When Western economists traveled to Eastern Europe and the former Soviet Union to advise reformers at the onset of market reforms, they consistently emphasized instituting capitalism by designing sweeping changes in the formal rules governing property rights, capital and labor markets. Such emphasis on writing and legislating new rules of economic action overlooked the realities of power and interests vested in existing institutional arrangements and longstanding personal relationships of the political élite. By contrast, the trial-and-error approach taken by reformers in China has allowed for a more evolutionary approach to economic transition (Lin, Cai and Li, 1996; Tsui and Lau, 2002). We infer from China's greater success that institutional change is driven not so much by new formal rules but by bottom-up realignment of interests and power as new organizational forms, private property rights and new market institutions evolve in an economy shifting away from state control over economic activity. In China, changes in the formal rules governing the emerging market economy have tended to follow ex post changes in the informal economic practices and the competitive environment. A parallel process has occurred in Eastern Europe and the former Soviet Union, where following failed attempts at designing capitalism in one fell swoop, a more incremental bottom-up approach tacitly replaced the big-bang approach of top-down legal and regulatory changes, as political and economic actors grappled step-by-step with the concrete problems of their emergent market economy (Elster, Offe and Preuss, 1998; Sachs and Pistor, 1997; Stark and Bruszt, 1998).
Market transition theory builds on the basic premise that state socialist redistribution and a market economy entail fundamentally different institutional mechanisms facilitating, motivating and governing economic behavior (Nee, 1989b, 1996). The theory's propositions are formulated in a series of articles that specify the repertoire of mechanisms involving both political and economic actors. The theory advances three core arguments about the nature of institutional change in departures from state socialism:
(1) Institutional innovations in economic reform are initiated and implemented by the state, and that the course of subsequent institutional change arises from the interaction between the revenue-maximizing interests of political actors and constraints imposed on rulers by organizational and economic actors (Nee, 1989a, 1992, 2000; Nee and Lian, 1994).

(2) Instituting market exchange as the dominant coordinating mechanism for an economy involves a de-institutionalization of core features of state socialist redistribution, contributing to a decline in the relative advantages of the political élite.
(3) Institutional change promoting reliance on the market mechanism alters the structure of incentives through changes in the informal and formal rules governing property rights, expanding pay-offs to market-oriented performance for economic agents and firms (Nee, 1989b, 1992, 2000; Nee and Su, 1996).

Clearly, market transition theory does not rule out an independent, ongoing, causal effect of the state in shaping the post-Communist stratification order in urban China. Specifying a political economy approach that posits two sources of causal mechanisms – state and market – shaping the stratification order in departures from state socialism, it claims:
As long as major productive assets are owned or controlled by the state, officials will pursue power-conversion the more political capital is diminished relative to the appreciation of economic capital . . . Political capital is likely to persist as a strong predictor of advantage in the sectors of the transition economy that are state-owned. It is also likely to persist as a stronger predictor than human capital where structural holes at the boundaries of the state and nonstate sectors of the economy provide opportunities for political actors to serve as middlemen in economic transactions. (Nee and Cao, 1999, pp. 806–807)
In sum, the intact ruling power of the Communist Party combined with the concentration of state-owned productive assets in urban China provide favorable conditions that perpetuate the power and advantages of the political élite, both through conventional redistributive means and through market-based opportunities to convert positional power into private financial advantage.
Walder's (2004) political theory takes a different tack from market transition theory's focus on the causal effects of market penetration and state intervention in departures from central planning. Walder focuses solely on political variables centering on régime change (or not) and constraints (or not) on asset appropriation by the political élite to explain the course of market transitions in post-Communist economies. His state-centered approach builds on his earlier claim that the emergence of a market economy per se has no causal consequences for the constitution of the post-communist stratification order (Walder, 1996). In specifying the political variables shaping the post-Communist stratification order, his recent work offers an important advance over the prior formulation of causal mechanisms centered on the action of political élites and the state (Bian and Logan, 1996; Nee, 1992, 1996; Nee and Cao, 1999; Nee and Lian, 1994; Parish and Michelson, 1996; Rona-Tas, 1994; Zhou, 2000). However, we disagree with his claim that the emergence and expansion of a market economy has no independent causal effect on the allocation of power and income.
