The Interaction of the State and the Market. in a Developing Transition Economy: the Experience of China

Transcription

1 Asia-Latin America Cooperation ALAC An ECLAC-UN, IDE/JETRO-Japan and IE/UFRJ-Brazil initiative International Seminar Promoting Growth And Welfare: Structural Changes And The Role Of Institutions In Asia Santiago, Chile and Rio de Janeiro, Brazil April 29 - May 03, 2002 The Interaction of the State and the Market in a Developing Transition Economy: the Experience of China Chunlin Zhang Senior Enterprise Restructuring Specialist The World Bank Office Beijing czhang2@worldbank.org First Draft April

2 Abstract This paper assesses how the particular form of interactions between the state and the market may have had its impact on China s overall success in the past two decades in reform and development. In doing so, the question of development economics, i.e., the role of the state and the market in development, is discussed in a context of transition economics. Some key components of China s reform experience are examined to illustrate the way the state interacts with market forces, including ownership transformation of people s communes and state owned enterprises, regional decentralization, price reform, financial sector reform and capital market development, and bankruptcy regime. In interacting with market forces in some transition institutions, the state seemed to have played positive roles in ensuring political and social feasibility of reform measures, supporting the market to achieve initial improvement in efficiency, and maintaining the momentum of reform. 2

3 I. Introduction The role of the state and the market in economic development has been constantly a subject of debate. While the debate is far from concluded, it seems clear that at least in the field of economic policy making, the meaningful question is not one of state or market. Instead, it is much about how the two can interact in such a way that economic development is promoted. From this point of view, China s experience of transition and development in the last two decades provides an interesting case to study with. China differs from most other transition economies in that it is still a developing country, and from most developing countries in that it has experienced a transition from a Soviet-style central planning economy to a market economy. Since the transition in China has been a process of crossing the river by touching the stones, in which both the old and new mechanisms are partial but have a role to play, it is interesting to see how the state and the market interact during the transition process and how such interactions impact on development. China s modern history started in 1840 when the door of the closed empire was broken up by British warships. In the following one hundred years or so, China suffered miserably from great internal strife and foreign exploitation. When the Chinese Communist Party (CCP) succeeded in its revolution and came to power in 1949, China s share in global GDP was only about 5%, down from 33% in 1820 (Dahlman and Aubert, 2001, p1). Blaming private ownership, market forces and international capitalism for the hardship, the CCP decided to rebuild the country by adopting a Soviet-style central planning model based on public (state and collective) ownership. From , the whole urban economy (accounting for around half of GDP and less than 15% of population) was nationalized, while the rural economy was collectivized. Market mechanism was no longer allowed to exist in any significant sense. A wide range of central planning mechanisms was transplanted from the Soviet Union, which also assisted China to start its post-war industrialization with 156 key investment projects in early 1950s. Despite the cooperation with Soviet Union in early 1950s, overall international environment was such that the country s door was closed again to most parts of the world, from both inside and outside. In terms of the role of state and market in development, China went to one of the extremes: only state, no market. The implementation of this model, however, was never straight forward. Mainstream communists led by late Chairman Mao Zedong had to engage in political and ideological campaigns year after year to crackdown oppositions and fight against human nature. The Cultural Revolution in marked the extreme of such efforts. The Disastrous Ten Years awakened those communists who put the welfare of the people and development of the country before ideology, such as Deng Xiaoping, the paramount leader of China in It was recognized that economic development was not going to be achieved before the traditional communist ideology is abandoned to allow a role of market forces and private initiatives. The idea that later guided China s reform 3

