Chinese leadership: Provincial perspectives on promotion and performance

Abstract

This article addresses whether the promotion of China’s provincial leadership can be explained through economic growth or alternative factors (experience, education or having previously held a central government post). We use promotion definitions, based on official administrative rankings and a more accurate reflection of political-governmental power relationships, and five econometric models. Analysis is also located within specific scholarship about promotion in the Chinese bureaucracy, bureaucratic cultures and transitions from socialism. Findings suggested some support for the idea that economic growth records affected promotion in terms of the second definition and additional evidence to imply impacts from non-economic factors.

Keywords

Economic development governance

Introduction

This article considers whether promotion of provincial Chinese leaders (Governors and Communist Party of China (CPC) Provincial Committee Secretaries) can be explained through their economic growth records or whether alternative explanations offer plausible narratives. We address themes significant to understanding administrative contexts of the world’s second largest economy and aim to augment a small existing scholarly literature. Our starting point is the development of a market economy following the third plenary session of the 11th CPC Central Committee in December 1978.

In locating this analysis within transitions from socialism, we also connect to debates about bureaucratic responses to such reforms. This issue has been discussed widely, for instance Walder (2003) studied capacity of Communist-era elites to appropriate public assets through reform processes. Similarly, Brown et al. (2009) considered participation of Russian bureaucrats in the privatization of state assets. We also place our study within a wider literature about why people are promoted, particularly the importance of performance (Blau and DeVaro, 2007; Weber, 1947) and the emergence of public sector performance and improvement agendas (McLaughlin et al., 2002), thus connecting this study with substantive international scholarly and practitioner perspectives.

Our analysis is shaped through a distinctive contemporary delegation in China of much of ‘the governance of the national economy … to sub-national governments’ (Xu, 2011: 1078), which facilitated territorial competition for example through rivalry for inward investment (Yeung, 2000). Competition occurred, however, within centrally-imposed agendas, a process that has accelerated since 2012 under Xi Jinping (Lee, 2017), and which has been delivered, partly, through central control over sub-national, or at least provincial, appointments. These arrangements have been characterized as a ‘regionally decentralized authoritarian (RDA) regime’ (Xu, 2011: 1078).

The authoritarian element also reflected imposition of performance agendas, which could be interpreted as Chinese adaptations of western public sector improvement programmes (Afonso et al., 2005) or Maoist planning legacies (Fan, 2006). These developments have been chronicled by scholars such as Tsui and Wang (2004: 89), who observed ‘a target-based vertical control system’ that could be used to affect sub-national expenditure.

This top-down performance agenda has long emphasized economic growth, both the tenth (2001–2005) and eleventh (2006–2010) 5-year plans incorporated objectives to double the level of gross domestic product (GDP) recorded in the year 2000 by 2010. Similarly, the twelfth 5-year plan (2011–2015) included a 7% annual GDP growth target. Sub-national competition operated within a framework of quantifiable economic targets and territories competed ‘to generate the highest growth rates’ (Chien and Gordon, 2008: 38). This context shifted behavior of sub-national bureaucrats, as Nee (1992: 14) observed, such ‘cadre-entrepreneurs’ sought success in the market-place ‘to advance their careers, to gain higher bonuses, and to expand their organizational power and influence’.

Our core research question concerns whether promotion of provincial leaders to national level was affected by their economic growth records or a range of alternative explanations – experience; education; or central connection, the latter defined through previous employment at the national tier. We also ask what implications our findings have for shifting organizational cultures amongst the Chinese government and changes to the ethos and values of the Chinese bureaucracy. We make, therefore, clear connections between our empirical analysis, wider debates about bureaucratic leadership in China and transitions from socialism, whilst also connecting to the extensive international literature about promotion and performance.

In the remainder of this article, we consider relevant scholarly literature before discussing our data sources and outlining the methodology. Subsequently, we specify and apply two definitions of promotion and five econometric models and evaluate the results. Finally, conclusions are highlighted and future research directions sketched.

Promotion and performance

There is a substantive relevant academic literature about why individuals are promoted within public sector bureaucracies. Scholars such as Weber (1947) recognized the importance of seniority and performance, but ‘stopped short of stating explicitly the conditions that might result in emphasizing one criterion at the expense of another’ (Halaby, 1978: 466). Limitations that were evident in wider scholarship about promotion decisions (Blau and Scott, 1962; Gouldner, 1957).

More recent decades have witnessed the emergence of a rigorous empirical literature investigating why promotions occurred. Halaby (1978) studied US financial bureaus and drew conclusions about how organizational characteristics affected promotion criteria, for example agencies with a relatively high proportion of employees with advanced accounting degrees gave less importance to seniority. Similarly, Baron et al. (1986) considered variations in the internal promotion regimes of 100 organizations, while Kalleberg and Van Buren (1996) discussed effects derived from organizational scale.

Alternatively, writers such as Groot and Van Den Brink (1996), Booth et al. (2003), and Blau and DeVaro (2007) considered impacts arising from gender, while scholars such as Ruderman et al. (1995), and Pudney and Shields (2000) assessed promotion within wider diversity contexts. Others such as Spilerman and Lunde (1991) studied the importance of educational backgrounds. Promotion prospects have also been connected to wider social themes, for example Kim (2002) associated promotion with individuals’ social capital, while Lemons and Jones (2001) and Garcia-Izquierdo et al. (2012) discussed promotion decisions through procedural justice.

A substantial theme of this literature has been explicit expression of the idea of promotion as a merit-derived process reflective of performance (Alonso and Lewis, 2001; DeVaro, 2006; Fairburn and Malcomson, 2001). An issue that has also been contexualized through the emergence of the New Public Management (McLaughlin et al., 2002) and an invigorated and enhanced emphasis on public sector performance (Bale and Dale, 1998; Brown, 2004; Horn, 1995). Overall, scholarship indicates that promotion can reflect achievement of specific job-related goals and or other forms of personal attainment such as educational qualifications. It also highlights the importance of personal attributes unrelated to performance for example, seniority, gender or ethnicity, and notes impacts from broader organizational characteristics.

Promotion, performance, the Chinese bureaucracy and economic growth

The positions of authority in the government or the Communist Party mentioned here comprise part of a cadre system, with its origins in the revolutionary struggle. During the Maoist period (1949–1976), cadres were chosen from the peasantry and at the start of the period typically had revolutionary credentials forged in the civil war and or the Japanese conflict. Maoism regarded intellectuals as bourgeois and, in contrast to European communist regimes, excluded them from leadership roles. Cadre recruitment was ‘based on class background and social relations rather than on talent and specialization’ (Wang, 1981: 4). Attempts by more pragmatic leaders, such as Liu Shaoqi, to co-opt intellectuals as cadres, so that their skills could be used for economic development, were resisted (Lee, 1991: 77). Furthermore, recruitment of peasants meant that levels of educational attainment amongst the cadres were typically low. Absence of a retirement system made breaking into the communist elite difficult. Established hierarchies were disrupted during the Cultural Revolution, when many in senior positions were purged, although almost everyone was eventually reinstated, provided they survived.

During the 1980s, the cadre system was overhauled and recruitment and promotion criteria shifted from revolutionary zeal towards technical competency. Changes that can be dated to a speech made by Deng Xiaoping to the Politburo on 15 August 1980, which acknowledged a new policy to improve the education and competence of the cadre corps, and lower its age profile. An agenda that, in part at least, was motivated through aims to promote economic growth.

In order to remove cadres without the technical skills required for economic growth, retirement rules had to be addressed. In February 1982, the CPC Central Committee issued a document (CPC Central Committee Decision on the Establishment of a Veteran Retirement System) which commenced the process of abolishing life-tenures for senior officials.¹ Certain veteran cadres were eligible for a special form of retirement (lixiu), which effectively translates as ‘leaving a job to recuperate’ (Lee, 1991: 235). Retirement was combined with ‘unchanged political treatment’ and ‘slightly better economic treatment’ (Lee, 1991: 236).

