Classic Microfinance Players

According to the Chinese Association for Microfinance, there are five-levels at which classic microfinance is provided.

Informal Suppliers: Informal suppliers include pawnbrokers, high-interest rate lenders, family and friends. Informal suppliers typically charge much higher interest rates than banks, especially in southern China and in eastern seaboard areas of the country. The annual interest rate they charge range from 23 percent to 27 percent on average, but have been known to rise up to 50 percent. Family and friends are also a common source of financing for micro-entrepreneurs; while these sources may not charge interest, borrowers often pay for the favor with other in-kind compensation (such as helping in the fields during busy season).

Non-Governmental Organization Microfinance Institutions (NGO MFIs): NGO MFIs refer to projectbased microfinance generally aimed at alleviating poverty or development. Most of these are operated either by NGOs, international organizations or social organizations based in less-developed middle and western regions of China. The expansion potential of these organizations is limited due to the lack of legal financial institution or corporate status, rendering them unable to mobilize deposits or take on debt. As such, many of these organizations still operate under cash donations. Many of these programs are looking to transform themselves into Microcredit Companies, or stay as project offices operating under a local government or development association. The largest classic microfinance program is the microfinance arm of the China Foundation for Poverty Alleviation (CFPA) which began in 1996. Another large player in this category is the Funding for the Poor Cooperatives (FPC), first introduced in 1993 and modeled after the Grameen Bank group lending methodology. Examples: PLAN International, Heifer Project International, Mercy Corps International

Rural Mutual Credit Cooperatives (RMCCs): Rural mutual credit cooperatives are community-based mutual financial institutions jointly funded by the village/town residents and local enterprises. The main sources of capital for RMCCs are members' deposits, external endowments and/or funds from other banking institutions. This capital may be used to issue loans to members, make deposits in other banking institutions, or purchase government bonds and financial bonds to satisfy members' funding demand. RMCCs can also engage in settlement business, but they are not allowed to receive deposits from or grant loans to non-member persons or entities, or to provide other kinds of financial services, or furnish security for other persons or entities with their own assets. Example: Association for Rural Development of Yilong County Group Lending Initiative(Sichuan Province)

Commercial Banks: Commercial banks refer to downscaled financial institutions. Many financial institutions that do not traditionally operate within the microfinance space have created business lines to cater to this market by downscaling the size of their loans. These institutions can be both traditional banks and non-banking institutions. As financial institutions, commercial banks offer traditional banking services. Example: Harbin Bank (Heilongjiang Province)

Rural Credit Cooperatives (RCCs): Rural credit cooperatives, first launched in the 1950s, account for 32 percent of all SME lending by the end of 2010. RCCs usually have a presence in townships and are thus one of the most pervasive of all financial institutions in China. RCCs have begun to RCCs have begun to issue loans of up to RMB 50,000 (EUR 5,500) without collateral. This is due in large part to use of previously untried forms of guarantee such as joint-guarantees (often with a local salaried government employee) or land management right certificates in lieu of a traditional guarantor. However, farmers continue to face complicated application procedures and long processing periods; in some cases, applicants must have good relations with the loan officer to secure approval. Recent reforms have improved the utilization of deposits for credit business. Some RCCs are going through a transformation process to become rural commercial banks or rural cooperative banks. Example: Gansu Rural Credit Cooperative (Gansu Province)

Page 1 Administrative Decentralization and Government Size in China: An Empirical Evaluation Dr. Xin ZHANG Associate Professor of Public Policy School of Public Administration Renmin University of China Beijing 100872, China Tel: 86-10-62516308 E-mail: zhangxin1996@yahoo.com.cn Paper to be presented at the Public Administrative Specialist Group Panel (Panel 42): Contemporary Challenges to Public Administration in Conference 2010: 60 years of political studies sponsored by the Political Studies Association March 29-April 1 st , 2010 Edinburgh, UK Copyright PSA 2010 Page 2 Abstract China's unitary administrative system under a command economy, in response to marketoriented reforms ever since the late 1970s, has been decentralized through several stages such as organizational streamlining, functional retreat, and services provision. However, it is theoretically and empirically argued that if the size of government in terms of performance evaluation is effectively

constrained through time. Based upon the Leviathan hypothesis by James M. Buchanan, this paper tries to evaluate empirically through multiple regression models if administrative decentralization leads to downsize government with the case of China in transition. The findings based upon the time-series and cross-section data suggest the robustness of the Leviathan hypothesis. Keywords: administrative decentralization; Leviathan hypothesis; government size; economic development level 2 Copyright PSA 2010 Page 3 Introduction Ever since the late 1970s China has taken an institutional transition from a command economy towards a market economy, and meanwhile the unitary system of administrative state has been adaptively decentralized through several stages of reforms. According to the broad theoretical political economy framework of fiscal federalism, public choice, and price mechanism, it makes sense on the one hand that market-oriented reforms should correspond with administrative decentralization (Weingast, 1995; Qian and Weingast, 1997; Davoodi and Zou, 1998). On the other hand, it is argued that government size should be constrained for sustainable economic growth through the emerged private sector as the engine (Peden, 1991; Karras, 1996; Lane, 2000). Thus, such a derived research hypothesis with respect to the case of China in transition is as follows: whether or not administrative decentralization leads to a constrained (smaller) government size. As a matter of fact, there are several relevant studies done. However, their findings are not consistent to some extent. For instance, Wallace E. Oates in his empirical study of searching for Leviathan (1985) did not find any strong, systematic relationship between the size of government and the degree of decentralization on the basis of the data from the world sample, industrial countries, and developing countries. Moreover, the Leviathan model seems revived in terms of the growth of government under fiscal federalism (Oates, 1999; Rodden, 2003). As far as the case of transitional China is concerned, a negative relationship between fiscal federalism and economic growth is significantly found (Zhang and Zou, 1998), while it is empirically found a positive impact of increased fiscal autonomy of Chinese provinces with economic growth (Lin and Liu, 2000). Precisely, it is argued that the size of government growth rather than fiscal decentralization has a positive effect on the rate of economic growth. To put it differently, if government size is regarded as the independent variable (cause), and economic growth rate as the dependent variable (effect), then fiscal decentralization is treated as the interfering variable. Thus, the relationship between government size and fiscal decentralization needs to be spelled out empirically as well as theoretically. This paper tried to relate the Leviathan hypothesis 1 put forth by Brennan and Buchanan (1980 and 1985) with the case of China in transition. Both time-series and cross-section data are separately analyzed through multiple regression models to warrant the derived empirical proposition that government size can be constrained by decentralization. Administration Reforms of China in Transition 1 Brennan and Buchanan argue that fiscal decentralization has an effective constraint on Leviathan (unitary big government), ie, competition among multiple governments can put limits on the monopolized revenue maximization of a centralized government system. 3 Copyright PSA 2010 Page 4

China's unitary administrative system under a command economy, in response to market-oriented reforms ever since the late 1970s, has been decentralized through several stages. The first stage was started in 1982, focusing on organizational streamlining for all the government levels through a topdown approach, to adapt the national policy of reforming and opening-up (modernization). Consequently, representative ministries and agencies of central planning and economic administration were canceled, changed, or reorganized during this period, and meanwhile some of them were transformed into group corporations of the state-owned enterprises. Unlike the 1982 administrative reforms, the second stage since 1988 emphasized the great transformation of the functions and modes of (central rather than local) governments. Specifically, economic administrative organizations of (central rather than local) governments were required to change from direct administration to indirect regulation, to strengthen macro-regulatory functions but weaken micro-managerial functions, and to legalize the administrative process on the basis of the so-called scientific allocations of governmental functions in the context of market-oriented reforms. Year 1993 marked that establishing a market economy was officially set up for the future of China. Correspondingly, the third stage of administrative reforms lied in the institutional as well as functional separation between government and state-owned enterprises towards the market-based allocation and utilization of resources. During this period, most of economic administrative ministries and agencies were transformed into group corporations, industrial associations, or state-assets trusts, or regulatory agencies, and meanwhile the governmental function was separated from civil society organizations of local communities, various associations, and public services institutions. To some extent, the 1998 reforms were influenced by the movements worldwide of reinventing government as well as by the 1997 World Development Report titled The Role of the State in the Changing World. Just as the process of marketization went on further, professionalizing the state administration was advocated by means of establishing the system of civil servants nationwide, limiting the governmental scope of economic as well as social functions, and legalizing the process of the governmental system through constitutional and administrative laws. It seems that the mixed economy of government and market has been recognized based upon the experiences of institutional reforms. The latest stage was the 2003 reforms, defining the government functions in the context of a marketoriented economy as economic regulation, market supervision, social administration, and public services, based upon three essential principles: (1) the consistency principle of rights, obligations, and responsibilities; (2) the 4 Copyright PSA 2010 Page 5 competency principle of streamlining, unification, and efficacy; (3) the coordination principle of decision, implementation, and supervision. However, it is argued that administrative system reforms in China did not get rid of the vicious circles (1) from streamlining to expanding and from re-streamlining to reexpanding, (2) from consolidating to separating and from re-consolidating to re-separating, and from centralizing to decentralizing and (3) from re-centralizing to re-decentralizing. To put it differently, the argument here lies in the limits to government size through administrative system reforms for the sake of economic growth. As a matter of fact, the government size in terms of the total expenditure share in gross domestic product (GDP) went sharply downwards from the late 1970s to 1994 and went steadily up since then to 2007, while the government size in terms of the government staff per 10,000 citizens

over time climbed up from 1978 to 1992 and was stabilized 1993 to 2007 (see Figure 1). In other words, the trends of government size through time in Figure 1 can be interpreted as follows: both indicators of government size are shown decreased during the Jiang-Li Administration (1992-1998), and in contrast both indicators are shown increased during the Hu-Wen Administration (2002-2008). In addition, during the Jiang-Zhu Administration (1998-2002), one indicator of staff-based size is shown decreased, and the other expenditure-based size shown increased. Finally, during 1978 to 1992 (before building a marketoriented economy was officially legislated), the staff-based indicator is shown increased, and the expenditure-based indicator shown decreased. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 Time T o ta le xpe ndit u r e s h a r e in GD P ( % ) 0.00 20.00 40.00 60.00 80.00 100.00 120.00 Gove r nm e nt s t a f f pe r 10, 0 00 c itiz e ns ( N o. ) Expenditure-based size Staff-based size Figure 1 Government size through time: 1978-2007 Obviously, 1993 and/or 1994 was the point of demarcation with respect to the change trends of the two government size indicators. The opposite trends downwards and upwards were showed before the demarcation point (expenditure-based size down 5 Copyright PSA 2010 Page 6 and staff-based size up), while the parallel trends upwards were shown after the point (both expenditure-based and staff-based size ups). Significantly, 1993 was the official year towards a market economy, and then 1994 was the official year towards a compound fiscal system. Thus, it seems to conclude that market-preserving federalism makes sense accounting for the government size and its change of China in transition (Weingast, 1995; Qian and Weingast, 1997). 4.40 4.30 4.20 4.10 4.00 3.90 3.80 Provincial expenditure in the total (in the ln form) 3.60 3.40 3.20 3.00 2.80 2.60 2.40 T o tal expenditure in GDP (in the ln f o rm) 2007 2006 2005 2004 2002 2001 2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 1988 1987 1986 1985 1984 1983 1982 1981 1980 1979 1978 Figure 2 Scatter plot of expenditure-based size and decentralization The Leviathan hypothesis revisited Just as Wallace E. Oates said that the proper goal of restructuring the public sector cannot simply be decentralization, actually it needs to understand which functions and instruments are best centralized and which are best placed in the sphere of decentralized levels of government (Oates, 1999). As a matter of fact, unlike the former Soviet Union and Eastern European countries, socialist China since its founding has taken on administrative decentralization in the context of a dual governmental system of central-provincial relations. 6 Copyright PSA 2010 Page 7 4.40 4.30 4.20 4.10 4.00 3.90 3.80 Provincial expenditure in the total (in the ln form) 4.60 4.50 4.40 4.30 4.20 4.10 4.00 3.90 3.80 T o tal government staf f per 10,000 citizens (in the ln f o rm) 2007 2006 2005 2004 2002 2000 1999 1998 1997 1995 1992 1991 1990 1988 1987 1986 1985 1984 1983 1982 1981 1980 1979 1978 Figure 3 Scatter plot of staff-based size and decentralization Over past 30 years of reforming and opening up, China showed remarkably further decentralization in the need of adapting to market-oriented reforms. Given the two indicators of government size, one is based on the total expenditure share in the gross domestic product (GDP), and the other is on the total government staff number per 10,000 citizens, together with the indicator of administrative decentralization in terms of

