Interprovincial Effects of Fiscal Decentralization on Financial Decentralization in China
1 National Academy of Economic Strategy, Chinese Academy of Social Sciences, Beijing, China
2 International Business School, Shaanxi Normal University, Xi’an, China
By motivating local governments to fight for financial resources, China’s tax sharing reform has affected the structure of financial decentralization and inflated local financial systems, thus spawning regional financial risks. Based on theoretical analysis and empirical evidence, this paper has arrived at the following findings: due to different policy objectives, central and local governments exhibit different fiscal and financial behaviors; public finance and financial sector have become financing instruments with certain convertibility under local economic growth framework; fiscal decentralization inevitably affects financial decentralization and lays the foundation for provincial fiscal disparities, resulting in a certain spatial effect of interprovincial fiscal variable; financial explicit centralization/implicit decentralization and fiscal centralization have fueled local competition for financial resources and resulted in correlation between the spatial effects of provincial financial and fiscal variables, and moreover, their mismatch has also spawned fiscal and financial risks on various fronts. Hence, setting clear boundaries of financial centralization and decentralization and ensuring local government fiscal accountability is the key to the prevention and mitigation of fiscal and financial risks in China.
fiscal decentralization, financial decentralization, financial risks, spatial effect JEL Classification: E61, G28, H11
Economic power structure induces change in economic system and affects the efficiency, risks and volatility of economic operation. As key elements of economic decentralization, fiscal/ financial power structures and centrallocal decentralization have defined the course of China’s financial system evolution and served as a unique angle for observing China’s financial operation and risks. Scholars believe that the institutional framework of fiscal and financial decentralization between Chinese central and local governments formed since 1978 has affected local government behaviors
and r esource a l locat ion and improved macroeconomic operation (Ding and Fu, 2012). *
Corresponding author: He Dexu, Institute of Finance and Banking, Building No. 2, Compound 2, Yuetan North Street, Beijing, China (100028). Email: email@example.com.
This paper is a result of National Social Science Foundation Key Project “Strategic Study on China's Financial Security in the 13th Five-Year Plan Period” (Grant No. 15AJY017) and National Social Science Foundation General Project “Study on Regional Risks, Moderate Decentralization and Local Financial System Reform” (Grant No. 14BJY192). We acknowledge the suggestions of anonymous reviewers and the technical support of Professor Zhang Xuelan from Zhongnan University of Finance and Economics and Zhou Chao from Zhangye Central Branch of the PBoC but take sole responsibility for this paper.
However, given local government interventions in the al locat ion of f inancial resources , China’s financial power structure cannot be simply defined as “financial centralization.” In fact, China’s current financial system can be described as explicit centralization with implicit decentralization, i.e. explicit centralized approval, regulation and relief coexist with implicit decentralized financial operation. Among various reforms, fiscal decentralization represented by the tax sharing reform is a key factor of regional financial risks.
Existing studies have laid the foundation for analyzing the correlation between China’s fiscal decentralization and financial explicit centralization/ implicit decentralization. One relevant literature is concentrated in the intrinsic correlation between local government financial intervention and financial decentralization. It is generally believed that fiscal decentralization increased local government budgetary engagements and influenced their financial intervention and decentralization, as evidenced in local governments competing for financial resources to ease budgetary constraints after financial decentralization through an array of methods - such as direct administrative interventions, attempts to influence bank decisions and evasion of bank loan repayment ( Ba, et al., 2005). Contrary to centralization, financial decentralization will lessen local government budgetary constraints ( Qian & Roland, 1998). In the course of China’s marketoriented economic reform, fiscal decentralization experienced three rounds of reform, i. e. the tax sharing reform implemented in 1993, the central-local fiscal transfer payment reform since 1995 and the income tax sharing reform carried out in 2002. Each and every round of reform increased the independence of local interests but also reduced the share of local fiscal revenues (Hu, 2001).
With the rising share of local spending, attracting state- owned bank loans became an alternative to asking for more fiscal allocations from central government. Meanwhile, local governments acquired controlling shares of city commercial banks as an alternative or supplement to state- owned banks - and even turned them into a “secondary source of fiscal revenue.” From this standpoint, tax sharing reform increased local government intervention in the financial sector ( Chen et al., 2002) for the purpose to acquire sufficient and even more than sufficient financial resources to spur local economic growth.