In order to adjudicate empirically the competing claims of market transition theory and Walder's state-centered theory, we examine the pattern of earnings inequality in urban China. Walder's theory predicts that given the concentration of party power in urban China and the new opportunities for rent-seeking and profit-making provided by economic reform, the cadre élite can be expected to have significantly augmented their power and privileges relative to non-party economic actors, including skilled employees, professionals, managers and private entrepreneurs. We argue that the emergence and expansion of a market economy in urban China has similar effects on the stratification order to those found in tests of the theory using data from rural China. But if Walder's argument that the shift to markets per se has no consequences for the allocation of power and income is true, hypotheses adapted from market transition theory to the urban context will not find empirical confirmation. Market transition theory does not advance the 'notion that opportunities for old regime elites eventually decline with the extent of reform' as portrayed by Walder (2004, p. 913). Rather, it is a claim that the rise of a market economy favors economic actors – skilled workers, private entrepreneurs, professionals and managers – relative to political élites. The expansion of a market economy unleashes a repertoire of mechanisms causing the demise and transformation of institutional arrangements upholding state monopoly control over the allocation of resources.
Although our empirical study of urban China might not definitively resolve this debate, we would be satisfied if we can demonstrate that the emergence and growth of an urban market economy unleashes causal mechanisms, altering the allocation of power and income giving rise cumulatively to transformative change in the post-communist stratification order, not accounted for by a theory that only specifies political variables. Below we elaborate three such mechanisms linked to the rise of a market economy, which we claim causes a relative decline in the significance of positional power and political capital in the course of market transition in urban China. This claim does not rule out augmentation of opportunities enhancing the power and privileges of members of the Communist élite, but instead insists that the expansion through market-driven economic growth of opportunities for economic actors – i.e. skilled employees, professionals, managers and entrepreneurs – is likely to outpace in relative terms the gains secured purely through rent-seeking defined as allocation of rewards and income not tied to increases in marginal productivity.
In modern urban societies, most people achieve economic well-being through employment, suggesting that abstract concepts used to characterize macroeconomic institutional arrangements, e.g. redistribution and market, should be elaborated in terms of how they substantively affect individuals' employment choices and their associated pay-offs. Just how have mechanisms of institutional changes accompanying the rise of a market economy brought a new dynamic to the urban employment relationship?
The Causal Effect of Higher Marginal Productivity of the Private Enterprise Sector
A major component of the Chinese economic reform is the change of the structure of property rights, evidenced in the diversity of ownership forms in the economy. In the Maoist era, virtually all productive assets and capital were public property. Most organizations were classified as either state-owned or collective, and private ownership had become a negligible fraction after waves of nationalization following the founding of the PRC. State ownership possessed a more central position in the national economy, while its collective counterpart was relegated to a peripheral role (Whyte and Parish, 1984). As a result, although both were integrated in the state planning system, state-owned organizations enjoyed better access to state resources and at the same time were subject to tighter governmental supervision.[1]
Starting in the early 1980s, the Chinese government began to allow limited growth of non-state ownership forms. The original plan was not to encourage privatization of state-owned assets but to allow private enterprise to generate competitive pressure and thus stimulate performance among public enterprises. However, by the mid-1990s, private and hybrid property forms had become an increasingly significant and perhaps the most dynamic component of the national economy. According to the State Statistical Bureau of China (1995), in 1994, private and hybrid organizational forms accounted for 25% of the nation's total industrial output, compared to 0.74% in 1982. During the same time period, the proportion of the labor force employed in the private/hybrid sector had risen from 1.29% to 13.8%. Thus we calculate that labor productivity in the private/hybrid sector had grown at a rate over three times as fast as in the public sector. Take the example of Guangzhou, a large city located in southern China. In 1995, labor productivity in Guangzhou averaged 51,993 yuan in the private/hybrid sector, compared to 36,865 for state enterprises and 18,705 for collective enterprises (Guangzhou Statistical Bureau 1996, p. 128).
The private/hybrid sector is mainly comprised of four types of economic entities: self-employment businesses (getihu), domestic private firms, Sino-foreign joint ventures and branch companies solely owned by foreign capital. Among these four types, self-employment poses the least ideological challenge to the Communist régime because it involves little or no labor exploitation. The Chinese government originally encouraged it, partly as a solution to increasing urban unemployment in the early to mid-1980s. Domestic private firms, on the other hand, suffer from a less comfortable fit with the Marxist doctrine. Officially, they are classified under the category of 'privately operated enterprise' (siying qiye), a label that does not declare outright their distinctive ownership. As a result of financial constraints and governmental protection of public ownership in industries of strategic importance, both self-employment and domestic private firms tend to concentrate in commerce, service and light industries.
