Using the Government Financial Reporting Framework to Redraw the State and Market Boundary in China: A Two-Step Approach Christine Wong and Min Zhao1 First draft, March 2017 Updated, June 2018 Abstract After four decades of remarkable economic achievement under market reforms, the leadership has called for a reset in the boundary between the State and the Market as an important corrective to help China sustain rapid economic growth, by imposing hard budget constraints on government and insulating SOEs from local government predation. This could start with revealing and reviewing the current operation and finance of the government through the new Government financial reporting framework (GFRS). The sheer size of SOEs and their engagement in provision and finance of public goods and services poses great challenge for China to immediately adopt international standard for GFRS. Given their huge size and diverse characteristics, it is neither correct nor practical to include all SOEs in the public sector. We therefore proposed a two-step approach for using the GFRS to redraw the boundary of the state and market. The first step is to adopt an accounting framework that aims to provide a comprehensive count of government operation and finance, focuses on the fiscal impact of entities, and simplifies the reporting requirements for the vast majority of SOEs. The second step is to review the government operation and finance with an economic framework. It is also hoped that the exercise itself will stimulate further reform of SOEs and a rethinking of the division of responsibilities between government and market. While one should not expect to reach a clear and ideal division between the state and market overnight, with successive iterations, the exercise will lead incrementally to greater clarity and improvements, as the process of implementing the GFRS sets off a beneficent cycle for China’s economic transformation to a higher quality and sustainable growth. 1 This policy note is prepared at the request of China Ministry of Finance. We thank China MOF for facilitating the meetings with representatives of local government finance officials and state-owned enterprises, and for their helpful comments. Remaining mistakes are our own. Questions and feedback can be addressed to Min Zhao (mzhao1@worldbank.org). 1 I. Introduction In November 2013, the Chinese government mapped out an ambitious reform program that anchored its vision of economic transformation on modernizing the system of governance, to make government more effective in delivering public services and reduce opportunities for corruption. The program was endorsed at the third Plenum of the 18th and the 19th Central Committee of the Communist Party of China. 2 A prerequisite for the modernization of the governance system is to clarify the boundary of the state and market to solve the twin problems of both excessive market intervention and inadequate regulation by government. Fiscal reform has played a leading role in this reform program, starting with passage of the revised Budget Law in August 2014, followed by State Council Document No. 63 in December. Together the documents laid down a mandate for Governments at all levels to compile and release to the public a comprehensive Government financial report that provides an accurate and comprehensive reflection of Government financial outcomes including public assets, liabilities, revenues, expenditures and cash flows. The Budget Law also called for the reporting of budget information on an accrual basis. Together, these requirements essentially called for the creation of a new Government financial reporting system (GFRS)3 that differs sharply from the current practice of reporting only on-budget revenues, expenditures and direct debts. The new GFRS will be an important foundation for building a modern system of fiscal management that will promote fiscal sustainability and support good governance. Selecting and classifying the reporting entities are critical first steps in building a GFRS, especially in China, where the transition to a market economy is not yet complete, and the boundary between the public and nonpublic sector remains fluid. In particular, the state- owned enterprise sector is huge. As seen in Annex 1, nationwide, in 2013 there were more than 18,000 “above scale� state owned enterprises (SOEs), with assets of 34 trillion yuan, and they are scattered throughout the whole country, albeit unevenly. Chinese Ministry of Finance (MOF) developed an Experimental Approach for Compiling Government Comprehensive Financial Report in late 2015, and piloted it in selected government agencies and subnational governments (MOF Decree, CaiKu No. 212, 2015). A more detailed guideline was issued in 2016 and marginally updated in 2018. This experimental approach doesn’t specify the scope and classification of entities to be covered in the government comprehensive financial report. Instead, it stipulates that the report shall consolidate the financial statements of all accounts under the direct control of the Treasury including the financial statements of Fiscal General Budget Accounts, Agriculture Comprehensive Development Fund, Departments’ Account, Land Reserve Fund and Resources Reserve Fund. The intersectoral transactions between these financial statements shall be offset. The government equity investment in and investment income from SOEs, that currently are not reflected in the government general ledger, shall be derived from the SOEs’ financial statements. This implies that China’s GFRS will 2 See “Decisions regarding key issues on deepening reforms by the central committee of the Communist Party of China.� 3 To distinguish this new system in China from the Government Finance Statistics (GFS) system widely adopted internationally, this paper will refer to the Chinese system of Government finance reporting as GFRS. 2 cover only government departments and some of public service units which are budget financed and whose accounts are included in the government general ledger. This paper proposes that China reveal and review the operation and finance of the whole of government in its GFRS, drawing on international standards, good practices in selected countries and analysis of China’s public sector institutional structure. This requires expanding GFRS coverage to include the public benefits SOEs in addition to the administrative units and public services units. Inclusion of public benefit SOEs in GFRS could help put the spotlight on the accounting practices of SOEs and bring more transparency to their reporting. The mandate that government financial reports be released to the public will promote transparency and invite public scrutiny of the public sector. This will enhance the subnational governments’ accountable for the use of public resources, and help mitigate the risk of off-budget subnational borrowing through UDICs and other public benefit SOEs. In this paper, we propose first applying an accounting framework to decide how entities shall be included and classified in GFRS. Taking into account the institutional structure of China’s public sector and keeping in mind the government’s exhortation for quick implementation, our framework focuses on the fiscal impact of entities, and simplifies the reporting requirements for the vast majority of SOEs. It is hoped that the framework is straightforward and easy to apply, and will bring a significant improvement in government financial reporting quickly. GFRS, if revealing the true operation of the whole of government, will provide a great opportunity for China to make this process a powerful driving force for reform not only for public financial management, but for other areas and sectors as well. The process will shed more light on the financial interactions between SOEs and Government and clarify the boundary between them. This paves the way for the next step in public financial management reform, to query the justification for state intervention in each entity, toward realigning public expenditures to focus more tightly on Government policy objectives. In the process of implementing the new GFRS, each budgeting authority faces a series of decisions it has to make over which entities to include in the financial report, and how to include them. At each step, the choices spur consideration about the function and role of government, and the process sets in train a dynamic that pushes toward clarifying the boundary between the state and the market. This logical next step is to examine whether the entities currently included in the public sector can be justified on the grounds of economic efficiency or social equity. As introduction of the new GFRS provides more comprehensive and accurate information on the use of public resources and the Government’s financial status, the Government can begin to undertake the task of cleaning up and adjusting public expenditures to improve efficiency and putting Government budgets on a more sustainable basis. This task is especially critical for Government at subnational levels, where more than 85 percent of public expenditures and 95 percent of public investments take place. In this paper, we also propose applying an economic framework to examine the justification of government intervention in the current context of China. While we propose a two-step approach to redrawing the boundary of state and market, these two steps should not necessarily be sequential. In many countries, the transition to comprehensive accrual accounting alone involves a process that can take up to a decade (Liu, Pradelli and Zhao 2015). In China, the size and difficulty of the project of building the new GFRS 3 is magnified multiple-fold