Abstract
China has been held up as a modern-day exemplar of ‘market-preserving federalism.’ This article challenges this popular belief by showing that its local governments face soft budget constraints. Fiscal indiscipline among subnational governments, which risks national indebtedness and macroeconomic instability, can pose serious dangers to federations. A large body of literature which proposes solutions to fiscal indiscipline through electoral incentives and political party structure cannot be applied to China. The Chinese Communist Party’s cadre-evaluation and dual accountability systems make it an imperative for local officials to augment fiscal revenue and allow them to tap resources at local credit institutions. This has resulted in mounting local government debt, the lion’s share of which is unrepaid loans owed to local credit institutions. To harden budget constraints, political institutions need to be reconfigured to allow the central government more effectively to hold local authorities accountable for resources deployed in achieving their job-performance targets.
1. Introduction
All around the world fiscal federalism is in vogue. What are its attractions? Traditionally, it has been argued that decentralization improves the efficiency of public goods and services provision since local authorities that are closer to the citizens are more in tune with their tastes and preferences. Based on the model proposed by Tiebout (1956), the proponents of this view contend that fiscal decentralization promotes competition among local governments for better use of public resources. A particular strand of fiscal federalism, ‘market-preserving federalism’ (MPF), proposed by Weingast (1995), purports that fiscal federalism promotes economic growth by protecting economic rights and preserving the market. Along with the United Kingdom during the 18th century and the USA during the 19th century, China has been held up as a prime example of how fiscal federalism has created the market-preserving institutional structure that underlies its impressive growth – despite the absence of political reforms.
Weingast argues that a de facto political arrangement has emerged in China whereby, even in the absence of constitutional guarantees, fiscal federalism promotes economic growth when subnational governments face hard budget constraints and the predatory powers of the center are restrained. Nonetheless, the experiences of a wide range of countries, especially those in Latin America and Eastern Europe which have enthusiastically embraced fiscal federalism, suggest the need for caution. A key assumption of MPF is that ‘all local governments face hard budget constraints.’ 1 Fiscal indiscipline among subnational governments, underpinned by a lack of accountability in political institutions or by the inability of the center to hold local officials responsible for their behavior (or both), can pose serious dangers to federations (Rodden et al., 2003a). When fiscally irresponsible subnational governments are allowed to raise and spend revenue freely, they may overfish from common waters. When this happens, the center is often forced to absorb the costs of local officials’ fiscally irresponsible behavior, which risks national indebtedness and macroeconomic instability. As some observers warn, this is a real danger posed by decentralizing the power of the purse to local governments.
A large body of literature argues that solutions to fiscal indiscipline can be provided through the design of political institutions, and especially through electoral incentives and the political party structure (see Blanchard and Shleifer, 2001; Dillinger and Webb, 1999; Khemani, 2007; Rodden, 2006; Samuels, 2000; Stepan, 2000; Wibbels, 2003). The central idea advocated by this body of literature is that a dominant national party in the central government will be held accountable by voters for any unfavorable macroeconomic outcome. As a result, the central government would have incentives to discipline regional authorities for fiscally irresponsible decisions. The case of China is important for advancing our understanding of the theory. Because China has been held up as a modern-day exemplar of MPF, and since recent enthusiasm about the theory has to a large extent been generated by its impressive economic growth, whether the theory really applies to China has widespread implications beyond its border.
China has embraced fiscal federalism despite being one of the most durable authoritarian states in the world. Under Deng Xiaoping’s rule in the mid-1980s, fiscal power was decentralized to regional and local authorities so as to encourage local leaders to promote economic growth even though constitutionally China remains a unitary state. According to the World Bank, when measured by local governments’ contribution to funding public goods and services provision, China is currently the most decentralized country in the world (Fock and Wong, 2007). Municipal, county, and township authorities are responsible for financing a wide range of public services, from infrastructure and healthcare to education. Notwithstanding this, China remains stubbornly a one-party state. Even though there has been some degree of political decentralization allowing regional leaders to exercise control over lower-level officials, the central state still maintains a strong grip on party officials at all administrative levels through various political instruments and incentive systems.
Political institutions, namely, the cadre-evaluation system and dual (tiao-kuai) accountability system of the Chinese Communist Party (CCP), have made it an imperative for local officials to augment fiscal revenue while allowing them to tap resources at local credit institutions. This has resulted in undisciplined local government borrowing for the purpose of investment in industrialization projects. The outcome was mounting local government debt, the lion’s share of which was unrepaid loans owed to local credit institutions, thereby putting these credit institutions at risk of bankruptcy.