4 was expressed by Deng in late 1970s using a thousand-year old Chinese proverb: a cat is a good one so long as it catches mice, never mind it is white or black. The reform and opening up launched by Deng in 1978 has proved a phenomenal success. As Stiglitz (1999) noted, China s growth record since 1978 is nearly unparalleled in human history. As shown in Figure 1 and Figure 2, during a period of 23 years since 1978, China s GDP increased by more than 26 times, from RMB362.4 billion (USD43.66 billion) to RMB billion (USD billion). Per capita GDP in 2001 reached RMB7537 (USD908.07), 19 times higher than that of 1978 (RMB379 or USD45.66). More importantly, China s long period of rapid growth has been achieved without making dramatic inroads into poverty. It managed to reduce the number of people living in absolute poverty (defined as living on USD1 a day or less) by over 200 million in a period of two decades (Stiglitz, 1999). The improvement of living standard of ordinary Chinese has been unprecedented in Chinese history. In 2000, 51% of China s GDP was produced by secondary industry, 33% by tertirary industry, and 16% by primary industry. In terms of employment, however, primary industry accounted for 50%, secondary industry and tertiary industry shared the rest (22.5% and 27.5%, respectively). China is clearly on its way in the transition from an agricultural economy to an industrial and service economy. The ownership structure of the corporate sector is shown in table 1. The roughly equal shares of state owned enterprises (SOEs) sector and private sector highlighted the fact that China is about half way in its transition from a central planning economy to a market economy. The extent to which China has integrated into the world economy is indicated by a 44.5% ratio of total value of imports and exports to GDP in 2000, a sharp contrast to a ratio of 9.8% in Totally USD40.7 billion foreign direct investment (FDI) was utilized in China in 2000, equivalent to 10.3% of its total fixed assets investment. Table 1: Ownership Structure of China s Corporate Sector Ownership type Description Share in 1998 GDP (%) SOEs Industrial and other enterprises that are wholly owned by the state and supervised by either central 37 Collective enterprises Domestic private enterprises Foreign funded enterprises Joint stock companies or local government departments Collective Township and Village Enterprises (TVEs) funded by township or village government and urban collectives funded by urban communities Privately owned companies, individual proprietorships, self-employed businesses, some of them are created and owned by farmers and called TVEs as well. Mainly joint ventures and foreign enterprises 6 Typically jointly owned by the state, other enterprises of various ownership, and individuals, most of them were transformed from SOEs

5 Source: IFC, most of them were transformed from SOEs China s success is undoubtedly a success of the market: the reform and opening up have been a constant movement from the state to the market. Indeed, since the pre-reform system was at one extreme of state-market spectrum, the reform had to be marketoriented. The important fact, however, is that China is not the only economy that has undertaken such a movement, but seems to have been more successful than others in terms of economic development. This suggests that the direction of the movement is not the only determinant of success. In the past decade or so, a huge body of literature has built up as economists tried to find out what are the others. For example, many tend to attribute China s success to its gradualist approach of reform, while others emphasize difference of initial conditions between China and other transition economies. Without intending to be involved in this ongoing debate, this paper will try to assess how the particular form of interactions between the state and the market may have had its impact on China s overall success. In doing so, the question of development economics, i.e., the role of the state and the market in development, is discussed in a context of transition economics. Specifically, some key components of China s reform experience are examined to illustrate the way the state interacts with market forces, including ownership transformation of people s communes and state owned enterprises (section II), decentralization(section III), price reform (section IV), financial sector reform and capital market development (section V) and bankruptcy regime (section VI). Key findings are summarized in Section VII. II. Partial Transformation of Ownership At the outset of reform, the Chinese economy was dominated by public ownership. In the rural economy, land and other production assets were owned by People s Communes (based in townships) and Production Brigades (based in villages). Each People s Commune (PC) consisted of dozens of Production Brigades. Party secretaries and managers of Production Brigades were selected from local peasants by PCs, while that of PCs themselves were appointed by county Party organs and governments. Production Brigades organized production activities according to government plan received from PCs, sold their outputs to and bought their inputs from distribution entities run by PCs or other levels of the government. This system enabled the state to fully control the rural economy, in particular, to set both amounts and prices of agricultural products procurement to support its overall development goals. In the urban area, state owned enterprises dominated. By 1978, SOEs accounted for 80% of China s industrial output. Unlike a firm in the market economy, a pre-reform SOE assumed multiple functions. It was foremost a political institution, a cell of the party-state that served as the carrier of the Party s presence at the grassroots level. It was also a level of state administration, exercising administrative control on behalf of the party-state. In terms of economic and social functions, a SOE was nothing more than a production factory of the super company owned and managed by the central or local planning 5