Another area existing literature is concerned with the effects of local government financial competition and decentralization. Local government financial competition has led to the direct result of financial decentralization and the indirect result of inflation and financial risks. By seeking financial support for non- SOEs, local governments promoted decentralized financial development (Zhang, 2006). Yet under the joint effects of fiscal and financial decentralization, competitive pressures have induced local governments to increase real estate loans of city commercial banks, reduce risk provisioning coverage and increase concentration, giving rise to non- performing loans ( NPLs) ( Qian, et al., 2011).
Local government financial competition in recent years has been reflected in the surging numbers of local banks, local financial markets, local government financing vehicles and liabilities1. Due to poor supervision and accountability, local financial institutions and multi- tiered financial markets have spawned regional financial risks, instead of serving their intended purpose to optimize financial structure.
As can be seen from the above literature, some fundamental questions remain unanswered: ( 1) how should financial power structure be defined? Is China’s financial 1
According to the National Audit Office data, by the end of June 2013, liabilities for which governments at all levels are responsible to repay amounted to 20.698 trillion yuan, liabilities guaranteed by governments at all levels reached 2.925 trillion yuan and liabilities for which governments at all levels assume certain remedy obligations totaled 6.65 trillion yuan. In total, these numbers add up to 30.27 trillion yuan. Liabilities for which local governments are responsible to repay stood at 10.88 trillion yuan. Source: NAO website Announcement 32, 2013: Audit Result of National Government Liabilities, http://www.audit.gov.cn.
system a decentralized or centralized system? ( 2) Does fiscal decentralization significantly affect financial centralization in the financial aggregate? (3) Is China’s fiscal power structure compatible with its financial power structure? What are the consequences if not? This paper will individually examine the above questions. The rest of this paper is structured as follows: Section 2 discusses the intrinsic relationship between fiscal and financial decentralization; Section 3 conducts an economic theoretical analysis and proposes empirical propositions to be tested through modeling; Section 4 explains empirical method, variable design and data sources; Section 5 explains the estimation result and verifies modeling and theoretical propositions; Section 6 offers concluding remarks.
2. Theories and Propositions
For the government as an economic sector in a market- based economy, public finance is an instrument to ensure the performance of government functions, including resource allocation, income distribution, economic s tabi l i ty and developmen t through tax collection, financing and procurement whenever market failure occurs. Fiscal system defines the authority and procedures for government tax collection, procurement and financing behaviors (Guo, 2003). Financial intermediation is an instrument for market entities to realize the temporal and spatial allocation of funds, payment settlement and other functions. Financial system is an institutional arrangement that enables resource allocation in a less costly and more efficient manner (Zhang, 1998). In a government- dominated market economy, both fiscal and financial systems are the means for the government to promote economic growth and achieve sectoral objectives. Under the two- level or multi- level government structure, administrative decentralization for the organization and control of fiscal revenues based on the administrative affiliations of enterprises and economic decentralization based on tax sharing system ( Jia, 2013) inevitably affects financial decentralization through the correlation of market funds.
Under the system of fiscal decentralization, local and central governments have different policy objectives and methods.
( 1) For the central government, although economic growth is a top priority, inflation must be reined in as well. Local governments, however, are only concerned with how to maximize the economic output within their respective jurisdictions.
( 2) While central government may obtain seigniorage income by manipulating money supply, it must also curb inflation to maintain social stability, increase seigniorage rate without reducing the total amount of seigniorage income, and refrain from transferring SOE liabilities to the banks in the interest of financial stability. However, these concerns are not shared by local governments. By insatiably fighting for bank loans, local governments fuel money supply; by acquiescing local businesses to leave liabilities to the banks, they contribute to mounting NPLs; by resorting to all means, they strive to maximize financial resources at their disposal ( Miao, 2013). Financial decentralization provides local governments a convenient method to fight for financial resources.
( 3) Ideally, the alignment between fiscal and financial decentralizations will propel economic development. Yet when the two are at odds with each other, central and local objectives will diverge, giving rise to structural distortions and imbalances. While the boundary of fiscal decentralization is clear, financial decentralization is embodied in the explicit centralization of industry access, regulation and rel ief on one hand and the impl ici t decentralization of financial resources allocation on the other. Inconsistencies between fiscal and financial decentralizations will to some extent spawn financial risks.