Two other new property forms – Sino-foreign joint venture (zhongwai hezi and hezuo) and foreign company (waiqi) – resulted mainly from China's 'open-door' policy, in which the government made foreign investment an anchor of its overall developmental plan and enacted a series of preferential policies to attract international capital. By 1995, international capital accounted for 11.2% of the nation's total investment and 12% of industrial output (State Statistical Bureau of China, 1997, p. 33). Most joint ventures are joint stock companies founded on the basis of pre-existing public enterprises. The Chinese partners usually contribute the land, the buildings and other forms of fixed investment, and the foreign partners provide one or more of the following: technology, equipment, financial capital and brand name. This type of joint ownership often leads to hybrid organizational structures that reflect both foreign investors' financial interests and the concerns of local governments with maintaining employment and social stability. Foreign companies, on the other hand, are essentially overseas branches of foreign corporations that seek to take advantage of the business opportunities in China. Operation and management are structured in parallel to that in the parent companies. Employees are recruited on a competitive basis, and enjoy wage levels far above that offered by domestic employers.
Employment in this expanding private/hybrid sector has become an institutionalized alternative to the traditional career paths under the state socialist régime. In the early days of the reform, a majority of the jobs in the private/hybrid sector were manual jobs in very small firms in labor-intensive industries. Though undesirable in many respects, these jobs nevertheless constituted valuable opportunities for unemployed urban residents and peasant workers who migrated to urban areas in search of a better life. As the private/hybrid sector expanded in both size and scope, more professional and technical employment opportunities began to emerge, and it has become a trend that competent public sector employees leave behind their 'iron rice-bowls' to participate – a career change metaphorically termed 'plunging into the sea' (xiahai) by urbanites.
How do the emergence and expansion of the private/hybrid sector affect overall inequality? If the private/hybrid sector generated mostly undesirable low-paying jobs, which in turn attracted only people with no prospects in the state-controlled sectors of the economy, then we should not expect any significant change in reward distribution. The reason is simple: the least privileged remain at the bottom of the pyramid, while the rest are unaffected. However, such a scenario is inconsistent with the more than two-decade-long record of rapid growth and competitive dynamism of the private/hybrid sectors in China. So long as the private/hybrid sector as a whole can achieve higher marginal productivity relative to state-owned firms, and state asset appropriation is not dangerously predatory, new opportunity structures for economic mobility are bound to continue shifting incentives in the transition economy, eroding the previously unchallenged dominance of the state-controlled economy. The state may be able to maintain a comparable or even higher wage structure in the public sector artificially in the short term (Zhou, 2000). But if state-owned firms are less productive, as a result of inefficient property rights and weak corporate governance, than it is difficult to do so in the long run in light of the higher marginal labor productivity and robust economic performance of private/hybrid firms, which include foreign branch firms of multinationals. Thus:
Hypothesis 1: In equilibrium, the higher marginal productivity of workers in the private/hybrid firms relative to state-owned enterprises causes a growing wage gap favoring employees in private/hybrid firms relative to employees in state-owned firms and organizations.
Firm-based Competition in Labor Markets for Skilled and Semi-skilled Labor
A second important institutional change stemming from economic reform is increasing competition by firms for skilled and semi-skilled laborers in emergent labor markets and the end of the state's monopoly on labor allocation. In pre-reform China, labor mobility was under strict bureaucratic control. The government not only monopolized job provision, but also dictated all personnel transfers across work organizations, or work units. Without permission from the government's labor department, work units could not establish employment affiliation (renshi guanxi) for any job candidate or accept his/her personal dossier. Voluntary job change was in general rare, as it was not encouraged by the government and required multiple stages of petition and approvals from current work unit, labor department and receiving work unit. The process could become even more complicated when change in the locality of household registration was involved. As a result of these constraints, many workers remained in one work unit until they retired, while others were forced to relocate to peripheral areas where the government perceived their labor could be best utilized.
