niiina Urban Land Management: Options for an Emerging Market Economy (In Two Volumes) Volume l: The Main Report October 1, 1992 Environment, Human Resources and Urban Development Operations Division China and Mongolia Department East Asia and Pacific Regional Office FOR OFFICIAL USE ONLY MICROFICHE COPY Report No.:10692-CHA Type: (SEC) Title: URBAN MANAGEMENT : OPTIONS Author: HAMER, ANDREW Ext.:84083 Room:A7 111 Dept.:EA2EH GREEN AND YELLOW COVERS ARE 2 V. Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contenWnay not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS Currency - Renminbi; Currency Unit - Yuan 100 fen $1.00 = Y 5.46 Y 1.00 - $0.19 FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric System ACRONYMS AND ABBREVIATIONS CABR - China Academy of Building Research CASS - Chinese Academy of Social Sciences CSCEC - China State Construction Engineering Company CSHREDC - China State Housing and Real Estate Development Company CAUPD - China Academy of Urban Planning and Design CBD - Central Business District CBTDC - China Building Technology Development Center DPA - Development Permission Area DSL - Da-She Lan District, Beijing FAR - Floor Area Ratio GDP - Gross Domestic Product GIS - Geographic Information System KLDC - Korea Land Development Corporation LAB - Land Administration Bureau LDC - Land Development Corporation, Hong Kong LIS - Land Information System MoA - Ministry of Agriculture MoC - Ministry of Construction MOF - Ministry of Finance NBSM - Flational Bureau of Surveying and Mapping NEPA - National Environmental Protection Agency O&M - Operations and Maintenance PC - Portable Computer PCBC - People's Construction Bank of China REAB - Real Estate Administration Bureau REDC - Real Estate Development Company SHCLAB - Shanghai County Land Administration Bureau SHMLAB - Shanghai Municipal Land Administration Bureau SIS - Strategic Information Systems SLA - State Land Administration SMG - Singapore Municipal Government SPC - State Planning Commission SREC - Shanghai Real Estate Company SUCF - Shanghai Urban Construction Fund THG - Tianjin Municipal Government TREDC - Tianjin Real Estate Development Company TVE - Township and Village Enterprise TVIE - Township and Village Industrial Enterprise URA - Urban Redevelopment Authority, Singapore ZGC - Zhong-Guan-Chun District, Beijing FOR OFFICIAL USE ONLY CHINA URBAN LAND MANAGEMENT: OPTIONS FOR AN EMERGING MARKET ECONOMY Acknowledsements This report was 'ritten on the basis of a preparation mission, undertaken during July/August 1991, and a main mission, which took place in January/February 1992. The report was discussed with local and national authorities in August/September 1992 and revised in response to comments made at that time. Mission members, some of whom participated in only one mission, includeds Andrew Hamer (EA2EH), task manager; Alain Bertaud (INURD); Paul Stott (RMC) Nena Manley (LEGEA); David Dowall, Peter Fong, Daniel Lam, and Youxuan Zhu (Consultants); as well as local consultants Wang Yukun, Ren Zhiyuan, and Sun Chongwu. The report was prepared by Hamer and produced by Meredith Dearborn (EA2DR), who also provided valuable editing assistance. The team's work would have been impossible to accomplish without the help of over 100 Chinese colleagues, working within central agencies and in the cities of Shanghai, Guangzhou, Chengdu, Hangzhou, Fuzhou, and Shenzhen. The team extends its gratitude to Vice-Minister Zhou Ganshi, Ministry of Con- struction; Director Wang Xiangj in, State Land Administratior; and Director Zhu Fulin, Ministry of Finance; for facilitating the mission's access to informa- tion and advice. At the central level, grateful thanks go to the State Land Administration team organized by Deputy Director Zou Yuchuan and led by Sun Yinghui, Qi Mingshen, and Li Ruanjun. Within the Administration's working- level group, special thanks are due to He Zhan, Zhao Long and Li Ling. At the Ministry of