But because the principal source of local credit institutions’ capital is rural households’ savings, their insolvency risks the widespread loss of such savings and ensuing social instability. Thus, these credit institutions are ‘too big to fail’, forcing the central government to bail them out. The bailout of the local credit institutions also, in effect, financially rescued the indebted grassroots governments, resulting in a classic case of local government soft budget constraint.
This article is organized as follows. Section 2 briefly reviews the MPF literature and the body of literature on the danger of local government soft budget constraints stemming from the expectation of central government bailouts. Section 3 describes the Chinese central government’s bailout of the largest local credit institutions nationwide, the Rural Credit Cooperatives, whose major debtors were the grassroots authorities. Section 4 analyzes how the design of political institutions has contributed to the problem of local government fiscal indiscipline in China. Lastly, Section 5 addresses the normative question of creating sustainable hard budget constraints in China.
2. Market-preserving federalism, soft budget constraints, and bailouts
Montinola et al. (1996) argue that Chinese-style federalism offers ‘a strong and credible political foundation’ that limits the power of the central government and provides political protection for China’s economic reforms. The theory has been subjected to various criticisms. Some criticize that the theory lacks underlying political foundations to make it work (Sinha, 2005). Others argue that China’s growth did not stem from fiscal federalism (Cai and Treisman, 2006), and in the absence of constitutional protection, the central government could take away the economic rights of local governments, like it did with the fiscal recentralization policy in 1994 (Yang, 2006). This study contends that the validity of MPF critically hinges on the assumption that ‘all governments face hard budget constraints’; that if violated, fiscal federalism does not induce growth, and it may even cause macroeconomic instability.
Fiscally undisciplined subnational governments overdrawing from the common resource pool and shifting the costs to other jurisdictions is a real danger of fiscal federalism. When fiscal power is decentralized, local officials facing soft budget constraints have an incentive to overspend and overborrow. Mounting local government debts can result in a fiscal crisis, insolvent financial institutions, or macroeconomic instability which compels the central government to intervene.
The experience of a range of federalist states highlights three conditions leading to local government’s soft budget constraint. First, the central government lacks a credible commitment to withhold bailouts when local governments or institutions are ‘too big to fail’. 2 This occurs when local government behavior produces negative spillover effects bearing financial or political costs too significant for the center to ignore. Second, the expectation of a bailout is heightened in the case of regional governments with strong representation in the national legislature. Third, the expectation of a bailout is increased when local governments feel they will be realistically able to demand such bailouts from the center by logrolling or exchanging favors with other legislators, as is common in Brazil and increasingly so in India (Rodden et al., 2003b: 16).
A prominent example of local government soft budget constraint is provided by the fiscal crisis in Brazil in the early 1990s. At the time, Sao Paolo, the largest state in the country, relied heavily on a few state-owned banks to finance its public expenditure, incurring a debt amounting to almost half that of all Brazil’s state governments. When Sao Paolo ceased to service its loans and forced the major state banks into bankruptcy, the center was forced to bail out the banks – and, by implication, absorbed Sao Paolo’s debt. Brazil’s experience illustrates the central government’s inability credibly to commit itself to not intervening because negative externalities from a potential failure were too enormous to turn a blind eye to (Wildasin, 1997).
The existing literature presents two approaches to hardening local government soft budget constraints. In countries with vertically integrated party systems, such as Germany and Australia, the fact that the career prospects of elected local officials are in the hands of the national parties can provide electoral incentives for more responsible fiscal decisions. When voters hold national political parties responsible for bailout costs, those local governments controlled by the national government’s party that press for bailouts will be punished in the electoral system (see Rodden, 2006; compare Khemani, 2007). Alternatively, the central governments of Norway, Canada, and Hungary successfully employ hierarchical mechanisms, such as tight monitoring of local budgets and strict controls over local spending and borrowing, to regulate local fiscal decisions (Rodden and Eskeland, 2003). Nevertheless, this existing literature is entirely drawn from the experiences of countries with multiparty structures and competitive elections. As such, it has little practical application for authoritarian countries, such as China, even though such countries are also vulnerable to the problem of soft budget constraints in local government.