the provincial expenditure share in the total, the relationships between government size and decentralization is able to be elucidated (see Figure 2 and Figure 3). As for the relationships between expenditure-based size and decentralization, four periods were categorized to catch the essence of correlation, which were the first period from 1978 to 1984, taking on government size downwards but administrative decentralization downwards (rurual reforms started in 1978); the second period from 1985 to 1995, taking on government size downwards and administrative decentralization upwards (urban reforms started in 1984); the third period from 1996 to 1999, taking on government size upwards and administrative decentralization downwards; and the last period from 2000 to 2007, taking on government size 7 Copyright PSA 2010 Page 8 upwards and administrative decentralization upwards. By the same token, with respect to the relationships between staff-based size and decentralization, five periods were categorized to catch the essence of correlation, which were the first period from 1978 to 1984, taking on government size upwards and administrative decentralization downwards (rural reforms started in 1978); the second period from 1985 to 1989, taking on government size upwards and administrative decentralization upwards (urban reforms started in 1984); the third period from 1990 to 1995, taking on government size downwards and administrative decentralization upwards; the fourth period from 1996 to 1999, taking on government size upwards and administrative size downwards; and the last period from 2000 to 2007, taking on government size upwards and administrative decentralization upwards. Research design and data Based upon the Leviathan hypothesis, an empirical proposition is then derived that government size can be constrained by decentralization. The general multiple regression models are set up accordingly (see below), in which government size are indicated by the expenditure-based (size 1) and the staff-based (size 2), economic development level (EDL) by per capita gross domestic product (per capita GDP), and administrative decentralization (AD) by the provincial expenditure share in the consolidated (national) expenditure. More specifically, the expenditure-based size is measured with the provincial/national expenditure share in the provincial/national gross domestic product (GDP), and the staff-based size is with the government staff number per 10,000 citizens. Size 1/ Size 2 = 0 (constant) + 1 *AD + 2 *EDL + error Table 1 Overview of dependent and independent variables by region Region 2000 2007 Size 1 (%) Size 2 (staff no.) AD (%) EDL (RMB yuan) Size 1 (%) Size 2 (staff no.) AD (%) EDL (RMB Yuan) National 16.01 87.11 65.3 7,858 19.95 97.72 77 18,934 Beijing 17.87 178.73 2.79 22,460 17.64 200.24 3.31 58,204 Tianjin 11.41 120.88 1.18 17,993 13.35 118.39 1.35 46,122 Hebei 8.17 95.64 2.62 7,663 10.99 104.57 3.03 19,877 Shanxi 13.69 127.69 1.42 5,137 18.31 134.10 2.11 16,945 InnerMongolia 17.65 125.00 1.56 5,872 17.77 131.39 2.17 25,393 Liaoning 11.10 95.33 3.26 11,226 16.00 110.98 3.54 25,729 Jilin 14.31 92.01 1.64 6,847 16.72 108.06 1.78 19,383 Heilongjiang 11.74 105.99 2.40 8,562 16.80 104.60 2.38 18,478 8 Copyright PSA 2010 Page 9 Shanghai 13.37 90.20 3.83 34,547 17.90 96.88 4.38 66,367 Jiangsu 6.89 62.38 3.72 11,773 9.92 75.54 5.13 33,928 Zhejiang 7.14 74.41 2.71 13,461 9.62 100.00 3.63 37,411 Anhui 10.65 68.33 2.04 4,867 16.89 71.59 2.50 12,045 Fujian 8.27 83.26 2.04 11,601 9.85 80.42 1.83 25,908 Jiangxi 11.16 83.09 1.41

4,851 16.45 87.91 1.82 12,633 Shandong 7.18 86.13 3.86 9,555 8.71 105.48 4.54 27,807 Henan 8.67 87.29 2.80 5,444 12.46 103.21 3.76 16,012 Hubei 8.62 87.59 2.32 7,188 13.84 92.12 2.57 16,206 Hunan 9.42 82.76 2.19 5,639 14.75 102.91 2.73 14,492 Guangdong 11.18 80.65 6.80 12,885 10.16 93.24 6.35 33,151 Guangxi 12.61 76.63 1.63 4,319 16.55 76.13 1.98 12,555 Hainan 12.37 111.82 0.40 6,894 20.04 101.78 0.49 14,555 Chongqing 11.81 64.08 1.18 5,157 18.64 77.06 1.54 14,660 Sichuan 11.27 71.92 2.85 4,784 16.75 80.60 3.53 12,893 Guizhou 20.29 77.73 1.27 2,662 29.01 91.71 1.60 6,915 Yuennan 21.18 84.19 2.61 4,637 23.94 93.04 2.28 10,540 Tibet 51.06 217.56 0.38 4,559 80.47 260.56 0.55 12,109 Shaanxi 16.36 106.24 1.71 4,549 19.28 114.99 2.12 14,607 Gansu 19.14 90.16 1.18 3,838 24.99 131.45 1.36 10,346 Qinghai 25.90 119.69 0.43 5,087 36.01 137.68 0.57 14,257 Ningxia 22.91 112.10 0.38 4,839 27.20 122.95 0.49 14,649 Xinjiang 14.00 130.39 1.20 7,470 22.57 148.93 1.60 16,999 Mean 14.43 99.67 2.15 8,593 19.47 111.56 2.48 21,973 SD 8.35 32.66 1.32 6,523 12.87 38.14 1.40 13,987 CV 0.58 0.33 0.61 0.76 0.66 0.34 0.56 0.64 Based on the national data through time (see Figure 4), it seems there exits a weak positive relationship between administrative decentralization in terms of the provincial (subnational) expenditure share in the consolidated total and economic development level in terms of per capita GDP. Primarily, it can be confirmed that administrative decentralization in such a unitary system as China to some extent leads to economic growth through time. However, taking them into consideration with the two government size indicators (see Figure 1), the relationship between government size and administrative decentralization seems complicated, so does the relationship between government size and economic growth. Moreover, with respect to the provincial indictors of dependent and independent variables (see Table 1), staff-based government size showed less disparity across provinces than expenditure-based government size in 2007 as well as 2000, while the differences across provinces of both indictors increased a little bit from 2000 to 2007 in terms of the coefficient of variation. Moreover, independent variables (AD and 9 Copyright PSA 2010

Page 10 EDL) showed more differences across provinces than dependent variables (Size 1 and Size 2) in terms of the coefficient of variation. In contrast, the differences across provinces of independent variables decreased a little bit from 2000 to 2007. 0 5000 10000 15000 20000 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Time Pe r c a p i t a GD P ( R M B y u a n ) 0 10 20 30 40 50 60 70 80 90 P r ovi nc i a l e xp e ndi t ure s ha r e in t h e c o n s o l id a t e d to ta l ( % ) EDL AD Figure 4 Decentralization and development through time: 1978-2007 Results and discussions Actually, eight multiple regression models are analyzed empirically; two is based upon the time-series data with the equation without the natural logarithmic form, and other six are based upon the cross-section data with the natural logarithmic form. The results of the time-series multiple regression models are shown in Table 2. Concretely, it shows a significant negative relationship of administrative decentralization with the expenditure-based government size through time, and a strong significant positive relationship of administrative decentralization with the staff- based government size through time. It is worthy to mention it is not shown any significant relationship of economic development level with any indicator of government size. Table 2 Time-series multiple regression coefficients Size1 Size2 Constant 33.02*** (5.44) 16.99*** (1.20) AD -0.22* (-2.19) 0.97*** (4.01) EDL 6.63E-005 (0.27) 0.00 (0.79) ARMAAdj. (0,5)

(0,2) Note: t-values are in the parentheses. * denotes a significant level at 0.05, while *** a significant level at 0.00. 10 Copyright PSA 2010 Page 11 With respect to the cross-section data, six multiple regression models were conducted with respect to the 2000 data, the 2007 data, and the change data from 2000 to 2007. The results are shown in Table 3. In detail, as for the 2000 data and the 2007 data, the regression coefficients spell out the strong significant negative relationships of administrative decentralization with both expenditure-based and staff-based indicators of government size. Thus, to great extent, the empirical results confirm the research proposition derived from the Leviathan hypothesis that decentralization can limit government size. Table 3 Cross-section multiple regression coefficients 2000 2007 2007-2000 Size 1 Size 2 Size 1 Size 2 Size 1 Size 2 Constant 2.96* (2.39) 2.84*** (3.92) 4.89*** (4.23) 2.84*** (3.05) 0.93*** (2.96) 0.01 (0.08) AD -0.44*** (-3.91) -0.27*** (-4.06) -0.42*** (-4.42) -0.24*** (-3.17) 0.93* (2.21) 0.12 (0.70) EDL -0.02 (-0.13) 0.21* (2.52) -0.18 (-1.47) 0.21* (2.10) -0.84* (-2.36) 0.08 (0.61) F 10.10*** 8.47*** 17.78*** 5.25* 3.37* 0.95 R 2 0.42 0.38 0.56 0.27 0.19 0.06 Adj. R 2 0.38 0.33 0.53 0.22 0.14 0 Note: tvalues are in the parentheses. * denotes a significant level at 0.05, while *** a significant level at 0.00. Moreover, the results of multiple regression models based upon the 2002 data and the 2007 data show the weak significant positive relationship of economic development level with the staff-based government size, and no relationship with the expenditure-based government size. In paraphrasing the Wagner law, it says that the more economic development, the more government staffing. However, it seems to infer the no or negative relationship of administrative decentralization with economic development level in order to harmonize the empirical results above, and/or the contradictory relationships between fiscal decentralization and economic growth found in other empirical studies reviewed in the introduction section. Finally, with respect to the change data from 2000 to 2007, the results show that the change rate of the expenditure-based government size corresponds positively with the change rate of administrative decentralization but negatively with the change rate of economic development level. And meanwhile, it shows no significant relationships of the change rate of the staffbased government size either with the change rate of administrative decentralization or with the change rate of economic development level. Thus, the change model suggests that the increased change of administrative decentralization through expenditure equalization across provinces corresponds to the larger size in terms of the expenditure-based indicator, while the increased rate of economic growth does to the smaller size. 11 Copyright PSA 2010 Page 12 Conclusion China's unitary administrative system under a command economy, in response to marketoriented reforms ever since the late 1970s, has been decentralized through several stages such as organizational streamlining, functional retreat, and services provision. However, it is theoretically and empirically argued that if the size of government in terms of performance evaluation is effectively constrained through time. Based upon the Leviathan hypothesis by James M. Buchanan, this paper tries to evaluate empirically if administrative decentralization leads to downsize government with the case of China in transition. Two indicators of government size are chosen: one is the share of government

expenditure over gross domestic product, and the other is the number of government employment per 10,000 persons. Moreover, administrative decentralization is indicated by the share of provincial (subnational) government expenditure over the expenditure, while provincial per capita GDP indicates provincial economic development level. The general multiple regression models are set up taking government size as dependent variable, and administrative decentralization and economic development level as independent variables. The time-series models based upon the national data suggest that administrative decentralization leads to the smaller government size in terms of the expenditure-based indicator, but the larger government size in terms of the staff-based indicator. In addition, the crosssection models based upon the provincial data suggest generally that administrative decentralization leads to the smaller government size in terms of both expenditure-based and staff-based indicators. However, the change models suggest specifically that the expenditure-based government growth can be constrained by faster economic growth rather than by increased administrative decentralization through expenditure equalization. It is necessary to mention that the interfering variable of economic development level has no effect on government size are with respect to the time-series models. However, there are not consistent effects of economic development level on government size with respect to the cross-section models. Specifically, with respect to the 2000 and 2007 models, there are weak positive effects of economic development on the staff-based government size, but no effects on the expenditure-based size. And with respect to the change models, there is weak negative effect on the expenditure-based size, but no effect on the staff-based size. To conclusion, the findings with the case of China in transition do demonstrate the robustness of the Leviathan hypothesis. References 12 Copyright PSA 2010 Page 13 Brennan, Geoffrey, & Buchanan, James M. (1980). The power to tax: Analytical foundations of a fiscal constitution. New York: Cambridge University Press. Brennan, Geoffrey, & Buchanan, James M. (1985). The reason of rules: Constitutional political economy. New York: Cambridge University Press. Davoodi, Hamid, & Zou, Heng-fu. (1998). Fiscal decentralization and economic growth. Journal of Urban Economics, 43: 244-57. Karras, G. 1996. The Optimal Government Size: Further International Evidence on the Productivity of Government Services. Economic Inquiry, 34:193-203. Lane, Jan-Erik. (2000). New public management. London: Routledge. Lin, JY and Z.Liu. (20000). Fiscal Decentralization and Economic Growth in China. Economic Development and Cultural Change, 49:1-23. Oates, Wallace E. (1985). Searching for Leviathan: An empirical study. American Economic Review, 75: 748-57. Oates, Wallace E. (1999). An essay on fiscal federalism. Journal of Economic Literature, 37(3): 1120-1149. Peden, Edgar. (1991). Productivity in the United States and Its Relationship to Government Activity: An Analysis of 57 Years, 1929-1986. Public Choice 69:153-73. Qian, Yingyi, & Weingast, Barry R. (1997). Federalism as a commitment to preserving market incentives. Journal of Economic Perspective, 11(4): 83-92. Rodden, Jonathan. (2003). Reviving Leviathan: Fiscal federalism and the growth of government. International Organization, 57: 695-729. Weingast, Barry R. (1995). The economic role of political institutions: Marketpreserving federalism and economic development. Journal of Law and Economic Organization, 11: 1-31. Zhang, Tao, & Zou, Heng-fu. (1998). Fiscal decentralization, public spending, and economic growth in China. Journal of Public Economics, 67: 221-4

The Size of Government over Time

This graph shows the expenditure components of GDP, consumption, investment, government spending, and net exports, as a percentage of GDP:

Expenditure Components as a Percentage of GDP

[These data are from the last slide of a PowerPoint presentation for chapter 18 of Economics by McEachern at this site . The textbook is not one I've used.]

The graph in the textbook that is the source of these data presents this as though the yellow area is the government spending percentage and makes the point that the percentage of GDP arising from government expenditures declines slightly over time while the consumption percentage increases slightly over time. But that's not quite correct, though the picture does not change much when this is accounted for.

The red area at the top is net exports and this value is negative when it is above the 100% line. For example, let C=5, I=1, G=3, and NX=-2. Then GDP=5+1+3-2=7. In the graph, the part above the line is 5+1+3=9 and the red area would be 2 in magnitude. So the government spending percentage is actually the yellow area plus the red area not just the yellow area when net exports are negative (the yellow area is only 1/7 while government expenditures are 3/7). Here's the same graph with both areas colored the same (though on the occasions when net exports are positive this would be misleading, but as is evident in the top graph these areas are very small in magnitude when they exist at all):

Expenditure Components as a Percentage of GDP

According to these data the size of government as measured as a percentage of GDP has remained relatively stable over time rather than falling. However, these data do not support the claim that government has grown substantially larger over the last four decades as is often claimed...