Finance, Xu Pangming and Chen Zhigang provided valuable logistical support, information, and advice. The Ministry of Construction's Deputy Directors of the City Planning Department and the Real Estate Department (Zou Shimeng and Gu Guangeheig, respectively) helped to clarify many issutes and provided early reaction to preliminary mission findings. At the local level, special thanks go to the late Vice Mayor Ni Tianzeng (Shanghai), and to Vice Mayors Shi Anhai ((uangzhou), Su Luanyi (Chengdu), Li Zhixiong (Hangzhou), and Lin Zhongxion (Fuzhou). In each of those citios, along with Shenzhen, the directors or bureau heads in charge of the local Construction Commission, State Land Administration Bureau, and Finance Bureau were very supportive. Important information was also provided by the City Planning Bureaus, Real Estate Management Bureaus, Environmental Protection Bureaus, and the Housing Reform Groups in the various cities. In Hong Kong, Singapore, and Seoul, local government authorities dealing with urban land use management patiently accommodated the demands of team members. Consultant George Tolley provided a "desk report" that proved extremely useful. In addition, special acknowledgement should be made of the role played by Bank staff, including Shahid Yusuf (EA2DR), Paul Cadario (EA2CO), Anthony Ody (EA2EH), Bertrand Renaud (EMTIN), Gregory Ingram (RAD), Stephen Mayo (INURD), Michael Rutkowsky (EC3HR), Natalie Lichtenstein (LEGRA), This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. - ii - Songsu Choi (ASTIN), Williar )illinger (INURD), Catherine Farvarque (INURD), Johannes Linn (FPRVP) and Fitz Ford (INURD). The Urban Management Program, locally administered by Mr. Ford, provided critical financial support to the project; his United Nations' Nairobi-based colleague, Patrick McAuslen, also assist d the mission members with very useful advice. As always, Professor Lin Zhiqur., at the Ministry of Construction, deserves special acknowledgement for providing the team with unstinting advice and data. The team also benefited greatly by previous work done on the sub- ject, particu_arly by the Chinese Academy of Social Sciences and Chreod Con- sultants, working in Shanghai. Several individuals played "supporting roles" that nevertheless deserve acknowledgement. Wang Yuan (RMC) proved an invaluable source of advice, while Wu Zhiyong and Sok Lee deserve special thanks for translating otherwise unpenetrable written materials in Chinese and Korean. Anthony Yeh, at the University of Hong Kong, ant Edwin Lehman, head of the Chreod consult- ing firm, proved generous with their advice. Although already acknowledged as part of the sector study team, researcher Youxuan Zhu's contributions in the field and at headquarters deserve reemphasizing. He personally prepared sev- eral annexes. Without his help, bookshelves full of Chinese-language journals and unpublished information obtained from continuous personal contact with local officials would have been unavailable to the sector team. - iii. CHINA URBAN LAND MANAGEMENT: OPTIONS FOR AN EME.LGING MARKET ECONOMY VOLUME I: THE MAIN REPORT Table of Contents Page No. EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . vii I. THE RATIONALE FOR URBAN LAND MANAGEMENT SYSTEMS REFORM . . . 1 A. Creating Competitive Cities: Objectives of Urban Land Management . . . . . . . . . . . . . . . . .. . . . . 1 B. The Urban Sector's Decade of Achievement . . . . . . . . 4 C. The Challenge of Restructuring the Urban Economy . . . . 9 D. The Urban and Peri-Urban Environmental Challenge . . . . 12 E. The Issue of Industrial Redevelopment . . . . . . . . . 18 Industrial Relocation in Shanghai . . . . . . . . . 19 Industrial Relocation in Hangzhou . . . . . . . . . 21 Industrial Relocation in Fuzhou . . . . . . . . . . 22 Industrial Relocation in Guangzhou . . . . . . . . . 23 F. Residential Redevelopment Issues . . . . . . . . . . . . 24 G. An Overview of Redevelopment Projects . . . . . . . . . 