by the large size of the country and the highly decentralized fiscal system - the GFRS will need to be adopted by subnational Governments down to the county and district level, a group that includes more than 3200 budgeting authorities with widely varying capacities.4 Given the arduousness of this first task of building the GFRS, it may be tempting to put off the second step of examining the economic justification for each entity. We will argue in this paper that these steps should be iterative. And the iterative use of the two-step approach will lead to an incremental adjustment of the boundary between the state and the market. As this process progresses, a new framework for the governance and management of public benefit SOEs may also be introduced.5 Dadukou district of Chongqing Municipality, under the World Bank’s support through development policy lending, has adopted a transitory arrangement to restructure UDICs by splitting the accounting of governmental activities and commercial activities. With the success of this transitory arrangement in Dadukou, Chongqing later scaled up this practice to all districts. The paper is organized as follows: after this introduction, Section II provides a brief description of existing international standards (IPSAS 35 and the IMF’s GFS framework) and practices of selected countries. Section III describes the current Chinese institutional structure of public entities and offers some estimates of their size and structure. Section IV proposes an accounting framework for classifying these institutions and for including them in the GFRS. Section V will present an economic framework for examining the justification for Government intervention in the provision of public services and the different policy choices available. To illustrate the challenges and opportunities in implementing these reforms, in Part VI we present the experience of Dadukou, a district government in Chongqing Municipality, in adopting this framework. Part VII will provide some concluding comments. II. International Standards and Practices in Selected Countries Both IPSAS and GFS define the coverage of reporting entities, and the classifications of the reporting entities and their presentation in the report. IPSAS is the only international standard for public sector accounting and financial reporting that currently exists in the world, and GFS is the internationally accepted system of government finance statistics. The IPSASB and the IMF have made many efforts to harmonize the principles of IPSAS and GFS, and to eliminate unnecessary differences in terminology and coverage, but significant differences remain between IPSAS and GFS due to their different aims and objectives. Some countries have developed their own standards while using these two frameworks as references. A brief description of IPSAS 35: Consolidated Financial Statements The key principles defining the coverage of entities are documented in IPSAS 35, which was released in January 2015 and supersedes IPSAS 6.6 It requires that a government consolidate, on 4 This includes 31 provincial governments, 333 prefectural level governments, 2854 counties and districts, and, of course, the central Government itself (Chinese Statistical Yearbook). 5 Australia is one country that has introduced a regime for the governance of government business enterprises. http://www.finance.gov.au/gbe-directors-guide/ 6 The IPSAS Board revised the definition of control criterion and made it more consistent with GFS. According to the International Federation of Accountants, “This standard still requires that control be assessed having regard to 4 a line-by-line basis, all of its controlled entities and present them in a whole-of-government general-purpose financial statement. The consolidated financial statements combine like items of assets, liabilities, net assets/equity, revenues, expenses and cash flows, and eliminate intra-entity assets, liabilities, net assets/equity, revenues, expenses and cash flows. A complete set of financial statements comprises: 1) A statement of financial position; 2) A statement of financial performance; 3) A statement of changes in net assets/equity; 4) A statement of cash flow; 5) When the entity makes publicly available its approved budget, a comparison of budgeted and actual amounts either as a separate additional financial statement or as a budget column in the financial statements; and 6) Notes, comprising a summary of significant accounting policies and other explanatory notes. IPSAS determines that the government controls an entity if it meets all of the three criteria: 1) Power over the entity (such as voting, right to appoint senior management, right to approve or veto the entity’s business decision); 2) Exposure, or rights, to variable benefits (financial or non-financial) from its involvement with the entity; and 3) The ability to use its power over the entity to affect the nature or amount of the benefits from its involvement with the other entity. IPSAS 35 introduced a concept of investment entity and provided that these investment entities controlled by government units be exempted from consolidation requirement. A controlling entity shall measure the investments of a controlled investment entity at fair value through surplus or deficit. Investment entities are defined as an entity with exclusive objective of investing for returns from capital appreciation, investment revenue (such as dividends or similar distributions, interest or rental revenue), or both. If the entity has additional social or public policy objectives such as improving employment outcomes in the jurisdiction, it would not meet the definition of an investment entity. A brief description of the GFS framework The GFS is an internationally accepted system for collecting and reporting government finance statistics developed by the IMF. The GFS framework has evolved over time. The 2014 GFS Manual (pre-publication draft) shows the framework has been harmonized with IPSAS to a greater extent, and is based on accrual accounting. 7 It is designed to measure the economic activities of the public sector, including revenues, expenditures, the resulting surplus or deficit, assets and liabilities and net worth or net debt position. The core of the analytic framework is a set of four financial statements, and two supplementary statements: benefits and power, but the definition of control has changed and the standard now provides considerably more guidance on assessing control. The definition of control focuses on an entity’s ability to influence the nature and amount of benefits through its power over another entity.� http://www.ifac.org/publications-resources/ipsas-35- consolidated-financial-statements. 7 IMF (2014). Government Finance Statistics Manual 2014. Washington, DC. International Monetary Fund. 5 1) The statement of government operations; 2) The statement of other economic flows; 3) The statement of balance sheet; 4) The statement of sources and uses of cash; 5) The statement of total changes in net worth (supplementary); 6) The summary statement of explicit contingent liabilities and net implicit obligations for future social security benefits (supplementary). In comparison to IPSASs’ control and benefit criterion for determining reporting entities and consolidation, a key advantage of the GFS is that it provides additional information for analysing public sector activities by making a distinction between the general government sector and the broad public sector. In the GFS, “general government� is defined as comprising the set of resident entities that perform the functions of government as their primary activity, including the central government, state governments, local governments, and social security funds. The “public sector� consists of all units of general government and resident public corporations. The GFS framework pertains to the general government and public sectors as defined in the 2008 System of National Accounts. Corporations are recognized public corporations where government is in a position to exercise control. The key factors to be considered are: 1) Ownership of the majority of the voting interest; 2) Control of the board or other governing body; 3) Control of the appointment and removal of key personnel; 4) Control of key committees of the entity; 5) Golden shares and options; 6) Regulation and control; 7) Control by a dominant customer; and 8) Control attached to borrowing from the government. Figure 1 shows the relationship between the general government sector, the public sector, and the other main sectors of the domestic economy. Figure 1 The Public Sector and its Relation to Other Institutional Sectors of the Economy General Nonfinancial Financial Households Sector Nonprofit Government Corporate Sector Corporate Sector Institutions Sector Serving Households Sector Central government Public corporations Public corporations State government Private Private Private corporations Private corporations Local government Public Sector Source: GFS manual, IMF, 2014, page 19 The critical characteristic in determining which entities are classified as general government and which are public corporations is the economic nature of the entity as a market producer, in that it produces goods and services and charges economically significant prices for them. 6 The GFS Manual defines economically significant prices as prices that have a significant effect on the amounts that producers are willing to supply and on the amounts buyers wish to buy. By extension, a market producer is a unit that provides all or most of its output to others at prices that are economically significant while a nonmarket producer provides all or most