6 authorities, and a state agency providing social welfare to employees. SOEs had little independence and no separate legal identity. They received production quotas, guaranteed outlets for products and the necessary resources from the budget to implement the plan. Whatever profits were made had to be remitted to the state budget. As the basic cell of the command economy, the SOE was responding to administrative orders. There was a direct state control over factory employees salaries, together with permanent employment. (Tenev, Zhang and Brefort, 2002). The systems of PCs and SOEs were designed to facilitate state planning and suppress the role of market forces. Ubiquitous state control over economic activities in PCs and SOEs meant practical impossibility for them to respond to price signals in any significant sense. And incentive structures were such that they could not be profit-driven suppliers in the market. It follows, therefore, that market forces could play no role before ownership transformation takes place in PCs and SOEs. The interesting question is what kind of ownership transformation to take place. The simple answer would be privatization. Private ownership would certainly be sufficient to allow market forces to replace the state. However, political and social constraints that prevailed in China in the 1980s dictated otherwise. Since a complete privatization was simply not acceptable and feasible, China was thus forced to search for compromise, i.e., the kind of ownership transformation that allows a larger role of market forces on one hand, and is politically and socially feasible on the other. The reform that did take place in PCs and SOEs in the 1980s was such a kind of ownership transformation that might be conveniently viewed as partial and informal privatization. Household Contract Responsibility System. In rural China, ownership transformation was implemented in early 1980s through a household contract responsibility system, in which each rural household signed a contract with its Production Brigade to gain the rights of using certain pieces of collective land for certain years, and undertook to deliver certain amount of products to the state procurement system. This reform restored traditional household farming after two decades of collective ownership and PC system. In the sense of residual claim and residual control, the two key components of ownership as defined by Grossman and Hart (1986), agricultural production was privatized. Private farmers were allowed to decide on what to produce and how, and retain whatever was left after costs, taxes, and contractual contributions are deducted from total output. This was in sharp contrast with the PC system, which was characterized as one of eating rice in a shared pot in which it makes no difference if one works harder or not. Given the fact that agricultural production in China was still characterized by heavy labor inputs with traditional backward technology, such a change in incentive structure was sufficient to generate enormous increase in output and income. In a matter of four years, China s agriculture recovered from stagnation and reached historically high level of grain output in This allowed a gradual liberalization of agricultural prices and sharp increase in investment of farmers in industries, which led to the phenomenal rise of township and village enterprises (TVEs). While it is widely noticed that China s economic growth is more attributable to new entry of non-state firms, of which TVEs are the most important, than reform of existing firms, the rise of TVEs was clearly impossible without the success of agricultural reform. 6

7 However, despite the success, the privatization of agriculture was, and is still partial: land has not been privatized. And this is not insignificant. Without legally protected private ownership of land, under-investment by farmers in their lands has been an unsolved problem for over two decades. The government tried to alleviate the negative impact by extending the contracts to 15 to 30 years, without visible success. Since the population of villages are changing, the government has been forced to redistribute land use rights among households periodically to reflect the demographic changes, which made it even more difficult to encourage long term investment in land. On the other hand, the absence of a market for land prevented free flow of land among farmers, which tends to prevent technological and structural improvement from taking place in private farms. (Wu, 1999, p131-32) However, these distortions and inefficiencies were not large enough to offset the positive impact of the reform. In addition, a rigid and government controlled distribution of land use rights served to limit the scope of competition and its impact on income distribution by preventing concentration of land ownership to a small number of households, which happened frequently in the Chinese history. Politically, privatization without changing land ownership has never been regarded as privatization even by the most conservative fractions of the society, and therefore contributed to political stability. Managerial Autonomy and Earning Sharing. China s SOE reform has been the least successful part of its reform program. In 2000, there were SOEs in China, of which 51% was losing money according to official statistics. However, given the dominance of SOEs in the urban economy at the beginning of the reform, and the fact that the pre-reform SOE system did not allow a role of market forces in any significant sense, it is also beyond any doubt that no reform could have succeeded had the SOE reform been a complete failure. So, how did the seemingly unsuccessful SOE reform contribute to the overall success of the reform? China s SOE reform in the 1980s was characterized by managerial autonomy and earning sharing between the state and the enterprise, which essentially meant managers and workers, or insiders, of SOEs. This reform strategy is better known in China as delegating power and conceding prifit (fang quan rang li). It was developed with the political and social constraint that no privatization was acceptable. However, the strategy was also derived from the diagnosis that over-centralization of decision making power and weak incentives of SOE managers and workers were responsible for their poor performance. It was therefore believed as essential to transfer more decision making power to managers, and allow managers and workers to share the economic outcomes of their efforts with the state. In terms of changes in incentives, this was of the same nature as the contract responsibility system in the rural economy: part of the residual control rights and cash flow ownership rights were privatized. The start of the reform was marked by an experiment of enlarging enterprises autonomy carried out by Sichuan Provincial Party Committee in October 1978 in six SOEs. In May 1979, six central government ministries carried out their own experiment in 8 large SOEs, which led to a State Council decision in July to enlarge managerial autonomy of all SOEs. Together with the enlargement of managerial autonomy was a 7