Page 1 THE RISE AND FALL OF CHINA'S GOVERNMENT REVENUE LIN Shuanglin EAI Working Paper No. 150 ISSN 0219-1318 ISBN 978-981-08-3371-8 All rights reserved Date of Publication: 3 June 2009 Page 2 1 THE RISE AND FALL OF CHINA'S GOVERNMENT REVENUE LIN Shuanglin * I. Introduction China's fiscal capacity is of great interest to academia and policymakers worldwide. This paper discusses China's tax reforms, analyzes the rise and fall of government revenue, compares the size of China's government revenue with other countries, examines the structural problems in the current tax system, and provides suggestions for further tax reform. China's government revenue includes three major parts: budgetary revenue, extra-budgetary revenue, and social security contributions. Budgetary revenue consists of tax revenue and state enterprise contributions. Extra-budgetary revenue includes various fees and charges. Social security is run by local governments and social security contributions are included in neither budgetary nor extra-budgetary revenues. There exists also off-budgetary revenue in China, most of which is illegal, consisting of unreported revenue. There is no official data on this government revenue. China's government budget is managed separately by governments at different levels. There are five levels of governments in China: central government, provincial governments (including autonomous regions), prefecture governments, county governments, and township governments. Each level of government has its own budget. The national budget composes of the central government budget and local government budgets. As stipulated by the Budget Law of the People's Republic of China , the budget draft and budget implementation of the central and local governments must be reviewed and approved by the National People's Congress of the People's Republic of China (PRC) and local People's Congresses respectively. 1 The central and local budget adjustment plans and final budgets must be reviewed and approved by the Standing Committee of the National People's Congress and local People's Congresses. The budget year corresponds with the calendar year, starting on January 1 and ending on December 31. * Prof Lin is Professor and Chair, Department of Public Finance, School of Economics, Peking University, Beijing, China 100871; Phone: 8610 8252 4699; Fax: 8610 8252 4675; Email: linpku@gmail.com. Noddle Distinguished Professor, Department of Economics, CBA 512, University of Nebraska, Omaha, NE 68182-0048, USA; Phone: (402) 554-2815. Fax: (402) 554-2853; E-mail: slin@mail.unomaha.edu. He is also Research Associate, East Asian Institute, National University of Singapore. The author thanks Professor John Wong for his valuable suggestions and Di Kang, Jackie Lynch, Yanyan Qi, and Xiaoxi Wang for their assistance. 1 See the National People's Congress of PRC (1994), The Budget Law of the People's Republic of China, passed by the Eighth National People's Congress Second Meeting on March 22. Page 3

2 Under a centrally planned economic system before 1978, the government controlled the production and allocation of resources. Budgetary principle was called uniform collection and uniform spending ( tongshou tongzhi ). State enterprises, which produced most of the output, were required to turn in all of their profits to the state. Profits from state enterprises were the major source of government revenue. 2 There existed no competition among firms and the government set prices for almost all products. The government also taxed farmers explicitly through the agricultural tax and implicitly through lowering its purchasing prices of agricultural products. In 1978, government budgetary revenue share in GDP was 31.2%. China launched a series of tax reforms in the 1980s. As a result, government revenue share in GDP declined significantly. By 1994 government revenue share in GDP declined to 10.8% of GDP. Economists were deeply concerned about insufficient government revenue for infrastructure construction in China. Bahl and Wallich (1992, p. 20) argued that although it was difficult to judge the overall revenue adequacy of China in the absence of concrete projections of expenditure needs, it could be safely concluded that public service levels were deficient in all parts of China and the infrastructure gap was likely to be an especially critical problem in the future. Chen (1996) and Huang (1996) also advocated an increase in the Chinese government budgetary revenue. Stiglitz (1998) believed that the Chinese government revenue share in GDP is too small compared with other countries and is not adequate to fulfill China's ambitious development plan. Brean (1998) argued that Chinas economic development and social stability needed more government revenue, and warned that if the government revenue remained low the end result could be a serious weakening of the government's ability to provide essential public goods and services, which could threaten macroeconomic stability and jeopardize the economic transition. In 1994, China implemented another round of major tax reforms which laid the foundation for increasing government revenue in the late 1990s and in the coming century. The target share of government budgetary revenue in GDP was set at 20% [Wang (1998), former Chinese Minister of Finance]. Government revenue has increased much faster than GDP in recent years. As a result, government budgetary revenue share in GDP has increased to more than 20% of GDP in 2007. Despite the financial crisis and economic slowdown in 2008, government fiscal revenue still grew at 17% in nominal value. 3 Meanwhile, China ranked third in 2007 and fifth in 2008 in "Forbes Tax Misery Index." Recently, economists became concerned with excessive tax burden in China and called for a tax reduction [see Lin (2006)]. What are the reasons for the decline in government revenue in the 1980s and early 1990s? What are the reasons for the increase in government revenues since the late 1990s and recent years? Is the Chinese government collecting too much revenue? How can the current tax system be improved? This paper intends to address these important issues. Section II discusses China's market-oriented tax reforms; Section III analyzes the rise and fall of government revenue; Section IV examines the sources 2 In 1978, tax revenues only accounted for 46% of total government revenue. See National Bureau of Statistics of PRC (1998, p. 273). 3 See National Bureau of Statistics of PRC (2009), Statistical Bulletin on National Economy and Social Development in 2008, February 26, 2009. The inflation in 2008 was 5.9%. So the real growth rate of government revenue was 11.1%, still higher than the GDP growth which was 9% in 2008. Page 4

3 of government budgetary revenue; Section V discusses extra-budgetary revenue; Section VI compares the size of China's government revenue with other countries; Section VII examines China's tax structure and provides suggestions for further tax reforms; and Section VIII concludes the paper. II. Establishing a Tax System for the Market Economy In 1978, China started market-oriented economic reforms. The essence of the reforms was to improve production efficiency through economic decentralization. Tax reform is an important part of economic reform, which is aimed at providing state enterprises with production incentives, cutting off fiscal dependence of state enterprises on government, equalizing tax burdens among enterprises, and providing adequate revenue for the government. Tax reforms experienced the following stages. 4 1. Early Tax reforms Experiments in tax reform began in 1979 by allowing state enterprises to keep part of their profits in order to expand production and to issue bonuses and awards to workers. In 1983, the government decided to pursue further fiscal reform by subjecting state enterprises to income taxes, instead of contributing profits, a reform commonly called substituting taxes for profits ( li gai shui ). This reform was completed in 1984. The tax revenue share in total government revenue increased from 46% in 1978 to 58% in 1984, and has been steadily increasing since then. To increase government revenue, a Contract Responsibility System (CRS) was introduced in December 1986. 5 Under the CRS, enterprises were contracted to pay income tax and adjustment tax at a specific level of profit. 6 If enterprises could not achieve that level of profit, they were to make up the tax shortfall from their own resources. If they exceeded the contract level of profit, they paid taxes at a lower rate on their additional profits. The contract levels were usually based on the previous year's profits plus increment. The CRS provided strong incentives for state enterprises to pursue profits. However, since the contracted profits were not set to grow at a sufficiently high rate, the CRS did not solve the revenue shortage problem for the government. In 1989, the government launched a new tax reform called profit sharing (li shui fenliu) . State enterprises were required to submit a portion of their profits to the government after paying corporate income taxes. All these tax reforms failed to increase the share of total government budgetary revenue in GDP, which declined from 31% in 1978 to only 12.3% in 1993. In addition, since 1978, the central government revenue share of total revenue has decreased significantly, down to only 22% in 1993. The central government was determined to reverse this trend. 4 For a detailed discussion of the early tax reforms, see Lin (2000). 5 The CRS was first introduced in agriculture, then expanded to industry, and finally used in the area of taxation. 6 Contracts do not cover product tax, value-added tax, business tax, resource tax, and other taxes. Page 5 4 2. Establishing the Tax Sharing System (fen shui zhi) In 1994, a new tax system, called the tax sharing system, was established. Several significant changes in the tax system took place, including reduction of the number of taxes from 37 to 23; unification of the income tax rate for all enterprises to 33% (joint ventures preserve their preferential tax rates); division of taxes into three categories, namely, national taxes which were paid to the central government, joint taxes which were shared by the central and local governments, and local taxes which were paid to local governments; and the establishment of a central tax bureau ( guo shui ju ), a local tax bureau ( di shui ju ) and a central to local tax rebate ( shuishou fanhuan ) system. Central government taxes include (1) tariffs (customs duties) and import- related VAT and consumption tax collected by the customs; (2) consumption tax (similar to excise tax in the US); (3)

corporate income tax from central government- owned enterprises (eg, Offshore Petroleum and Natural Gas Corp, PetroChina Company Limited and China Petroleum & Chemical Corporation); (4) income tax from local banks, foreign banks and other financial institutions; (5) income tax from railroad, bank headquarters (Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, Construction Bank of China, Development Bank of China, Agricultural Development Bank of China, Import and Export Bank of China), insurance company headquarters, including business tax, income tax, profit and urban construction and maintenance tax, and (6) profits from central government-owned enterprises. Local government taxes include (1) business tax (excluding business tax from railroad, bank headquarters, and insurance company headquarters); (2) income tax from local enterprises (excluding income tax from local banks, foreign banks and other financial institutions); (3) profits from local enterprises, personal income tax, urban land tax, fixed investment direction adjustment tax; (4) urban construction and maintenance tax (excluding the tax from railroad, bank headquarters, and insurance company headquarters); (5) housing tax; (6) vehicle and vessel usage license plate tax; (7) urban real estate tax; (8) vehicle and vessel usage tax; (9) stamp tax; (10) land appreciation tax; (11) slaughtering tax; (12) agriculture tax and animal husbandry tax (including agriculture special tax) and state-owned land use tax; (13) cultivated land occupation tax, and (14) contract tax. Taxes shared by the central and local governments include (1) value-added tax (the share of the central government is 75% and the share of the local government is 25%); (2) natural resource tax (only ocean oil resource tax belongs to the central government and the rest goes to the local governments), and (3) stamp tax from stock transactions. State tax agencies and local tax agencies were established separately. In principle, state taxes and shared taxes are collected by state tax agencies while local taxes are collected by local tax agencies. The central government returns part of its tax revenue to all local governments and provides additional assistance in the form of fiscal transfers to less developed regions and regions inhabited by ethnic minorities. Page 6 5 The 1994 tax reforms did not raise government tax revenue immediately. In fact, the real total tax revenue actually declined from 1993 to 1994. The decrease in total tax revenue was a result of the reduction of the number of taxes (from 37 to 23) and the reduction of the corporate income tax rate for large and medium size enterprises from 55% 33%. Corporate income taxes from both state enterprises and collective enterprises declined from 1993 to 1994. However, the 1994 tax reforms laid the foundation for the government tax revenue growth in the coming years. 3. Recent Tax Reforms Tax reforms have continued after 1994. The significant tax reforms in recent years are as follows. A. Increasing Tax Share of the Central Government Stamp tax from stock transactions was allocated in favor of the central government gradually, from 50% initially to 80%, 88% and then 97%. Only 3% goes to local governments now. From January 1, 2002, personal income tax and corporate income tax of local government-owned enterprises, which were pure local government taxes, were shared by the central and local governments at a ratio of 50% to 50%. Since January 1, 2003, the share of the central government increased to 60% and the share of the local governments decreased to 40%. 7 Interest income also became the revenue of the central government. B. Abolishing Agricultural Tax Facing growing rural and urban income disparity, the government decided to reduce the tax burden on

farmers. On January 1, 2006, China abolished the Regulations on Agricultural Taxes , and eliminated all agricultural taxes. The agricultural tax regulation was passed in 1958. Based on this regulation, the government collected taxes from all individuals and collectives engaged in agricultural activities. The tax base was the agricultural output. A standard output was set. To encourage farmers to produce more, the tax base was fixed, leading to increased agricultural output. Tax rates varied from province to province, at an average of 15.5%. Due to the large increase in agricultural productivity, the effective tax rate was much lower and the revenue from the agricultural tax was small. The agricultural tax revenue only accounted for 3.2%, 1.4%, and 0.9% of total tax revenue in 2003, 2004, and 2005, respectively. C. Unifying Corporate Income Taxes For many years, the corporate income tax rate for domestic firms was 33% (30% to the central government and 3% to the local governments), while it was as low as 15% for foreign-invested firms. The unequal tax burdens have drawn a lot of criticism from domestic firms and economists. On January 1, 2008, the government changed the corporate tax rates for domestic firms and foreign firms to a uniform rate of 25%. 7 State Council of PRC (2001), The Notice on the Announcement of the Reform Plan on the Sharing Income Taxes , State Council Document No. 37. Page 7 6 D. Excluding Investment from the Base of VAT In 2009, value-added tax (VAT) was converted from production type to consumption type. Unlike the European countries, China did not exclude investment from the tax base, which made the effective tax rate in China much higher than in other countries. The experiments of VAT reforms, excluding investment from the tax base, took place in three northeastern provinces and ten cities in the provinces in central China. Beginning January 1, 2009, investments have been excluded from the value-added tax base. China's VAT rate is 17% for most products. It is expected that the VAT reform will reduce tax revenue by 120 billion yuan in 2009. III. The Fall and Rise of Government Revenue As tax reforms proceed, China's government revenue has experienced large fluctuations. We will discuss the ups and downs of the level of real government revenue and then analyze the decline of the share of government revenue in GDP from 1978 to 1995 and the increase from 1995 to the present. Total government budgetary revenue in current yuan has been increasing since 1978 (113.2 billion yuan in 1978, 293.7 billion yuan in 1990, 1339.5 billion yuan in 2000, and 5.13 trillion yuan in 2007). However, the level of real government revenue measured at the 1978 constant retail price went through ups and downs in the 1980s. Figure 1 illustrates government revenue and its growth rate based on the 1978 constant price. Real budgetary revenue was 113.23 billion yuan in 1978, decreasing every year to 106.21 billion yuan in 1981. Real budgetary revenue increased consecutively to 156.5 billion yuan in 1985, then, declined continuously to 131 billion yuan in 1989. Since 1989 real budgetary revenue has gone up steadily, reaching 141.4 billion yuan in 1990, 168.22 billion yuan in 1994, 378 billion yuan in 2000, and 1362.4 billion yuan in 2007. Government revenue grew more slowly from 1978 to 1994, but much faster from 1994 to 2007, particularly in recent years. For example, the growth rate of real tax revenue (adjusted by the retail price index) was 18.25% in 2004, 18.93% in 2005, 21.25% in 2006, and 27.56% in 2007, much faster than the growth rate of real GDP (adjusted by the retail price index), which was 13.83% in 2004, 13% in 2005, 13.89% in 2006, and 12.84% in 2007. 8 The Chinese government reports growth rates of government revenue based on current values. Thus, the official growth rate of total government revenue for 2003, 2004, 2005, 2006, and 2007 was 20.4%,