of its output to others for free or at prices that are not economically significant. It is important to differentiate the entity classification in GFS from the entity’s legal and social status. In the GFS Manual, three types of legal and social entities are described and defined for inclusion in government financial accounts: government units, public corporations, and non- profit institutions (NPI). Government units are legal entities established by political processes, and have legislative, judicial, or executive authority over other institutional units within a given area. Corporations are defined as entities that are capable of generating a profit or other financial gain for their owners, are recognized by law as separate legal entities from their owners, and are set up for purposes of engaging in market production. Nonprofit institutions (NPIs) are legal or social entities created for the purpose of producing or distributing goods and services. They differ from public corporations in that while their activities can generate surpluses, the surpluses cannot be appropriated by the institutional units that establish, control, or finance them. Government units typically are uncontroversial part of general government, while it is challenging to classify corporations and NPIs. NPIs may engage in market production and charge economically significant prices for their services – i.e. sufficiently high to have an impact on the demand for their services, and they may also be able to raise additional funds through donations. In these cases, these NPIs are market producers, and must be treated as public corporations if they are controlled by government units. NPIs controlled by government, engaged in nonmarket production and charging fees that are not economically significant are treated as government units. Given this classification of NPIs, it can be seen in Figure 2 that within the GFS framework, NPIs “disappear� as a distinct type of entity, leaving the public sector with only two types: general government and public corporations. In the real world, NPIs are common in all countries, and often play important roles in the provision of public services – e.g. schools, universities, clinics, hospitals, etc., and they are commonly considered “broader government institutions.� Within the government financial reporting framework, however, NPIs are classified (and included) as either government units or public corporations. This distinction follows from the extent to which the entity depends on government or market funding. As for corporations, we can use as illustrative examples an electric power company and a water company. Electric rates are “economically significant� in that they are high enough to affect household demand for power, while at current levels, water prices are not yet significantly affecting household water use. The former – the electric power company, would be classified as a public corporation, while the water company would be classified as a government unit in the GFS framework. (And both would be included as controlled entities under the IPSAS standards.) Because of the special status and economic characteristics of the central bank and state-owned 7 commercial banks, the GFS further divides public corporations into two subtypes: financial and nonfinancial. Figure 2 provides a schematic “decision tree� for classifying public entities. Figure 2 A Decision Tree for Classification of Public Entities Source: Government Finance Statistics Manual, IMF (2014) Practices in selected countries Liu, Pradelli and Zhao (2015) reviewed the government financial reports of four states in three countries, namely the state of Victoria in Australia, the state of Maryland and the Commonwealth of Virginia in the USA, and Ontario Province in Canada. All of them apply the control criteria (in modified form) in determining which entities should report through GFRS. 8 But these subnational governments classify public entities in sectors by using different methodologies. There are no substantive differences in the methodologies. The differences mainly affect how public entities are classified and reported - the financial activities of a public entity may be consolidated with the primary or general government, or may not be consolidated but reported in a different column, schedule, or in a separate volume. What is important to note is that even if the activities of a particular public entity are not consolidated, their inclusion in the GFRS (through a column, schedule, or column) shows that the activity of this public entity is an integral part of the GFRS. Victoria (Australia) takes control criteria for disclosure and consolidation, and then within all the reporting entities that meet control criteria, Victoria further classifies the reporting entities into primary government, and public corporations (financial and non-financial). The classification is similar to the GFS framework. Victoria reports the financial statements of the primary government, public non-financial corporations, and public financial corporations in three parallel columns, respectively; the data in the three columns are then consolidated into the fourth parallel column. Maryland and Virginia (USA) also follow the control criteria in determining which entities should be disclosed in GFRS, and classify the entities into two major sectors—primary government and components units. Their respective financial statements are presented in different schedules, and the two sectors are not consolidated as required by IPSAS. The primary government itself is further divided into governmental activities and business activities. Ontario (Canada) also follows the control criteria in determining which entities should be disclosed in GFRS. However, government business enterprises controlled by government are not consolidated in the financial statement line by line as required by IPSAS. Further, Ontario classifies all reporting entities into Ministries, governmental organizations and government business enterprises. The main financial report reveals the consolidated financial statement, while the financial statements for ministries, governmental organizations and government business enterprises, respectively, are released in supplementary volumes. III. Public Sector Institutions in China In China, the government is organized in five tiers: 1) the central government, 2) 31 provinces (including four metropolitans and 5 autonomous regions), 3) 333 municipalities, 4) more than 2800 counties and districts, and 5) 40,000 towns and townships. Under each government are ministries, departments, bureaus and agencies. These form what is commonly called “core government� (政 府机关), comprising “government administrative units� (政府行政��). Under each unit of government are a large number of entities outside core government, which provide public services such as education, health care, culture, agricultural services, as well as research institutes. As a group these entities are called public service units (PSUs, 事 业��), different from SOEs. PSUs number roughly 650,000, employ nearly 31 million people, and had assets of RMB 5.3 trillion (compared to 7.2 trillion in core government) at year-end 2012.8 To the extent that they are government owned, and provide services at government-stipulated 10 This section is based on Wong and Zhao (2015). 9 prices, they are like the nonmarket NPIs described in the GFS Manual, and would belong under the classification of government units. Some government units and PSUs manage a large number of SOEs. During the SOEs reform in mid-1990s, most SOEs have been corporatized and many non-financial SOEs were moved out from the control of functional departments to a newly created government unit —State Assets Supervision and Administration Commission (SASAC). By 2013, only for SOEs under the supervision of SASAC at all levels of governments, the number totalled 109 thousand, with assets of RMB 85.4 trillion, equity of RMB 30 trillion, and operating revenues of RMB 42 trillion (equivalent to 71.4 percent of GDP).9 If we include all PSUs as government units and treat SOEs as public enterprises, the public sector in China could be constructed as a tripartite structure, with (1) core government (政府 行政��), (2) PSUs (事业��), and (3) SOEs (国��业). Estimates for the size of these three components are shown in Table 1. A salient feature of the public sector in China is that both the PSU and SOE sectors are extraordinarily large compared to those in the countries presented in Liu, Pradelli and Zhao (2015). Table 1 Estimated Assets and Liabilities of the Chinese Public Sector, 2012-2013 (RMB trillion) Core government (1) PSUs (2) (1)+(2) SOEs (3) (1)+(2)+(3) Gross assets 72.1 53.3 125.4 170.1 295.5 Total debt 16.8 16.3 33.1 155.4 188.5 Net assets 55.3 37 92.3 14.6 106.9 As percentage of GDP Gross assets 123% 91% 213% 289% 503% Total debt 29% 28% 56% 264% 321% Net assets 94% 63% 157% 25% 182% Source: Estimates from Du Jinfu (2015), “Scientifically Reflecting Government Control Capacity and Ability: Chinese Government Balance Sheet Statement, 2012-2013�, Volume 3, Journal of China Finance. The complex and diverse universe of SOEs The ownership and control structures are extraordinarily complex for SOEs, as are their financial and functional structures. This is illustrated by the situation in Beijing Municipality. As shown in Table 2, data from four authoritative sources give significantly different statistics on the numbers of enterprises and their aggregate assets and liabilities. Table 2 Differing Estimates of SOEs in Beijing Municipality, 2013 (RMB billion) Number of Operating Total Assets Total Debt Equity Profits Enterprises Revenues A 790 2,333.5 1,202.1 1,131.4 1,068.0 70.9 B 549 573.3 458.6 395.7 436.2 33.1 C -- 2520.9 1724.5 796.4 949.9 47.2 10 This section is based on Wong and Zhao (2015). 