Obviously, fiscal and financial decentralizations are closely related variables important to stability and development. With fiscal decentralization as an entry point, this section will uncover the performance and characteristics of financial decentralization, as
well as their mutual relationship, demonstrate the conclusions of model analysis in Section 2, and identify propositions for the econometric empirical study on the relationship between fiscal decentralization and financial decentralization in Section 4.
2.1 Fiscal Decentralization and the Spatial Effect of Public Finance
For a large and p opu lou s count ry, government administration must be properly decentralized to ensure information symmetry and economic efficiency. A representative economic theory is fiscal decentralization under market federalism. “Market- preserving federalism” requires imposition of stricter limits on local fiscal budgets in central- local fiscal arrangements (e.g. transfer payment) and embed formal and informal systems to incentivize public decision- makers in allocating fiscal functions among various levels of government ( Oates, 1999). According to the hypothesis of “Chinese federalism”, administrative decentralization since the 1980s has enhanced local government economic decision- making power and concurrent fiscal decentralization allowed local governments to share fiscal revenues with the central government, both of which create important incentives for local governments (Montinola et al., 1995; Blanchard & Shleifer, 2001; Jin et al., 2005). On such basis, some scholars investigated the internal governance of the Chinese government from the standpoint of official promotion incentives and examined their effects on interprovincial economic operation (Maskin et al., 2000; Zhou, 2007), believing that fiscal decentralization is partially responsible for interprovincial economic and fiscal disparities. As proven in reality, the economic variables of provinces have spatial spillovers and are not independent from each other. Thus, under fiscal decentralization, the fiscal variable of provinces should exhibit differentiated but correlated characteristics of spatial structure, i. e. spatial effect. Hence, we arrive at the following proposition:
Proposition 1: Fiscal decentralization is the foundation of interprovincial fiscal disparities and responsible for the spatial effect of interprovincial fiscal variable.
2.2 Financial Decentralization and the Spatial Effect of Financial Intermediation
China’s financial institutional arrangement results from the compromise of various parties in the course of changing political, economic and cultural systems, a process largely shaped by central- local decentralization. Financial decentralization provides a unique perspective for unravelling the patterns of China’s financial operation. According to Ding and Fu ( 2012), China has developed an institutional combination of fiscal decentralization and financial centralization after many rounds of central-local power adjustments. Yet this paper considers that the current definition of China’s centralized financial system is not accurate: financial decentralization should be the division of rights to control and supervise financial resources between central and local governments, including the rights of financial development ( i. e. the right to develop financial institutions, markets and infrastructures), the rights of financial control (including ownership control, operational control and personnel control) and financial regulation ( including market access, daily supervision and remedies). Specifically, (1) local governments may acquire controlling shares in local and national institutions ( e. g. rural credit cooperatives and city commercial banks) despite central government control of financial licensing; (2) local governments have the right to grant administrative approval for certain types of financial institutions ( such as microcredit companies, small- sum guarantee companies , pawnshops, online lending companies and crowdfunding platforms), which means that financial industry access is not entirely monopolized by the central government; ( 3) while national banks and markets are centralized, local governments still have significant power to intervene in state- owned banks and shareholding banks, requiring them to satisfy local government financing needs in exchange for administrative conveniences. By the degree of financial decentralization, China’s
financial system should be classified as explicit centralization with implicit decentralization, i . e . explicit centralization of approval, supervision and remedies coexists with implicit decentralization at the operational level.
Pressured by growing expenditures despite falling revenue, local governments have intervened in the development of financial sector, endowing it with fiscal functions. The central government, seemingly in control of financial market access, has to be fully responsible for systemic financial stability and relief under current legal framework. Due to lack of accountability, local governments fight for control over national financial resources without assuming financial risks accordingly. Tax sharing reform has increased local government intervention and influence in the lending decisions and behaviors of local banking systems, giving rise to interprovincial spatial disparities in financial resources. Due to implicit financial centralization and the spillover of regional economy, financial industry should exhibit differentiated and correlated spatial characteristics across provinces. Thus, we have:
Proposition 2: Under explicit centralization and implicit decentralization of financial sector, fiscal decentralization has spurred local governments to fight for financial resources and the variable of interprovincial finance has a certain spatial effect.
