 (1988) |Model 2 (1994) |

| | | |

| | | |

|Age |.003 (.013) |-.040 (.013)* |

|Women |-.260 (.398) |.098 (.333) |

|Minority ethnicity |-.165 (.273) |-.081 (.284) |

|Timing of party entry |.055 (.016)* |.048 (.011)* |

|College education |-.038 (.196) |.229 (.201) |

|Technical training |.978 (.225)* |.401 (.179)* |

| | | |

|-2 Log likelihood |879.493 |1,011.386 |

|Model Chi-Square |67.102* |80.621* |

| | | |

|No. of cases |740 |848 |

| | | |

* P ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download