10 D* 7040 3513.5 -- 1098.6 1106.6 65.3 Sources: A – China Statistical Yearbook 2014, Table 13-4; B – Beijing Statistical Yearbook 2014, Table 11-8; C – Beijing Municipal SASAC website (http://www.bjgzw.gov.cn/qt/forepageview/gzhy/gzhy15.jsp), accessed 6/9/15; D – Ministry of Finance, China Accounting Yearbook 2014, p.605-606. One cause of the differences may be that some of the data refer to all SOEs while others only to above-scale SOEs, but a bigger cause of the conflicting numbers may be the complex ownership and control structures that characterize many SOEs. To illustrate, there are 46 SOEs that report directly to the Beijing Municipal State-owned Assets Supervision and Administration Commission (SASAC), but most of these 46 are actually enterprise “groups� comprising many subsidiary companies, so that the number of municipal-owned SOEs is many multiples of 46. For example, the Beijing Municipal Drainage Group, one of the 46, has 28 subsidiaries that engage in different parts of the water business, from supplying fresh water to disposal of waste water and pipeline management, many of them independent legal entities. Taking another example, the Beijing Municipal State-owned Asset Management Corporation has 23 second-tier and more than 80 third- and fourth-tier subsidiaries, for a total of 114 in the “family� that includes two listed companies. Many of the SOEs, but not all, engage in some public goods provision alongside their market activities. Indeed, the distinction among the three types of public sector entities – government units, public corporations, and NPIs - is fuzzy in China, and some SOEs may include all three types among their subsidiaries. For example, the Beijing Municipal Infrastructure Investment Corporation is responsible for the financing and construction of subways - a task that has clear and important public goods characteristics, but it also owns 11 subsidiary companies that engage in for-profit land and real estate development. The mixed nature of SOE businesses is shown in Annex 2 for a small sample of municipal-owned SOEs in Beijing, along with their complex web of financial interactions with the municipal government. The huge number of SOEs presents some special problems for government financial reporting in China. In accordance to the control criterion in IPSAS standards, all SOEs must be consolidated line-by-line in the GFRS, which would impose an enormous and perhaps impossible task for immediate adoption. There is some rationale for arguing that including all SOEs in the GFRS would exaggerate the size of the public sector since the vast majority of SOEs are not engaged in the production of “public goods� in the economic sense. Moreover, after nearly two decades of reform, the finances of commercial SOEs are largely delinked from the government. However, excluding all SOEs would obviously underestimate the true size of resources that the government mobilizes to finance the delivery of public services and, more importantly, the liabilities subnational governments incurred to finance infrastructure investment. Yang et.al (2015) estimated the balance sheet of the whole-government and classified government in central government and subnational governments and SOEs in local government financing vehicles (LGFVs) and other SOEs. According to this estimation, the public-sector gross liabilities in China is about RMB 40 trillion in 2014, or 62 percent of GDP. Of these, 16.6 trillion are LGFVs’ debt. 11 Table 3. The Whole-Government Balance Sheet, 2012-2014, RMB Trillion 2012 2013 2014 2012 2013 2014 Government's Deposit at Central Bank 2.1 2.9 3.1 Central Fiscal Domestic Debt 7.7 8.6 9.5 Soverign External Debt 2.8 3.2 3.4 Land & Natural Resources Assets 57.2 62 65.4 Non-LGFVs Subnational Debt 6.5 8.6 10.6 Administrative Units' State Assets 9.6 11.8 13.4 Subnational LGFVs' Debt 13.5 15.5 16.6 Non-Financial SOEs' Total Assets 82.8 96.4 116.2 Non-financial SOEs' Debt (excluding LGVFs) 44 51.6 65.4 Financial SOEs' Total Assets 15.2 20.3 27.7 Policy Banks' Financial Debt 7.9 8.9 10 Banks' Non-Performing Assets 0.5 0.6 0.8 Contingent liabilities from Disposing Bank's 4.2 4.2 4.2 Non-Performing Assets National Social Security Funds' State Assets 1.1 1.2 1.5 Hidden Debt of Pensions 3.6 3.6 3.6 Assets Total 168 194.6 227.3 Liabilities Total 90.7 104.8 124.1 Net Worth 77.3 89.8 103.2 Source: Estimates from Yang Li et al. (2016) “China State Balance Sheet 2015�, page 29. IV. A Two-Step Approach to Redraw the Government Boundary The presence of the sheer size of SOEs and their engagement in the delivery and finance of public goods and services in the complicated manner brings two challenges for China to draw a clear boundary of the government. The government financial reporting framework could help redraw the boundary of the government if 1) it reveals the true performance of the government operation and finance; and 2) the authorities capture the opportunity to review and justify the government’s interventions. We therefore propose a two-step approach to redraw the government boundary, the first is to adopt an accounting framework for GFRS with an aim to provide a comprehensive depiction of the government operation and finance; and in the second step, an economic framework is recommended to decide which public goods and services shall be included in the government provision, and how they should be delivered and financed. 4.1 Step One: the Accounting Framework for Government Financial Reporting10 To move from the starting point of the current budget reporting which shows little or no information on public service units (PSUs), state-owned enterprises or the Government’s financial interaction with them, the critical and obvious first step is to build a new GFRS that provide comprehensive information of government operation with appropriate selection and classification of the reporting entities. The accounting framework proposed here is to provide some simple guidelines for which entities to include, with the view toward facilitating a quick adoption of the new GFRS by local Governments. In this framework, we suggest that for selecting entities for inclusion in the GFRS, China use the control and benefit criterion developed by the Board of International Public Sector Accounting Standards (IPSASB); and for developing the classification of entities and their presentation in the GFRS, China use the Government Finance Statistics (GFS) system outlined in the manual published by the International Monetary Fund (IMF). Because of the large and diverse character of SOEs – with so many subsidiaries and different activities, our framework proposes a two-step procedure for classifying SOEs. The first step is to apply to each SOE the principle that government financial reporting must include all 10 This section is based on Wong and Zhao (2015). 12 entities that have material impact on the government’s fiscal position . That is, the enterprise must be included in the public sector if its activities have, or could have, any effect on government revenues, expenditures, cash flow, and/or net assets and liabilities. To be outside the public sector, an SOE must demonstrate that it: 1. Receives no transfers from the government and has no expectation of doing so, 2. Receives no income or benefits from the use of government assets, and 3. Has incurred no debt for which the government has any direct, indirect, or contingent liability. A strict application of these three conditions will let very few – if any – SOEs through the net. As long as they are state-owned, their assets belong to the government, and by definition they are benefitting from the use of these assets (including land) in the firms’ productive activities. For the SOEs that are in the public sector, the next step in sorting through how to report them in the GFRS is to examine their economic character. If the SOE is a market producer – charging economically significant prices for its products or services, it should be counted as a commercial SOE irrespective of its profit or loss status. (The SOE is a market producer even when the prices are set by government, as long as it is selling the goods and services on a voluntary basis, free from government compulsion.) If it is not a market producer, it should be counted as a government business corporation. Given the complex and diverse nature of activities described in the last section, these two classification steps should be performed not only on SOEs but also on their subsidiaries . In Figure 3 we outline a modified decision tree framework for Step Two of the sorting exercise. Figure 3 A Proposed Decision Tree for the Classification of SOEs in China SOE Step B: Does it have a Government Step A: Is it a no no subsidiary unit that is a Business market producer? market producer? Corporation yes yes Classify the subsidiary as a Commercial SOE quasi-commercial SOE The Beijing Municipal Drainage Group (BDG, 北京城市排水集团) provides an illustrative example. The BDG is an SOE that performs public services in the disposal and treatment of waste water at government-set prices but also owns a for-profit business that supplies fresh water to the city of Yangzhou in Jiangsu province. At Step A, BDG is not a market producer – the disposal and treatment of waste water is undertaken at government-set prices for which the firm receives a fiscal subsidy. At Step B, the subsidiary that owns a for-profit business supplying fresh water to 13 Yangzhou is a market producer, and would be classified as a quasi-commercial SOE. This part of BDG goes into the Commercial