25 An Overview of Compensation and Relocation Policies 26 H. Looking Back; The Government's Rationale for Urban Land Management R4form During the Transition to a Regulated Market Economy . . . . . . . . . . . . * . . . . . . . . 28 II. URBAN LAND PRICE BEHAVIOR IN A MARKET ECONOMY: AN INTRODUCTION 33 A. Economic Model of Land Price Behavior . . . . . . . . . 33 B. Land Price Behavior Across Socialist and Market Economies: Evidence from Large Cities . . . . . . . . . . . . . . . 38 III. FIRST THINGS FIRST: PRICES AND PROPERTY RIGHTS . . . . . . 42 A. Do Urban Land Prices Exist in China? . . . . . . . . . . 42 B. The Abolition of Urban Land Prices and Property Rights: A Retrospective Look .i .o......................... 42 C. The Search for a Third Way: Evaluating Price and Property Reforms to Date . o s . . a . . e .. . .. . . .. . 51 D. Implementing Comprehensive Reforms in Urban Land Price and Property Rights Reforms: A Suggested Strategy . . . 66 Recommendations for Improving tLe Feasibility of Residential Redevelopment Projects . . . . . . . 76 Broadening the Base of Compensation--Paying for Land 78 - iv - Paae No. E. Land Revenues: Will Thf' Prove Significant in Financing City Expenditures? . . . . . . . . . . . . . . . . . 82 F. The Legal Underpinnings of Price and Pzoperty Rights Reforms . . . a . . . . . . . . . . . . . . . . . . . . 91 IV. REINVENTING THE CHiNESE CITY: PLANNING WITH PROPERTY MARKETS IN MIND . . .. . . . . . . .6 . . . .a. . . . . . . 93 A. Introduction . . . . . . . . . . . . . . . . . . . . . . 93 B. The Orderly City: Local Urban Development Planning in China . . . . . ....... .... ....... ................ 93 C. Chinese Planning Practices in Reeponse to Emerging Land Markets: A Reassessment . ..... ....................... 98 V. INFORMATION AS INFRASTRUCTURE: ITS ROLE IN DEVELOPING AN INTERNATIONALLY COMPETITIVE CITY . . . . ... . . . . . . . . 120 A. Introduction . . . . . . . . . . . . . . . . . . . . . . 120 B. Land Administration and Registration . . . . . . . . . . 120 Cadaster Administration . . . . . . . . . . . . . . 122 Land Use Rights . . . . . . . . ... 122 Building Rights . .... ........... 123 Information Access . . . .. .. .. ..... . 124 C. Geographic Information Systems . . . . . . . . . . . . . 124 Issues and Constraints in Developing GIS . . . . . . 126 D. Strategic Information Systems . . . . . . . . . . . . . 127 E. Planning Without Facts: The Case of the Temporary Population of Urban China . . . . . . . . . . . . . . . 128 VI. MOVING FORWARD. . . . . . . . . . . . . . . . . . . . . . . 130 TABLES IN TEXT 1.1 1985 End-Year Housing and Nonresidential Structure Census: Vintage of Existing Stock . . . . . . . . . . . 6 1.2 Shanghai: Changing Floor Area in City Districts (1980-90) 6 1.3 Hong Kong: Current Value of Property and its Relation to GDP ....... 8 1.4 Shanghai: Change in Cultivated Land in Peri-Urban Tcwnships, 1980-88 . . . . . . . . . . . . . . . . . . . 15 1.5 Shanghai: Distribution of Reduced Cultivated Land, 1980-88 16 1.6 Seventh Five-Year Plan Industrial Relocation Costs, Shanghai . . . . . . . . . . . 20 1.7 Profile of Redevelopment Case Studies . . . . . . . . . 26 1.8 Pre- and Post-redevelopment Floor Area Ratios . . . . . 27 -v- Page No. 3.1 Change in the Size of Built-up Areas in Selected Cities: 1949-57 & a . . . ................................ ......... 45 3,2 Farmland Expropriation i2 Beijing: 1949-85 . . . . . . 46 3.3 Urban Land Requisition in Cuangzhou (1951-90) . . . . . 46 3.4 Land Use Rights Transfer in Shenzhen: 1987-91 . . . .. 52 3.5 Land Lease Sales in Shenzhen by Bidding: 1991 . . . . . 53 3.6 Land Charges Collected from Users Who Lease Land . . . . 53 3.7 Land Charge Collected from Users of Administratively Allocated Land . . . . .. . . . . . . . . . . . . . . . . 54 3.8 Shanghai Land Leases: 1988 - June 1992 . . . . . . . . 54 3.9 Land Lease Transfer in Guangzhou: 1990/91 . . . . . . . 