8 profit retention scheme introduced in July 1979 by the State Council. All SOEs which were independent accounting units and profitable were eligible to adopt the scheme. A fixed retention rate was set for three years based on parameters of past financial record, and the retained profit was supposed to be split into three envelopes (later known as three funds ): production development fund for re-investment, employees welfare fund for public welfare of employees, and employees reward fund for bonus and other rewards. Although profit retention was intended to be only a limited experiment, it spread rapidly, and by the mid-1980, 6,600 state-owned industrial enterprises, accounting for 60 percent of total industrial output and 70 percent of total profit of industrial SOEs, had instituted some form of profit retention. In late 1986, the government made further decision to implement contract responsibility system in SOEs. By the end of 1987, 80% of SOEs adopted one or another form of contract responsibility systems. The specific terms of the contracts varied substantially from case to case. The most commonly used ones are the following, expressed as formula of total remittance requirement: a base amount + a sharing ratio x (total profit realized target profit) a base amount x (1 + a progressive annual growth rate) a base amount fixed for certain years a fixed target of loss reduction two guarantees and one linking : enterprises guarantee a target amount of profit and tax remittance, guarantee the implementation of technical innovation projects, and their total wage bill can be linked with total profit and tax. SOE managers were given greater control over enterprise operations in return for meeting profit remittance targets. Many contracts also gave the enterprises greater autonomy over sales and permitted managers to grant employees bonuses and hire contract workers. In view of the terms of contracts, the contract responsibility system is an extreme form of the managerial autonomy and earning sharing approach. Subject to the legal ownership of the state and the desire of government agencies for power, contracting responsibility system probably maximized the independence of SOEs from their owner in terms of managerial autonomy and residual claim. To the extent that ownership of a firm can be defined as the combination of residual claim, or cash flow rights, and residual control, what happened in Chinese SOEs in the 1980s was indeed privatization. However, it was again partial and informal. It was partial in the sense that what was privatized was only part of the cash flow plus a subset of control rights; only in rare cases did insiders manage to gain the entirety of cash flow rights and control rights. It was informal in the sense that the cash flow rights and control rights received by insiders were granted by the other player of the game, the government, through administrative order or contract. The division of cash flow rights and control rights between the state and insiders has never been explicitly defined by a law as a division of legal ownership rights. In legal terms, it is still the state that owns the entirety of the enterprise. Reflecting this fact, no matter how much of cash flow and control insiders were allowed to share with the state, the state always had the power to decide 8

9 with whom to share, i.e., the state has firmly retained the power of appointing and dismissing managers of SOEs. This partial and informal privatization may have been efficiency enhancing at least in one sense. Suppose there was a strong demand for certain product, and it was surely profitable for factories to increase their production because of the potentially high price generated by short supply. Without earning sharing and managerial autonomy, managers and workers of factories would never have the incentives to increase their production, because it would add nothing to their private income under the sharing big rice pot system. And even if they had such incentives, they were handicapped in that they did not have the rights to make the decision and the access to the needed inputs. The result could be nothing but the continuation of shortage economy and loss of income growth. The reform of earning sharing and managerial autonomy made it possible to avoid such efficiency loss. This is probably the most important success of China s SOE reform in the 1980s. SOEs could have performed much worse in the competition with non-state enterprises, and economic growth in China could have been much slower given the large share of resources occupied by SOEs, had the original system been kept intact. However, this approach could not have been an end of SOE reform because it was seriously flawed in some fundamental aspects. First, it could provide incentives for managerial performance only in an imperfect way. Managerial performance can not be observed directly. And commonly available indicators such as profit and sales are not a sufficient statistic. They reflect many other factors in addition to managerial performance, and can be manipulated. Second, this strategy could privatize only profit, not loss. The primary reason for this failure is the wealth constraint of insiders. Various attempts were made to alleviate this problem in late 1980s to improve the contracting responsibility system. For example, managers and workers were required to provide deposit before being granted a contract. However, it was never sufficient to free the state from a situation of head I win, tail you lose, because the potential loss of state assets that could be brought about by actions of insiders was far too large in comparison with their personal wealth. This resulted in a mismatch of residual claim and residual control when insiders were granted substantial control rights, and encouraged insiders to undertake riskier actions that a rational private owner of the enterprise would not. These problems point to the main dilemma that reformers faced. If the state kept granting autonomy and sharing profit with insiders, it could lead to insiders control of a considerable degree, and in effect an informal and partial privatization of control rights and profit and a complete socialization of loss. On the other hand, if the state wanted to put insiders under control, the only means it could employ was still the old system, i.e., re-centralization, or reinforcement of administrative control, because legal ownership was not changed. The actual course of reform in China during the period of was indeed one of hitting and reflecting in a corridor of two hard walls: the old system of state run enterprise on one side, and insider control on the other. This gave rise to a corporate governance structure in most SOEs that was characterized by a combination of insider control and administrative intervention. 9