2.3 Consistency and Contradiction of Financial and Fiscal Spatial Effects
Structural adjustment of public finance from centralization to decentralization requires consistent change in the degree of financial centralization. Otherwise, the two will contradict with each other and compromise policy effects and economic equilibrium. To some extent, fiscal decentralization influences financial centralization. Yet due to lag effect and mismatch, they often conflict and contradict with each other. By investigating central- local relationship, Zhou ( 2007) concluded based on the history of fiscal reforms that China has experienced many “fiscal decentralizationcentralization cycles” since the People’s Republic’s founding in 1949. Similar cycles may also be found in financial reforms, which coincided with fiscal reforms. Public finance and financial sector, both endogenous to local economic system, are subject to local economic and fiscal strength. Given the effects of fiscal decentralization on financial decentralization, financial sector and public finance should have similar interprovincial spatial effects.
From the reality of economic operation, fiscal decentralization structure is sometimes at odds with financial decentralization structure. In the 1970s, China’s central-local relationship was characterized by fiscal decentralization and financial centralization; since the 1980s, China’s fiscal system evolved from “division of revenue and expenditure between the central and local governments and with contracts at different levels” to tax sharing reform with an increasingly decentralized fiscal structure. However, the boundary between financial centralization and decentralization was vague. Local governments had the authority to appoint the senior management of branches of the big four state- owned commercial banks and national shareholding commercial banks in their jurisdictions. Obviously, the degree of financial decentralization increased. Since 1997, the PBC and the four state- owned commercial banks implemented vertical management; the PBC replaced provincial branches with nine regional branches - an indication of increased financial centralization. This round of centralization was followed by the rapid emergence of local financial institutions, including city commercial banks, township and village banks and microcredit companies; local governments indirectly obtained funds from commercial banks through bond issuance and acquisition of commercial banks using local financing vehicles. These developments are a symbol of explicit financial centralization and implicit decentralization with a more significant tendency of decentralization at the local level. Overall, after many rounds of adjustments, China’s fiscal and financial power structures are characterized by fiscal decentralization with explicit financial centralization and implicit decentralization. The
uncertainty of whether financial sector should be centralized or decentralized is in mismatch with the certainty of fiscal decentralization. Thus, we have:
Proposition 3 : Given that fiscal decentralization affects financial decentralization and both are subject to regional economy, when coordinating with each other, financial variable and fiscal variable share similar spatial effects and when in conflict, they exhibit differentiated spatial effects.
3. Variables, Methodology and Data 3.1 Variables
First, fiscal decentralization: the shares of fiscal revenues for various localities, denoting the abilities of local governments to divide national income. In this paper, fiscal decentralization is denoted by fiscal revenue of various provinces / national
fiscal revenue, i. e. . It needs to be noted that the shares of provincial fiscal revenues do not add up to 1 but are equal to the shares of local fiscal revenues in national total fiscal revenue.
Second, financial decentralization: it is more difficult to depict financial decentralization because the qualitative indicators of local government influence must be identified for various financial departments and functions and financial indicators are varied ( e. g. bank loans, number of listed companies and capitalization, size of local financing vehicles) and cannot be simply added up. Thus, this paper uses the share of bank loans at various localities to denote financial decentralization for the following reasons: ( 1) bank loans are the dominant asset type for banks and provide important clues for bank operations. Meanwhile, commercial bank loans are still the most important form of financing and financial resources and the shares of bank loans may reflect financial decentralization structure; ( 2) one of local government priorities is to increase lending support from major banks and national shareholding banks; ( 3) normally, bank loans are positively correlated with other financial indicators and move in the same interaction with the change in total financial resources. Of course, the downside is obvious as well. Shares of bank loans are influenced more by the level and structure of local economic development and the business decisions of bank headquarters and less by the central-local power structure. They reflect both centralization and decentralization. Yet before a better indicator of financial decentralization is found or designed, this paper still adopts the shares of bank loans as an imperfect indicator. Thus, financial decentralization is denoted by the share of bank loans for various localities, i.e.