SOE box for counting while the main part of BDG goes into the box for Government Business Corporations. Government business corporations (GBCs) are functionally very similar to PSUs, and can be treated similarly in budgeting and in the GRFS. They have a much closer relationship to government budgets, and they would be incorporated into the government’s statements of consolidated operations and cash flows (Table 4). In contrast, commercial SOEs should have no recurrent interaction with government finances. The ad hoc subsidies or capital injections they might receive would be reported as part of government expenditures, but reporting on these SOEs can be confined to the balance sheet of assets and liabilities. As holders (and users) of state assets, the activities of all SOEs – commercial and GBCs alike, have an impact on the government’s net assets and liabilities, which affect the government’s future fiscal position: an increase in assets has the potential to produce future incomes (revenues) for the budget, while an increase in liabilities may increase future costs (in debt servicing expenditures). It is therefore important for all SOEs to file detailed annual financial statements with the government to be included in the GFRS. The different reporting requirements for the two types of SOEs are outlined in Tables 4 and 5. Table 4 Consolidated Statement of Government Operations Public Sector General Government Government Public Corporations Intersectoral Government Organizations business Total (Commercial SOEs) elimination corporations Revenues Revenues Revenues Revenues tax revenues .. .. nontax revenues business incomes .. .. business receipts .. .. .. Expenditures Expenditures Expenditures Expenditures .. .. .. .. contractual payment for services transfers 14 Table 5 Classification of SOEs and Their Reporting in GFRS SOEs Government Business Corporations (GBCs) Commercial SOEs  Full disclosure  Full disclosure in a separate schedule in  Fully consolidated in government financial government financial report statements  Assets and liabilities are not included in  Included in government’s statements of government balance sheet operations, cash flow and balance sheets.  Net assets are recognized as government investment in government balance sheets. 4.2 Step Two: The Economic Framework for Inclusion in Government Provision11 To review the economic principles for this next step in public financial management reform, we start with two basic questions: (i) should the Government be involved in the delivery of a particular service? If the answer is yes, the second question is (ii) how should Government be involved – on the supply side through direct provision by a Government department, a Government corporation (public service unit), or through a provider operating on market principles (either a state-owned enterprise or a private enterprise)? Or should the Government intervene on the demand side by enabling the service’s delivery by subsidizing its consumption – through subsidies or cash transfers to the consumer? This framework draws from both economic theory and international experiences. Theory helps to provide some broad principles for distinguishing between public goods and private goods, or identifying market failures; but in the real world, there are few “pure� public goods and total market failures, but many grey areas. Moreover, Governments face finite budgets, and may have to choose among the many areas in which public intervention is warranted. As a result, the coverage of Government services varies across countries, and international experiences are helpful for revealing the common choices in prioritization, and some discernible “best practices�. 11 This section draws from Zhang, Wong and Mountfield (2005). 15 Figure 1: Policy Decision Tree for Government Interventions IVA1b (i). IIIA1a. Commercial Government State-owned Departments enterprises IIA1. Provide IIIA1b. Government IVA2a(ii). directly in For-profit public Corporations private sector enterprises IA. IIIA2a. Intervene IIA2. Fund Other IVA2a (iii). Providers NPOs IIIA2b. IIA3. Consumers IVA2b(i) Regulate Income IB. D n’t Support Intervene IVA2b(ii) Vouchers Government Intervention in Service Delivery: Scope and Priority The question of what the Government should do in service delivery may have two aspects. The first is about the scope of Government intervention: where to draw the boundary between Government and the market - where does Government stop and allow the market to take over? The second is about priority: given the finite amount of fiscal resources, how should Government prioritize among competing demands for services - what should it do first? For a given service for which there is a demand in the economy, the Government’s decision on the first question of whether it should intervene is represented as the choice in Level I in Figure 2. According to economic theory, Government should intervene (IA) only in cases where such intervention improves efficiency or equity, or both. Otherwise the provision of goods and services should be left to the market. This is fairly straightforward in mature market economies, where the private sector is well-developed. In China, the starting point in 1978 was from the opposite situation, and the process of finding the appropriate balance requires the withdrawal of Government from many sectors. Nevertheless, the inquiry is essentially the same, and economic theory advocates Government intervention only in sectors that are not well served by the private sector. There are four common causes of such “market failures�: public goods, externalities, natural monopolies, and asymmetric information. In these cases, the market fails to produce the optimal outcome, and Government intervention is efficiency enhancing. In addition to the cases of market failure, Governments also intervene sometimes for equity reasons, to improve outcomes for the 16 less fortunate citizens by ensuring they have access to vital public services, such as education and basic health care. For goods and services that involve neither market failures nor significant impact on equity outcomes, there is generally no justification for Government intervention (IB). While the existence of a market failure or inequality creates a rationale for Government intervention, the benefits and costs of the intervention must be assessed. This requires a hard look at the realities of the political and economic situation, and especially at the Government’s resources and institutional capacity. This is one reason why the boundary between Government and the market varies significantly across countries. The challenge is to ensure the most equitable, efficient and adequate service delivery by appropriately handling the trade-off between market failure and the possibility of Government failure, when a poorly designed intervention fails to bring an improvement or even worse. The next consideration is how to prioritize among competing demands for the Government’s finite resources. The process of setting priorities, as well as setting the ceiling on how much Government will do, is best handled through a comprehensive budgeting process for debating and reconciling the allocation of fiscal resources among the key policy objectives. Table 1 presents a matrix of functions usually performed by Governments around the world. The functions are organized in the table to illustrate both scope and priority issues. The functions are divided into two types: those addressing market failures, and those improving equity. The rows of the matrix divide these functions into three categories: minimum functions, intermediate functions, and activist functions. These are presented in a descending order of priority - for example, the minimum functions are necessary in providing for a functioning state, and may be used as a rough guide for matching these functions to the Government’s revenue and management capacity. Moving down the table from the first to the second row extends Government into functions that require more resources and greater management capacity, as the solutions required for addressing the problems become more complex, difficult, and costly. The third row presents functions that are normally beyond what most Governments undertake, to a much more active role in economic management. Even though market failures and inequality occur for the same reasons in all countries, there is great variation in the role of Government across countries. This is due not only to variations in administrative and fiscal capacity, but also the differing views of the trade-off between market failure and Government failure in each country. Table 1. What Do Governments Do? Addressing Market Failures Improving Equity Minimal Functions Providing public goods: such as Protecting the poor: Such as Defense Law and order Property rights protection Antipoverty programs Contract enforcement disaster relief Public health Macroeconomic management Intermediate Addressing Regulating Overcoming imperfect Providing social Functions externalities: Such monopolies: information: Such as insurance: Such as 17 as Such as Utility Insurance (health, life, Redistributive pensions Basic education regulation pensions) Family allowances Anti-trust Financial regulation Environmental Unemployment policy Consumer protection protection insurance Activist functions Redistribution: Coordinating private activities: Such as Such as Industrial policies Asset redistribution Source: adapted from World Bank (1997). Means of Government Intervention in Service Delivery When Government intervention in the delivery of a particular service is justified, it does not mean the Government must provide it directly. Other