10 The informal and partial nature of the privatization of ownership rights, and therefore the potential economic and social value of putting insider control under control, has led to a fundamental failure of the SOE reform in the 1980s: the reform has basically failed to address the problem of politicization (in a similar sense as Boycko, Shleifer and Vishny 1993) or the multiple functions of SOEs. Despite all kind of reform efforts to enlarge the autonomy of managers, SOEs remain cells of the party-state economy. Since the managers of SOEs are appointed and dismissed by the Party as cadres, it is unrealistic to expect them to concentrate on profit maximization and economic efficiency, even though they themselves are sharing economic profit with the state. The failure to depoliticize SOEs helped maintain an environment in which even those who really have entrepreneurial talents find it difficult to survive until they train themselves to become bureaucratic entrepreneurs or entrepreneurial bureaucrats. To summarize, the ownership transformation of PCs and SOEs can be viewed as one case of interaction of the state and the market. In the form of partial privatization, it improved incentives, hardened budget constraints, and unleashed standard market forces, which combined to generate growth and raise efficiency. In the meantime, none of them was done in a complete way. State ownership and control was retained to certain degree to ensure that the reform was acceptable to all fractions of the society, and the political and social shocks the reform brought about was within a manageable degree. The partial nature of the reform had its own cost, of course. However, in retrospective, the critical point does not seem to be the fact that there was a cost. Instead, what has been critical is that on the balance, the net gain from reform was large enough that further reform was both justified and supported. III. Federalism with Political Centralization As market-oriented reform started from a point of only state, no market, there is a natural question of incentives of the state. Given the fact that everything was under government control, reform could only be launched and carried out by the government. On the other hand, however, complete government control over economic, political and social life meant no reform would be possible without loss of power on the government side. So, how could the government be expected to take the lead in reform? One way out of the dilemma is a political revolution to replace the existing government with one that is somehow pro-reform. China experienced a similar political change in 1976 when Chairman Mao died and the Gang of Four, the leaders of the left-wing extremlists who were the driving force of the Cultural Revolution, were arrested. However, it was managed as an internal political event within the CCP rather than a political revolution comparable with what happened in late 1980s in the former Soviet Union and Eastern Europe. When reformers led by Deng Xiaoping took the power in , the issue of incentives of government became one of how the central government incentivizes local governments (provinces, prefectures, counties, and municipalities at each of the three levels). Given the size and regional diversity of the country, it is not difficult to see how crucial this is to the success of reform. China s approach to this problem was similar to 10

11 the one to ownership transformation of PCs and SOEs. A fiscal contracting system was implemented in 1980s to make local governments residual claimants, while more decision making power was delegated to them as part of the decentralization process. This reform led to the emergence of a system of federalism with political centralization, as Blanchard and Shleifer (2000) put it. Under the fiscal contracting system, better known in China by the nickname of eating from separate kitchens as opposed to the pre-reform system of eating from a shared pot, local government entered into long-term fiscal contracts with higher level governments (Qian, 1999). The nature of the contract varied across regions and over time. Table 2 shows the main variants of the contracts between central and provincial governments. All the formula were applied after the division of expenditure responsibilities and the scope of current year total revenue were clarified. Despite the complexity in some cases, there is clearly a common feature that local governments were made responsible for the revenue consequence of their economic policies. This is especially comparable with the system of managerial autonomy and earning sharing that the government implemented in SOEs. Table 2: Revenue Sharing under China s Fiscal Contracting System, Formula. Expressed as the local government s budget Remarks. provinces include revenue after a remittance is made to, or a subsidy is provincial level receive from, the central budget. TR=total revenue municipalities, autonomous collected by the local government regions, and some local cities that were treated separately in the fiscal contracting system system, prevailed until TR x fixed coefficient of local share 1 province. In implementation, the coefficient was renegotiated a few times 2. TR fixed remittance requirement Guangdong and Fujian provinces, the two opened to FDI first 3.(TR 1979 TR) x fixed coefficient of tranfer to or 15 provinces from the central budget 4. (TR 1979 TR) x 100% + (1+10%) x central government subsidy last year 1988 system, prevailed until provinces with high concentration of ethnic minorities 1. [last year s TR x (1+ a fixed coefficient of growth 10 provinces. Eg. Beijng: rate)] x a fixed coefficient of local share + [TR (last fixed coefficient of growth year s TR x (1+ a fixed coefficient of growth rate))] rate=4%, fixed coefficient fo x100% local share=50% 2. TR x fixed coefficient of local share 3 provinces. Eg. Shanxi: the 3. (last year s TR x fixed cofficient of local share) + (total revenure last year s TR) x another coefficient of loca share fixed coefficient=87.55% 3 provinces. Eg. Dalian: fixed coefficient of local share=27.74%; another