, which indicates the abilities of local governments to influence and utilize national funds.
To inves t igate the effects of fiscal decentralization on financial decentralization, this paper defines the shares of bank loans as one of the explained variables and the shares of fiscal revenue as explanatory variable.
3.2 Research Methodology
This paper first utilizes cross- section data regression method to estimate the effect of fiscal decentralization ( share of fiscal revenue) on financial decentralization ( shares of bank loans) and considers the share of bank
deposits as a variable controlling for the l evel o f economic development. The reasons are twofold: first, the data of bank deposits for various provinces are more accurate than output data; second, the data of bank deposits provide a more realistic picture of economic situations in various provinces. We create Model (1):
Yet in reality, independent observations do not universally exist, not least when it comes to
spatial data. Consequently, two spatial effects are often overlooked in economic research: spatial dependence and spatial differentiation (Anselin, 1994). When spatial differentiation coexists with spatial correlation, the classic method of econometric estimation is no longer valid. Spatial dependence (i.e. spatial autocorrelation) is often denoted and depicted by the following models: spatial error model if the error terms of model are spatially correlated; spatial lag model if the spatial dependence of variables appears to be critical to the model to the extent that results in spatial correlation ( Anselin, 1995). Spatial heterogeneity reflects the instability that widely exists in the relationship of economic behaviors among spatial observation units in the economic practice. A sound solution is the geographically weighted regression ( GWR) model of spatial coefficient.
With spatial dependence brought under consideration, this paper first conducts a spatial correlation pre- test. Among various methodologies for spatial statistics and spatial econometrics, the most common are Moran’s I index, Geary’s C ratio and Getis index. The difference between Moran’s I index and Geary’s C ratio is that Moran’s I is intended for global spatial correlation analysis while Geary’s C index is applicable to regional spatial correlation analysis. Hence, this paper adopts Moran’s I index to estimate the geographical correlation of China’s interprovincial public finance and financial sector, employs spatial error model ( SEM) to estimate the direction and degree of the effects of financial decentralization observations of adjacent regions on the financial decentralization observations of local region, and verifies the spatial effects of fiscal decentralization on financial decentralization through GWR model.
3.3 Data Sources
Provincial and national fiscal revenue and bank loan data are derived from China Statistical Yearbook and China Financial Statistical Yearbook. Panel data of relevant variables are calculated according to the above indicators. Combining both definitions, the geographical weight matrix in this paper acquires GIS latitude and longitude coordinates of various provinces with latitude as X axis and longitude as Y axis to calculate the distance between two points ( centers of both regions) based on Euclidean distance. Then, threshold distance is automatically configured in GeoDa 1.4.0 spatial econometric analysis software pack, defining a certain distance as adjacent observations for the creation of an adjacent structure to generate geographical spatial weight matrix. Relevant variable data of sample provinces are processed into tertile data, i. e. developed regions, moderately developed regions and less developed regions. Sample provinces include Beijing, Shanghai, Tianjin, Chongqing, Heilongjiang, Jilin, Liaoning, Hebei, Inner Mongolia, Shandong, Shanxi, Henan, Hubei, Hunan, Jiangsu, Anhui, Zhejiang, Jiangxi, Fujian, Guangdong, Guangxi, Hainan, Sichuan, Yunnan, Tibet, Guizhou, Shaanxi, Ningxia, Gansu, Qinghai and Xinjiang.