From the mid-1980s onwards, there was also evidence of the development of a more rigorous cadre management system (Manion, 1985). This evolved into the Cadre Responsibility System (Edin, 2003), which involved accountability for the delivery of ‘goals on a scale of importance: soft targets, hard targets and priority targets’ (Collins and Cottey, 2012: 86). Failure to implement those priorities negating other achievements. Accountability being (of course) through party-state structures at the next level in the hierarchy, political cultures that also facilitate swift implementation of new agendas. This effect can be illustrated through reference to how the decision by the national government to make climate change ‘a high priority of provincial and prefectural governments’ (Collins and Cottey, 2012: 86) quickly affected cadre performance targets and actions (Qi et al., 2008). The cadre system also incorporated procedures to constrain the decentralization reforms through limiting the extent to which cadres identified with their locality, for example through rotation of individuals between different areas (Zheng, 2009).

Importance of Chinese national policy agendas, evidenced particularly through the 5-year plans and capacity of national government to quickly impose its policies, might suggest that leadership of territories experiencing strong economic performance facilitates promotion. This interpretation, nevertheless, assumes similar reorientations in bureaucratic cultures (O’Reilly et al., 1991) to reward performance with promotion and so the emergence of organizational cultures and processes reflective of the reform agenda. It is, therefore, assumed that these changes have effectively challenged the traditional ‘mode of conduct’ (Rokeach, 1973: 5), dating from the Maoist period.

Such changes can be illuminated through relevant typological distinctions. Adopting Cameron and Freeman’s (1991) classification, constructed through internal-external and organic-mechanistic axes, we can postulate that cultural shifts reflecting performance agendas are consistent with movements from hierarchy-based to market-based cultures. The former coalescing around prioritizing internal maintenance, for example through integration and smoothing actions, the latter stressing external competition and differentiation. Both orientations are, however, focused on the mechanistic axis, reflective of values such as control and order.

Literature about China’s transition from socialism has also addressed shifting organizational cultures. Writers such as Tsui et al. (2006) considered variations amongst state-ownership, private domestically owned enterprises and foreign-owned firms. Others, such as Lau et al. (2002), connected Chinese organizational cultures to change agendas and employee attitudes towards their firm and job. Similarly, our study has implications for understanding organizational cultures of the Chinese bureaucracy. Regarding the hierarchy and market distinctions specified by Cameron and Freeman (1991), evidence that promotion reflected economic growth performance could be interpreted as indicating market cultures. Central control imposed to encourage provincial leaders to respond to external stimuli. Alternatively, dominance of other factors, particularly experience and central connection, might be appreciated as indicating hierarchy-based and inward-focused cultures.

Although there is an absence of scholarship connecting organizational cultures to bureaucratic promotions, the career trajectories of territorial Chinese leaders have been addressed through several studies. Chen et al. (2005) argued that promotion prospects of provincial leaders depended on relative economic performance compared to their immediate predecessors. Li and Zhou (2005) asserted that promotion and dismissal of provincial leaders reflected their economic records, while Blanchard and Shleifer (2001: 175) claimed that the Chinese Communist Party supported ‘governors whose regions have performed well economically’ and disciplined ‘governors who have followed anti-growth policies’. Similarly, Wei (2002: 1733) commented that ‘local government officials who have succeeded in achieving local economic development can be promoted’. Guo (2007) associated promotion with revenue growth; while Landry (2003) observed positive but limited impacts from economic performance on mayoral careers.

Alternatively, Su et al. (2012) concluded that economic growth records didn’t alter the career paths of provincial bureaucrats. This finding echoed those of Bo (1996), who expressed some scepticism about the idea that good economic records facilitated the promotion of provincial leaders. Similarly, Shih et al. (2012: 166) concluded that there was ‘no evidence that strong growth performance was rewarded with higher party ranks’ and Zhang (2014: 263) observed that good ‘economic performance’ did not increase probability of a Politburo or Secretariat appointment.

Regarding alternative explanations, studies addressed the impact of age. Bo (1996: 149) commented that amongst provincial leaderships, younger bureaucrats were ‘less likely to be demoted’; while Li and Zhou (2005) noted age-related impacts on promotion and termination. Scholars also considered educational attainment, Bo (1996: 147) concluded that provincial leaders with no college education were more likely to be promoted than those with college education. Similarly, the idea that promotion reflected previous employment at national level was addressed. Bo (1996: 145) argued that provincial ‘leaders of central origins are more likely to be promoted because they have stronger connections with the center’, while Li and Zhou (2005: 1755) commented that ‘previous or current work experience in the central government increases … the probability of promotion’.

This scholarship has, however, been characterized by significant methodological weaknesses, implying requirements for further analysis. First, some studies were limited in scope for example Su et al. (2012) was restricted to re-coding data used by Li and Zhou (2005) and Chen et al. (2005); while Landry’s (2003) study was confined to mayors, ignoring other echelons of the Chinese bureaucracy. Furthermore, this focus on mayors meant that the study was beset with intensive difficulties in obtaining accurate and comprehensive data.

Second, there were timeframe deficiencies in several studies. Specifically, both Landry (2003) and Guo (2007) used data from relatively short 11 and 8 year periods, respectively. Bo (1996) covered a substantive 1949–1994 period but failure to differentiate between sub-periods meant that impacts derived from post-1978/1980 reform processes were obscured.

Third, the literature has, to some extent, been characterized by definitional weaknesses. Studies, such as those undertaken by Chen et al. (2005) and Li and Zhou (2005), have been diminished through failure to recognize and model valid alternative definitions of promotion. Similarly, deficiencies in the treatment of economic growth were widespread (see Bo, 1996; Chen et al., 2005; Guo, 2007; Landry, 2003; Li and Zhou, 2005; Su et al., 2012; Zhang, 2014). For example, Guo (2007) omitted to situate economic growth records within contemporary economic performance of the relevant regions. This issue matters given uneven spatial development (Li and Wei, 2010) and substantive economic cycle variations (Zhang and Wan, 2005).

Fourth, potential connections with wider themes such as changes to bureaucratic leadership in China (Lin, 2008), or more generally transitions from socialism (Ferrero, 2001), have not been properly addressed through this existing literature. Our study avoids these methodological weaknesses, it covers a substantive timeframe relevant to modernization; uses an extensive quantitative dataset; deploys precise and robust definitions of promotion, economic growth and the other explanatory variables; and makes connections to the wider themes.

Methodology and data

Our data comprises biographic details about Governors and CPC Provincial Committee Secretaries and economic growth statistics. Information about the leaders was collated from a diversity of authoritative sources, including governmental and party websites and media outlets, such as xinhuanet.com and people.com.cn. Our decision to study provincial governmental and party leaderships reflected common scholarly approaches (Bo, 1996; Chen et al., 2005; Li and Zhou, 2005) and the importance of this tier. Particularly, through economic decentralization, which rendered ‘each provincial leader’s performance individually distinguishable and comparable and thereby allows for a sensible link between performance and turnover’ (Chen et al., 2005: 422).

We included individuals from 31 territories – the 22 provinces, four municipalities (Beijing, Tianjin, Shanghai and Chongqing) and five autonomous regions (Inner Mongolia, Guangxi Zhuang, Tibet, Ningxia Hui and Xinjiang Uyghur). We excluded the two special administrative regions (Hong Kong and Macau), which operate under the one country, two systems regime and were, partly, insulated from national governmental authority. We incorporated leaders, whose tenure ended after the Third Plenary Session of the 11th Central Committee of the CPC (December 1978). This timeframe reflects commencement of economic reforms and shifting bureaucratic perspectives from revolutionary agitators to technocrats (Lee, 1991). Data were obtained about 487 postings undertaken by 350 individuals, 345 (98.6%) of which were male and 309 (88.3%) Han Chinese. We treated each posting as a separate case because key characteristics, such as age, central connection and education, of each official, and their economic growth records, differed amongst postings.