12 loca share 4. TR [last year s remittance to the central budget x (1+ a fixed coefficient of remittance growth)] share=27.74%; another coefficient of local share=27.26% 3 provinces. Eg. Guangdong, 1987 remittance=rmb1.4 billion, the fixed coefficient of remittance growth=9% 5. TR fixed amount of remittance to the central budget 3 provinces. Eg. Shanghai, the fixed amount of remittance= RMB 10.5 billion 6. TR + fixed amount of subsidy 16 provinces. Eg. Tebit, fixed amount of subsidy= RMB898 million Source: Wu, 1999, p This system is believed to have had positive impact on economic growth in the following aspects. In terms of efficiency of decision making, local governments have information advantage over the central government. This is so because a large part of the information that is needed in making economic decisions is idiosyncratic in the sense that it is about changes in particular circumstances of time of place and by its nature cannot enter into statistics and be conveyed to the central authority (Hayek, 1945). In the sense of incentives, the reform strengthened local governments incentives and hardened their budget constraints, in a similar way as the managerial autonomy and earning sharing system did in the case of SOEs. With decentralized decision making and strengthened incentives, inter-jurisdictional competition was generated (Qian, 1999). In other words, market forces were brought into play. The contribution of local governments to China s economic success has been widely recognized. One of the most significant aspects has been its role in promoting and facilitating the growth of non-state enterprises. China s rapid economic growth has been made possible not much by privatizing existing firms but by new entry of non-state firms, especially TVEs. From 1978 to 1997, China s industrial output grew 27 times, while that of SOEs grew only 9 times. More than three quarters of the output growth was attributable to non-soes. In the same period, China s total employment and urban employment increased by 294 million and 107 million, of which the contribution of nonstate sector was 89% and 67%, respectively (NBS, 1998). While SOEs and PCs dominated Chinese economy at the outset of the reform, in 1998, the non-state sector accounted for 63% of GDP (IFC, 2000). However, inter-jurisdictional competition could not have had such a positive impact without the interaction with the state. A recent comparative study of China and Russia (Blanchard and Shleifer, 2000) highlighted this issue. It is recognized that the difference in economic growth between the two countries has come mostly from the growth of the new private sector, and the behavior of local governments in both countries are important to explain the difference. In China, local governments have actively contributed to the growth of new firms, while in Russia, local governments have typically stood in the way, 12

13 be it through taxation, regulation, or corruption. What made Chinese local governments behave differently, according to this study, was political centralization. In China, the CCP and the central government has been in a strong position both to reward or to punish local governments, as all the senior local officials are de facto appointed by the central authority. By contrast, transition in Russia has come with the emergence of a partly dysfunctional democracy, in which the central government has been neither strong enough to impose its views, nor strong enough to set clear rules about the sharing of the proceeds of growth. In this sense, inter-jurisdictional competition has been more like a contest, in which the state acts as a referee. Reward and punishment were not exercised automatically by market forces, as is the case in competition among firms. They were exercised by the central authority of the state. Without such a role of the state, market forces might have led local governments to maximize their welfare through different means, such as blocking the development of private sector. In the words of Blanchard and Shleifer (2000), for federalism to function and to endure, it must come with political centralization. However, even in China, federalism with political centralization has come with such a heavy cost that it has been criticized by some prominent Chinese reform economists and public finance experts. And indeed the criticisms are well justified. For example, despite political centralization, inter-jurisdictional competition has led to strong local protection and fragmented domestic market, even when the country is opened to international market. Fighting against local protection is still high on the agenda of the central government, as is evidenced in its report to the National Peoples Congress (the parliament) in early March, The fiscal decentralization was also seen as highly problematic and dysfunctional by public finance experts, who believed that it distorted resource allocation, generated regional inequality, and undermined the central government s fiscal policy (Qian, 1999). In 1994, a major package of comprehensive reform changed the central-local fiscal relationship again. However, although eight years has passed since then, China s fiscal system reform remains an unfinished agenda. From this point of view, federalism with political centralization is best viewed as a transitional institution rather than best practice. The crucial point to note is that the particular way of interaction between the state and the market made it fit much better to the particular circumstances of time and place than any other best practice institution. IV. Dual-Track Pricing: Growing out of the State In terms of interaction of the state and the market, China s unique approach to market liberalization, the dual-track pricing, provides another interesting case, in which the form of interaction can be characterized as growing out of the state, following Barry Naughton s (1995) famous book Growing out of the Plan. In the absence of market, it was the state that set prices in the central planning economy. In the case of China, state pricing accounted for 94.4% of agricultural products and 97% of total retail sales in 1978 (NBS, 1998b, p71). Obviously, market forces would not play any role despite of ownership transformation in PCs and SOEs, and decentralization, if the pricing mechanism remained the same. However, similar to complete privatization, 13