25.7%, 20%, 21.4%, and 31.9%, respectively. Due to inflation, the official growth rates exaggerated the real growth of government revenue. For example, the growth rate of government revenue was 10.1% based on current value in 1988, but it was -9.58% based on constant price. Government revenue grew rapidly only after the 1994 tax reform, while the GDP growth has been high both before and after 1994. From 1978 to 1994, the average annual growth rate of budgetary revenue was only 2.47%, while the average annual growth rate of GDP was 9.06%! From 1994 to 2007, the average annual growth rate of government budgetary revenue was 16.1%, while the average annual 8 National Bureau of Statistics of PRC (2008), Statistical Yearbook of China . Page 8 7 growth rate of GDP was 11.15%. This caused a decline in the government budgetary revenue share in GDP before 1995 and an increase after 1995. In 1978, total government budgetary revenue accounted for 31.2% of GDP, down to 22.36% in 1985, and 10.7% in 1995. Since 1995 the government budgetary revenue has been steadily increasing. In 2005, total government budgetary revenue accounted for 15% of GDP, and in 2007, it reached 20.6% of GDP. Due to economic slowdown, the share of government budgetary revenue in GDP declined slightly in 2008 to 20.4%. Figure 2 shows the ratio of government budgetary revenue to GDP for the period of 1978 to 2008. 1. Reasons for the Decline in the Share of Budgetary Revenue in GDP 1978-1995 What are the reasons for the fluctuation of government revenue? Tax revenues depend on tax rate, as well as the size of taxable income. Factors which contributed to the decline in government budgetary revenues from 1978 to 1995 include lowered corporate income tax rates, limited tax base, and tax evasions. 9 Before the economic reforms in 1978, state enterprises submitted all their profits to the government, which implies that the corporate profit tax rate was 100%. Government revenues depended largely on profits from state enterprises. The corporate profit tax rate for small enterprises was lowered from 55% to 35% in 1989 and to 33% in 1991. In 1994 the corporate income tax rate for all domestic enterprises was lowered to 33%. The share of corporate income taxes in GDP for state enterprises declined from 6.65% in 1985 to 1.68% in 1993. The share of corporate income taxes in GDP for collective enterprises declined from 1.11% in 1985 to 0.28% in 1993. 10 In China, although tax rates might be high, the tax bases were small. One example is the agricultural tax. Agricultural tax rates set in 1958 were different among different provinces, and were around 15%. The tax base is the quantity of agricultural products per unit of cultivated land in an average year. For more than 40 years (particularly since the collapse of the people's communes), despite a large increase in agricultural productivity, the average product per unit of land ( mu ) was never changed. Thus, the actual tax rate was less than 2.5%! 11 A small tax base was also caused by tax exemptions and reductions. For example, the decrease in tariffs was caused by tax exemptions and increase in rebate rate on exports, as well as decrease in tax rate on imported goods [see Li (1997)]. Also, for many years, foreign investors have enjoyed tax exemptions for two years, tax reduction for three years, and only paying taxes after the fifth year when they have earned profits. 9 See Lin (2000) for a detailed discussion of the decline in China's government revenue. 10 Calculated by the author based on the data from National Bureau of Statistics of PRC (1999), Statistical Yearbook of China . 11 See Ministry of Finance of PRC (1996). Finance Yearbook of China . Page 9

8 0 200 400 600 800 1000 1200 1400 1600 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Year Billion Yuan -15 -10 -5 0 5 10 15 20 25 30 % Real Government Revenue Growth Rate Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2000-2008. Figure 1 Real Government Revenue at 1978 Constant Price and its Growth Rates (1978-2007) 0 5 10 15 20 25 30 35 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Year Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2000-2008. People's Bank of China, 2009, The Report on Monetary Policy Implementation for the 4th Quarter of 2008 , China.com.cn, February 23. National Bureau of Statistics of PRC, 2009, Statistical Bulletin on National Economy and Social Development in 2008. Figure 2 The Ratio of Government Budgetary Revenue to GDP 1978-2008 % Page 10 9 Tax evasion, one of the most intractable problems of economic policy in developing countries, was widespread in China during the transition period. It was estimated that 30% of state enterprises, 60% of joint ventures, 80% private enterprises, and 100% individual street vendors evaded taxes [Yu (1997)]. Many enterprises had two accounts, one was real, and the other was fake [see Liu (1998)]. The fake account was used for evading tax payment. Personal income tax evasion was also a serious problem. A part of an individual's income was non-wage income (cash awards, subsidies, and in-kind transfers). When people receive payment-in-kind instead of money, it is legally a taxable transaction. However, this part of the income is not reported as income and personal income tax cannot be applied. 12 2. Reasons for the Rise in Tax Revenue from 1995 to 2007 The rise in the level of government budgetary revenue from 1995 to 2007 was closely related to the rapid economic growth in the same period. Figure 1 shows that government budgetary revenue has been growing at a high rate since 1995. As a result, the level of government revenue has greatly increased. Almost all the taxes are related to output levels, such as value-added tax, business tax, corporate income tax, personal income tax, etc. Figure 3 shows China's per capita real GDP adjusted by the 1978 constant retail price and the growth rates of real GDP reported by the government. In the 1980s, China's GDP growth experienced large fluctuations. According to the official data, GDP growth was 10.2% in 1978, 3.9% in 1981, 13.7% in 1984, 2.5% in 1989, and 2.3% in 1990. The fluctuations were larger based on per capita GDP adjusted by the retail price index. The Chinese economy entered the steady high growth track after 1991. The per capita GDP growth reached 12.8% in 1992, and has never had a significant drop since then. Growth rate of per capita GDP was 6.8% in 1998, 6.7% in 1999, 7.6% in 2000, 7.5% in 2001, 8.4% in 2002, 9.3% in 2003, 9.4% in 2004, 9.8% in 2005, 11% in 2006, and 11.4%in 2007. 13 Per capita GDP at the 1978 constant retail price was 381.2 yuan in 1978, 791.5 yuan in 1990, 2217.2 yuan in 2000, and 5026.3 yuan in 2007. Per capita GDP in 2007 was 13.2 times as high as that in 1978! Thus, economic growth is undoubtedly the major reason for the growth of total tax revenue. The rise of the government revenue share in GDP (see Figure 2) implies that the government budgetary revenue increases faster than GDP. Theoretically, if the tax system is progressive and the tax rate remains unchanged when income increases, then an increase in the GDP will increase the share of taxes in GDP. Economic growth will push people to higher tax brackets, making them subject to higher tax rates. In this case, tax revenue will increase faster than income growth,

resulting in a higher ratio of tax revenue to income. However, most of China's taxes are proportional and non-progressive, such as VAT, business tax, corporate income tax, etc. Personal income tax is progressive, but it only accounts for a very small proportion of tax revenue. For example, in 2007 personal income tax revenue 12 There may be several reasons for the high proportion of tax evasion: (1) Lack of rule of law; (2) Non-existence of the legal concept of taxation in the Chinese society, and tax payers do not have the habit of paying taxes; (3) Grant of reductions or exemptions without proper authorization from the central government by some local officials; (4) Lack of efficiency in tax administration, and (5) Acceptance of bribes by tax collectors in some cases [See Li (1991)]. 13 National Bureau of Statistics of PRC (2008), Statistical Yearbook of China . Page 11 10 accounted for 6.98% of total tax revenue, and only 1.27% of GDP. Thus, economic growth is the reason for the increase in the growth rate of budgetary revenue, but not the main reason for the increase in the government revenue share in GDP. Increased tax revenue share in GDP is attributable to the following: A. Broadened Tax Base A broadened tax base was the major reason for the increase in government revenue share in GDP. First of all, the tax base for the value-added tax was largely broadened. In 1984, the State Department formally mitigated The Preliminary Rules on Value-Added Tax of the People's Republic of China, and thereafter, the VAT became an independent tax in China. However, the VAT was imposed only in the production process of a small proportion of products, with a variety of tax rates. Starting in 1994, VAT has been imposed universally on production, wholesale, retail and import of goods. The VAT rate is 17% for most products, 13% for some products (eg, agricultural products), and 6% for small tax payers. Exporting enterprises receive VAT refunds as an export incentive, with refund rates ranging from 9% to 17%. For example, the refund rate for electronic products is 17%. 14 To prevent tax evasion, a nationwide standard receipt is designed. Usually an invoice method is used each firm is liable for taxes on total sales but taxes already paid by suppliers as a credit against this liability are claimable. Special VAT invoices were adopted. VAT is self-policing and it is difficult to avoid taxes. Thus, VAT became a money machine for the government. B. Introduction of New Taxes Introduction of new taxes was another reason for the increase in tax revenue share in GDP. In 1994, consumption tax was introduced. This tax is imposed on some specified consumer goods, which is an extra tax imposed on some selected consumer goods after VAT is levied. Currently, consumption tax is imposed on 11 items including tobacco, wine and alcohol, cosmetics, fireworks, jewelry, gasoline, tires, motorcycles, cars, etc. The volume of consumption tax is computed by the ad valorem method or the specific duty method, and the tax is mainly collected at the stage of production and import. The consumption tax rate based on values of goods ranges from 3% to 45%. Some consumption taxes are based on quantities of goods. Also introduced were land value-added tax, stamp tax on stock exchange, inheritance tax, and gift tax. The government also increased the land use tax. C. Strengthened Law Enforcement The government launched the Gold Tax Project in 1994. The objectives of the project were to prevent tax evasion, reduce corruption, and improve the efficiency of tax collection. The government made great efforts to establish a computerized database of the records of both taxpayers and tax collectors. However, in the first stage (1994-1995), only 50 cities were involved in the project. The government could only check the VAT invoices of 50 cities, and the VAT invoices outside the 50

cities were not easy to check. In 1998, the second stage of the project began. All the provinces were involved in the project. A computer network connecting the National 14 Other countries, such as European countries that adopted VAT, also rebate VAT on exported goods. Page 12 11 Tax Administration, provincial national tax collectors, city national tax collectors, and county national tax collectors was established. Collectors of local taxes were also hooked up to the network. An accelerated pace in tax computerization was achieved to ensure information integration, promote extensive application of integrated administrative software and improve the database focusing on individual taxpayers. Tax-monitoring meters were widely applied in the industries of retailing, services, entertainment, catering and transport to strictly exercise oversight of the instruments of counterfalsification of VAT invoices. Along with the wide-use of tax-monitoring meters, the practice of prizewinning invoices will be actively promoted. The third stage of the project started in 2005, aimed at establishing more comprehensive information and management system. D. One-time Factors Some onetime factors contributed to the rapid growth of tax revenue. For the high growth of tax revenues in 2007, there are other policy-related and one-time factors. First , the increase in transactions on the securities market and in the stamp tax rate on securities transactions raised tax revenue by 182.6 billion yuan. Stamp tax on securities exchanges increased drastically, 500%, in 2007. Its share in total tax revenue increased from 0.48% to 3.69%. This is a result of the large increase in the value of stocks and the increase in the stamp tax rate. The Shanghai Stock Index increased from 2715 in the beginning of 2007 to 6124 on October 16, 2007! Meanwhile, on May 23, 2007, the Ministry of Finance of the PRC increased the stamp tax rate on stock exchanges from 0.1% to 0.3% on the purchasing and selling of stocks. Second , tax rebates were reduced by 40 billion yuan due to policy changes, equivalent to an increase in revenue by the same amount. In China, subsidies are treated as negative revenues, instead of spending. Third , railway transportation enterprises turned over to the central government 19.6 billion yuan from the proceeds of their sale of state assets, and that was a one-time only source of revenue. Fourth , special surcharges on oil sale for the whole year totaled 50.1 billion yuan, an increase of 12.9 billon yuan over 2006 when they were collected for only two quarters. 15 These factors increased government revenue by 305.3 billion yuan, a one-time increase. Fifth , the second fastest growing tax is the tax on urban land use, with a growth rate of 118%. The local governments are in severe revenue shortage and have largely relied on sales of land for real estate development. The rapidly rising housing prices inspired real estate developers to allocate more and more land for housing construction. These factors caused a large increase in the tax revenue on urban land use. IV. Composition of Government Budgetary Revenue China's government revenue is classified under several large categories: tax revenue, contributions by state-owned enterprises, subsidies to loss-making stateowned enterprises, contributions for education, and other revenue. The data of each category can be traced back to the 1950s. Major components of government revenue from 1950 to 2007 will be discussed here. 15 Ministry of Finance of PRC (2008) , On the Draft of Central and Local Budgets for 2008 , First Session of the 11 th National People's Congress , March 5. Page 13