4. Econometric Analysis
4.1 Spatial Distribution
This paper selects the cross- section data of six years for the comparison of spatial distribution and change of fiscal and financial decentralizations in various years, including 1978 ( watershed year for China’s launch of reform and opening up program); 1988 (outburst of serious inflation); 1993 ( development of socialist market economy); 1998 ( creation of the PBC’s regional branches to replace provincial branches and the eruption of Asian Financial Crisis); 2003 ( launch of China Banking Regulatory Commission and a new round of rural financial reform) and 2012 ( end of US financial crisis and formation of a new generation of China’s leadership). In this manner, we generate the tertile spatial distribution maps for fiscal decentralization , financial decentralization and provincial economic status for the above years. Due to limit of length, we only provide the characteristics of tertile map distribution of fiscal decentralization, financial decentralization
and economic variable. The higher share of local fiscal revenue in national fiscal revenue, the stronger local fiscal strength will be after fiscal decentralization, and vice versa. Share of local fiscal revenue depicts the structure of national fiscal decentralization. In comparison of fiscal decentralization structure in six years, Shanghai, Beijing, Guangdong, Jiangsu, Zhejiang, Shandong and Liaoning have the highest shares of fiscal revenue; Tibet, Qinghai and Ningxia have the lowest shares of fiscal revenue and other provinces are in the middle. Bank deposits in various provinces denote the levels of local economic development, household income and savings. In comparison of the shares of bank deposits in the six years, in addition to Shanghai, Beijing, Guangdong, Jiangsu, Zhejiang, Shandong, Sichuan and Liaoning that boast high shares of bank deposits, Hebei, Henan and Hubei have also increased to a relatively high level; Tibet, Qinghai, Ningxia, Xinjiang, Guizhou and Gansu are at the bottom in terms of the share of bank deposits and other provinces are at the medium level. Among them, Inner Mongolia experienced wild volatility. Share of provincial bank loans in national bank loans denotes the amount of national financial resources shared by the province. While Shanghai, Beijing, Guangdong, Jiangsu, Zhejiang, Shandong, Sichuan and Liaoning rank high in terms of financial resources distribution, Fujian and Inner Mongolia have also jumped in the ranking. In comparison, share of provincial bank loans is consistent with provincial fiscal decentralization and provincial shares of bank loans.
4.2 Regional Financial Spatial Correlation
This paper calculates the spatial correlation of bank loans and fiscal revenue for various provinces, as shown in Figure 1:
Similar spatial correlation exists for both the share of provincial fiscal revenue ( value range 0.2-0.3) and the share of provincial bank loans ( value range 0.1- 0.3). During 19922008, the two shared very similar spatial correlation in terms of the value, tendency and the band and frequency of volatility. During 2009- 2012, however, the spatial correlation increased for the share of provincial fiscal revenue ( fiscal decentralization), i. e. increased spatial dependence, but reduced for
the share of provincial bank loans ( financial decentralization). This may have to do with the introduction of a host of prudent monetary policies in China in the wake of US sub-prime mortgage crisis in 2008. Strained by the dearth of financial resources, local governments resorted to financing vehicles and the form of financial resources shifted from loans to government bonds, resulting in the reduced correlation of loans across regions. Meanwhile, Beijing, Shanghai and some other prosperous coastal regions outpaced inland provinces in terms of Internet finance, which further reduced regional correlation.
4.3 Comparison of Cross-section Estimation
To further analyze the quantitative relationship between fiscal decentralization and financial decentralization, this paper conducts an estimation using simple cross- section regression, spatial error model estimation and geographical weighted regression (GWR). Here, the number of cross-section observations is 31 for all major years, which is not small sample in a strict sense (small sample means fewer than 30 samples) but is vulnerable to be questioned in robustness test. In creating cross- section regression model, Jeffrey et al. (1996) employed the data of 20 emerging market economies with six explanatory variables but still arrived at conclusions that passed the test, while this paper has employed more cross- section samples. In addition, the estimation results in the following paragraphs have verified the conclusions of Sections 2 and 3 and are consistent with China’s economic realities. The estimation results are summed up in Schedule 1. Adjusted for goodness of fit, R2 test value increases from 0.7 for 1978 and 1988 to above 0.95 after 1998, which shows increasingly sufficient power of fiscal decentralization and share of bank deposits to explain financial decentralization. As can be found from the observation of the natural logarithm of the likelihood function, Akaike Information Criterion and Schwarz Criterion, the higher the logarithm value of the likelihood function, the higher the natural logarithm value of the likelihood function will be for the model estimation after consideration of spatial effect and the better result of goodness of fit. This indicates a certain degree of spatial correlation and agglomeration for both fiscal and financial decentralizations.