14 given the distortions in relative prices resulted from decades of government pricing, the constant threat of inflation, and limited progress in terms of ownership transformation in PCs and SOEs, a complete and quick price liberalization in early 1980s was regarded as politically not acceptable and economically not feasible, mostly for fear of inflation. Instead of totally abandon the old regime, reform had to be initiated from somewhere in it. During , while agricultural reform was the top priority of the government, and SOE reform was in the early stage, the government concentrated its efforts in adjusting prices, i.e., to reduce distortion within the framework of government pricing. For example, as part of the agricultural reform, state procurement prices of agricultural products was raised by 25% on average in However, as ownership transformation and decentralization unleashed the natural forces of growth especially in the non-state sector, by mid-1980s, it became clear that mere adjustment of prices was not adequate. From 1978 to 1984, the share of non-soes in industrial output has increased from 22% to 31%. Even SOEs found it increasingly difficult to operate with administratively set prices, as the managerial autonomy and earning sharing arrangement injected profit motivation, albeit limited, into SOEs, which tended to drive them to buy and sell in response to profitable market supply and demand conditions. Started from 1985, a dual track pricing system was implemented. In the rural sector, the farming household was assigned the obligation to sell a fixed quantity of grain to the state procurement agency under the contracting responsibility system, and the price remained administratively determined. However, subject to fulfilling this obligation, it was free to produce and sell whatever it considered profitable, at prices determined in the market. Thus, two prices were created for one product: one determined by the state, another by the market. In the urban sector, a quota of production materials of industrial enterprises was set up based on 1983 actual. SOEs were secured to receive production materials supply from the state at planned prices so long as their demand was within the quota. For any above quota demand, they had to purchase in the market at prices determined by supply and demand. For producers, the same regime applied. Within-quota output was purchased by the state at planned prices, and above-quota output could be sold in the market. The emergence of this market track made it possible for non-soes to survive and grow without being covered in the state-run distribution network. What happened in the following years was that the coverage of the planned track stayed the same, while that of the market track grew. For example, the state procurement of domestically produced grains between 1978 and 1988 remained essentially fixed, while there was almost a one-third increase in grain output. In the industrial sector, the same thing happened. For coal, the principal energy source, the planned delivery was increased somewhat from 329 million tons in 1981 to 427 tons in 1989, but the market track increased from 292 million tons to 628 tons in the same period. In case of steel, the share of planned allocation fell from 52% in 1981 to 30% in Since the planned track was more or less frozen, the economy was able to grow out of the plan on the basis of market track expansion fueled by both non-state and state sector growth. (Qian, 1999). 14

15 While dual track pricing was adopted, full liberalization was also implemented wherever it was feasible. For example, prices for meats, sea foods, eggs and vegetables were liberalized in 1985, though the degree of liberalization varied across regions and products. For some product in some region, price ceiling was imposed to curb inflation. In 1986, prices for 7 major industrial consumer goods, including bicycle, black-&-while TV set, refrigerator, washing machine, and textile products were also liberalized. (NBS, 1998b, p68-69). The most significant advantage of dual track pricing was that it introduced real market prices and market allocation of resources in the very early stage of reform, and created conditions that allowed rapid growth of the non-state sector, without incurring the cost of political and social shocks that would be brought about by a complete and quick price liberalization. While economic growth was generated, the dual track system itself evolved as the market track outgrew the planned track. However, again, this benefit did not come as free lunch. The cost of long-lasting dual track system was particularly high in that it created a vast hot-bed for corruption in the 1980s. The opportunities of corruption is obvious, as the plan track for inputs conferred instant profits upon the favored purchaser upon reselling quota inputs in the free market (Woo, 2002). For example, at some point of time in late 1980s, the market price for steel was six times higher than the planned one, which made some storekeepers of steel companies millionaires. In 1988, Deng Xiaoping himself attempted to crash through the gate of price reform by liberalizing prices in a big-bang manner, but was forced to abort it by panic purchases in the market. The widespread corruption contributed to the deterioration of the political and social atmosphere of reform, which eventually led to the tragic event in Tiananmen Square in June After reform was re-launched in 1992, price liberalization was immediately implemented in By 1997, 85% of agricultural outputs, 95% of retail sales, and 96% of production materials had been traded at fully liberalized market prices. (NBS, 1998b, p71). V. The Financial System with the State as a Primary Intermediary The combined reform efforts in the field of ownership transformation, regional decentralization, price reform, together with the growth of the non-state sector led to the emergence of a market for goods and services (goods market). Even SOEs changed their behavior in 1990s from going to the mayor (asking government for help) to going to the market (finding a market niche to survival). However, goods market alone does not make a market economy. For market forces to allocate resources, goods market must function jointly with factor markets, for example, capital and labor markets. In this regard, interaction of the state and the market is again an important feature of China s reform. While the goods market has been largely developed, though not fully yet, the allocation of financial resources is still dominated by the state. China s financial system is characterized by the strong role of the state as a primary intermediary. In a simple setting, a market economy can be viewed as two sectors, households and firms, linked by three markets: goods market, labor market and capital market. Capital 15