12 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 197 8 197 9 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 Yuan -10 -5 0 5 10 15 20 Year % Real GDP per Capita (RPI adjusted) Growth Rate (RPI adjusted) Growth Rate (Official) Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2000-2008 . Figure 3 Per Capita GDP Adjusted by RPI, Per Capita GDP Growth, GDP Growth Table 1 shows total government revenue and shares of tax revenue, contributions by state-owned enterprises, subsidies to loss-making state-owned enterprises, additional contributions to education, and other revenues in China's government revenue. From 1950 to 1957, tax revenue was the major part of government revenue, accounting for 51% of total government revenue in 1957. In 1957, China completed socialist economic reforms, eliminated all the private enterprises and established state-owned enterprises and urban collective-owned enterprises. In 1958, China launched a Great-Leap Forward movement to catch up with Great Britain and the United States in a short period of time. Following the example of the Soviet Union, enterprises were required to submit their profits to the government. Contributions from SOEs became the major part of government revenue in all the years before 1978, except in 1962 (end of the Great-Leap-Forward movement), 1968 (the third year of the Cultural Revolution), 1975-1977 (the end of the Cultural Revolution) in which enterprises suffered severe losses. Since 1978, tax has become the major part of government budgetary revenue. After the tax-for-profit reforms in 1984, tax revenue became the predominant component of total government revenue. The share of tax revenue in total government revenue was 102% in 1985, 96% in 1990, 94% in 2000, and 89% in 2007. 16 16 Since the subsidies to loss-making enterprises are considered negative revenue for the government, tax revenue share can go beyond 100%. Page 14 13 Table 1 Government Budgetary Revenue and its Composition Tax Revenue (%) Year Total Govt. Revenue (billion yuan) Total Tax Industrial and Commer- cial tax Tariffs Agri- culture and Related Taxes Corpor. Income Tax Rev. from SOEs (%) Subs. to Loss- making SOEs (%) Rev. from Add. Contr. for Edu. (%) Other Rev. (%) 1950 6.22 78.8 42.3 5.7 30.7 14.0 7.2 1951 12.50 64.9 40.7 5.6 18.7 24.4 10.6 1952 17.39 56.2 37.7 2.8 15.7 32.9 10.9 1953 21.32 56.1 40.9 2.4 12.9 36.0 7.9 1954 24.52 53.9 38.7 1.7 13.5 40.6 5.5 1955 24.93 51.1 36.9 1.9 12.3 44.9 4.0 1956 28.02 50.3 37.8 1.9 10.6 47.9 1.8 1957 30.32 51.1 39.4 1.9 9.8 47.6 1.4 1958 37.96 49.4 39.1 1.7 8.6 49.8 0.8 1959 48.71 42.0 33.8 1.4 6.8 57.3 0.7 1960 57.23 35.6 29.6 1.0 4.9 63.9 0.5 1961 35.61 44.6 36.8 1.7 6.1 53.7 1.7 1962 31.36 51.7 42.9 1.5 7.3 46.6 1.7 1963 34.23 48.0 39.8 1.2 7.0 50.5 1.5 1964 39.95 45.6 38.0 1.1 6.5 53.3 1.2 1965 47.33 43.2 36.5 1.2 5.4 55.8 1.0 1966 55.87 39.7 33.3 1.2 5.3 59.7 0.6 1967 41.94 46.9 39.0 0.9 6.9 52.1 1.0 1968 36.13 53.0 43.0 1.8 8.3 46.2 0.8 1969 52.68 44.7 37.9 1.2 5.6 54.4 0.9 1970 66.29 42.4 36.5 1.1 4.8 57.2 0.4 1971 74.47 42.0 37.2 0.7 4.1 57.5 0.5 1972 76.66 41.4 37.0 0.7 3.7 58.1 0.5 1973 80.97 43.1 38.2 1.1 3.8 56.4 0.5 1974 78.31 46.0 40.4 1.8 3.8 52.0 2.0 1975 81.56 49.4 43.9 1.8 3.6 49.1 1.5 1976 77.66 52.5 46.8 1.9 3.8 43.5 3.9 1977 87.45 53.5 47.2 3.0 3.4 46.0 0.4 1978 113.23 45.9 40.8 2.5 2.5 50.5 3.6 1979 114.64 46.9 42.1 2.3 2.6 43.2 9.9 1980 115.99 49.3 44.0 2.9 2.4 37.5 13.2 1981 117.58 53.6 46.6 4.6 2.4 30.1 16.3 1982 121.23 57.7 51.4 3.9 2.4 24.5 17.8 1983 136.70 56.7 50.4 3.9 2.4 17.6 25.7 1984 164.29 57.7 49.3 6.3 2.1 16.8 25.5 1985 200.48 101.8 54.7 10.2 2.1 34.7 2.2 -25.3 21.3 1986 212.20 98.5 56.7 7.1 2.1 32.6 2.0 -15.3 14.8

Page 15 14 Table 1 Government Budgetary Revenue (Continued) Tax Revenue (%) Year Total Govern- ment Revenue (billion yuan) Total Industrial and Commer- cial tax Tariffs Agri- culture and Realted Taxes Corpo- rate Income Tax Rev. from SOEs (%) Subs. to Loss- making SOEs (%) Rev. from Add. Contr. for Edu. (%) Other Rev. (%) 1987 219.94 97.3 58.3 6.5 2.3 30.2 1.9 -17.1 17.8 1988 235.72 101.4 63.0 6.6 3.1 28.7 2.2 -18.9 15.4 1989 266.49 102.3 66.1 6.8 3.2 26.3 2.4 -22.5 17.7 1990 293.71 96.1 63.3 5.4 3.0 24.4 2.7 -19.7 21.0 1991 314.95 94.9 62.9 5.9 2.9 23.2 2.4 -16.2 0.9 18.0 1992 348.34 94.6 64.4 6.1 3.4 20.7 1.7 -12.8 0.9 15.5 1993 434.90 97.8 73.5 5.9 2.9 15.6 1.1 -9.5 1.0 9.5 1994 521.81 98.3 75.0 5.2 4.4 13.6 7.0 1.2 7.5 1995 624.22 96.7 73.5 4.7 4.5 14.1 -5.3 1.3 7.2 1996 740.80 93.3 71.1 4.1 5.0 13.1 -4.6 1.3 10.0 1997 865.11 95.2 75.8 3.7 4.6 11.1 -4.3 1.2 7.9 1998 987.60 93.8 77.2 3.2 4.0 9.4 -3.4 1.1 8.4 1999 1144.41 93.3 77.6 4.9 3.7 7.1 -2.5 1.1 8.1 2000 1339.52 93.9 77.4 5.6 3.5 7.5 -2.1 1.1 7.1 2001 1638.60 93.4 69.3 5.1 2.9 16.1 -1.8 1.0 7.4 2002 1890.36 93.3 69.5 3.7 3.8 16.3 -1.4 1.0 7.0 2003 2171.53 92.2 70.5 4.3 4.0 13.4 -1.0 1.1 7.8 2004 2639.65 91.5 69.2 4.0 3.4 15.0 -0.8 1.1 8.1 2005 3164.93 90.9 67.7 3.4 3.0 16.9 -0.6 1.1 8.6 2006 3876.02 89.8 65.9 2.9 2.8 18.2 -0.5 1.2 9.5 2007 5132.18 88.9 2.8 17.1 2008 6137.70 Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2000-2008. People's Bank of China, 2009, The Report on Monetary Policy Implementation for the 4 th Quarter of 2008, China.com.cn, February 23. Note: Agriculture and related taxes include agriculture tax, animal husbandry tax, and the tax on cultivated land, tax on special agricultural products and contract tax; other revenue includes funds raised for energy, transportation, budget adjustment, etc. Subsidies to loss-making SOEs increased immediately after the completion of the tax-for-profit reform in 1985. Many SOEs suffered tremendous losses after paying taxes and the government had to subsidize the lossmaking enterprises. In 1985, the subsidies were as high as 25% of total government revenue! As SOE reforms proceeded, the government subsidies declined gradually, down to only 0.5% of total government revenue in 2006. 17 17 Over the years, the government has privatized many small lossmaking SOEs. Also, the government has transferred SOEs' social security burden to the society. In the past, SOEs took care of the retirement and healthcare of their workers, and now, local governments take care of the social security of employees of the SOEs. All firms must join the social security program. New firms without retirees actually subsidized the SOEs (see Lin (2008)). Page 16 15 Revenue from additional contribution to education was introduced in 1991 to finance education. Enterprises and individuals who pay product tax (later converted to consumption tax), VAT, and business tax must make additional contributions to education. The contribution rate is 1% of the value of the above taxes. Those who make contributions to rural education based on the 1984 State Council regulation are exempted from making new contributions. 18 Over the years, this contribution has been around 1% of total government revenue. -40 -20 0 20 40 60 80 100 120 197 8 197 9 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 Year % Tax Revenue Revenue from Enterprises Subsidies to Loss-making Enterprises Revenue from Additional Contribution for Education Other Revenue Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2000-2008. Figure 4 Composition of Government Budgetary Revenue Tax Revenue Other Revenue Revenue from

Enterprises Subsidies to Loss-making Enterprises Revenue from Additional Contribution for Education The last item includes other revenue, such as funds raised for energy and transportation, budget adjustment, etc. The other revenue has varied widely over the years, reaching its peak in 1983-1984 at about 25% of total government revenue and declined gradually at less than 10% of total government revenue in recent years. Figure 4 shows graphically the changes in the shares of tax revenue, contributions by SOEs, subsidies to loss-making SOEs, additional contributions to education, and other revenues in total government revenue. It can be seen that, before 1985, government revenue composed of tax revenue, contributions by SOEs, and other revenues. After the tax-for-profit reforms, SOEs started to pay corporate income tax, instead of submitting profits, and the government began to subsidize lossmaking enterprises. 18 State Council of PRC (1984), The State Council's Notice on Raising Revenues for Rural Education, State Council Document, No. 174. Page 17 16 V. Extra-Budgetary Revenue An important part of China's government revenue is the extra-budgetary revenue, which mainly includes non-tax revenues collected by local governments, government agencies and institutions, and state enterprises. Fees are officially classified under two large categories: administrative fees ( xingzheng xing shoufee ) and operating fees ( shiye xing shoufee ). Administrative fees are those collected by government branches such as ministries at the central government level and bureaus at the local government level. Operating fees are collected by government operating units such as public schools, public health institutes, and public infrastructure maintenance units. Fee collection was widespread in China in the late 1980s and 1990s. Table 2 shows extra-budgetary revenue, its composition, its ratios to budgetary revenue and GDP from 1952 to 2006. The ratio of extra-budgetary revenue to budgetary revenue was 8% in 1952, 20.6% in 1960, and 31.1% in 1978. The ratio of extrabudgetary revenue to budgetary revenue increased to 76% in 1985 and reached 111% of budgetary revenue in 1992. The coverage of the extra-budgetary revenue has been adjusted since 1993 by excluding the extra-budgetary revenue of state-owned enterprises. Thus, the data after 1993 are not comparable with the earlier data. Usually, local governments rely more on fees as a source of revenue than the state, provincial, or regional governments, while the state government, provincial and regional governments rely more on fees as a source of revenue than the central government. In most countries, fees are only a small part of government revenue. In the US in 2000, the revenue from fees, service charges, non-industrial and incidental sales, fines and forfeits accounted for 1% of the central government's revenue, 8.5% of total revenue of the state government, and 19% of total revenue of the local governments. In Mexico in 2000, it accounted for 1% of the central governments revenue, 7.4% of total revenue of the state government, and 11.2% of total revenue of the local governments. In Malaysia in 1997, it accounted for 5% of the central government's revenue, 17% of total revenue of the provincial governments. In Thailand in 2000, it accounted for 1.3% of central government revenue and 2.5% of local government revenue. In Russia in 2000, it accounted for 0.4% of the central government's revenue and 0.5% of the total revenue of the local governments. 19 As shown, China's fee collection was much higher than all of these countries. The main reason for the increase in fee collection was the rapid increase in government expenditure caused by the large increase in government administrators and the slow growth of government budgetary revenue due to fiscal reforms. 20 19 International Monetary