55 3.10 Commodity Housing Sale in Selected Chinese Cities: 1989 57 3.11 1984 Guangzhou Land Ilse Fees . . . . . . . . . . . . . . 59 3.12 Star.dard Land Price ror Negotiated Land in Shenzhen: 1991 61 3.13 Off-Site Infrastructure Fee Schedule in Shenzhen: 1991 62 3.14 Combined Land Price and Off-Site Fee: Shenzhen, 1991 . 62 3.15 Cash Compensation and On-Site, Near-Suburban and Far- Suburban Commodity Housing Prices . . . . . . . . . . . 77 3.16 Purchase and Financing of On-Site, Near-Suburban and Far- Suburban Replacement Housing for Various Levels of Cash Compensation . . . . . . . . . . . . . . . . . . . . . . 78 3.17 Residual Land Values for Various Commodity Housing Prices, by FAR . ..*9too*o..*9 . . .. ..... 79 3.18 Estimates of Potential Land and Building Compensation . 80 3.19 Land Revenue in Shanghai: 1987-91 . . . . . . . . . . . 85 3.20 Land Revenue Retained by Shanghai: 1987-90 . . . . . . 85 3.21 Land Revenue in Fuzhou: 1986-91 . . . . . . . . . . . . 86 3.22 Land Revenue in Guangzhou: 1984-91 . . . . . . . . . . 87 3.23 Land Revenue in Chengdu: 1987-91 . . . . . . . . . . . 88 3.24 Land-Related Revenue in Hangzhou: )988-90 . . . . . . . 89 3.25 Revenue from Land in Hong Kong . . . . . . . . . . . 90 4.1 Typical Land Use Conversion in City Centers . . . . . . 108 4.2 Shanghai City: 23 Redevelopment Projects l . . a . . . 109 4.3 Redevelopment Projects in Fuzhou . . . . . . . . . . . . 110 4.4 Housing Redevelopment in Hangzhou . . . . . . . . . . . 110 4.5 Shanghai: Changing Industrial Floor Space (1987-90) . . 111 4.6 New Housing Projects in Shanghai (1980-90) . 112 4.7 New Housing Development in Hangzhou . . . . . . . . . . 112 4.8 New Housing Development in Fuzhou . . . . . . . . . . . 113 4.9 Shanghai: Basic Situation in Minghang and Wujing: 1980-90 116 - vi - Page No. BOXES IN TEXT 1.1 Land Use Per Capita, by Size of Urban S-ttlement . . . . 1 1.2 lnternationAl Urban Development Densities . . . . . . . 2 1.3 The Redevelopment of Baltimora's Inner Harbor . . . . . 3 1.4 Singapore: Tanjong Paga;- !.istoric Preservation Project 3 1.5 Amsterdam: Rediscovering the Value of Historic Preservation . . . . . . . . . . . . . . . . . . . . . . 4 1.6 Selected Time Series for Temporary Residents in Eight Chinese Cities . . . . . . . . . . . . . . . . . . . . 8 1.7 The Role of Suburban County Autonomy in Hindering Metropolit:-- Guided Land Development Strategies . . . . 15 1.8 Peri-Urbun Development Regulation in Hong Kong: A New Strategy . . . . . . . '. . . . . . . . . . . . . . . ............................ 17 1.9 The Fuzhou Redevelopment Model of Residential Neighbor- hoods ..................................................... 27 2.1 The Relative Advantages of CBD and Suburban Locations . 38 2.2 Stockholm's Housing Rental Price Gradient . . . . . 40 2.3 Standard Land Prices in Chinese Cities . . . . . . . . . 41 3.1 Guangzhou: A Case Study in Rising Land Requisition Costs 57 3.2 The French Experience with Land Fees and Exactions . . . 60 3.3 What Do Leases Entail? . . . . . . . . . . . . . . . . . 67 3.4 Hong Kong Experience With Leasehold Regulation . . . . . 69 3.5 Can Industrial Redevelopment be Financed Through Differ- enntial Rent? . . . . . . . . . . . . . . . . . . . . . . ............................ 74 3.b Urban Construction Financing in Shanghai . . . . . . . . 83 4.1 Land Requisition Regulations in China . . . . . . . . . 95 4.2 Shanghai: An Alternative Approach to the Master Plan? . 100 4.3 Hangzhou: The Urban Planning Programs of the Northern Extension . . . . . . . . . . . . . . . . . . . . . . . 105 4.4 Tianjin's Third Ring Road: When Planners and Market Forces Collite.................... 106 4.5 Guangzhou: Increasing Density in New Development . . . 114 4.6 Guangzhou: Suburban County Urban Land Development Policies . . . . . . . . . . .. . . . . . . . . 118 5.L Title Issuance Procedure in Guangzhou . . . . . . . . . 