3. Local government indebtedness and the center’s bailout of the local credit institutions: The case of China
Ballooning local government debt
In China, mounting levels of local government debt are in fact a direct result of fiscal indiscipline by subnational authorities faced with soft budget constraint. On paper, there are numerous rules intended to make sure local government budget constraints are ‘hard’. China’s Budget Law (yusuanfa) requires subnational governments at all levels (provincial, prefecture, county, and township) to maintain balanced budgets. This law also forbids them from engaging in debt financing of any kind (Yang, 2003: 4). 3
Nevertheless, local governments in China have numerous ways of skirting this law. One such way was to set up companies that could borrow from banks, and in turn contributed to local government coffers in various ways. Grassroots government, including township and village authorities, 4 had commonly evaded the legal restriction on borrowing from credit institutions by setting up collective township and village enterprises (TVEs) and pressuring credit institutions into lending to these enterprises. As I have shown elsewhere, collective enterprises pay taxes to local coffers and some even help local governments finance public goods and services provision. 5 When these enterprises defaulted on loans on a massive scale in the mid- to late 1990s, township and village governments (the de facto enterprise owners) had to assume the enterprises’ debts and liabilities as their own. 6
The other more popular channel nowadays is to set up local investment platforms (difang rongzi pingtai) to act as borrowers in order to circumvent the legal prohibition against local government borrowing. Local investment platforms, which are essentially corporations with empty shells, have multiplied quickly because of the 2008 central-government-initiated ¥4 trillion fiscal stimulus package aimed at pump-priming the economy. However, the center was only committed to funding one-third of the package, leaving the remaining two-thirds to local authorities. Local governments that were given directives by the center to finance local infrastructure projects turned to financial institutions for funding via these investment platforms.
Studies indicate that China’s local government debts, including the liabilities of enterprises under local government control, are on the rise and show no signs of abating. A 2007 World Bank study underscores the unsustainability of these debt levels: even at the modest estimate of village liabilities amounting to ¥250 per villager (equivalent to ¥240 billion in total village debt), the debt level is almost twice the ¥140 villager income per capita (Fock and Wong, 2007). This financial indebtedness incapacitates local governments’ ability to carry out basic governmental functions, turning them into what some observers described as ‘empty shells’ (kongke) (Oi and Shukai, 2007). This has deprived many rural residents of basic public goods and services, such as healthcare, basic infrastructure, and the central-government-guaranteed, nine-year basic education plan. 7 Recognizing the severity of grassroots funding constraints, the central government has transferred to higher-level authorities some financing responsibilities for education and healthcare. However, this has not solved the problem entirely because many higher-level authorities, particularly county governments, are also under fiscal stress.
Undoubtedly, the recent borrowing binge has augmented local government’s indebtedness, particularly at the municipal and county levels. During 2010 an estimated 8000 investment platforms were set up nationwide, raking up a total of between ¥7 trillion and ¥11 trillion. 8 Local governments have been servicing interest payments using proceeds from land sales. These loans are used for building roads, highways, airports, high-speed rail links, and other infrastructure projects that have dubious investment returns or bring no immediate benefits to local residents. This gargantuan debt estimate may dwarf the magnitude of the grassroots authorities’ financial liability discussed earlier; both instances highlight the problem of the local government soft budget constraint and the similar mechanisms that allow local government to raise funds from financial institutions. Similarly, TVEs and local investment platforms are channels that allow local governments to borrow from banks. The former was prevalent during the 1980s and 1990s, while the latter has become increasingly common since the mid-2000s.
The Rural Credit Cooperatives
The Rural Credit Cooperatives (RCCs) are semi-public savings and loans institutions prevalent in rural China. Though their shares are owned by farmer households, their personnel appointments and loan-making decisions are heavily influenced by local authorities. The RCCs were established during the 1920s and 1930s to provide credit support to the members of farmer cooperatives by absorbing savings from rural communities. However, after the communist takeover in 1949, the RCCs were transformed into the financing arms of communes and brigades, which later became townships and villages, respectively. Consequently, farmers’ savings were effectively controlled by grassroots Communist Party leaders.
Between the early 1980s and 1996 the central government subordinated the credit cooperatives to the state-owned Agricultural Bank of China (ABC) (Tam, 1988). During this period, the ABC formally controlled the RCCs’ personnel and loan allocations, though local governments exercised stronger influence over such decisions de facto, for reasons discussed in Section 5. Between 1996 and 2003 the RCCs were directly managed by county credit unions, which in turn reported to the regional offices of the central bank, the People’s Bank of China (PBoC). Since 2003 the RCCs have been managed by provincial credit unions, which report to the provincial governments.
Throughout the 1980s and 1990s about half of the RCCs’ loans were allocated to the TVEs, which were de facto owned by local governments (see Figure 1). 9 The rapid growth of the TVEs in the 1980s and early 1990s has been well documented in the literature. Indeed, it is generally recognized that the TVEs underlay the take-off of rural industrialization in China during this period (Oi, 1992). Fiscal decentralization in the early 1980s enhanced local governments’ responsibility for financing the provision of essential public goods and services. Hard pressed for revenue, township and village governments enthusiastically developed collective TVEs that contributed enormous taxes to local coffers, initially through a value-added tax and subsequently through an enterprise income tax. Aside from formal fiscal revenue, collective enterprises sometimes contributed their profits to local authorities’ extra-budgetary funds, which were not subject to sharing with higher levels of government. In addition, some collective enterprises even assisted local governments in providing public services directly. Some built schools in the rural community for enterprise workers to send their children to; others constructed roads to link villages to nearby towns in which their products were sold (Ong, 2006). However, the late 1990s saw the collapse and privatization of the collective TVE sector, and the assumption of the TVEs’ debts by their guarantors, that is, by township and village authorities.