Fund (2002), Government Finance Statistics Yearbook . 20 See Lin (2005) for a detailed discussion of fee collection in China. Page 18 17 Table 2 Extra-budgetary Revenue 1952-2006 Extra-Budgetary Revenue (billion yuan) Year Total Administra- tive and Institutional Units Local Govts SOEs and Super- visory Ministries Fund- raising of Town- ship Govts Govern- ment Funds Others Share in GDP (%) Share in Budgetary Revenue (%) 1952 13.62 12.53 1.09 2.01 21.90 1953 8.91 2.07 1.40 5.44 1.08 7.10 1954 14.23 3.34 2.07 8.82 1.66 8.20 1955 17.02 3.68 3.27 10.07 1.87 8.00 1956 21.42 3.85 5.00 12.57 2.08 8.70 1957 26.33 3.80 5.66 16.87 2.47 10.60 1958 55.99 9.29 17.59 29.11 4.28 20.00 1959 96.55 11.78 35.39 49.38 6.71 31.80 1960 117.78 23.13 23.39 71.26 8.08 31.00 1961 57.40 15.61 13.29 28.50 4.70 11.80 1962 63.63 15.21 20.50 27.92 5.54 11.10 1963 51.85 11.99 7.19 32.67 4.20 14.60 1964 65.86 16.07 8.87 40.92 4.53 21.00 1965 75.56 18.74 9.47 47.35 4.40 22.10 1966 81.13 20.00 10.36 50.77 4.34 20.30 1967 83.61 22.00 9.72 51.89 4.71 17.70 1968 77.44 24.00 9.96 43.48 4.49 13.90 1969 87.42 26.00 12.19 49.23 4.51 20.80 1970 100.94 28.00 13.45 59.49 4.48 27.90 1971 118.56 30.00 14.72 73.84 4.89 22.50 1972 134.24 31.66 23.28 79.30 5.33 20.30 1973 191.29 32.57 24.14 134.58 7.03 25.70 1974 219.72 34.60 22.65 162.47 7.88 28.70 1975 251.48 42.30 27.86 181.32 8.39 31.10 1976 275.32 48.81 28.35 198.16 9.35 35.20 1977 311.31 56.84 30.76 223.71 9.72 38.20 1978 34.71 6.34 3.11 25.26 9.52 30.66 1979 45.29 6.87 3.99 34.43 11.15 39.50 1980 55.74 7.44 4.09 44.21 12.26 48.05 1981 60.11 8.49 4.13 47.49 12.29 51.12 1982 80.27 10.12 4.53 65.63 15.08 66.21 1983 96.77 11.39 4.98 80.40 16.23 70.79 1984 118.85 14.25 5.52 99.07 16.49 72.34 1985 153.00 23.32 4.41 125.27 16.97 76.32 1986 173.73 29.42 4.32 139.99 16.91 81.87 Page 19 18 Table 2 Extra-budgetary Revenue (Continued) Table Extra-Budgetary Revenue (billion yuan) Year Total Administra- tive and Institutional Units Local Govts SOEs and Super- visory Ministries Fund- raising of Town- ship Govts Govern- ment Funds Others Share in GDP (%) Share in Budgetary Revenue (%) 1987 202.88 35.84 4.46 162.58 16.82 92.25 1988 236.08 43.89 4.89 187.29 15.69 100.15 1989 265.88 50.07 5.44 210.38 15.65 99.77 1990 270.86 57.70 6.06 207.11 14.51 92.22 1991 324.33 69.70 6.88 247.75 14.89 102.98 1992 385.49 88.55 9.09 287.86 14.32 110.67 1993 143.25 131.78 11.47 4.05 32.94 1994 186.25 172.25 14.00 3.86 35.69 1995 240.65 223.49 17.17 3.96 38.55 1996 389.33 339.58 22.47 27.29 5.47 52.56 1997 282.60 241.43 11.59 29.58 3.58 32.67 1998 308.23 198.19 5.47 33.73 47.84 23.00 3.65 31.21 1999 338.52 235.43 5.01 35.89 39.65 22.54 3.77 29.58 2000 382.64 265.45 5.92 40.33 38.35 32.58 3.86 28.57 2001 430.00 309.00 6.00 41.00 38.00 36.00 3.92 26.24 2002 447.90 323.80 7.20 27.20 37.60 52.10 3.72 23.69 2003 456.68 333.57 5.23 29.31 28.71 59.85 3.36 21.03 2004 469.92 320.84 6.41 21.31 35.13 86.23 2.94 17.80 2005 554.42 385.82 4.78 19.29 35.93 108.59 3.03 17.52 2006 640.79 421.68 4.49 22.13 37.65 154.84 3.02 16.53 Sources: Ministry of Finance of PRC (2000, 2005, 2007, 2008) Note: The scope of extra budgetary revenue and expenditure was adjusted between 1993-1995 and 1996, not comparable with previous years. Since 1997, the extra budgetary revenue and expenditure do not include the intra- budgetary government funds (fees), and thus, the figures are not comparable with the previous years. First, the large increase in government administrative expenditure. One might think that more government employees are needed under a centrally planned system to formulate and implement

government plans, while fewer government employees are needed under a market-oriented system. Unfortunately, this is not the case in China. The number of government employees has increased dramatically since economic reforms began in 1978. Workers and staff in government agencies, party agencies and social organizations (in state units) increased by 158.8%, from 4.17 million in 1978 to 10.86 million in 2000. 21 In the meantime, the population increased 21 See National Bureau of Statistics of PRC (2002), Statistical Yearbook of China , p. 129. Page 20 19 by 29.4% from 962.59 million to 1267.43 million. 22 The government was under strong pressure to raise the compensation (salaries and bonuses) of government employees. The share of government administrative and operating expenditure in total government expenditure increased from 4.71% in 1978 to 17.42% in 2000! 23 Administrative and operating expenditure in total extra-budgetary expenditure increased from 6% in 1982 to 32% of total extra-budgetary expenditure in 1995, to reach 48% in 1997 and 63.1% in 2000! Second, the low growth of budgetary revenue. Budgetary revenue accounted for 31.1% of GDP in 1978; it declined to 23.4% in 1980, 14.7% in 1992, and 10.3% in 1995. Local government budget situations have been more serious than the central government situation. After economic reforms, the central government has transferred many duties of public goods and services provisions to the local governments. Since the tax reform in 1994, provinces have reported budget deficits. Chinese provincial and local governments have no right to issue debt (neither do they have the right to make tax laws). Local governments have to look toward the central government for money. Even the rich provinces needed large rebates to cover their budget deficits. The prefecture, county and township governments usually request for money from their higher level of government. There is less and less money available for the lower levels of government. The fiscal difficulties are especially severe for township governments. Many local governments have yet to pay their employees and are desperately in need of more revenues. 24 Third, the self-collection and self-utilization policy. Facing limited revenue and growing demand for expenditure, the central government could not provide its agencies, institutions and local governments with the needed fiscal transfers. All it could do was to give them the right to raise their revenue on their own. 25 In 1980, except three major cities controlled by the central government (Beijing, Shanghai and Tianjin), the government started the fiscal responsibility system (caizheng baogan) in all provinces, calling for government agencies and institutions with potential for fund-raising to actively expand services and raise their own revenue. 26 In 1989, after four consecutive years of fiscal deficits, the Ministry of Finance announced that government institutions should rely fully on the government for some expenditure, rely partially on the government for some expenditure, and rely on themselves for other expenditure. 27 The policy of self-collection and self-utilization provides strong incentives for government agencies and institutions, as well as local governments, to 22 It was estimated that the ratio of officials to the population was 1:3900 in Tang Dynasty, 1:2900 in Ming Dynasty, and 1:297 in the early 1950s, 1:50 in 1978, and 1:29 in 1994 (Zhouwei Liu, 2000). 23 National Bureau of Statistics of PRC (2003), Statistical Yearbook of China . 24 See Lin (2005) for a discussion of local government fiscal difficulties. 25 Government institutions refer to state universities, research institutes, etc. 26 Chen (1996, p. 291). 27 See Wu (1997, p.228). Page 21

20 collect more revenue. Many of them began to start their own businesses. 28 Poor management skills and harsh competition from private enterprises soon drove them out of business. They finally figured out an effective way to raising revenue: collecting fees from the business enterprises and the general public using their monopoly power! Facing increasing complaints about excessive fee collection, in the late 1990s the government launched a tax-for-fee reform to lower the number of fees and reduce the size of extra-budgetary revenue and expenditure. Entering the new century, although extra-budgetary revenue is increasing in absolute value, the ratio of extra- budgetary revenue to budgetary revenue and the ratio of extra-budgetary revenue to GDP have declined. Extra-budgetary revenue accounted for 38.55% of budgetary revenue in 1995, 28.57% in 2000, 17.52% in 2005, and 16.56% in 2006. Extrabudgetary revenue accounted for 4.12% of GDP in 1995, 3.43% in 2000, and 3.01% in 2005 and 2006. VI. Size of China's Government Revenue in an International Perspective The calculation of China's government revenue is slightly different from that of many other countries. For example, in many countries, such as the United States, Great Britain, and Russia, social security contributions are included in total government revenue. However, in China the social security system is run by the local governments and social security contributions are not included in government revenue. Also, in China the government subsidies to loss-making enterprises have been treated as negative revenue (deducted from total government revenue), instead of expenditure. This has reduced the size of government revenue. For a more accurate comparison of China's government revenue size with the rest of the world, it is necessary to include social security contributions and subsidies to loss-making enterprises in China's total government revenue. Social security contribution includes contributions to basic pension insurance, unemployment insurance, medical insurance, work injury insurance, and maternity insurance. The social security account revenues have grown rapidly, from 18.68 billion yuan in 1990 to 100.6 billion yuan in 1995, 264.45 billion yuan in 2000, 697.52 billion yuan in 2005, and 864.32 billon yuan in 2006. As can be seen in Table 1, the subsidies to loss-making state-owned enterprises were large in the 1980s and the early 1990s, but declined substantially in recent years, along with the privatization of unprofitable SOEs and reduction of social security burden of existing SOEs. For example, in 1985 the subsidies to loss-making SOEs were 50.702 billion yuan, accounting for 25% of total tax revenue; in 1990 the subsidies to loss-making SOEs were 57.89 billion yuan, accounting for 20.5% of total tax revenue, while in 2006 the subsidies to loss-making SOEs were 18.02 billion yuan, accounting for only 0.5% of total tax revenue. Adding up budgetary revenue, extra-budgetary revenue, subsidies to loss- making enterprises, and social security contributions, China's total government revenue in 2006 should be 5399.15 billion yuan, accounting for 25.5% of GDP in 28 Even the People's Liberation Army had its own businesses (hotels, restaurants, etc.) for a long time. Page 22 21 2006. Figure 5 shows the ratios of government budgetary revenue, extra-budgetary revenue, social security contributions, and total government revenue to GDP from 1978 to 2007. It can be seen that the share of total government revenue in GDP, as well as the share of tax revenue to GDP, declined sharply from 1978 to 1995, and has been increasing steadily since 1995. The share of extra-budgetary revenue was high in the 1980s but has declined significantly since the early 1990s. Also, the share of social security contribution in GDP is low but increasing. 0 5 10 15 20 25 30 35 40 45 1 978 1 979 1 980 1 981 1

982 1 983 1 984 1 985 1 986 1 987 1 988 1 989 1 990 1 991 1 992 1 993 1 994 1 995 1 996 1 997 1 998 1 999 2 000 2 001 2 002 2 003 2 004 2 005 2 006 2 007 Year % Budgetary Revenue Share in GDP Extrabudgetary Revenue Share in GDP Social Security Contribution Share in GDP Total Government Revenue Share in GDP Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 20002008. Ministry of Finance of PRC, Finance Yearbook of China , 2000-2007. Figure 5 Ratios of Government Budgetary Revenue, Extra-budgetary Revenue, Social Security Contributions, and Total Government Revenue to GDP 1978-2007 Total Government Revenue Share in GDP Budgetary Revenue Share in GDP Extra-budgetary Revenue Share in GDP Social Security Contribution Share in GDP Table 3 shows the shares of government revenue in GDP in other countries with data available. It can be seen that industrial countries on average have higher share of government revenue in GDP, with Norway, Sweden, Demark, Finland, and France among the highest. Asian countries had lower shares of government revenue in GDP in 2006, with Maldives being an exception (61.2%, the highest in the world). China's share of government revenue in GDP is higher than Singapore (19.9%), Hong Kong (20.04%), and Thailand (21.95%), but lower than some other Asian countries. With a 25.5% government revenue share in GDP, China's government revenue share in GDP ranks low in the world, lower than those of any industrial countries, and lower than most developing countries. However, when looking at the size of government, one should never overlook the large number of SOEs owned by the Chinese government. Page 23 22 Table 3 Shares of Government Revenue in GDP Country Year % Country Year % Industrial Countries Asia Japan 2005 32.32 Singapore 2005 19.90 United States 2006 33.99 Hong Kong (China) 2005 20.04 Australia 2006 36.26 Thailand 2006 21.95 Ireland 2006 37.12 Vietnam 2004 25.07 Switzerland 2005 37.32 Macau (China) 2006 25.42 Greece 2006 39.53 China, People's Republic 2006 25.50 Luxembourg 2006 39.69 Malaysia 2003 26.07 Spain 2006 40.40 Bhutan 2000 35.89 Canada 2006 41.67 Mongolia 2003 38.73 United Kingdom 2006 41.93 Maldives 2006 61.24 New Zealand 2006 42.31 Europe (East) Portugal 2006 42.46 Armenia 2006 19.91 Germany 2006 43.82 Kyrgyz Republic 2006 22.04 Italy 2006 45.62 Kazakhstan 2006 22.38 Netherlands 2006 46.66 Albania 2005 24.89 San Marino 2002 46.95 Georgia 2006 27.92 Austria 2006 47.82 Romania 2005 31.69 Belgium 2006 48.71 Lithuania 2006 33.75 Iceland 2006 49.09 Slovak Republic 2006 35.06 France 2006 50.81 Latvia 2006 35.53 Finland 2006 52.55 Estonia 2006 36.60 Denmark 2006 55.88 Czech Republic 2006 38.09 Sweden 2006 56.55 Moldova 2005 38.58 Norway 2006 58.47 Poland 2006 39.07 Middle East Bulgaria 2006 39.99 Iran, IR of 2004 30.72 Russia 2006 40.55 Israel 2006 45.05 Malta 2006 41.53 Kuwait 2006 45.88 Hungary 2006 42.64 Western Hemisphere Cyprus 2006 42.69 El Salvador 2006 18.21 Ukraine 2006 43.59 Peru 2005 18.65 Slovenia 2006 44.72 Paraguay 2006 22.49 Croatia 2006 44.81 Costa Rica 2006 25.12 Belarus 2006 48.21 Chile 2006 27.84 Africa Argentina 2004 29.43 Mauritius 2006 21.11 St. Vincent and the Grenadines 2004 33.57 Swaziland 2003 29.34 Jamaica 2005 35.68 Congo 2003 31.80 St. Kitts and Nevis 2003 37.88 South Africa 2006 37.25 Barbados 2004 43.86 Lesotho 2005 51.07 Bolivia 2006 50.87 Seychelles 2005 58.08 Sources: International Monetary Fund (2007), Government Finance Yearbook 2007 . Page 24