Promotion

We used two definitions. First, an official version constructed through the Chinese administrative rankings; that codifies equivalence of Ministers, Provincial Governors and CPC Provincial Committee Secretaries (see Table 1). This classification has been recognized by scholars, specifically Li and Zhou (2005: 1745) commented that a ‘province ranks at the same level as a ministry in the central government’. Second, criteria that is more reflective of political and governmental power relationships. This aligns us with Li and Zhou (2005), in terms of classifying shifting from Provincial Governor to CPC Provincial Committee Secretary as promotion; assumptions reflexive of the central role of the CPC in delivering national agendas. Similarly, and in contrast to the definition adopted by Li and Zhou (2005), we interpreted moving from provincial leadership to a ministerial appointment as promotion; participation in national decision-making being the key factor.

Table 1.

Chinese leaders at national and provincial tiers.

Level	Sub-level	Positions
National (Tier 1)		CPC General Secretary, President, Premier, Central Military Commission (CMC) Chair, NPC Standing Committee Chair, Chinese People’s Political Consultative Conference (CPPCC) Chair, and the Politburo Standing Committee
National (Tier 2)		Vice-Chair of the CMC, Member of the Politburo, Alternate Members of the Politburo, Secretary of the Central Commission for Discipline Inspection, Secretary of the CPC Central Committee, Vice-Chair of the CPPCC National Committee, Vice-Chair of the NPC Standing Committee, Vice-Premier of the State Council, State Councillor, President of the Supreme People's Court, Attorney/Procurator General of the Supreme People's Procuratorate, Members of Standing Committee of the Central Advisory CommissionNote: The first and second grade cadres constitute the national leadership
Ministerial and provincial (Tier 3)	Ministerial	National government ministers
	Provincial	CPC Provincial Committee Secretaries
	Provincial	Provincial Governors
	Provincial	Provincial officials of equal ranking to the provincial leaders cited above but with less actual authority e.g. Provincial CPPCC Chairs and Mayors, Directors of the Provincial People’s Congress
	Provincial and ministerial	Prestigious honorary positions e.g. Deputy Secretary of the CPC Central Committee; and Director of the NPC Standing Committee. Members of the Central Advisory Committee.Officials from other key state institutions e.g. the Supreme People’s Procuratorate; and leaders of state owned businesses.
Ministerial and provincial (Tier 4)	Provincial	Provincial leaders with substantive power e.g. Deputy Provincial Governor; Deputy CPC Secretary
	Provincial	Provincial leaders with some significant power e.g. Provincial CPC Vice-Chairs, Deputy Director of the Provincial Peoples’ Congress
	Ministerial	Vice-Ministers
	Provincial and ministerial	Officials of equal ranking e.g. chiefs of provincial propaganda/publicity departments and departments of organization (HR). These departments must be chaired by members of the Provincial Standing Committee or the Deputy Provincial CPC Committee Secretary. Other examples include some leaders of state-owned businesses and universities.
Prefecture-level, cities (Tier 5)		Prefecture Party Secretaries and Mayors; Directors of the Municipal People’s Congress, Directors of the Municipal CPPCC

Economic growth

Economic performance of a given provincial leader is measured by the time weighted geometric growth rate during their tenure. Denote r_i,t as the economic growth rate of province i at year t. Suppose the Governor of province i began their tenure in March 2003 and departed in July 2005. We ignore the month when the tenure started, a decision reflexive of inadequate data about the precise appointment date; similarly we ignore the end month. Here, tenure would be recorded as 27 months. The average economic growth rate is calculated as follows

{\bar{r}}_{i} = {[\frac{9}{12} (1 + r_{i, 2003})] \cdot [\frac{12}{12} (1 + r_{i, 2004})] \cdot [\frac{6}{12} (1 + r_{i, 2005})]}^{1 / (\frac{9}{12} + 1 + \frac{6}{12})} - 1

The geometric average is used because the arithmetic average tends to overstate actual average growth. Our analysis incorporates 20 cases without a valid growth measure. This occurs when the tenure started and ended in the same year, or when there were no valid provincial economic statistics. However, partial information was available for these leaders; hence these cases are included in the tabulation but omitted from the model building.

The economic growth measure of the province alone cannot capture effectively performance of the provincial leadership, because we need to address the issue of national government benchmarking the performance of individuals against their peers. Furthermore, economic growth rates include many shocks, both domestic, such as regional economic policy; and international, for instance the severe economic recession in 2008–2009. To address these issues, we developed and improved the approach used by Chen et al. (2005), who included provincial GDP growth of the immediate predecessor and GDP growth of the neighbouring provinces as benchmarks.

First, we used a 5-year geometric average provincial growth rate for the period before the start of a given tenure to measure predecessor(s)’ performance. This approach reflects the length of a normal tenure and concern to prevent wide variations in the tenure length of immediate predecessor(s) undermining our comparison benchmarks. Calculation is relatively straightforward, it is identical to the growth measure of a hypothetical predecessor who has had a 60 months tenure.

Second, we grouped provinces into regions using the standard geographical classification of Chinese regions. The regional groupings are as follows: North (Beijing, Tianjin, Hebei, Shanxi and Inner Mongolia), North-East (Liaoning, Jilin and Heilongjiang), East (Shanghai, Jiangsu, Zhejiang, Anhui, Fujian and Shandong), Central (Jiangxi, Henan, Hubei and Hunan), South (Guangdong, Guangxi Zhuang and Hainan), South-West (Chongqing, Sichuan, Guizhou, Yunnan and Tibet) and North-West (Shaanxi, Gansu, Qinghai, Ningxia Hui and Xinjiang Uyghur). This approach is adopted because growth rates reflect regional policy variations. Chinese economic policy has long had a regional focus, for example from 1979 onwards central government implemented preferential economic policies to stimulate the economic growth of the coastal regions. More recently, several specific regional economic policies have been introduced; an example being the North East Area Revitalization Plan (China State Council, 2007). Furthermore, to isolate the impact of the province on the regional economic growth rate, we excluded the province while calculating the average regional economic growth. This is particularly important for smaller regions, or regions with uneven economic growth amongst the provinces. Suppose there are N provinces in the region, we index them as 1, … N. We need to calculate the rate of returns for all the other provinces during the tenure, as follows

\begin{array}{l} {\bar{r}}_{1} = {[\frac{9}{12} (1 + r_{1, 2003})] \cdot [\frac{12}{12} (1 + r_{1, 2004})] \cdot [\frac{6}{12} (1 + r_{1, 2005})]}^{\frac{1}{(\frac{9}{12} + 1 + \frac{6}{12})}} - 1 \\ {\bar{r}}_{2} = {[\frac{9}{12} (1 + r_{2, 2003})] \cdot [\frac{12}{12} (1 + r_{2, 2004})] \cdot [\frac{6}{12} (1 + r_{2, 2005})]}^{\frac{1}{(\frac{9}{12} + 1 + \frac{6}{12})}} - 1 \\ \dots . \\ {\bar{r}}_{i - 1} = {[\frac{9}{12} (1 + r_{2, 2003})] \cdot [\frac{12}{12} (1 + r_{i - 1, 2004})] \cdot [\frac{6}{12} (1 + r_{i - 1, 2005})]}^{\frac{1}{(\frac{9}{12} + 1 + \frac{6}{12})}} - 1 \\ {\bar{r}}_{i + 1} = {[\frac{9}{12} (1 + r_{2, 2003})] \cdot [\frac{12}{12} (1 + r_{i + 1, 2004})] \cdot [\frac{6}{12} (1 + r_{i + 1, 2005})]}^{\frac{1}{(\frac{9}{12} + 1 + \frac{6}{12})}} - 1 \\ \dots . \\ {\bar{r}}_{N} = {[\frac{9}{12} (1 + r_{N, 2003})] \cdot [\frac{12}{12} (1 + r_{N, 2004})] \cdot [\frac{6}{12} (1 + r_{N, 2005})]}^{\frac{1}{(\frac{9}{12} + 1 + \frac{6}{12})}} - 1 \end{array}

The regional economic growth benchmark is then generated as

\frac{({\bar{r}}_{1} + {\bar{r}}_{2} + \dots + {\bar{r}}_{i - 1} + {\bar{r}}_{i + 1} + {\bar{r}}_{N})}{N} .