Composition of the RCCs’ Loans Nationwide (Various Years)
The official nonperforming loan rate of the RCCs reached as much as 50 percent in the late 1990s (Yu, 2003). The actual bad loan rate was likely much higher, as local officials had a variety of ways of ‘hiding’ nonperforming loans on their books. Loans to households generally had higher repayment rates than those to the TVEs (International Fund for Agricultural Development, 2001; Park et al., 2003). The nonperforming loan (NPL) rate of loans to TVEs during the 1990s is not publicly available. My own estimate suggests that it was as high as 80 percent in the late 1990s. 10 As Section 5 will illustrate, local governments are able to influence the loan allocation pattern because local credit institutions are also accountable to local authorities, which is an intrinsic feature of China’s political institutional design. Since many of these loans were not allocated based on purely economic considerations, it is not surprising that most of them went sour.
The Rural Cooperative Foundations
The Rural Cooperative Foundations (RCFs), established by township governments, were another local credit institution that had lent heavily to the TVEs. Though established by local governments, the RCFs were informal financial institutions, not regulated by the PBoC. As a result, the RCFs were able to offer higher savings rates than formally regulated institutions such as the RCCs. This helped attract rural household savings to the RCFs, which by the mid-1990s totaled ¥150 billion, or about one-fifth of RCC-held savings at that time. The rural household savings absorbed in this way were in turn channeled to the TVEs, to other local-government-related projects such as the development of public infrastructure, and sometimes even directly to cover government fiscal shortfalls (dianshui) or to line the pockets of local officials. 11
The central government ordered the 40,000 RCFs nationwide to be shut down in 1999, when the RCFs in some rural locales were unable to honor savings, resulting in bank runs and eventually in widespread protests and rural political unrest. Following the closure of the RCFs, government audits revealed that many of them suffered from negative net worth, with liabilities well in excess of assets. The central government made honoring the savings of household depositors (whom it perceived most likely to pose risks to social stability) a responsibility of township governments. Meanwhile, some of the RCFs’ loan portfolios were transferred to the parallel local RCCs, which were made to pay RCF depositors on behalf of township authorities with limited financial resources. Most township governments have failed to service these debts to the RCFs and RCCs, which therefore remain one of the largest items on local government books.
There are few systematic studies on the sources of local government debt in China. Nevertheless, available data do suggest that financial liabilities to the RCCs and RCFs account for a substantial share of grassroots government debt. Figures 2 and 3 summarize the levels of village and township debt, based on World Bank data which were collected from sixteen townships in eight counties in Chongqing municipality and Hubei, Jiangsu, and Jilin provinces (Fock and Wong, 2007). These data show that the RCCs and RCFs, combined, account for almost half of village debts and one-fifth of township debts. Figure 4 presents data from my own research, collected during 2005–06 from six townships of four counties in Sichuan, Shandong, and Hebei provinces, where the RCFs were more densely populated. These data show that about 60 percent of township debts were owed to local RCCs and the now defunct RCFs. These two sets of data (drawn from six provinces and a municipality) provide a reasonably accurate picture of the degree of local government indebtedness at that point in time. This is particularly true in consideration of the fact that local government debts are illegal and inherently non-transparent as local governments try every means to conceal them. 12
Sources of Village Debts
Sources of Township Debts
Township Debts
The bigger proportion of debts owed to the credit institutions indicated in my data (see Figure 4) can be explained by a higher concentration of the RCFs in the provinces from which my data were collected, such as Sichuan and Shandong. Given the lack of nationwide representative data on the sources of local government debt, we can infer from the two sources we have that, on average, the RCCs and RCFs, combined, account for between one-fifth and one-half of total grassroots government debt. The data also show that the next largest lenders to local governments are private lenders and corporations. This is despite the fact that private moneylending (from institutionalized private money houses to less-institutionalized village lenders) is considered strictly illegal by the communist regime. This highlights the fact that local governments are similarly able to circumvent the legal restriction against local government fundraising by borrowing from illegal underground moneylenders. 13
The RCCs are ‘too big to fail’
The primary reason for the RCCs being ‘too big to fail’ is their massive holdings of household deposits in rural China, where three-quarters of China’s population still lives. As Figure 5 shows, 80 percent of RCCs’ capital comes from rural household savings. If the central government let the RCCs fold, hundreds of millions of rural depositors would lose their savings. From the perspective of the central government, bank runs on a massive scale would have inconceivable consequences for social stability in the countryside, the priority concern of the communist regime. Indeed, the current Hu-Wen administration’s rhetoric of creating a ‘New Socialist Countryside’ reflects the central government’s concern for rural residents getting the short end of the stick in the country’s dizzying economic growth. The increased frequency of popular protests, totaling more than 90,000 in recent years, further underscores the risks to social stability posed by unequal wealth distribution. Xie Ping (2003), a former leading central banker in Beijing, concedes that ‘with a large amount of deposits from low-income farmers and the absence of explicit deposit insurance, closures (of the RCCs) are expected to cause serious political and social unrest. Therefore, RCC bankruptcy is almost “out of the question.”’