Bureaucratic control over labor mobility has loosened considerably during the reform period. Officially, employment affiliation is now defined by a labor contract between the two parties. Product markets replace the work unit's ration system, which used to distribute grain and consumer goods. Personal dossiers can be registered under the newly created talent exchange centers, and more and more employers no longer enforce restrictions related to the household registration system. These changes together greatly weakened individuals' dependence on their work unit, which Walder (1986) argued was the basis of a neo-traditionalist organizational culture of Communist Party patronage. Voluntary job change thus has become not only feasible but also fairly commonplace, especially in large cities and coastal regions. Although some government regulations over labor mobility continue to exist, they are often either reduced to record-keeping procedures or simply bypassed by employers.
The emergence of demand-driven competition for labor implies a fundamentally different logic in matching individuals with positions. On the one hand, as employment opportunities become available through labor market channels, individuals with skills are in a better bargaining position. Competition for skilled and semi-skilled labor thus makes it increasingly difficult for state-owned organizations and other employers to acquire or retain valuable employees with less than competitive compensations. On the other hand, to the extent that their survival and success is determined by their employees' performance, employers also become more selective when recruiting new workers and renewing labor contracts. For instance, job interviews – virtually non-existent in pre-reform China – are now common. Professional and managerial positions in the private/hybrid sector usually require not only a college diploma from relevant disciplines, but also competitive scores in specifically designed competency exams. In sum, instead of being dictated by the state, labor mobility in the reform period is increasingly based on the mutual choice of potential employers and employees.
How does competition for skilled and semi-skilled workers affect the structure of incentives and relative earnings? Our analysis suggests that market-coordinated labor mobility leads to more meritocratic distribution of material compensation. Returns to human capital should increase, as individuals with talent and expertise can now secure a high salary through bargaining and voluntary job changes, while those with little are likely to be forced to settle for a minimum. As to political credentials, e.g. Communist Party membership, we expect an overall decline of significance because they are unlikely to be rewarded in settings where party loyalty and political activism fail to contribute to employer success. That is:
Hypothesis 2a: The more developed the competition for skilled and semi-skilled labor in labor markets, the higher the returns to human capital.
Hypothesis 2b: The more developed the firm-driven competition for labor in labor markets, the less advantage to party membership.
The effect of labor markets should also be more pronounced in the private/hybrid sector. Most private/hybrid businesses were established in the reform period, and employees thus were more likely to have acquired their jobs through labor markets, instead of state allocation. Hence:
Hypothesis 3a: The higher marginal productivity of private/hybrid firms allows for higher returns to human capital in these firms than in public-sector firms.
Hypothesis 3b: Advantage to party membership is smaller in the private/hybrid firm than in firms in the public sector because labor productivity is more important in the private enterprise sector whereas party membership is more salient in the public sector.
Rising Meritocracy in Public Firms
Despite rapid growth of the private/hybrid sector, public ownership has remained dominant as an organizational form in China's industrial economy. By 1994, state and collective enterprises still accounted for 75% of the nation's industrial output, and over 85% of the urban labor force worked in the public sector. Reward allocation within public-ownership organizations thus has important implications on the overall structure of inequality. This leads us to consider a third mechanism – the strengthening of meritocracy in public enterprises. In particular, we highlight the difference in the institutional environments for non-profit and for-profit organizations in the reform era and examine how their growing resource dependency on markets affects reward allocation.
As in classic accounts of state socialism, both non-profit organizations and public firms in pre-reform China were under strict control of the overseeing government jurisdictions and party branches (Schurmann, 1968). Both types of organizations depended entirely on state budget and allocation of other resources for organizational inputs; production and operation were carried out according to state policies or plans; and outputs and revenue were channeled to and then redistributed by the state. Consequently, the distinction between non-profit organizations and public enterprises mainly reflected functional differences (cf Hansmann, 1987), as both were governed by the same state socialist redistributive system.
Such an institutional environment imposed three major goals. First, as extensions of the state apparatus, these organizations were required to secure the political conformity of their employees. Second, as functional subsidiaries, they were required to perform functional duties, e.g. producing certain amounts of goods and services. Third, as embodiments of the Communist ideology, they were required to ensure the welfare of the employees. Given strict control by the Chinese Communist Party/state, these goals became highly institutionalized, and organizations were pressured to pursue all of them simultaneously in order to avoid severe sanctions (Bian, 1994; Schurmann, 1968; Walder, 1986).