23 Although the share of government revenue in GDP is not large, China has recently been ranked high in terms of tax burden. According to the 2007 "Forbes Tax Misery Index," China ranked third in the world on tax burden, only behind France and Belgium. According to the 2008 "Forbes Tax Misery Index," China ranked fifth in the world on tax burden, only behind France, Netherlands, Belgium, and Sweden. Some Chinese experts criticize the index for using the maximum statutory tax rate as a standard for its tax index calculation while ignoring the number of individuals who are subject to the highest tax rate (this information is unavailable to the public), and focusing on statutory tax rates and ignoring the tax waivers. In fact, all countries have some kind of tax waivers. Thus, high tax rates are a fact in China. VII. China's Tax Structure and its Further Reforms With the size of China's government revenue in mind, we now examine China's tax structure, compare China's tax structure with other countries, and discuss options for future tax reforms in China. 1. China's Tax Structure After completing the tax-for-profit reforms in 1984, tax revenue became the major component of government revenue. The tax structure has changed over time. In 1985, corporate income tax was the major revenue-generating tax in China, followed by business tax and value-added tax (VAT). The 1994 tax reforms completely changed the landscape of China's tax structure. The most important tax is the value-added tax levied on all stages and spheres of industrial processing and is shared by the central and local governments. Currently, there are two rates: for products related to agricultural production (such as fertilizer, feeds, and machines) and basic consumption goods (such as foods, cooking oil, water, and natural gas) the tax rate is 13%, and for most other goods the rate is 17%. Business tax ( yingye shui ) is the major local tax. It is levied on the transfer of intangible assets or sale of immovable properties, primarily service industries, including transportation, communications, financial services, real estate sales, and entertainment. Educational institutions and hospitals are exempt from this tax. The tax rate varies from 3% to 20% according to tax regulations passed in 1993. Consumption tax is the tax imposed on special products, such as tobacco, alcoholic beverages, and luxuries. Consumption tax and VAT on imports are other major taxes for the government. Corporate income tax is another important source of government revenue. Until recently, the tax rate on domestic enterprises was 33%, while the rate for foreign-invested enterprises in special areas was as low as 15%. As mentioned earlier, the rate was unified at 25% in 2008. Personal income tax was also introduced in 1980 and was reformed in 1994. The tax rate on personal income is not low, with the highest marginal tax rate being 45%. Resource tax, which was initiated in 1984 and reformed in 1994, includes taxes on resources on crude oil, natural gas, coal, iron ore, precious metal, salt, and so on. Resource tax is a specific tax, based on the quantity of the product. As resource prices increased in recent years, there were calls for increasing resource tax by Page 25 24 changing the tax base to the value of the product. The revenue from recourse tax in China has been low, varying from 0.89% to 0.5% of total tax revenue. Urban land use tax was introduced in 1988 and reformed in 2006. This tax is collected from individuals, enterprises, and governmental organs and institutions for using urban land. The tax rate, based on square meters of land used, varies from 0.6 yuan/square meter to 30 yuan/square meter. The central government gives local governments the limited right to adjust the rate, ie, the rate cannot be lowered by more than 30% and any increase in rate must be approved by the Ministry of Finance. As urban real estate develops, the revenue from this

tax has increased significantly in recent years. Stamp tax on security exchange was introduced in 1990 by the Shenzhen City government, with a rate of 0.6% on the value of sold stocks. In 1992, the National Bureau of Taxation and National Committee on Economic System Reforms set the tax rate at 0.3% on both purchasing and selling of stocks. Over the years, the rate has been adjusted many times. The revenue from stamp tax on security exchange fluctuates along with the stock index. Consumption tax and VAT on imports are collected on imported products. Meanwhile China rebates VAT to exporting enterprises at various rates. China considers the tax rebate as negative tax revenue, instead of expenditure. In 2007, the government lowered the tax rebate to curb China's increasing trade surplus with the United States and Europe. However, facing an economic slowdown, the government restored the tax rebate for some industries in 2008. Other Taxes 9% Tariffs 3% Comsumption Tax and VAT on Net Exports 1% Consumption Tax 5% Urban Maintenance and Construction Tax 3% Stamp Tax on Security Exchange 4% Personal Income Tax 7% Company Income Tax 19% Tax on Resources 1% Business Tax 14% Value Added Tax 34% Data Sources: National Bureau of Statistics of PRC, Statistical Yearbook of China , 2008. Figure 6 Composition of Tax Revenue in 2007 Page 26 25 Urban maintenance and construction tax is collected from the taxpayers of VAT, business tax, and consumption tax. The tax base is the amount of VAT, business tax, and consumption paid. The rate varies from 1% to 7%, depending on the residency of the taxpayers. Foreigners and foreign companies do not pay this tax. Figure 6 shows the revenue share of major taxes in total tax revenue in 2007. The revenue from VAT accounted for 34% of total tax revenue, corporate income tax 19%, business tax 14%, consumption tax 5%, and VAT and business tax from net exports 1% (taxes from VAT and business tax on imports minus rebates from VAT and business tax on exports), personal income tax 7%, urban maintenance and construction tax 3%, tariff 3%, resource tax 1%, and other taxes 9%. 0 5 10 15 20 25 30 35 40 45 50 1 985 1 986 1 987 1 988 1 989 1 990 1 991 1 992 1 993 1 994 1 995 1 996 1 997 1 998 1 999 2 000 2 001 2 002 2 003 2 004 2 005 2 006 2 007 Year % Value Added Tax Consumption Tax Business Tax Company Income Tax Personal Income Tax Data Sources: Ministry of Finance of PRC, Finance Yearbook of China , 2005, for data 1985 to 2004. National Bureau of Statistics of PRC, Statistical Yearbook of China , 2008, for data 2005 to 2007. Figure 7 Share of Major Taxes in Total Tax Revenue Value Added Tax Company Income Tax Business Tax Consumption Tax Personal Income Tax Figure 7 shows the revenue share of major taxes in total tax revenue from 1985 to 2007. The share of value-added tax in total tax revenue increased from 7.2% in 1985 to 45% in 1994, and then decreased to 33.9% in 2007; the share of business tax increased from 10.3% in 1985 to 13.07% in 1994, and to 14.4% in 2007; the share of corporate income tax decreased sharply from 34.4% in 1985 to 13.8% in 1994, increased to 20.2% in 2006, and declined to 19.2% in 2007. Personal income tax was negligible before 1999, and its share in total tax revenue was 3.9% in 1999 and 7% in 2007. Page 27 26 2. Tax Structure of Other Countries We now examine tax structures in other countries. Table 4 shows the share of various taxes (taxes on income, profits, and capital gains, taxes on property, and taxes on goods and services), as well as non-tax contributors (social security contributions, grants, and others) in

GDP for all countries in the world with data available in 2006. If the data for 2006 is not available, the data for the closest year is used. Table 4 also shows the share of various taxes in total tax revenue. Other taxes are small in generating revenue, and therefore, are not shown in the table. Several observations can be made based on Table 4. First, developed economies heavily rely on taxes on income and capital gains. For example, taxes on income, profits, and capital gains accounted for 63.4% of total tax revenue in the US, 63.6% in New Zealand, 63.1% in Hong Kang, 62.1% in Norway, 61.4% in Denmark, 59% in Australia, and 57.7% in Canada. In developing countries, South Africa, Chile and Singapore have relatively higher shares of taxes on income and capital gains in total tax revenue (52.9%, 48.3%, and 47.3%, respectively). It should be mentioned that personal income tax plays an important role in developed countries as well as developing countries. In 2006, personal income tax accounted for 47.8% of total tax revenue in the US, 45.2% in New Zealand, 42.5% in Canada, 39.4% in Germany, 35.6% in the UK, 27.6% in Hungary, 21.4% in Poland, and 13% in Russia. 29 Second, developing countries rely more on taxes of goods and services. For example, taxes on goods and services accounted for 71.7% of total tax revenue in Croatia, 69.4% in Bulgaria, 50.2% in Thailand, 68.7% in Bolivia, 52.3% in Peru, and 43.7% in Chile. Among industrialized countries, Greece, Portugal, and Netherlands have high shares of taxes on goods and services in total tax revenue (56.3%, 56.7%, and 49.5%, respectively). The share of taxes on goods and services in total tax revenue was only 14.9% in Japan, 21.1% in the US, lowest among all the countries listed. Third, property tax is levied in all developed countries and almost all developing countries, and the revenue from property tax is significant. For example, the revenue from property tax accounted for 14.5% of total tax revenue and 3.1% of GDP in the United States, 16.5% of total tax revenue and 4.6% of GDP in France, 11.5% of total tax revenue and 2.6% of GDP in Argentina. Overall, revenue from property tax is higher in developed countries than in developing countries. Fourth, social security contribution is higher in developed countries than in developing countries. For example, in 2006 the share of social contribution in GDP was 15.96% in Austria, 15.71% in Belgium, 15.13% in Netherlands, 8.39% in the UK and 7.03% in the US. In Australia and New Zealand, social security is financed by general taxes, and therefore, social contributions are negligible. 29 Calculated by the author based on International Monetary Fund (2007), Government Finance Statistics Yearbook . Page 28 27 Table 4 Composition of Taxes and Other Government Revenue for Countries with Data for 2005-2006 (%) Taxes Income, Profits, Capital Gains Property Goods and Services Country or Region Year As % of GDP As % of Total Tax Revenue As % of GDP As % of Total Tax Revenue As % of GDP As % of Total Tax Revenue Total Taxes as % of GDP Social Contri- butions as % of GDP Grants as % of GDP Other Revenue as % of GDP United States 2006 13.46p 63.37p 3.08p 14.50p 4.49p 21.13p 21.24p 7.03p p 5.73p Canada 2006 17.04p 57.74p 3.49p 11.82p 7.95p 26.94p 29.51p 5.40p p 6.76p Australia 2006 17.57 59.16 2.58 8.69 7.76 26.13 29.70 6.56 Japan 2005 8.23 46.66 0.31 1.76 2.62 14.85 17.64 10.64 4.04 New Zealand 2006 22.70p 63.62p 1.99p 5.57p 10.32p 28.92p 35.68p 0.07p p 6.57p Austria 2006 12.43 45.85 0.57 2.10 11.82 43.60 27.11 15.96 0.16 4.60 Belgium 2006 15.97p 52.55p 3.18p 10.46p 11.20p 36.85p 30.39p 15.71p 0.13p 2.48p Finland 2006 16.58p 53.29p 1.10p 3.53p 13.39p 43.04p 31.11p 12.23p 0.26p 8.96p France 2006 10.85 39.44 4.55 16.54 11.09 40.31 27.51 18.31 0.08 4.92 Germany 2006 12.10 52.56 0.83 3.61 10.10 43.87 23.02 17.27 0.21 3.32 Greece 2006 7.46p 37.28p 0.97p

4.84p 11.26p 56.27p 20.01p 13.22p . . Ireland 2006 12.91p 47.01p 0.85p 3.09p . . 27.46p 6.25p . . Italy 2006 14.02 47.87 0.83 2.83 12.54 42.81 29.29 13.02 0.29 3.03 Luxembourg 2006 12.48 49.27 1.43 5.65 11.37 44.89 25.33 10.78 0.03 3.55 Netherlands 2006 10.74 43.34p 1.69p 6.82p 12.27p 49.51p 24.78p 15.13p 0.09p 6.66p Portugal 2006 8.51p 35.07p 1.06p 4.36p 13.75p 56.67p 24.26p 12.47p . . Spain 2006 . . . . . . 24.11p 12.86p 0.59p 2.84p Denmark 2006 29.30p 61.43p 1.86p 3.90p 16.33p 34.24p 47.69p 1.92p 0.08p 6.19p Iceland 2006 18.71 47.96 2.16 5.54 17.57 45.04 39.01 3.36 0.12 6.59 Norway 2006 21.66p 62.13p 1.21p 3.47p 11.89p 34.10p 34.86p 8.72p p 14.89p Sweden 2006 19.33p 52.95p 1.39p 3.80p 12.61p 34.54p 36.50p 12.89p . . Switzerland 2005 13.00 58.56 2.33 10.50 6.67 30.05 22.20 6.98 8.13 UK 2006 14.82p 48.95p . . . . 30.27p 8.39p . . Lesotho 2005 9.93p 22.83p 0.09p 0.20p 8.19p 18.83p 43.49p p 2.00p 5.59p Mauritius 2006 3.63 20.60 1.06 6.02 9.32 52.89 17.62 1.04 0.24 2.21 Seychelles 2005 7.28 20.26 19.57 54.47 35.93 11.24 1.35 9.55 South Africa 2006 16.19p 52.94p 1.72p 5.62p 10.77p 35.21p 30.58p 0.65p 0.01p 6.01p Hong Kong 2005 7.99 63.11 2.44 19.27 2.02 15.96 12.66 7.38 Macao 2006 1.26 5.73 0.77 3.50 19.74 89.77 21.99 0.12 3.31 Maldives 2006 0.77p 3.83p 0.07p 0.34p 4.67p 23.25p 20.08p p 17.66p 23.50p Singapore 2005 6.00 47.32 0.98 7.73 4.66 36.75 12.68 7.23 Thailand 2006 7.38p 39.78p 0.48p 2.59p 9.31p 50.19p 18.55p 0.95p 0.02p 2.43p Albania 2005 3.72 20.98 11.89 67.06 17.73 4.39 0.75 2.01 Armenia 2006 3.79 25.61 0.25 1.69 6.14 41.49 14.80 2.25 0.63 2.23 Page 29 28 Table 4 Composition of Taxes and Other Government Revenue for Countries with Data for 2005-2006 (%) (Continued) Belarus 2006 7.75 22.80 1.60 4.71 18.59 54.69 33.99 11.72 2.50 Bulgaria 2006 5.42 21.65 1.29 5.15 17.38 69.44 25.03 8.78 1.23 4.95 Croatia 2006 6.37 23.91 0.38 1.43 19.11 71.73 26.64 13.52 0.08 4.57 Cyprus 2006 10.16 35.39 0.95 3.31 16.03 55.83 28.71 7.90 . . Czech 2006 8.62p 43.60p 0.42p 2.12p 10.73p 54.27p 19.77p 14.65p 1.11p 2.56p Estonia 2006 7.15 34.95 0.25 1.22 9.61 46.97 20.46 10.31 . . Georgia 2006 5.27 27.45 0.62 3.23 12.18 63.44 19.20 3.65 1.41 3.66 Hungary 2006 9.07 37.14 0.86 3.52 14.22 58.23 24.42 12.62 0.53 5.08 Kazakhstan 2006 9.31 42.65 0.56 2.57 8.43 38.62 21.83 0.54 Kyrgyz 2006 3.10 18.53 0.48 2.87 10.68 63.84 16.73 0.24 5.07 Latvia 2006 8.09 38.93 0.59 2.84 11.82 56.88 20.78 8.51 2.38 3.87 Lithuania 2006 9.67p 46.20p 0.33p 1.58p 10.89p 52.03p 20.93p 9.26p 1.34p 2.22p Malta 2006 11.42p 41.80p 1.57p 5.74p 14.28p 52.26p 27.32p 7.77p 3.11p 3.33p Moldova 2005 4.44 19.04 0.62 2.66 16.20 69.47 23.32 7.83 1.22 6.21 Poland 2006 7.00 32.47 1.23 5.71 12.71 58.95 21.56 12.18 0.65 4.68 Romania 2005 . . . . . . 18.20 9.69 0.59 3.21 Russia 2006 10.32 38.75 1.01 3.79 6.69 25.12 26.63 5.65 8.27 Slovak 2006 5.72p 32.98p 0.43p 2.47p 11.19p 64.53p 17.34p 12.26p 0.42p 5.03p Slovenia 2006 9.16 37.85 0.62 2.56 13.21 54.59 24.20 15.17 1.07 4.27 Ukraine 2006 9.38 39.90 0.30 1.28 12.22 51.98 23.51 13.10 0.03 6.95 Israel 2006 13.84 45.08 2.97 9.67 11.96 38.96 30.70 6.75 2.20 5.40 Kuwait 2006 0.20 24.39 0.03 3.66 0.82 45.06 Argentina 2004 5.18p 22.61p 2.64p 11.52p 10.97p 47.88p 22.91p 3.01p 0.01p 3.49p Bolivia 2006 3.2 12.94 2.95 11.93 16.99 68.70 24.73 1.88 15.96 8.31 Chile 2006 10.6 48.25 0.58 2.64 9.61 43.74 21.97 1.36 4.51 Costa Rica 2006 3.48 23.61 0.67 4.55 9.35 63.43 14.74 7.41 0.06 2.91 El Salvador 2006 4.37 31.97 0.32 2.34 7.89 57.72 13.67 1.96 0.1 2.48 Jamaica 2005 5.28 19.02 0.23 0.83 11.76 42.36 27.76 2.54 0.12 5.26 Paraguay 2006 1.84p 14.45p 0.39p 3.06f 7.75p 60.88f 12.73p 3.29p 0.10p 6.38p Peru 2005 4.28p 31.01p 0.21p 1.52p 7.21p 52.25p 13.80p 1.54p 0.10p 3.21p Source: Calculated based on