Composition of the RCCs’ Savings Nationwide (Various Years)
Another reason that the RCCs are ‘too big to fail’ is that they were and still effectively are the only formal credit providers in the countryside. 14 Until the late 1990s, there were more than 45,000 RCCs scattered around the countryside, covering every single township and most villages. Today, there are still as many as 35,000 credit cooperatives in China, almost one to every township. Therefore, their closure would deny hundreds of millions of rural residents their only existing access to formal credit, forcing them to turn to informal moneylenders who charge usurious interest rates.
The importance of the RCCs to rural residents made it politically imperative for the central government to come to the RCCs’ rescue when their official nonperforming loan rate climbed to 50 percent in the early 2000s. The central bank handed out two major forms of financial subsidy (¥165.6 billion in debt-for-bills swaps (zhuanxiang piaoju) and ¥830 million in earmarked central bank loans (zhuanxiang daikuan)) to assist the RCCs with negative net worth in disposing of their bad assets and erasing their historical losses. 15 Upon the bills’ maturation in two to four years, the RCCs swapped them for cash, subject to certain required changes in corporate governance structure. The earmarked central bank loans had no conditions attached, and were intended to inject immediate liquidity into the cash-strapped RCCs. In addition, the central government also provided various forms of preferential tax treatment for the RCCs to help raise their profitability.
The central bank’s financial assistance amounts to a bailout because it does not harden the RCCs’ budget constraints in any fundamental way. In particular, the debt-for-bills swap plan has failed to yield any substantial improvement in corporate governance structure as it was intended to. Post-reform RCCs were required to set up three internal institutions required by the central bank: a board of directors (BoD), a board of supervisors (BoS), and members or shareholders representative meetings. Each of these is supposed to perform a unique function and check the others’ power. Members of the BoD, the supreme decision-making body, are supposed to be elected at the shareholders representative meeting. The BoS has a supervisory function over the BoD to prevent any abuse of power by the head or members of the board of directors.
However, as I have argued elsewhere, the separation of powers and checks-and-balances mechanisms are extremely difficult to institute effectively in an institutional environment onto which is superimposed a hierarchical party structure. 16 The BoS is typically a one-man team, consisting of a member of the Communist Party disciplinary committee (jilv jiancha weiyuan) who is also a deputy head of the BoD. He reports to the local party secretary, who is also the head of the BoD. Obviously, impartial execution of an oversight function is inconceivable when the overseer is subject to the top ‘overseen’ executive in both the party and corporate hierarchies. Put simply, this renders the overseer dependent on the oversight’s subject for career advancement. Moreover, the shareholders representative meeting is nothing more than a rubber stamp. Though formally described as the highest governing body of the RCC, it has no effective power to override personnel decisions made by the local party committee, nor does it have any prerogative to deliberate or to make any major decisions.
Because township and village governments are among the RCCs’ largest debtors, the central government’s rescue of the RCCs amounted to an indirect bailout of the local authorities. Given the massive scale of the local government debt problem, the rescue of local governments could have been in the central government’s calculus when devising financial subsidies for the RCCs. The rescue has a similar ‘moral hazard’ impact on local governments as it has on the RCCs: it does not harden local government budget constraints or deter government officials from borrowing from banks in the future even if they have little ability to repay their loans. 17
4. Design of political institutions and fiscal indiscipline
Two institutional arrangements politically underpin local officials’ fiscally undisciplined behavior in China. One is the Chinese Communist Party’s cadre-evaluation system, with its exclusive emphasis on maximizing economic growth and fiscal revenue. The other is the dual accountability system to which the credit cooperatives are subject. These two factors both provide local officials with the incentive to borrow beyond the financial means of their government and enable them to do so.