Third, to consider possible shocks from the global economy, we selected five countries; specifically, the United States, Japan, South Korea, Germany and Australia; and calculated the average economic growth rates during the tenure of the provincial leader. These are the top countries in terms of trading volume with China (based on 2014 figures), and are G20 members. The economic growth shocks of these countries were likely to have had spill-over effects on the Chinese economy.

Fourth, pre-existing prosperity is also important for evaluating impacts of individual leaders. Particularly this reflects the economic convergence literature (see Quah, 1996), which indicates that territories with higher income levels typically grow at slower rates, and vice versa. Suppose two provinces have the same economic growth; ceteris paribus, the leader from the more affluent province would effectively be outperforming his colleague from the poorer province, because it would be more difficult for the economy of the more affluent province to grow. There is also an interpretation through those with the most political leverage securing appointments in the more affluent provinces, rather than the poorer and less developed parts of the country. Thus, similar to Li and Zhou (2005), we included the lagged provincial per capita GDP (per capita GDP of the year before the start of the tenure) to control for pre-existing prosperity.

Central connection

Central connection enters the regression as a dichotomous variable (Li and Zhou, 2005). Leaders having previous experience in national government or holding contemporaneous joint-appointments in both central government and as provincial leaders, equating to 1, and 0 otherwise.

Information on central connection, was, however collected in an ordinal scale, detailing the exact tier in the bureaucratic system (see Table 2). This facilitates exclusion of cases that would otherwise generate over-counting of promotion and bias estimation results. Specifically, there are officials who possessed central connection at Tier 2, this cohort had effectively already been promoted to the national leadership, and would need to be removed from the first set of regressions. Furthermore, officials with central connections at Tier 3 level (e.g. ministerial posts) before appointment to the provincial leadership, would also need to be excluded. Such cases included Sun Zhengcai, who served as Minister of Agriculture before becoming the CPC Committee Secretary of Jilin province; and Jiang Zemin, later the General Secretary of the CPC Central Committee and then President of China, who served as Minister for the Electronics Industry before becoming the Mayor and then the CPC Committee Secretary of Shanghai. This cohort would need to be removed from the second set of regression analysis. Such job changes might be reflexive of personal desires to relocate back to Beijing or the natural culmination of temporary postings aimed at broadening administrative experience rather than a genuine promotion. Including those cases might, therefore, produce a false significant relationship between central connection and promotion. Similarly, we do not count posts below Tier 4, given their low ranking. Furthermore, our decision to investigate specific positions rather than individuals was confirmed because some officials had no central connection in an initial tenure but acquired one before a subsequent tenure. Most of them eventually acquiring central connection (see Table 3).

Table 2.

Central connection at start of the tenure.

Level of connection	Frequency	Percentage
Tier 2	19	3.9
Tier 3	56	11.5
Tier 4	71	14.6
Below Tier 4	21	4.3
No central connection	320	65.7
Total	487	100.0

Table 3.

Acquisition of central connections.

	Frequency	Percentage
Central connection before the tenure	167	34.3
Central connection during the tenure	16	3.3
Central connection after the tenure	213	43.7
No central connection	91	18.7
Total	487	100.0

Experience and education

Experience was addressed through age and length of service. Our age statistics reflect age at the end of each tenure (see Table 4). Age is (of course) important in determining bureaucratic career trajectories. Since our purpose is to identify contributing factors to promotion, retirement rules must be addressed, specifically the introduction of mandatory retirement. These rules specified retirement of provincial leaders at age 65; and were implemented in 1983 (Li and Zhou, 2005: 1759). Although there are dispensations to retain office until completion of current terms, tenures are seldom renewed for those aged above 63, and 68 is, with some very rare exceptions (see Table 7), usually the ceiling for joining the national leadership. Thus, to reflect impacts from implementation of mandatory retirement, we generated a dummy variable indicating whether the tenure commenced from 1 January 1983 and onwards (post-1982) and an age indicator (age > 65). We then included age, the post-1982 indicator and the multiplication of age with the post-1982 indicator, to capture the effects of age, the mandatory retirement rule and their interaction on promotion.

Table 4.

Age at the end of the tenure.

Age at end of the tenure	Whole sample		Start of tenureBefore 1 January 1983		Start of tenure1 January 1983 and after
Age at end of the tenure	Frequency	Percentage	Frequency	Percentage	Frequency	Percentage
≤50	23	4.7	0	0.0	23	5.2
51–55	58	11.9	1	2.4	57	12.8
56–60	131	26.9	1	2.4	130	29.2
61–65	189	38.8	16	38.1	173	38.9
66–68	56	11.5	15	35.7	41	9.2
≥69	30	6.2	9	21.4	21	4.7
Total	487	100.0	42	100.0	445	100.0

Length of service is considered through two aspects. First, the number of months an individual has served in the relevant post, cases where an official held their position for more than 5 years (one tenure) have been flagged out and effects captured as a dummy variable through regression analysis. Second, we recorded the number of provinces an individual has served in, as either Governor or CPC Provincial Committee Secretary. This, similar to central connection, captures the extent of their personal connections and administrative experience. We thus identified a variable overlooked by others such as Li and Zhou (2005) and Chen et al. (2005).

Educational attainment is assessed through a fourfold classification. Cadres were categorized in terms of whether their highest educational qualification was a PhD, master’s degree, a first degree or a high school graduation certificate or below. Levels were recorded for when each individual attained their first Tier 3 position and each subsequent tenure. Similar to central connection, educational attainment enters the regression as a binary variable (first degree and above = 1, lower = 0). Data on the educational attainment of 51 leaders was missing. However, it is unlikely that a provincial leader with higher education would omit those details from their resume. Alternatively, it is probable that those leaders had minimal formal education. Analysis strengthened this hypothesis. The oldest of these people is Liu Tianfu, born October 1908, and the youngest is Song Xiangrui, born October 1939. All were born and obtained maturity during the civil war or very shortly afterwards, when receiving higher education was unusual. We, therefore, classified them as high school graduates and below. We then compared the educational attainment of leaders whose tenure started before 1 January 1983 with those commencing on/after 1 January 1983 (see Table 5). This analysis suggested that educational attainment of the officials in the second sample exceeded substantially the first cohort. These findings might also be interpreted as reflexive of transitions from revolutionary cadres to party technocrats (Lee, 1991), given the rough proximity of our cut-off date to the start of reform processes.

Table 5.

Education level.

Start of tenure			High school graduate and below	First degree	Master’s degree	PhD	Total
Whole sample	Education level at the start of their first Tier 3 appointment	Frequency	199	271	15	2	487
		Percentage	40.9	55.6	3.1	0.4	100.0
	Current education level	Frequency	115	273	81	18	487
	Current education level	Percentage	23.6	56.1	16.6	3.7	100.0
Before 1 January 1983	Education level at the start of their first Tier 3 appointment	Frequency	31	11	0	0	42
		Percentage	73.8	26.2	0	0	100.0
	Current education level	Frequency	31	11	0	0	42
	Current education level	Percentage	73.8	26.2	0	0	100.0
1 January 1983 and after	Education level at the start of their first Tier 3 appointment	Frequency	168	260	15	2	445
		Percentage	37.8	58.4	3.4	0.4	100.0
	Current education level	Frequency	84	262	81	18	445
	Current education level	Percentage	18.9	58.9	18.2	4.0	100.0

Note: Current education level refers to their highest educational attainment at end 2015, reflective of the survey date.

Table 6.