As can be known from the parametric estimation result, fiscal decentralization has a significant explanatory power for financial decentralization. The correlation coefficient of the share of fiscal revenue for the share of bank loans is in the range of 0.5 ~ 2.8 and significant at 1% for 1978; this coefficient drops to 0.3 ~ 0.6 for 1988 and 1993 and is still significant at 1%; it further increases to 0.5 ~ 1 for 1998 and 2003 and is significant at 5%. For 2012, the correlation coefficient of the share of fiscal revenue for the share of bank loans is in the range of -0.2 ~ -0.35. A possible reason for this negative correlation could be that after the global financial crisis struck in 2008, the Chinese government beefed up infrastructure investments and supported bank loans to maintain economic growth, which resulted in a higher growth rate of bank loans for infrastructure investments such as road and bridge construction in less developed regions than in more developed regions, while the changing shares of bank loans across provinces were not accompanied by similar changes in the shares of fiscal revenues. These results are consistent with the change of Moran’s I stated in the foregoing paragraphs.
4.4 Panel Estimation
Based on Equation (1), this paper substitutes panel data for regression estimation with results shown in Schedule 2. As can be learned from the observation of Schedule 2, the share of fiscal revenue has an important explanatory power for the share of bank loans and the correlation coefficient is positive with the test value significant at 1%. Results of Akaike Information Criterion and Schwarz Criterion, D-W test and p value test are desirable. This indicates that financial sector is implicitly influenced by fiscal decentralization and also takes on the trait of decentralization.
Our econometric analysis further verified the views and propositions of the above
model analysis. Under fiscal decentralization and explicit financial centralization/ implicit decentralization, local governments have exhibited the following behavioral patterns: first, local governments have evolved into actors with special interests and utility preferences in the course of market-based economic operation, being able to pursue their own administrative objectives (He, 2007). China’s local legislatures and judicial systems are not fully- functioning accountability mechanisms; information asymmetry has undermined the ability of central government to hold local government officials accountable. Division of power without sufficient accountability expanded the space for local governments to act on their own will (Yu, Gao, 2012).
Second, the pursuit of maximal local economic interests to achieve growth as primary economic policy targets has driven local governments to use their power and influence to expand budgetary and non-budgetary revenues, including bank loans and other financial resources. The decentralized access to financial resources, coupled with centralized financial market access, supervision and risk relief, has led to a mismatch between accountability and access to financial resources. When financial resources are insufficient to meet their regional and personal interests, local governments will use their clout to expand into such sectors as financial and land resources, leaving the consequences of risk to the central government (Miao, 2013). Growth-centric behaviors of local governments and mismatch between power and responsibility have spawned financial risks.
5. Concluding Remarks
Based on the research, this paper has uncovered the difference of optimal choice and orientation of fiscal and financial behaviors between central and local governments due to different policy objectives. Public finance and financial sector have become financing instruments with certain convertibility under local economic growth framework. Fiscal decentralization, which inevitably affects financial decentralization, has laid the foundation for fiscal disparities across provinces and caused certain spatial effects of interprovincial fiscal variable. In comparison, financial sector is characterized by explicit centralization and implicit decentralization. For instance, the centralization of financial access approval, regulation and risk relief coexists with the decentralization of access to financial resources. The mismatch between financial explicit centralization/ implicit decentralization and fiscal decentralization has goaded local governments into fighting for financial resources and the interprovincial financial variable also has a certain spatial effect. Under the nonequivalence between financial return and risk, the mismatch gave rise to the division of power between central and local governments with limited accountability, allowing local governments to act on its own will over a broader range of issues. Unchecked economic expansion of local governments from fiscal revenues to financial resources including bank loans has spawned fiscal and financial risks. In view of this, He (2011) considers that at the local level, the explicit management of financial sector coexists with implicit intervention as a result of the GDP championship that prompted local governments to maximize their control of financial sector, inconsistencies between central and local government objectives, as well as limited financing channels available to local governments.
Under the financial system of explicit centralization and implicit decentralization, local governments have left the burden of risks to the central government, giving rise to financial risks. This problem cannot be resolved through complete financial decentralization. We believe that the solutions are as follows: first, the horizontal and vertical accountability mechanisms must function effectively to ensure that local governments assume risks proportionate to return; second, local governments should be granted certain approval authority for the access to regional financial markets through appropriate fiscal decentralization and held accountable by the
central government for their efficiency and risk assumption as part of local governance evaluation. In a nutshell, appropriate financial decentralization reform and the sharing of financial risks and responsibilities are among the fundamental solutions to the prevention and mitigation of regional financial risks.