means of intervention are available, and are presented as the second level of decision-making in Figure 2 as three basic options: provision (IIA1), financing (IIA2), and regulation (IIA3). The options are not exclusive. Government may wish to employ any one, or a combination of two or all of them. 1) Regulation This is often the least costly option where the Government leaves provision to private entities, but sets and enforces some rules to govern the market transaction. The choice to regulate is based on the premise that the private sector can efficiently and effectively provide the service, provided that Government sets some rules of conduct to prevent abuse and safeguard the interests of consumers. In many cases, no further Government action is needed. There are of course cases where regulation alone does not suffice to ensure satisfactory delivery of the service in question and one or all of other two means of intervention need to be employed in addition. For example, the Government may choose not only to regulate private schools, but also to pay for their services through education vouchers. At the same time, the Government may also decide to provide education services directly to some citizens through publicly owned schools, which will be regulated as well. Government regulation in service delivery may take a wide range of forms. First, entry can be regulated to cope with potential negative consequences caused by unqualified providers. Education and medical care are the most obvious examples. In areas such as some agriculture extension services, regulation on qualification of technicians is sometimes required as well to ensure adequate quality of service delivery. Second, prices can be regulated to prevent efficiency losses stemming from natural monopolies and protect consumers disadvantaged by asymmetric information. Third, Government regulation may aim to safeguard the quality of services. Many services are characterized by asymmetric information, in which consumers do not have sufficient information to assess the quality of the good or service – e.g. whether the meat sold in the market is safe to eat. 2) Provision 18 When regulation alone does not suffice to ensure services are delivered to citizens in a satisfactory manner, the Government may need to use other means. One common choice is “make or buy�: provide the service directly through public entities (IIA1), or have it provided through contracting out to non-Government entities (IIA2). These choices are primarily driven by considerations of how much Government needs to control the production processes to control the output or safeguard the information required for their production. The costs of this control have to be weighed against the benefits of efficiencies stemming from market competition, the flexibility of responding to changes in demand, the reduced administrative burden on Government, etc. Even within the public sector, there are options of ownership types (IIIA1a and IIIA1b), and a key difference among the various choices is likewise the relative balance between autonomy and accountability. Services provided directly by core Government departments represent one extreme of the choice of accountability over autonomy. At the other extreme is state owned enterprises, which have full legal autonomy to operate in the market as commercial entities, and accountability is achieved through market discipline and the exercise of shareholders’ rights by Government. In-between are Government-owned public service agencies, which vary greatly across countries in terms of their governance structure.12 In fact, many services do not require exclusive public or private ownership. Over the past half century, there has been growing evidence that many public services can be effectively and efficiently provided by both public and private providers, reducing the need for Governments to build and run growing public sectors. Public and private schools and hospitals coexist in many countries, including China. Indeed, the worldwide trend has been toward more private provision as Governments try to tap advantages of competition as well as mobilization of private resources – shifting from IIA1 to IIA2. 3) Funding – who should pay, and how? To fund the delivery of a particular service means to pay for it. As in the case of provision, that Government intervention is justified in the delivery of a particular service does not necessarily mean the Government should fund it. There is no free lunch: service delivery has costs and the costs must be recovered for the service to be sustainable. From the point of view of the providers, be they private or public, the costs of the service must be recovered, usually through charging the users. When the Government decides not to fund, consumers have to purchase the service at market price in the same way as other goods and services. When the Government decides to fund a service with tax revenues, the costs are no longer borne by the consumers but shared among taxpayers. 12 In some countries there is a sub-class of these entities called delegated public service agencies. These agencies remain legally indistinguishable from Government, but are given some managerial autonomy or independence and/or have a quasi-contractual relationship with their reporting ministry. One example is the UK Passports Agency, which for legal and security reasons remains part of the UK Home Office but for managerial efficiency reasons has been given operational flexibility through establishment as an “Executive Agency�. In the case of devolved public service agencies, the entity is legally separate from Government, although it remains part of the public sector and is often in a contractual relationship with the reporting ministry or department. Other examples are the British Broadcasting Corporation (BBC) and the National Health Service Trusts (NHS), which are established as “Non - Departmental Public Bodies�, are legally distinct from their reporting departments, but have a contractual relationship involving performance reporting and, in the case of NHS trusts, budgetary transfers. 19 The question is whether it is fair for all taxpayers to share the costs, and the answer depends on whether the benefits of the service are broadly shared, or enjoyed by only a few, and whether there is merit to Government intervening to ensure access to all. As in the case of provision, there are alternative ways of funding that are shown as the fourth level of decision-making in Figure 2. The Government can choose to fund a particular service by subsidizing either the supply side or the demand side (IIIA2a or IIIA2b). The most common approach in education and health is to subsidize or fund the supply side to induce providers to lower their charges to users. For this, the Government can subsidize Government or non- Government providers (IIIA1b and IIIA2a), or simply run Government owned public service agencies to provide free or subsidized services to users (IIIA1a). This supply side approach may be administratively straightforward, but could produce some undesirable outcomes. These include over-expansion of the service – if the subsidy is paid depending on the number of users; subsidizing the wrong people – if the rich are using disproportionately more of the services, or producing “gold-plated� services that cost more to produce than society would really be willing to pay (Bird 1999). To avoid such outcomes, an alternative is to subsidize the intended end users, to allow them to purchase the services at market prices (III2b). Such demand side funding has the added advantage of encouraging competition among providers to improve efficiency. Tax credits, transfer payments and other income support (IVA2b(i)), and school vouchers (IVA2b(ii)) are ways to provide subsidies to the demand side, while free or subsidized medical services provided by public hospitals and clinics are examples of supply side funding in many countries. The predominant trend in healthcare financing has been to move away from supply-side to demand- side subsidies in mature economies, even in national health systems such as in France and the United Kingdom, where many hospitals are publicly owned. The rationale for the switch is that it enables public funds to be targeted more carefully at the intended beneficiaries of public policy. Possible Combinations of Different Means Discussion in this section has made clear that deciding to intervene in the delivery of a particular service is indeed the start of a complex policy making process. A range of policy options are available from which the Government can choose to serve its policy goals most efficaciously. Making a distinction between provision and funding is one of the most critical choices. Too often it is automatically assumed that the need for Government action justifies Government involvement in both provision and financing, and the lack of distinction between provision and financing often leads to the over-use of Government ownership. The most notable example is the use of public ownership to address an affordability problem, assuming that public providers can provide the same services at lower prices. The illusion often comes from the fact that public providers enjoy hidden subsidies in the form of free capital or the rent-free use of public buildings. When these subsidies are made explicit and included in cost accounting, these public providers could face lower costs only if they are more efficient, which is not guaranteed by public ownership itself. V. The Chongqing-Dadukou Experiment in Restructuring LGFVs Throughout this paper, we have emphasized the logical progression that follows the two-step approach of adopting