Binary logistic regression on the promotion of the provincial leadership: Official administrative rankings.

	Model 1	Model 2	Model 3	Model 4	Model 5
Provincial growth rate	0.676(0.060)	−2.257(0.204)	−1.237(0.060)	−0.736 (0.017)
Regional growth rate (without the province)		3.624(0.463)	2.328(0.168)	3.972 (0.385)
Predecessor’s growth rate		−2.420 (0.200)	−2.889(0.282)	−2.922 (0.237)
Predecessor’s regional growth rate (without the province)		3.313(0.331)	5.055(0.689)	5.458 (0.612)
Lagged provincial per capita GDP (1-year lag) (1000 RMB)		0.066***(23.062)	0.089***(18.724)	0.096***(17.531)	0.069***(18.533)
Per capita GDP growth rate – United States			5.698(0.257)	−2.460(0.027)
Per capita GDP growth rate – Japan			−1.551(0.269)	−2.891(0.659)
Per capita GDP growth rate – South Korea			4.403(1.892)	7.837**(4.434)
Per capita GDP growth rate – Germany			1.378(0.177)	0.883(0.058)
Per capita GDP growth rate – Australia			−3.561(1.320)	−7.276(3.630)
Age at the end of the tenure				−0.117***(12.007)	−0.121***(15.335)
Post-1982 indicator (1 = post-1982, 0 = otherwise)				−2.972**(8.779)	−1.646**(6.913)
Post-1982 indicator × Age 65 indicator				1.652**(5.354)	1.302**(5.044)
Education (university = 1, lower = 0)				0.662(1.334)
Central connection indicator (1 = central connection at Tier 3, 0 = no central connection or central connection below Tier 3)				1.795***(20.495)	1.695***(20.126)
Serving more than one tenure in a specific provincial role (tenure $\geq$ 5 years)				1.319(12.473)	20.126***(12.768)
Number of provinces served				0.966***(14.189)	0.926***(14.390 )
Constant	−1.922(17.498)	−2.699***(15.328)	−9.802(0.652)	3.458(0.197)	4.268**(4.502)
Statistics	(1)	(2)	(3)	(4)	(5)
−2 Log likelihood	364.407	338.229	333.709	277.462	288.727
Cox & Snell R²	0.000	0.057	0.066	0.176	0.155
Nagelkerke R²	0.000	0.102	0.119	0.317	0.279
Hosmer and Lemeshow test	6.292	8.670	11.733	6.558	3.940

Note: Wald statistics values are in the parentheses. The significance levels of 1% and 5% are marked by *** and **. The Hosmer–Lemeshow tests the null hypothesis that predictions made by the model fit perfectly with the observed explanatory variables. Here, the statistic follows $χ_{(8)}^{2}$ under the null with a 5% critical value being 13.678.

Analysis

We explore connections between our explanatory factors and both promotion definitions. Our dependent variable is the turnover of provincial leaders, which is categorical or, more specifically, ordinal. Our dataset is panel data in nature, although we wouldn’t be able to specify a categorical panel model; the dataset is heavily unbalanced, because data about characteristics of specific leaders were collected for different years. Neither should a time-series framework be used, as individuals seldom spend more than a few years in senior provincial positions. Geographical categorization of our data implies relevance for a spatial model; however severe limitations concerning spatial categorical modelling, and the related econometric theory, mean that we eschew this approach. Nevertheless, spatial effects can be incorporated through specification of regional GDP growth rates as an explanatory variable. With regard to the overall analysis, there were another two crucial technical issues.

First, each category of explanatory variables included multiple measures, some of which are highly correlated for example growth and lagged growth. This matters, because such multicollinearity causes estimation problems and the consequential failure of the estimation algorithm. Thus, we first ran a separate linear regression of the independent variables on the dependent variable, examined the variance inflation factor (VIF) and only considered possible combinations of explanatory variables with a VIF less than 10.

Second, choice of the most appropriate categorical data model. Scholars such as Li and Zhou (2005) and Chen et al. (2005) applied ordinal regressions, however ordinal models are based on the parallel lines assumption. We were, therefore, obliged to undertake a full-likelihood ratio test on our dataset to compare the fitted ordinal model to a model with varying slope parameters, and found that the parallel lines assumption failed to hold. To resolve this problem, we could split the ordinal dependent variable into dichotomies and run binomial logistic regressions on each of these cumulative dichotomies, but that specification would not be parsimonious and the results would present interpretation difficulties. We decided, therefore, to specify a more parsimonious binary logistic model, with 1 representing promotion and 0 otherwise.

The binary logistic model predicts the probability of success given the set of explanatory variables. The model takes the following format

logit (\hat{p}) = l n (\frac{\hat{p}}{1 - \hat{p}}) = X β

where

(\frac{\hat{p}}{1 - \hat{p}})

is the predicted odds ratio,

\hat{p}

is the predicted probability of the event of promotion, which is coded as 1; X is the matrix of explanatory variables with the first column being 1, and β the vectors of coefficients.

Promotion: Official administrative rankings

Promotion is specified exclusively as elevation to Tiers 1 and 2.² After removing those with no economic growth measure and those having Tier 2 central connection, 450 cases remained. Using this definition, 63 (14%) were promoted, 250 (55.6%) remained at Tier 3 and 137 (30.4%) were retired or demoted.

Our regression analysis used five models to explain promotion from the provincial leadership. Each model is adequate, in the sense that the Hosmer–Lemeshow tests don’t reject. Because an important objective is to establish whether there is a causal relationship between economic performance and promotion, in the first three models the explanatory variables are restricted to economic growth indicators. The first model is confined to the provincial GDP growth measure. The second model extends the first model by including the current leader’s regional growth measure, both the provincial and regional growth measures of the predecessor(s) and lagged provincial per capita GDP. The third model incorporates possible international shocks by including the top five international trading partners’ corresponding per capita GDP growth rates.³

The results indicate that, except for lagged per capita GDP, which is used to capture existing levels of each province’s economic development (pre-existing prosperity), all economic growth-related variables are insignificant. Note that, apart from the provincial, regional, international economic growth rates during the tenure of the provincial leader, we examined the lagged 3-, 6- and 12-month economic growth rates. However, we dropped them from the tabulation as we found their coefficients were statistically insignificant. The calculation of these growth rates was the same as for those of a hypothetical provincial leader whose tenure was shifted backward by 3, 6 and 12 months, respectively.

We also examined non-linear interactions among different economic growth rates, for example ratios of the provincial economic growth rate plus 1 over the national economic growth rate plus 1. It is impossible to exhaust all non-linear interactions of growth rates, however we found that the economic growth rates and the non-linear interactions we attempted to include were all associated with statistically insignificant coefficients, and that there were no statistically significant reductions of the ‘−2 log likelihood’ by including them. These results were, therefore, omitted from the tabulation. The details are available from the authors upon request.

Next, we tried to include most of the explanatory variables in one regression, provided that the VIF would permit, and reported this configuration in the fourth model. We found that South Korea’s per capita GDP growth rate appeared to have a significant impact on the odds of promoting to the national leadership at the 5% level. However, this raises a question – why South Korea? South Korea has not been China’s largest trading partner. Moreover, such benchmarking would only make sense if the provincial economic growth rate matters first; on the contrary, the provincial economic growth rate itself is insignificant. Our results imply, therefore, that this significant relationship is spurious. Indeed, we found such significance disappeared after we removed some of the insignificant variables. We also reported the Cox & Snell R² and Nagelkerke R².⁴

Model 5 contains the explanatory variables which are statistically significant at the 5% level. Clearly, the fourth model can be interpreted as extending the fifth model by adding 10 explanatory variables. It has reduced the ‘−2 log likelihood’ by 11.265. However, such a reduction would be statistically insignificant, considering that 10 variables were added to the equation.