Cadre-evaluation system
The CCP’s cadre-evaluation-and-management system allows higher-level governments to exercise effective political control over lower-level party officials in spite of fiscal decentralization. 18 Through the nomenklatura system, an essential feature of the CCP’s political institutional design, the party manages and controls appointments, promotions, transfers, and removals of all leading cadres (lingdao ganbu). 19 The party committee at each level exercises authority over leading cadres one rank down the administrative hierarchy. The leading cadre of a township (typically, the party secretary, deputy party secretaries, and the head and deputy heads of the township) is managed by the party organization department (zuzhibu) of the county (one rank above). Leading cadres of the county are in turn evaluated by the prefecture; those of the prefecture by the province; and those of the province by the central party organization department.
The cadre-evaluation criteria are strongly biased toward economic performance and sway local officials to devote exclusive effort to activities that promote economic growth. The performance targets (kaohe zhibiao) in the evaluation criteria are ranked in terms of their significance, from ‘priority targets with veto power’ (yipiao foujue) and ‘hard targets’ (ying zhibiao) to ‘soft or ordinary targets’ (yiban zhibiao). Failure to achieve the ‘priority targets with veto power’ constitutes overall failure, even if the township scores high on other criteria. Nevertheless, the ‘priority targets’ themselves, though necessary, are insufficient for achieving excellence. Scoring high on ‘hard targets’ (as opposed to ‘ordinary targets’) is what really makes or breaks a local cadre’s careers.
The evaluation criteria provide evidence for the primary importance given to maximizing GDP growth. Table 1 presents the cadre-evaluation criteria of four townships from three provinces that differ vastly in income level and economic orientation. Across the various locales, the cadre-evaluation criteria are similarly focused on maximizing industrial growth and fiscal revenue, even though some townships are poorer and more agriculturally oriented than others. Though this evidence is only for selected locales, it confirms and matches research conducted by other scholars. For example, Susan Whiting (2001, 2004) and Maria Edin (2003b) find that the cadre-evaluation criteria are effective instruments for the central government to ensure local officials prioritize economic growth over other tasks. Studies by economists have shown the likelihood of local leaders’ promotion to be positively associated with the economic performance of their jurisdictions (Chen et al., 2005; Li and Zhou, 2005).
Basic Economic Conditions and Cadre-Evaluation Criteria in Four Townships
Sources: Social and economic indicators are from the respective townships’ websites, county yearbooks, and interviews with township finance bureaus. Cadre-evaluation criteria are from interviews with various township party secretaries, township heads, and organization departments.
As seen in Table 1, the CCP’s other stated primary policy objectives, such as population control (the one-child policy) and the maintenance of social stability and harmony, are ‘priority targets with veto power.’ As such, they are no more than the bare minimum for passing the evaluation system – absolutely necessary, but insufficient conditions for a high-flying career. What distinguishes outstanding performance from mediocre is how one scores on the ‘hard target’: the amount of fiscal revenue collected, the level of industrial production, and investment attracted. These three criteria are positively correlated with one another: the more investment dollars a township attracts, the more tax revenue it will have. Similarly, the higher the level of industrial production, the more revenue a township can extract from enterprises. Notably, these fiscal revenue targets are also annually set moving targets – once a level is achieved in the current year, the target in the following year will be adjusted upward by a certain percentage.
Achievement of ‘hard targets’ has direct impact on the career prospects of members of leading local cadres – those who do well get promoted faster, while others stagnate in their current positions. It also affects the cadre’s monetary rewards, though the year-end bonuses resulting from achieving the targets are relatively insignificant compared with the financial rewards brought about by a higher position. 20
The cadre-evaluation system is also highly outcome oriented, creating incentives for local political leaders to focus on results rather than on process. Local leaders are rewarded, politically and financially, for increasing fiscal receipts by x amount and attracting investment dollars by y amount, even though they may have achieved these targets by incurring z amount of debt, with z greater than the sum of x and y. In other words, local cadres are not assessed based on the process or resources utilized to bring about the outcome; they are judged solely by outcome, measured in quantifiable terms. This evaluation system therefore enables the central leadership to exercise control over local officials by formulating some quantifiable targets, but gives the center little control over how these targets are achieved. The process by which the targets are attained matters because local cadres have easy access to resources from credit institutions, which they can direct to fund enterprises or projects that contribute to their coffers. In other words, this evaluation system encourages fiscally irresponsible behavior among local officials.