This goal structure created multiple grounds on which rewards, both material and symbolic, were distributed. In his study of the pre-reform Chinese workplace, Walder found that both on-the-job performance and successful demonstration of political activism were rewarded (1986, pp. 132–140). Decisions on promotion and wage increases were made jointly by the management and party officials, who took both political activism and productivity into account. In fact, a crucial step in determining promotions was the so-called 'political inspection' (zhengshen). Only those politically qualified could pass onto the next stage, where competence and expertise became relevant factors. On the other hand, the strong influence of Communist egalitarianism substantially limited the magnitude of differentials in compensation (Walder, 1995), and negative sanctions were more often symbolic than material. For instance, almost all workers in state enterprises were permanently tenured. Firing was rare in general, and even rarer for inadequate on-the-job performance unless it could be interpreted as a sign of political non-conformity (Walder, 1986, p. 143).
Since its onset, a central theme of China's market transition has been to transform public enterprises from passive productive plants into relatively independent economic actors that could actively respond to prices and other market signals (Naughton, 1995). In fact, most urban reform policies implemented in early 1980s had been centered on the so-called 'separation of government and enterprises' (zhengqi fenkai). The state no longer guarantees all the material inputs and sales of products. Public enterprises thus must rely on market exchanges to acquire material supplies and realize profit. Meanwhile, state budgetary investment is replaced by bank loans, and instead of handing in all their revenues, public enterprises now can retain the surplus after fulfilling their tax obligations. Commodity prices, which were originally mandated by the state, were liberated in a step-by-step manner, and managers were also given greater power and freedom with regard to managerial decisions. Changes in the legal environment further make bankruptcy an imminent possibility for firms unable to maintain financial solvency.[2]
These reform policies significantly transform the organizational goal structure for public enterprises. To the extent that organizational survival and success depend on markets, public firms now face stronger incentive to reward on-the-job performance and relatively weaker ones to promote political conformity or ensure employee welfare. Public firms thus are likely to devise new compensation schemes and promotion criteria in favor of job qualification and performance. If true, this would suggest that reward allocation in public firms will become increasingly meritocratic and less virtuocratic (Cao, 2001; Shirk, 1982). Following the promulgation of the Company Law in 1994 which allowed public enterprises to incorporate, the shift to meritocracy is likely to be especially pronounced in companies listed in the new stock exchanges of China in which private ownership shapes corporate culture. In their study of listed companies in the Shanghai Stock Exchange, Opper, Wong and Hu (2002) show that the greater the share of private ownership, the less the Communist Party influences decision-making in the firm, including personnel decisions.
In contrast, China's reform measures have exerted relatively little impact on the institutional environment for non-profit organizations, in the sense that the government continues to play a dominant role in monitoring their operation and providing key organizational resources. With resource provision guaranteed by the state, organizational survival and success are not directly affected by the emerging market economy. Because of this, we expect that non-profit organizations are more likely than their for-profit counterparts to retain the pre-existing distributive practices. Consequently, returns to human capital should be higher in public for-profit firms, whereas party loyalty and political activism are rewarded more in non-profit organizations.
Although organizations are generally considered capable of adaptation, instant adjustment is far from realistic due to inertial forces inside. Organizations often retain features acquired at the time of founding for an extended period of time, as Stinchcombe (1965) observed. Fixed investment, information constraints, internal politics and risk aversion in decision-making all contribute to organizational inertia (Hannan and Freeman 1989, Chapter 4). In the case of China, public enterprises established before the reform tend to have developed organizational routines corresponding to the old governance structure and therefore cannot be expected to make adjustments immediately after the implementation of reform policies. Over time, however, more and more public enterprises should be able to overcome their internal inertia and accomplish systematic adjustments to the new institutional environment. From a cross-sectional point of view, the extent of organizational adaptation among public enterprises is in part a function of the time at which reform policies were implemented at the locality. This implies more pronounced sectoral distinction between public enterprises and their non-profit counterparts in places where reform assumed momentum earlier. Hence:
Hypothesis 4a: In localities where reform policies were implemented early, returns to human capital are higher in public-sector firms than in non-profit organizations.
Hypothesis 4b: In localities where reform policies were implemented early, Communist Party membership is rewarded more in non-profit organizations than in public firms.
Hypothesis 5a: In localities where reform policies were implemented late, returns to human capital are comparable in public firms and non-profit organizations.
Hypothesis 5b: In localities where reform policies were implemented late, Communist Party membership is rewarded similarly in public firms and non-profit organizations.