We, therefore, established the fifth model as being the parsimonious model, adequate for determining the key contributing factors to the odds of promotion. In summary, it is suggested that age, whether tenures ended after 1982, the interaction of the age dummy and post-1982 indicator, central connection, serving more than one tenure in a specific provincial role and the number of provinces served in as a leader all significantly affect the odds of promotion to the national leadership.

Overall, our findings imply that promotion to Tiers 1 and 2 from the provincial leadership is not much affected by performance as measured through the economic growth of the province. However, having leadership positions in the more affluent provinces appears to be an important positive factor in promotion to the national leadership, signalled by a significant and positively estimated coefficient for lagged provincial per capital GDP – measure for the initial income level. Had the economic growth rate been positive and statistically significant, a positive and statistically significant lagged per capita GDP would imply a promotion rule based on economic growth, given that it would be more difficult for a more developed economy to grow, as indicated by the economic convergence literature (e.g. Quah, 1996). However, given the insignificant impact of the provincial growth rate on the odds of promotion, the significant and positive impact from the initial income level may suggest that certain individuals possess the initial political leverage to secure postings in the more affluent areas rather than the poorer and less developed areas. Part of this political leverage can be measured by the depth and width of their political connections, specifically the central connection indicator and the total number of provinces served in as a leader by the end of the tenure. The first variable measures the connection with the top (depth), the second variable measures the broadness of the connection at the provincial level (width), both of which have statistically significant impacts on promotion; although such political leverage would also be related to qualities (e.g. family background) that would be very difficult to quantify.

Interpretation of the post-1982 indicator might not be straightforward. Superficially, it seems that cases in the post-1982 sub-sample are less likely to be promoted. This outcome might, however, reflect the fact that some officials have not yet realized their career potential as a consequence of relative youth (see Table 7). Most officials are promoted when aged between 56 and 65, while many cases in the post-1982 sample were under 55 years old. Nevertheless, we included a post-1982 indicator as an explanatory variable since it is reflexive of an important change in the age composition of the provincial leadership. We found that education has no significant impact on promotion, which is consistent with the findings in Li and Zhou (2005). Furthermore, an independent t-test was conducted to test the difference in the average levels of education between the cohort who were promoted to the national leadership and the cohort who weren’t; we found that there were no significant differences in levels of education.

Promotion: Political-governmental power relationships

Here, we also interpret moving from Governor to CPC Provincial Committee Secretary and from the provincial leadership to ministerial positions as promotion. We, therefore, specified promotion criteria more reflective of contemporary power relationships. Apart from removing cases with no valid end of tenure GDP, and those with central connections at Tier 2 and above, we eliminated cases with central connection at Tier 3 (see above), leaving 396 cases in total. Otherwise, we retained the same explanatory variables and our estimation results are reported in Table 8.

Table 7.

Age of those promoted: Official administrative rankings.

	All cases		Post-1982 cases
Age	Frequency	Percentage	Frequency	Percentage
≤50	5	7.9	5	8.6
51–55	9	14.3	8	13.8
56–60	19	30.2	18	31.0
61–65	19	30.2	19	32.8
66–68	9	14.3	6	10.3
≥69	2	3.2	2	3.4
Total	63	100.0	58	100.0

Note: Post-1982 cases, tenure started on 1 January 1983 or after.

Table 8.

Binary logistic regression on the promotion of the provincial leadership: Political-governmental power relationships.

	Model 1	Model 2	Model 3	Model 4	Model 5
	B	B	B	B	B
Provincial growth rate	5.104**(5.655)	4.703(1.373)	4.999(1.532)	5.799(1.770)	6.482**(6.796)
Regional growth rate (without the province)		−0.295(0.005)	−0.403(0.008)	2.980(0.337)
Predecessor’s growth rate		4.810(1.222)	4.614(1.094)	6.407(1.699)
Predecessor’s regional growth rate (without the province)		−1.110(0.057)	0.040(0.000)	−3.375(0.381)
Lagged provincial per capita GDP (1-year lag) (1000 RMB)		0.063***(17.033)	0.086***(13.930)	0.093***(14.380)	0.070***(20.233)
Per capita GDP growth rate – United States			6.325(0.508)	0.365(0.001)
Per capita GDP growth rate – Japan			2.110(0.685)	1.751(0.401)
Per capita GDP growth rate – South Korea			1.532(0.401)	3.628(1.982)
Per capita GDP growth rate – Germany			−1.904(0.519)	−5.238(3.322)
Per capita GDP growth rate – Australia			−1.502(0.389)	−3.936(1.950)
Age at the end of the tenure				−0.182***(29.918)	−0.162***(37.974)
Post-1982 indicator (1 = post-1982, 0 = otherwise)				−2.535***(12.829)	−1.639***(11.984)
Post-1982 indicator × Age 65 indicator				0.297(0.241)
Education (university = 1, lower = 0)				0.205(0.299)
Central connection indicator (1 = central connection at Tier 4, 0 = central connection below Tier 4, or no central connection)				−0.310(0.821)
Serving more than one tenure in a specific provincial role (tenure $\geq$ 5 years)				0.298(1.120)
Number of provinces served				−0.038(0.026)
Constant	−1.437***(15.479)	−2.259***(19.065)	−2.947**(9.841)	−4.151(0.386)	9.167***(27.431)
Statistics	(1)	(2)	(3)	(4)	(5)
−2 Log likelihood	507.523	479.077	475.539	418.966	432.705
Cox & Snell R²	0.014	0.083	0.091	0.212	0.184
Nagelkerke R²	0.020	0.114	0.125	0.292	0.253
Hosmer and Lemeshow test	10.107	3.027	14.644	11.395	7.159

Note: Wald statistics values are in the parentheses. The significance levels of 1% and 5% are marked by *** and **. Here, the Hosmer–Lemeshow statistic follows $χ_{(8)}^{2}$ under the null with a 5% critical value being 13.678.

For the same reason as in the previous case, concerning the official administrative ranking, we base our interpretation primarily on the fifth model. Comparison with findings from the previous case (see Tables 6 and 8), implies some shifts in promotion explanations. Specifically, coefficients for the provincial economic growth rate as well as for its initial income level (measured through lagged per capita GDP) are positive and statistically significant (Model 5). Our findings imply, therefore, that strong performance through good economic growth records facilitates promotion. Similarly, these findings suggest that, assuming similar growth records, possession of leadership roles in the more affluent provinces facilitates promotion (Model 5).

These results for provincial and lagged per capita GDP growth imply, therefore, a possible interpretation that aligns our findings with the economic convergence literature (e.g. Quah, 1996). As indicated above, this scholarship acknowledges that achieving economic growth is more difficult in a relatively affluent province and so equalling growth rates from the poorer areas means that leaders in those more prosperous provinces can be said to have outperformed their colleagues. In contrast with the first case, interaction between the age and post-1982 indicators, central connection, whether the leader served for more than one tenure in a specific provincial job and the number of provinces served in are insignificant (Model 4). In particular, findings about central connection are reflective of the fact that we have removed central connection at Tier 3 and above, leaving only those with central connection at Tier 4. As for the insignificance concerning the Post-82 indicator × Age 65 indicator, it can be seen (see Table 9) that the age composition of the post-1982 cases is not very different from the overall sample.

Table 9.

Age of those promoted: Political-governmental power relationships.

	All cases		Post-1982 cases
Age	Frequency	Percentage	Frequency	Percentage
≤50	8	5.8	8	6.3
51–55	29	20.9	28	21.9
56–60	57	41.0	56	43.8
61–65	37	26.6	30	23.4
66–68	8	5.8	6	4.7
Total	139	100.0	128	100.0

Note: Post-1982 cases, tenure started on 1 January 1983 or after.

Conclusions

We have evaluated whether promotion of provincial Chinese leaders can be explained through their economic growth records or several alternative factors. This emphasis means that our study contributes to wider debates about relationships between promotion and performance (see Alonso and Lewis, 2001; DeVaro, 2006) and deliberations about the extent to which Chinese governance is reflective of the emergence of the pronounced performance agenda (McLaughlin et al., 2002) observed across the public sector in a wide range of countries. Here, we utilized two sets of promotion criteria, official administrative rankings and an alternative definition more closely equated to actual politico-governmental power relationships.