Notwithstanding that, it is recognized that the fiscal recentralization policy in 1994 has recentralized major tax revenue sources from local to central government while leaving the expenditure responsibilities of the former unchanged. Scholars have argued that local governments face ‘unfunded mandates’, a term which indicates the fact that local governments are given many tasks and mandates to get things done, but are provided with few resources to do so (Bernstein and Lu, 2003). As I have shown elsewhere (Ong, forthcoming), local governments’ revenue restraints (along with the institutional pressure to perform) compel them to tap resources from outside the system, such as the RCCs described here, in order to perform their functions. 21
Dual accountability system
The other political institutional arrangement underlying local government fiscal indiscipline in China is the lack of separation between the party and financial institutions. Local credit institutions are subject to a dual accountability system: having to follow the directives of both institutional superiors (vertical; tiao) and party committees (horizontal superiors; kuai). This inherent feature of the Chinese political institutional design applies to all subnational bureaucracies and financial institutions alike, even though banks and credit cooperatives are not formally part of the bureaucratic apparatus. This gives rise to a perennial problem of conflicting control over local bureaucratic agencies, which contributes to a range of policy challenges, as has been shown in the literature with regard to environmental policy. While the Ministry of Environmental Protection may instruct a local environmental bureau to clean up polluted rivers, local party leaders may oppose this if, for example, the polluting enterprises are major contributors to local government’s fiscal coffers. 22 According to Kenneth Lieberthal (2004), such conflicts between vertical and horizontal authorities are a defining characteristic of China’s ‘fragmented authoritarianism.’ 23
All local financial institutions in China, including regional and local branches of the state-owned banks and the RCCs, are subject to the same dual accountability system. This enables local leaders to influence loan decisions in these institutions, severely curtailing the efficient allocation of financial resources. Indeed, the system was blamed for the creation of the massive money supply and the resulting runaway inflation in the early 1990s, when local branches of the state-owned banks (being held hostage by local governments) provided generous funding to government-related enterprises and projects (Girardin, 1997). This prompted the central government, under Premier Zhu Rongji’s iron reign in the mid-1990s, to strip local party committees of their veto power in personnel matters relating to the state-owned banks. 24
Nevertheless, even though the RCCs during this period were still under the authority of the state-owned Agricultural Bank of China, these sweeping reforms did not affect the RCCs, which therefore remained subject to the dual accountability system (depicted in Figure 6). All senior RCC credit personnel are party members, and an internal party committee can be found in every credit cooperative. This internal committee is led by a party secretary who is also a RCC manager, with the other committee members typically serving as assistant managers and as the accountant. In other words, all key personnel of a credit cooperative wear two hats: those of credit officers and of party officials. As party members, they are obliged to follow the party’s rules and toe the party’s lines. Moreover, local party committees had the final say over personnel matters in the RCCs – they could appoint and dismiss anyone at will. This gave local party leaders enormous suasion over loan decisions since, after all, they could easily replace any recalcitrant credit manager with a pliable one. As a result, local party committees had de facto ‘leadership command’ (lingdao guanxi) over the RCCs, superseding that of the banking superior.
The Accountability System of the RCCs before 1996
After 1996 and until today, the RCCs have remained accountable to local party committees. Though party committees no longer have veto power over personnel appointments, local party leaders still have huge informal influence over local credit institutions. In an institutional environment in which one’s personal relationships are critical in greasing the wheel of professional performance and career advancement, all agencies operating within an administrative area where the local party committee reigns are held hostage by local party leaders’ directives, at least to some extent. It is therefore unsurprising that local party committees had substantial leverage and still have considerable influence over loan allocations at the RCCs. While a banking superior may advise a township credit cooperative not to lend to an enterprise with a bad repayment record, the township party committee may instruct the credit manager to continue lending. Since the credit manager’s job performance and career prospects have been evaluated and determined by the local party boss, it is only logical that he will follow the local party boss’s instructions.
5. Conclusion
This article has illustrated how political institutional design in China underpins the problem of local government soft budget constraints. To harden budget constraints, political institutions need to be reconfigured so as to allow the central government more effectively to hold local authorities accountable for resources deployed in achieving their required targets. Effectual top-down control is critical in an authoritarian polity where bottom-up oversight is palpably absent.