the new GFRS and applying the economic framework to refocus public 20 expenditures, to strengthen financial sustainability and improve fiscal outcomes. Starting from the current situation in China where many subnational governments are highly indebted and have large baskets of SOEs, embarking on this path means facing the daunting task of sorting through and separating the tangled financial links between Government operations and many of their SOEs. The most intertwined are those with the local Government financing vehicles (LGFVs), which have borrowed with the implicit guarantee of local Governments and funded investments in local infrastructure. In this section, we discuss the case of a local government that, under the support of a World Bank development policy loan, undertook an experiment to redraw the boundary of the state and LGFVs by applying the two-step approach. While the experiment, which began in 2016, is far from completed, some of the practices adopted by Dadukou may hold useful lessons for the other local governments in China facing similar challenges. Dadukou is one of nine central districts in Chongqing Municipality, with 330,000 residents and GDP of RMB 16 billion in 2015. The district has undergone an abrupt economic transition. Its economy previously revolved around a large steel mill. However, in 2011, a tightening of environmental standards for the central part of Chongqing forced the relocation of the steel mill to another part of the Municipality. While this relocation is associated with significant environmental benefits for central Chongqing, it caused a sharp contraction in Dadukou’s GDP, by close to 30 percent. To offset this loss, the District Government took decisive steps to rebuild Dadukou’s economic base, and drove a strong recovery in 2013-2015. The land vacated by the steel mill was converted into the Jianqiao Industrial Park, and the District Government undertook an ambitious program of public investment and land development focused on promoting development in the IT sector, tourism, environmental protection technology, and residential housing. To finance these investments, the District Government relied heavily on its two urban development and investment corporations (UDICs) - Dasheng and Jianqiao. In addition to borrowing on behalf of the district government, Dacheng and Jianqiao carried out other public- sector operations. Jianqiao’s mandate is to carry out investment in infrastructure and land services in the industrial park, while Dacheng is responsible for investments mostly in urban roads and social housing as well as the management of some public assets including government office buildings in the district. This combination of the relocation of the steel mill and subsequent investments to revive the economy has left the District Government in Dadukou with a heavy debt burden. As of 2015, the government’s direct debt stood at 10.5 billion RMB, equivalent to 66 percent of local GDP, or 170 percent of local revenues. In addition, Dasheng and Jianqiao, the two UDICs, have significant commercial liabilities, many of which are tied to land development. Altogether, total liabilities in Dadukou, including direct government debt and the commercial liabilities of the UDICs, stood at RMB 25.4 billion in 2015, equal to 159 percent of the local GDP. In this set-up, Dadukou is typical of local governments throughout China. As China urbanized over the past three decades, local governments were hard-pressed to meet the burgeoning demand for infrastructure and urban facilities. Prohibited from borrowing on their own accounts, they turned increasingly from the early 1990s onwards to the use of a variety of local government financing vehicles (LGFVs) to provide funding for their capital expenditure needs (Wong 2013, World Bank 2014). Until 2015, LGFVs borrowed freely with the implicit guarantee of local 21 governments. Even today they operate as off-budget arms of local governments in managing the construction of physical infrastructure, hold and operate infrastructural assets and deliver public services. The case of Dadukou is illustrative of the immensely complicated task of untangling and unwinding local government debts from the commercial debts of SOEs. Because they are government-owned, LGFVs are by definition SOEs, and they are the most difficult to classify. An estimated 50 percent of LGFVs are UDICs, which have an explicit mandate to provide urban infrastructure and facilities. The exact count, though, is impossible because in many respects they are indistinguishable from other SOEs. As they provide services at economically significant prices, and often engage in profit-making land and real estate businesses, they are market producers under GFS and IPSAS rules. However, up to now, their financial interactions with local government have been ad hoc, informal and unregulated. They often provide uncompensated services to the public sector, are in turn given free use of public assets and often not required to remit profits to their owners – local governments. In clarifying the role of government, the Dadukou district Government has taken a phased approach by first focusing on the transformation of its two UDICs, with the main objective of restoring public sector sustainability. As early as 2014, Dadukou began the compilation of comprehensive government financial report to carry out a self-assessment of its overall financial health. The draft GFRS covered 122 government institutions (government administrative entities and public service units), three public-benefit SOEs13 and two commercial SOEs. The coverage of reporting entities and the consolidation approach adopted followed the 2012 Tentative Guidance of Accrual-Based Government Comprehensive Financial Report, and is largely consistent with the accounting framework recommended by Wong and Zhao (2015), whereby the assets, liabilities, incomes and expenses of the three public-benefit SOEs are consolidated line by line in the GFRS, and for the commercial SOEs only their net worth is consolidated. In 2015, the District Government made further efforts to improve its balance sheet. It froze the operational expenditures of the Government budget and used the savings to pay off some of the high-interest debt of the UDICs. It also used the proceeds of subnational government bonds to pay off the government debt held by Dasheng and Jianqiao. Finally, it spun off Gongli, one of the three public-benefit SOEs, from the government budget and transformed it into a commercial SOE. In 2016, with the support of the World Bank through a Development Policy Financing operation, the District embarked on a comprehensive reform program. The immediate measures included: 1) Developing a comprehensive public-sector debt sustainability analysis (DSA) which covers both Government debt and UDICs’ liabilities. 2) Developing a medium-term fiscal and public investment strategy anchored to the DSA; 3) Compiling a three-year rolling capital financing plan which includes a cap on the total public investment that is set to meet public sector debt sustainability requirements, with a comprehensive project-by-project itemization; 4) Disclosing to the public the aggregated assets and liabilities of the public sector including public-benefit SOEs. 13 The three public-benefit SOEs are Dasheng, Jianqiao, and Gongli Grain and Edible Oil Trading Company. Gongli was transformed to a commercial SOEs in 2014. 22 5) Separating LGFVs from government by dividing the assets and liabilities of its two LGFVs, Dasheng and Jianqiao, into government and commercial components. The Dadukou State- owned Assets Supervision and Administration Office issued a Decree to provide technical guidance to the two UDICs to carry out a clear division between government and commercial activities, incomes, expenses, assets, and liabilities. Assets with no commercial interest or that generate no revenues, and liabilities emanating from public investments that have received verification by the Government Audit Office, are brought onto the government’s account. After this accounting split, activities carried out by the LGFVs related to non-public investments (i.e. those booked as commercial) will be based on explicit contracts.14 This reform program has introduced an important institutional change in how the LGFVs function in practice. Starting from 2016, the financing of public investments by LGFVs will be carried out either explicitly on a commission basis, or recorded entirely on the government’s account. This shifts the financing of public investment and the acquired public assets onto the government budget, and the LGFVs can begin working according to explicit commissioned contracts with Dadukou District. In this endeavor, the commercial sides of the LGFVs thus created will be genuinely commercially viable entities. This will be a first step toward the spin-off of the commercial LGFVs as truly independent entities and opening them to outside investors. In 2017, Chongqing Municipal Government scaled up this practice to all districts in its jurisdiction15, and established a monitoring system of all liabilities associated with public benefit projects including those of LGFVs and PPPs. This is just the first step toward restoring fiscal health in Dadukou. Much more remains to be done to improve the classification of reporting entities for inclusion in the GFRS to ensure not only that all those with significant impact on the budget are included, but that things Government should do are not thrown out in the rush to slim down the budget and reporting requirements. This will be an incremental, iterative process. VI. Conclusion After three decades of remarkable economic achievement under market reforms, the leadership has called for a reset in the boundary between the State and the Market as an important corrective to help China sustain rapid economic growth, by imposing hard budget constraints on government and insulating SOEs from local government predation. This could start with revealing the current operation and finance of the government through the new Government financial reporting. Because the new GFRS will be so different from the previous Government financial reporting, building this new system will be a large and complex project with many steps and tasks, from the shift to accrual accounting, building a balance sheet that includes all entities with claims on public resources, to extending the scope of reporting to include all sources of public funds, etc. If consolidation of financial statements of all entities is not immediately feasible, it is essential to reveal all the entities that are involved in financing and delivering public benefits activities, and their interactions with the government’s budget. 