Concerning our first definition, promotion was facilitated through factors such as serving as a provincial leader beyond one 5-year tenure in a specific leadership role and or in leadership roles in multiple provinces; and age. Alternatively, promotion wasn’t facilitated through good economic growth records; although political leverage, as measured by the lagged per capita GDP of the province and central connection appeared to have an impact.

We identified, however, a positive and significant relationship between economic growth and promotion through the alternative promotion definition (see Table 8). This is interesting given the strategic authority exercised both directly from national government and indirectly through Communist Party structures, and supplies evidence to reinforce market and performance-driven interpretations of Chinese governance. There is, therefore, support for the interpretation that central government exercised its substantive control to reward bureaucrats with good economic records. Particularly, these findings are important in terms of understanding organizational cultures amongst the Chinese bureaucracy. It suggests that whilst promotion from Governors to CPC Provincial Committee Secretaries, and from the provincial leadership to ministerial positions, might be interpreted through market cultures reflective of adherence to values appreciative of a competitive ethos; promotion to higher levels appears to reflect a hierarchical and internally orientated organizational culture. Factors such as central connection, age, whether the leader has served more than one tenure in a specific provincial job and the number of provincial leadership tenures being important. Promotion can, therefore, be equated more with characteristics reflexive of status, and or seniority, rather than achievement.

Our analysis might, therefore, be interpreted to signify emergence of dual organizational cultures as expressed through bureaucratic promotions. With advancement to the less important levels reflective of market cultures, but promotion to more senior roles indicative of hierarchical and traditional values. We suggest, therefore, that after approximately 40 years of economic reforms, the senior bureaucracy seems reluctant to allow performance at provincial leadership level to affect directly composition of Tiers 1 and 2; insulating leadership selection for the upper echelons of the government and the CPC from market cultures, performance considerations and, specifically, the successful implementation of economic reform agendas.

Our findings about this dual organization culture concerning promotion raise questions about whether such cultural distinctions are sustainable or whether a convergence will occur at some stage. Of potential significance are contrasting organizational arrangements for promotion at different levels in the hierarchy. For example, the upper echelons are selected through three conferences, particularly the National People’s Congress (NPC), which has the supreme authority. The candidates for most senior positions being proposed by the presidium of the NPC, a temporary organization functioning during the duration of the NPC. Alternatively, ministerial positions are nominated by the Premier of the State Council, and approved by the NPC. When the NPC is closed, the NPC Standing Committee has the authority to appoint and dismiss ministers. CPC Provincial Committee Secretaries are usually selected by the Organization Department of the CPC Central Committee, or sometimes recommended by the Chinese national leadership, screened by the Central Political Bureau and its standing committee, and appointed by the central government.

Such institutional variations for selections at contrasting stages in the structure might, however, imply that erosion of this dual culture might have to await institutional reform and establishment of more uniform procedures. Alternatively, eventual promotion of individuals from CPC Provincial Committee Secretaries and ministerial positions with strong economic growth records, and perhaps acute concerns with such performance, might erode this dual culture.

Central is the issue of whether, despite its focus on economic growth, the system is reluctant to allow such performance criteria to determine elevation to the most senior levels, where the real power resides. This might also be considered within the emergence of a more nationalistic and authoritarian atmosphere (Lagerkvist, 2014) under Xi Jingpin since 2012. The core questions are whether this new authoritarianism might marginalize technocratic expertise, or if aspirations towards a more substantive global role might eventually facilitate promotion to the upper echelons of those with the stronger economic growth records. Economic strength being (of course) a key determinant of international influence and authority.

This discussion implies future research agendas through investigating changes to this dual preferment culture. Options might include interview-based projects to capture shifting attitudes and quantitative studies of promotion across a narrower range of posts using subsequent, and perhaps more detailed, datasets. The focus might also be on whether it was possible to detect movement in organizational culture through a more fine-tuned and micro-analysis of selections of a smaller cohort of cadres occupying a more distinctive position in the hierarchy.

Alternatively, our analysis might be extended through application to a wider group of lower-level cadres and or promotions of leaders of state-owned enterprises (SOEs), particularly with regard to the fact that the career trajectories of these individuals often incorporate similar roles in party and or governmental structures. This reality suggests some significant research questions, for example – does the economic performance of the SOE under their tenure affect their promotion prospects? Are the successful achievement of other goals, such as providing employment, more important for promotion?

Our approach might also be extended through use of much more complex datasets to consider promotion of certain political cliques, such as princelings, namely national leaders with family connections to previous members of that same elite. Prominent contemporary examples of Chinese princelings include current General Secretary, President and Chair of the Central Military Commission Xi Jingpin, who is the son of Xi Zhongxun, whose senior positions included Vice-Chair of the NPC Standing Committee; and Wang Qishan (the current Vice-President of China), who is the son-in-law of Tao Yilin (a former Vice-Premier). There are research questions concerning explanations for their promotion to the national leadership and how such explanations vary with those identified for the wider cohort of provincial leaders. Scholars could consider whether, or to what extent, those princelings, once established in provincial leadership roles, have an enhanced chance of promotion to the national leadership.

There is also potential to apply our models to other countries, either through promotion of governmental officials or, more specifically, to political systems with similar relationships between a dominant political party and governmental structures, and where electoral choice is heavily constrained.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was assisted through a National Natural Science Foundation of China grant, no.71401161.

Notes

Jiajing Sun, CFA, is a lecturer in financial econometrics at the School of Economics and Management, University of Chinese Academy of Sciences. She received her PhD from the Management School, University of Liverpool in 2011; and her MSc from the University of Edinburgh. She has a growing range of publications in economics and related fields.

Michael Cole received his BSc in Politics from the University of Southampton in 1985. He stayed at Southampton for his MPhil thesis on the accountability of quangos, which he received in 1991. He also has a MSc in Information Science from City University (1990) and a PhD on local government modernization from the University of Plymouth (2008). He has held full-time research appointments at the universities of Glamorgan, Northumbria, Exeter and Plymouth; and undertaken consultancy projects for several public sector organizations. He joined the University of Liverpool Management School in May 2007. He has published widely on governance-related theme.

Zhiyuan Huang received his BSc in Statistics from the University of Science and Technology of China in 2010 and his PhD in Management from the University of the Chinese Academy of Sciences in 2017. Since July 2017 he has been working for the China Central Depository & Clearing Co. Ltd.

Shouyang Wang is currently professor of Management Science and Vice President of the Academy of Mathematics and Systems Science, Chinese Academy of Sciences and Director of the Centre for Forecasting Science, Chinese Academy of Sciences. He received his BSc from Sun Yat-sen University in 1982, his MSc from the Institute of Systems Science, Chinese Academy of Sciences in 1983, and his PhD from the Institute of Systems Science, Chinese Academy of Sciences in 1986. Professor Wang has published over 25 monographs (including 10 published by Springer-Verlag) and over 200 papers in leading journals.

References

Afonso

Schuknecht

Tanzi

(2005) Public sector efficiency: An international comparison. Public Choice 123(3–4): 321–347.

Alonso

Lewis

(2001) Public service motivation and job performance: Evidence from the federal sector. The American Review of Public Administration 31(4): 363–380.

Bale

Dale

(1998) Public sector reform in New Zealand and its relevance to developing countries. The World Bank Research Observer 13(1): 103–121.

Baron

Davis-Blake

Bielby

(1986) The structure of opportunity: How promotion ladders vary within and among organizations. Administrative Science Quarterly 31(2): 248–273.

Blanchard

Shleifer

(2001) Federalism with and without political centralization: China versus Russia. IMF Staff Papers 48(S1): 171–179.