16 market in particular is responsible to mobilize and distribute savings from the household sector to the most efficient usages in firms. The central planning economy eliminated the three markets but could not eliminate the two sectors. While firms were all publicly owned, households remained private. A natural challenge to the central planning system was therefore the mobilization and allocation of savings by the state. In the case of China, the solution was found within the question. Since neither goods market nor labor market existed, the state took their place in allocating goods and labor, and this enabled the state to replace the capital market. Firstly, the replacement of the goods market by the state enabled it to distort the relative prices of agricultural and industrial products. The distorted relative price became thus a major vehicle to transfer savings from the rural household sector to the state owned industrial sector (World Bank 1997, p9). When the state purchased agricultural products from PCs at lower prices and sold industrial products at higher prices, the so called scissors difference resulted in an implicit transfer of income from rural households to the state. Part of the income that could have become private savings of rural households had relative prices not been distorted was transferred to the state sector in forms of lower cost and higher revenue of SOEs. Thus, as Lardy (1983, p101) observed, in the prereform era, prices were set by the government largely to generate industrial profit, and thus government revenue. His study found that after early 1950s, agricultural procurement and retail price policy were used to enable the state to generate significant resources through the profits of light industry, which was a source of 29% of all state budgetary revenue during (Lardy, 1983, p127, 126). The low relative prices of food grains he found (Lardy, 1983, p114, 119) also imply lower labor cost and higher profit in the whole SOE sector. Secondly, the replacement of the labor market by the state also enabled it to distort the prices of urban labor. This was done in the form of suppressed wage rates of SOE employees. It has been well known that workers in the state sector before the reform received only very low and hardly changed wages (Table 3), which left them unable to build up long term savings. The part of income that could have been distributed to workers and added to their private savings was collected by the government in the form of profit remittances from SOEs, which increased the government budget revenue and savings. Of course, various social services provided by the government or SOEs, such as housing, pension, medical care, education, job security, etc., tended to increase budget expenditure, decrease budget revenue and thus reduce government savings. However, this system also served to maximize government savings since it enabled the government to effectively control the level of expenses of social services and keep it as low as politically feasible. Table 3: Wages of Staffs and Workers of State Owned Units, Year Total Wage Bill ( RMB bn) Average Wage (RMB yuan per person) Total Wage Bill As A Percentage of Gross Value of Industrial Output of SOEs* (%) Bonus As A Percentage of Total Wage Bill (%) 16

17 Source: NBS, 1998b, p37, 38, 43; 1998a, p433. * The total wage bill and average wage are that of staffs and workers of all state owned units, including non-industrial SOEs and other state agencies, institutions, organisations. ** Including extra payments for above-quota performance under piece rate system. In doing so, the state played a role of intermediary in that it raised funds from household sector through distortion of prices of products and labor, and allocated funds through central planning and budget allocation. Private savings were suppressed to the minimum. In 1978, per capita bank deposit, by far the dominant form of private financial assets, was only RMB21.9, equivalent to 50 kg rice. The share of the government and enterprises in national savings, according to one statistical study (Guo and Han, 1991), was as high as 88% in However, the market oriented reform in the 1980s undermined the foundation of this system. Ownership transformation in PCs and SOEs and price reform substantially eliminated the distortion in goods and labor market, by allowing a sharp increase in agricultural prices in relation to prices for industrial products on one hand, and loosening government control over wages and particularly bonus on the other. The pattern of national income distribution among government, enterprises and households was changed in favor of households. In particular, budgetary income from SOEs kept falling, contributed to the overall decline of the share of fiscal revenue in GDP. From 1978 to 1995, fiscal revenue as a percentage of GDP dropped by 24 points, from 31.2% to a historical low of 10.7%. Of which about 15 points can be attributed to the decline of budgetary income from SOEs (NBS, 2000, pp256-7). As a result, savings were privatized in the sense that the share of private households in national savings rose sharply, from 12% in 1978 to 56% in 1995, according to some estimate (Wu, Xiaoling and others, 1998). A fundamental mismatch then emerged. On one hand, private households became the main source of external finance for the corporate sector, on the other hand, a substantial part of the corporate sector and the whole financial sector were owned by the state. How could the private savings be channeled to the corporate sector? One option would be privatization of both SOEs and state owned banks and non-bank financial institutions, and a complete liberalization of capital market. Again, this was not what happened. The government chose to maintain its ownership monopoly in the financial sector, mainly 17