14 Details can be found in Dadukou’s Decree No. 8, DuGuoZiWei (2016), dated April 17, 2018, “The Guideline on Dadukou State-owned Enterprises Implementing Separate Accounting Practice�. 15 Source: Chongqing Government Decree No. 74 (2017), dated June 6, 2017, “Notification on Enhancing the Management of Government Financing Vehicles.� 23 At the outset, it would be important to promote an understanding of the economic principles and agreed parameters for defining the role of government, along with the accounting principles and regulations for classifying reporting entities in the GFRS. As the process of adoption proceeds incrementally, by necessity, and as is the norm in China, the publication of government financial reports will proffer an increasingly detailed picture of the operation and financial status of the whole-of-government. This provides an opportunity that should be used to invite public scrutiny of the role of government and initiate an informed discussion that uses economic principles to shed light on areas where the government should do more, where the government should do less, and/or where the government should do differently. This discussion could in turn lead to reforms including the transformation of SOEs, redefining the role of public service units, designing appropriate governance arrangements for different types of public institutions, changing government procurement policy and introducing new modes of service delivery, etc. These reforms in changing the role of government will then be reflected in the subsequent government financial reports. The experience in Dadukou demonstrates the usefulness of this iterative approach, but also highlights the challenges in applying the approach at grass-roots governments. While one should not expect to reach a clear and ideal division between the state and market overnight, with successive iterations, the exercise will lead incrementally to greater clarity and improvements, as the process of implementing the GFRS sets off a beneficent cycle for China’s economic transformation to a higher quality and sustainable growth. In this process, maintaining incentives and managing expectations will be critical. 24 References Wong, Christine (2013). “Paying for Urbanization: Challenges for China’s Municipal Finance in the 21st Century,� R. Bahl, J. Linn and D. Wetzel, eds., Financing Metropolitan Governments in Developing Countries. Cambridge, MA: Lincoln Institute for Land Policy. Wong, Christine and Min Zhao (2015). “An Analytical Framework for the Selection and Classification of Entities in Government Financial Reporting in China�, a policy note for Chinese Ministry of Finance, the World Bank World Bank (1997). World Development Report: the State in a Changing World. New York: Oxford University Press. World Bank (2014). Urban China: Toward Efficient, Inclusive, and Sustainable Urbanization. Zhang, Chunlin, Christine Wong and Edward Mountfield (2005). “The Role of the Government in Service Delivery: An Analytical Framework,� Background Note for the World Bank Public Service Unit Study. 25 Annex 1 Selected Indicators for Above-Scale SOEs, 2013 (RMB billion yuan) Number of Operating Enterprises Total Assets Total Debt Equity Revenues Profits China 18,197 34,268.9 21,215.1 13,053.8 25,824.3 1519.4 Beijing 790 2333.5 1202.1 1131.4 1068.0 70.9 Tianjin 584 1054.3 699.7 354.6 922.2 72.9 Hebei 731 1533.3 987.7 545.5 1140.8 29.0 Shanxi 752 1755.2 1247.4 507.8 1041.1 34.6 Inner Mongolia 621 1223.7 752.9 470.8 628.8 50.0 Liaoning 643 1739.8 1194.6 545.1 1283.0 20.4 Jilin 361 825.3 513.4 311.9 855.4 55.0 Heilongjiang 460 909.4 524.7 384.7 666.9 73.2 Shanghai 777 1516.7 675.5 841.2 1382.2 131.4 Jiangsu 904 1655.0 1039.7 615.3 1456.1 80.3 Zhejiang 656 888.5 499.6 388.9 885.6 56.1 Anhui 643 1212.4 791.3 421.1 1007.8 36.8 Fujian 457 582.9 356.6 226.3 410.3 19.4 Jiangxi 475 516.2 333.0 183.2 599.0 22.9 Shandong 1178 2410.0 1519.2 890.8 2247.3 121.4 Henan 796 1299.6 877.8 421.7 1161.7 35.8 Hubei 689 1501.7 902.5 599.2 1196.6 69.5 Hunan 739 817.7 525.3 292.4 714.6 35.8 Guangdong 1028 1824.7 1061.0 763.7 1696.6 108.1 Guangxi 539 588.8 385.6 203.2 513.5 16.6 Hainan 83 69.1 30.3 38.8 40.6 4.7 Chongqing 482 625.3 412.8 212.5 445.0 19.6 Sichuan 906 1740.6 1215.9 524.8 986.0 54.4 Guizhou 493 643.0 420.7 222.3 347.2 27.8 Yunnan 552 1012.3 655.1 357.2 534.9 32.3 Tibet 26 40.4 11.9 28.5 5.3 -0.2 Sha'anxi 691 1671.6 955.1 716.5 1055.5 137.1 Gansu 381 788.9 519.5 269.5 677.8 21.4 Qinghai 98 305.7 210.5 95.2 107.6 10.1 Ningxia 106 281.0 186.4 94.6 167.1 9.6 Xinjiang 556 902.4 507.3 395.1 579.6 62.6 Source: 2014 Chinese Statistical Yearbook, Table 13-4. Note: “Above scale� enterprises have, since 2011, been defined as enterprises with annual revenue of more than RMB 20 million in their main activity. 26 Annex 2 A Sample of SOEs from Beijing Municipality Business scope Financial situation Special relationship with government Primary function is the planning and Year end 2014: Receipts from government: development of subways and rail, • Debt RMB 253.6 billion 1. RMB 10 billion p.a. Rail Transportation including investment and financing. • Equity 132.2 billion Special Fund to finance construction of a new Among the 41 subsidiary (second-tier) • Total assets 313.5 billion 1. Beijing line. This was increased to 15.5 billion in 2013 companies, 30 are involved in subway and of which: government portion Infrastructure with further increases promised in future. rail operations. The other 11 are • Debt RMB198.1 billion (77% Investment Co., 2. Fiscal subsidies to pay interest costs on businesses engaged in resource direct, 23% contingent) Ltd. government debt (unreported in accounts to development, equity investment including • Equity 155.4 billion avoid having to pay business tax on the receipt). in underground cables and rail-related Profits totalled 1.2 billion, of which 3. Annual ad hoc receipts to fund renovation communications, land and real estate 8-900 million came from resource projects: RMB1.5 billion in 2014 development. development and equity investments In 2013 it signed a franchise contract to Management of the city’s drainage, water undertake a three-year project with a total supply, sewage treatment sludge disposal, Debt about RMB 500 million, of investment RMB 40 billion to upgrade the 2. Beijing operation, maintenance and upgrading and which government debt is 100-200 central city sewage treatment system. The Drainage Group new construction of the network of water million. company is organizing equity financing, Co., Ltd. pipelines and related services (water Total assets of 45.5 billion including inviting external investors, and the government supply, equipment, engineering). Has 28 20 billion under construction. injected some land. Under the contract, subsidiaries, including foreign companies. government will guarantee full cost recovery, with a guaranteed profit. 27 • Fiscal subsidies RMB 457 million to pay for the gap between administrative price of 40 yuan /canister and market prices of 110-120 yuan, Storage, transport, filling of canisters, 3. Beijing plus transport costs. distribution, pipeline maintenance; liquefied • Assets RMB 1 billion • Since 2013 a new program to encourage rural capacity for 1.1 million users, but actual petroleum gas • Debt 800 million residents to use gas has brought 500,000 only 730,000; storage for 16,000 tons of company additional households. Municipal provides LPG in reserve. subsidies of 25 yuan, and the district 20 yuan (Tong County provides 45 yuan) on a reimbursement basis This is a local government financial platform company. it has four areas of operations: • International Trust, property rights • The SME Re-Insurance Company receives exchange, financing and leasing; Debt 50 billion, with interest-bearing subsidies (unspecified) 4. The Beijing • science and technology industry, capital liabilities of 30 billion (of which, 6.5 • Real estate companies operating in other State-owned information, industrial companies billion are in medium-term notes, 1.6 provinces receive subsidies from local Assets (Platform Company); billion in corporate bonds, trusts, companies there Management Co., • urban social services including real bank loans, asset-backed bonds). All • Cultural and sport events receive subsidies Ltd. estate, waste incineration power generation are commercial debt from the Cultural Office and the Cultural BOT, elder care, medical services group; (nongovernment). Resources Committee • culture and sports - Water Cube, Bird's Nest, and sports performances. Has 23 tier two companies and 80 third and fourth tier ones; total 114. Source: Beijing Municipal Finance Bureau. 28