Blau

DeVaro

(2007) New evidence on gender differences in promotion rates: An empirical analysis of a sample of new hires. Industrial Relations 46(3): 511–550.

Blau

Scott

(1962) Formal Organizations: A Comparative Approach. San Francisco: Chandler.

(1996) Economic performance and political mobility: Chinese provincial leaders. Journal of Contemporary China 5(12): 135–154.

Booth

Francesconi

Frank

(2003) A sticky floors model of promotion, pay and gender. European Economic Review 47(2): 295–322.

10.

Brown

Earle

Gehlbach

(2009) Helping hand or grabbing hand? State bureaucracy and privatization effectiveness. American Political Science Review 103(2): 264–283.

11.

Brown

(2004) Human resource management in the public sector. Public Management Review 6(3): 303–309.

12.

Cameron

Freeman

(1991) Cultural congruence, strength and type: Relationships to effectiveness. Research in Organizational Change and Development 5: 23–58.

13.

Chen

Zhou

L-A

(2005) Relative performance evaluation and the turnover of provincial leaders in China. Economic Letters 88(3): 421–425.

14.

Chien

Gordon

(2008) Territorial competition in China and the West. Regional Studies 42(1): 31–49.

15.

China State Council (2007) Northeast Revitalization Plan. Beijing: China State Council.

16.

Collins

Cottey

(2012) Understanding Chinese Politics. Manchester: Manchester University Press.

17.

DeVaro

(2006) Strategic promotion tournaments and worker performance. Strategic Management Journal 27(8): 721–740.

18.

Edin

(2003) Remaking the communist party-state: The cadre responsibility system at the local level in China. China: An International Journal 1(1): 1–15.

19.

Fairburn

Malcomson

(2001) Performance, promotion and the Peter principle. The Review of Economic Studies 68(1): 45–66.

20.

Fan

(2006) China’s eleventh five-year plan (2006–2010): From “getting rich first” to “common prosperity”. Eurasian Geography and Economics 47(6): 708–723.

21.

Ferrero

(2001) Political exchange in mass party regimes and the transition from socialism. Economic Systems 25(4): 365–379.

22.

Guo

(2007) Retrospective economic accountability under authoritarianism: Evidence from China. Political Research Quarterly 60(3): 378–390.

23.

Garcia-Izquierdo

Moscoso

Ramos-Villagrasa

(2012) Reactions to the fairness of promotion methods: Procedural justice and job satisfaction. International Journal of Selection and Assessment 20(4): 394–403.

24.

Gouldner

(1957) Cosmopolitans and locals: Toward an analysis of latent social roles.I. Administrative Science Quarterly 2(3): 281–306.

25.

Groot

Van Den Brink

(1996) Glass ceilings or dead ends: Job promotion of men and women compared. Economic Letters 53(2): 221–226.

26.

Halaby

(1978) Bureaucratic promotion criteria. Administrative Science Quarterly 23(3): 466–484.

27.

Horn

(1995) The Political Economy of Public Administration. Cambridge: Cambridge University Press.

28.

Kalleberg

Van Buren

(1996) Is bigger better? Explaining the relationship between organization size and job rewards. American Sociological Review 61(1): 47–66.

29.

Kim

(2002) Executive social capital and its impacts on job promotion. Academy of Management Proceedings 1: J1–J6.

30.

Lagerkvist

(2014) The Legacy of the 1989 Beijing Massacre: Establishing neo-authoritarian rule, silencing civil society. International Journal of China Studies 5(2): 349–369.

31.

Landry

(2003) The political management of mayors in post-Deng China. The Copenhagen Journal of Asian Studies 17: 31–58.

32.

Lau

C-M

Tse

Zhou

(2002) Institutional forces and organizational culture in China: Effects on change schemas, firm commitment and job satisfaction. Journal of International Business Studies 33(3): 533–550.

33.

Lee

(1991) From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley/Los Angeles: University of California Press.

34.

Lee

(2017) An institutional analysis of Xi Jinping’s centralization of power. Journal of Contemporary China 26(105): 325–336.

35.

Lemons

Jones

(2001) Procedural justice in promotion decisions: Using perceptions of fairness to build employee commitment. Journal of Managerial Psychology 16(4): 268–281.

36.

Zhou

L-A

(2005) Political turnover and economic performance: The incentive role of personnel control in China. Journal of Public Economics 89(9–10): 1743–1762.

37.

Wei

(2010) The spatial-temporal hierarchy of regional inequality in China. Applied Geography 30(3): 303–316.

38.

Lin

(2008) Demystifying the chameleonic nature of Chinese leadership. Journal of Leadership and Organizational Studies 14(4): 303–321.

39.

Manion

(1985) The cadre management system, post-Mao: The appointment, promotion, transfer and removal of party and state leaders. The China Quarterly 102: 203–233.

40.

McLaughlin

Osborne

Ferlie

(eds). (2002) New Public Management. London: Routledge.

41.

Nee

(1992) Organizational dynamics of market transition: Hybrid forms, property rights, and mixed economy in China. Administrative Science Quarterly 37(1): 1–27.

42.

O’Reilly

Chatman

Caldwell

(1991) People and organizational culture: A profile comparison approach to assessing person-organization fit. Academy of Management Journal 34(3): 487–516.

43.

Pudney

Shields

(2000) Gender, pay, race and promotion in the British nursing profession: Estimation of a generalized ordered probit model. Journal of Applied Econometrics 15(4): 367–399.

44.

Zhang

et al . (2008) Translating a global issue into local priority: China’s local government response to climate change. The Journal of Environment and Development 17(4): 379–400.

45.

Quah

(1996) Empirics for economic growth and convergence. European Economic Review 40(6): 1353–1375.

46.

Rokeach

(1973) The Nature of Human Values. New York: Free Press.

47.

Ruderman

Ohlott

Kram

(1995) Promotion decisions as a diversity practice. Journal of Management Development 14(2): 6–23.

48.

Shih

Adolph

Liu

(2012) Getting ahead in the communist party: Explaining the advancement of central committee members in China. American Political Science Review 106(1): 166–187.

49.

Spilerman

Lunde

(1991) Features of educational attainment and job promotion prospects. American Journal of Sociology 97(3): 689–720.

50.

Tao

et al . (2012) Local officials’ incentives and China’s economic growth: Tournament thesis reexamined and alternative explanatory framework. China and World Economy 20(4): 1–18.

51.

Tsui

Wang

(2004) Between separate stoves and a single menu: Fiscal decentralization in China. The China Quarterly 177: 71–90.

52.

Tsui

Wang

Xin

(2006) Organizational culture in China: An analysis of culture dimensions and culture types. Management and Organization Review 2(3): 345–376.

53.

Walder

(2003) Elite opportunity in transitional economies. American Sociological Review 68(6): 899–916.

54.

Wang

(1981) The problem of cadres in China. China Report, May/June: 3–12.

55.

Weber

(1947) The Theory of Social and Economic Organization. Edited by Henderson A and Parsons T. New York: Oxford University Press.

56.

Wei

(2002) Beyond the Sunan model: Trajectory and underlying factors of development in Kunshan, China. Environment and Planning A 34(10): 1725–1747.

57.

(2011) The fundamental institutions of China’s reforms and development. Journal of Economic Literature 49(4): 1076–1151.

58.

Yeung

(2000) Local politics and foreign ventures in China’s transitional economy: The political economy of Singaporean investments in China. Political Geography 19(7): 809–840.

59.

Zhang

(2014) Testing social ties against merits: The political career of provincial party chiefs in China, 1990–2007. Journal of Chinese Political Science 19(3): 249–265.

60.

Zhang

Wan

(2005) China's Business cycles: Perspectives from an AD-AS model. Asian Economic Journal 19(4): 445–469.

61.

Zheng

(2009) Can the communist party sustain its rule in China? In: Lee

Kim

J-H

Woo

(eds) Power and Sustainability of the Chinese State. London: Routledge, pp.186–209.