The CCP’s cadre-evaluation system is an effective instrument for the central authority to direct local political leaders to the areas in which they should spend their energies. The system has created powerful incentives for promoting local economic growth and developing industrial enterprises, consistent with the center’s priorities. This is attested to by the country’s impressive economic growth rates during the past three decades. Nevertheless, the system has been grossly ineffective in holding local officials accountable for the resources deployed (such as rural savings, highlighted here) in achieving those tasks. Notwithstanding that, in saying so, this article is not assigning blame to local governments as they are merely responding to institutional demands and incentives; it is the growth-oriented central government that sets the policy in the first place, and as some would argue, without giving full consideration to the resources at the disposal of local government officials. 25
Therefore, the cadre-evaluation system needs to shift from its exclusive emphasis on outcomes so as also to take into account process and accountability. The scope of evaluation criteria should be broadened to include the resources utilized in attaining an outcome, for example the magnitude of bank lending to local enterprises and other government-related projects, so that local leaders are not judged solely on their results. The system needs to incorporate the concept of resource efficiency: township leader A may have collected more fiscal revenue than township leader B, but A may have borrowed enormous loans and incurred high debt to do so, while B has not. The existing system will reward A for being an outstanding leader, but B may have in fact done better if the debts incurred by A are taken into account. Resources utilized to achieve growth targets should be made a ‘hard target’ or a determining factor for one’s job performance.
In non-democracies in which bottom-up supervision of local officials’ behavior is nonexistent, the need for the center’s supervision of local government is accentuated. Living in a democracy, it is easy to take for granted both the (relative) transparency of government bureaucracies and media freedom, which serve to check officials’ abuses and excesses. In authoritarian polities in which the feedback loop is visibly absent, the onus is on the central authority to hold local leaders accountable, and the cadre-evaluation system is the key mechanism to do so in China.
Furthermore, credit institutions should be freed from government interference. Specifically, this entails severing their accountability to local party committees while strengthening their accountability to the institution’s headquarters in Beijing or in the provincial capitals. Local party bosses should not have veto power over personnel matters, nor should they be consulted before any personnel appointment is made.
The long-term solution to the problem lies in improving the corporate governance structure of credit institutions, making them genuinely accountable to shareholders and members, rather than to party bosses. When a bank is paralyzed by mountains of bad loans, shareholders should be able to hold bank managers accountable for their lending decisions. This includes taking punitive measures such as reduction of performance bonuses or even, in serious cases, dismissal. However, the RCCs and the state-owned banks in China have failed to improve shareholder accountability despite having gone through numerous restructuring exercises.
Admittedly, this will be easier said than done in reality, as the tiao-kuai accountability system is an intrinsic feature of China’s political institutional design. It originated during the Maoist era to enable Communist Party institutions at all administrative levels to exert control over local bureaucracies and economies (Lieberthal, 2004: 186). In addition, owing to the authoritarian and all-encompassing nature of the CCP, even when the dual accountability system is formally abolished (as was the case with the four state-owned banks during the late 1990s) local bank managers still have informal obligations to local party leaders. Local banks operate on the turf of local leaders, who have control over all resources, ranging from land approvals and municipal permits to utility supplies.
Hence, genuine corporate governance is unattainable unless the magnetic power of the CCP is curbed. On paper, most banks are not responsible to the party nowadays, but this does not negate the fact that they are held ransom by informal obligations to local party bosses. This implies that bank restructuring is effective only to a certain extent. The rule of law is needed to curb the pervasive and deeply penetrating power of the ruling party. This calls for the establishment of institutions such as an impartial judicial system and a central bank independent from the CCP. These institutions are necessary to prevent power abuses by party leaders and to mediate disputes between the party and credit institutions. In their current form, the court system and the central bank are also responsible to the CCP, much like the credit institutions.
Democracies are, of course, not immune to bad loans and bailouts, as events during 2008–09 in the USA attest. Both cases vindicate the argument that when corporate governance is lacking (whether Wall Street insiders are able to clinch sophisticated deals without fully disclosing the details to shareholders, or bank lending is heavily influenced by local government), the outcome can be similarly disastrous. Nevertheless, as this article argues, strengthening the corporate governance of banks in authoritarian countries is a different ballgame: the fundamental rule of law is needed to help curb the power of the ruling parties in authoritarian systems.
This brings us to the question of whether China is an example of ‘market-preserving federalism’. Weingast and others argue that MPF helps protect economic rights and preserve the market against government encroachment, as the rule of law, the horizontal separation of powers, and democracy do in democratically competitive regimes (Montinola et al., 1996; compare Sinha, 2005). It is this article’s contention that until the rule of law is established so as effectively to restrain party leaders from tapping financial resources, local governments will always face soft budget constraints. The consequences range from macroeconomic instability and runaway inflation to fiscal strains and incessant taxpayer bailouts of local governments and financial institutions. China’s fiscal federalism does not preserve the market, nor is it a substitute for the rule of law. 26
Footnotes
Acknowledgements
I would like to thank Andrew MacIntyre, Stephan Haggard, Jonathan Unger, Erik Kuhonta, and five anonymous reviewers for helpful comments and suggestions. All remaining errors remain those of the author.