International Monetary Fund, Government Finance Statistics Yearbook, 2007. Notes: indicates that a figure is zero or less than half of a significant digit. . indicates the absence of data. f denotes forecasted or projected data. P denotes data that are preliminary or provisional. marks a break in the comparability of data; that is, data appearing after the symbol do not form a consistent time series with those for earlier years. Typically, break symbols will appear in the summary statements or detailed tables when, for example, changes have occurred in the coverage and classification of data or when the basis of recording has changed from cash to noncash. Page 30 29 Eastern European countries and former USSR countries have reasonably high ratios of social contribution to GDP. For example, the ratio of social contribution to GDP was 14.65% in Czech, 13.1% in Ukraine, and 5.65% in Russia. Most developing countries have relatively small social contributions in government revenue. For example, the ratio of social contribution to GDP was 1.54% in Peru, 1.32% in Chile, 3.1% in Argentina, and 0.95% in Thailand. 3. Prospects for Further Tax Reforms in China Economic development and social stability require governments to have adequate revenue for the provision of essential public goods and services. In the future, China's tax reforms would be structural, ie, reducing some taxes, increasing some taxes, and establishing new taxes. First, reduce tax burden on enterprises. Corporate income tax is high in China compared with other economies. Starting January 1, 2008, the corporate income tax rates for domestic firms and foreign firms have been unified at 25%. China's accounting system follows the Soviet style, and is different from the western style. It was estimated that China's effective tax rate was 50% higher than countries with high corporate income tax. 30 Thus, there is still room to lower the corporate tax rate. Recently, the government excluded investment from the tax base, effectively lowering VAT on firms in production areas. However, business tax for firms in the service industries remains unchanged. To stimulate the development of service industries, the government should consider lowering business tax. Second, collect more revenue from personal income tax. The top tax rate for personal income is high but the share of personal income tax revenue in total tax revenue is small. The top marginal tax rate for personal income is 45%, higher than that of many countries including the US (35%) and the UK (40%). The standard deduction is 2,000 yuan per month; there is no other deduction. Monthly personal income is taxed progressively at 5% - 45%, starting from taxable income 1-500 yuan. In the US, there are many other deductions, such as trade or business expense deductions, moving expenses, itemized deductions or standard deduction, deductions for the taxpayer's dependents, etc. The taxable income is much smaller than the gross income. Even though China's top marginal tax rate is high, only a very small group of individuals is subject to the top tax rate. A part of the compensation to employees is in-kind or in cash and is therefore not documented and taxable. Thus, tax evasion is still a serious problem. As a result, personal income accounted only for about 7% of total tax revenue and about 1.4% of GDP in 2007. The percentage of US federal personal income tax revenues in GDP from 1946 through 2006 was around 8%. In the US, many state governments also collect personal income taxes. China's revenue from personal income tax is not only substantially lower than that of the industrial countries, but also lower than many developing countries at the same level of economic development. Thus, China should consider lowering the maximum

marginal tax rate and strengthen its accounting system by collecting taxes on all taxable income. 30 Xu (2001). Page 31 30 Third, establish new taxes, such as property tax, inheritance tax, and gift tax. As can been seen from Table 4, almost all countries collect property taxes, and the revenue share of this tax is quite high. Property tax accounted for a substantially high proportion of tax revenue in US history. For example, in the early 1920s the revenue from property tax accounted for more than 45% of total government revenue. 31 As revenue from personal income tax increased, the revenue from property tax has declined. However, property tax is still the main source of local government revenue, accounting for 73.2% of local government taxes. 32 Property tax rates vary from state to state. For example, in 2007, it was 2.57% in Texas and only 0.4% in Hawaii in 2007. The rates of inheritance tax and gift tax are also high in the US. In 2007, the federal estate tax rate was 45%. Introducing property tax, inheritance tax, and gift tax can increase government revenue and redistribute income. VIII. Summary This paper has discussed China's tax reforms, analyzed the reasons for the ups and downs of budgetary revenue and the extra-budgetary revenue, compared the size of China's government revenue with that of other countries, examined the structural problems in the current tax system, and provided suggestions for further tax reforms. Tax reforms went through a number of changes: lowering taxes and allowing state enterprises to keep a part of their profits (1979); substituting taxes for profit (1983-1984), introducing the contract responsibility system that required SOEs to pay income tax and adjustment tax at a specific level of profit based on a contract with the government (1986); reviving a profit remitting system which required SOEs to submit a portion of their profits to the government after paying corporate income taxes; establishing the tax sharing system which divided taxes into central government taxes, joint taxes of the central and local governments, and local government taxes (1994); eliminating agricultural taxes (2006); merging corporate income taxes for domestic and foreign enterprises (2008); and excluding investment from the value-added tax base (2009). Early tax reforms provided great incentives for stateowned enterprises to make profits, but caused a large decline in government budgetary revenue and a huge expansion of extra-budgetary revenue. The 1994 tax reforms laid the foundation for China's modern tax system and government revenue has grown rapidly since then. Value-added tax, business tax, consumption tax, and corporate income tax become the major sources of government revenue. Rapid economic growth has resulted in rapid growth of government revenue. High tax rates, broadened tax base, and strengthened law enforcement have pushed the government budgetary revenue share in GDP from 10.7% in 1995 to 20.6% in 2007. As budgetary revenue rises, the relative size of extrabudgetary revenue has significantly declined in the past decade. 31 See Wallis (2000). 32 See US Census Bureau (2004), Local Governments--Revenue by State. Page 32 31 Including social security contributions and budgetary and extra-budgetary revenues, China's government revenue share in GDP is around 26% of GDP, which is lower than any industrial countries, and lower than most developing countries. However, China's government revenue share in GDP is higher than some Asian economies, such as Singapore, Hong Kong, and Thailand. The Chinese

government now heavily depends on value-added tax, business tax, and corporate income tax. Personal income tax is low in China, and personal property tax has not been established, which are important for revenue collection and income redistribution. Further tax reforms should aim at reducing taxes on enterprises, raising more tax revenue from personal income, and establishing new taxes, such as personal property tax, inheritance tax, and gift tax. Page 33 32 References Bahl, Roy, and Christine Wallich, 1992, Intergovernmental Fiscal Relations in China, Working Papers, WPS 863, the World Bank. Brean, Donald, 1998, Financial Perspectives on Fiscal Reform, in China's Tax Reform Options , Trish Fulton, Jinyan Li and Dianqing Xu, ed., World Scientific, Singapore/New Jersey, 47-56. Chen, Gong, 1996, Zhenxeng Guojia Caizheng de Jige Zhengce Wenti , Caizheng , 9. Also in The Finance Yearbook of China, Ministry of Finance, ed. , Publishing House of Financial Journals, Beijing China, 1997, 763-769. Huang, Kehua, Bawo Gaige Jiyu, Zhenxeng Guojia Caizheng , Caizheng , 9, 1996. Also in The Finance Yearbook of China, Ministry of Finance, ed. , Publishing House of Financial Journals, Beijing China, 1997, 763-769. International Monetary Fund, 2007, Government Finance Statistical Yearbook 2007 . International Monetary Fund, 2002, Government Finance Statistical Yearbook 2002 . Li, Jinyan, 1991, Taxation in the People's Republic of China , New York: Praeger. Lin, Shuanglin, 2008, Forced Savings, Social Safety Net, and Family Support: a New Oldage Security System for China, The Chinese Economy , forthcoming. Lin, Shuanglin, 2006, China's Public Finance Reform: Challenges and Options, The Peking University Journal, Vol. 43, No. 6, 102-110. Lin, Shuanglin, 2005, The Excessive Fee Collections in China: Reasons, Consequences, and Strategies, Contemporary Economic Policy , Vol. 23, No. 1, January, 91-106. Lin, Shuanglin, 2000, The Decline of China's Budgetary Revenue: Reasons and Consequences, Contemporary Economic Policy , Vol. 27, Number 4, 477-490. Liu, Zhouwei, 2000, Fei gai shui: shui you quanli shoufei (Replacing fees by taxes: who has the power to collect money?) Jia Kong, ed., Shuifei gaige yanjiujiu wenji (Readings in the research of fee and tax reforms), Beijing: Economic Science Publishing House, 179-184. Liu, Xinyi, 1998, Zhonggou Caishi Gaige de Lilun Sikao, Hubei Caishi , 4, 1998, 4- 7. Ministry of Finance of PRC, 2008 , On the Draft of Central and Local Budgets for 2008 , First Session of the 11 th National People's Congress, March 5. Ministry of Finance of PRC, 1996, Tax Division, Tax System in China , Beijing: Enterprise Management Publishing House. Page 34 33 Ministry of Finance of PRC, 1997-2007 , The Finance Yearbook of China, Publishing House of Financial Journals, Beijing, China. National Bureau of Statistics of PRC, 2009, Statistical Bulletin on National Economy and Social Development in 2008. National Bureau of Statistics of PRC, 1991-2008, Statistical Yearbook of China , China Statistical Publishing House, Beijing, China. National People's Congress of PRC, 1994, The Budget Law of the People's Republic of China , passed by the Eighth National People's Congress Second Meeting on March 22. People's Bank of China, 2009, The Report on Monetary Policy Implementation for the 4 th Quarter of 2008, China.com.cn, February 23. State Council of PRC, 2001, The Notice on the Announcement of the Reform Plan on the Sharing Income Taxes , State Council Document No. 37. State Council of PRC, 1994, Provisional Regulations of the People's Republic of China

on Value-Added Tax, in A Collection of Tax Laws and Regulations of the People's Republic of China, Tax Bureau of the People's Republic of China, ed., China Statistical Publishing House, Beijing, 1994, 102-108. State Council of PRC, 1984, The State Council's Notice on Raising Revenues for Rural Education, State Council Document, No. 174. Stiglitz, Joseph, 1998, China's Reform Strategies in the Second Stage , A Speech at Peking University, People's Daily , November 13. Xu, Shanda and Zhang Xuerui, 2001, The Background of China's Tax Reforms in 1994, http://www.macrochina.com.cn/zhtg/20010904018551.shtml. US Census Bureau, 2004, Local Governments--Revenue by State. Wallis, John J., 2000, American Government Finance in the Long Run: 1790-1990, Journal of Economic Perspective, Vol. 14, No. 1, 61-82. Wang, Bingqian, Renzhen Guanche Shwuda Jingshen Kaichuang Caizheng Keyan Xinjumian, Caizheng Yanjou , 3, 1998, 2-11. Wu, Shi-an, 1997 , China's Fee Collection Research, Zhongguo Caizheng Jingji Chubanshe (China Finance and Economics Publishing House. Yu, Dingcheng, 1997, Jingji Zhuhuo, Shuishou Zhu Kong, Yiangzhou Daxue Shuiwu Xueyuan Xuebao , 1, 5-10.