123 - vii - CHINA URBAN LAND MANAGEMENT: OPTIONS FOR AN EMERGING MARKET ECONOMY Executive Summary Background: What Drives Urban Land Reform: The Chinese Rationale i. During the 1980., nconomic activity within China's cities and statu- tory towns, which accounts for more than half of the gross domestic product (GDP), grew at well over 10 percent per annum. In addition, the urban la- tion grew at a rate of 5 percent per yasr between the census of 182 an.' .iat of 1990. By 1990, 26 percent of China's pcpulation, or 301 million, lived in urban areas. In absolute terms. China now has the world's larRest urban Popu- lation. Urban employment rose at roughly similar rates. The size or the urbanized, built-up areas of cities and towns expanded at 6 percent a year in the aggregate. The scale of urban residential and nonresidential investment grew at equally dramatic rates, adding 6 percent to the existing stock each year and accounting for 10 to 15 percent of .urban GDP. ii. In spite of this vigorous sxpansion, the built-up areas of cities totaled only 12,000 square kilometers .c2) in 1990, with an additional 8,251 km2 under development in statutory towns. By contrast, total cultivated land in China totaled at least I million kma and the b-ilt-up area of rural villages and townships occupied an additional 140,000 =m2 (1985). Cities and statutory towns cover an area equivalent of only 2 percent of China's culti- vated land area. This corresponds to the experience of other Asian countries, as well as land-abundant North America and Western Europe. iii. This report's main theme is that though the land area involved is small, urban land management reform is required to accelerate the efficient expansion of a regulated market economy, where land is treated as any other commodity, and where many complementary measures involving planning, legal and regulatory institutions, taxation and information infrastructure are in place. As Chapter VI makes clear, that is now emerging as a view at the central and local level. However, during the first 6-7 years of overall urban reform, beginning in 1984, the central and local authorities often viewed reform as a set of disarticulated changes that did not have a comprehensive content or an ultimate objective of promoting the development of land market transactions. The initial rationale was to either increase administrative controls over land or generate some revenue from its use. iv. For example, because of a national commitment to maintain highly localized urban foodstuffs self-sufficiency, the perceived need to control urban spatial expansion became paramount by the mid-1980s. That urge to con- trol expressed itself at the macrospatial and project level as different tiers of government, particularly the central authorities, tried to limit the over- all expansion of localities, while eliminating "wasteful" and "unnecessary" allocation of land to individual urban investment projects. Control. ra ther tLan the introduction of market mechanisms. became a Primarv drivina force behind urban land "reform." That limited objective was quickly achieved. - viii - V. Revenue-generation concerns, as well as a desire to assert control over wasteful urban land uses, enco"raged goverrnent at the central and local level to charge domestic land users for the use of land, beginning as early as 1982. Yet this initiative was tempered by an overriding concern that intro- ducing market "prices" would prove "urxffordable" to state-owned enterprises. vi. the pressure for change derived further impetus from the reintroduc- tion of foreign investmeuts into China. Between 1979 and 1991, 15,000 fo;:eign-funded firms invested $22 billion in China. These investors, accus- tomed to well-definsd rights over property, found prevailing arrtangements unsatisfactory. For exampla, land contributed by Chinese partners as equity proved hard to value, impossible to use as collateral to obtain loans, and subject to unreasonable occupancy time limits. Local governments, especially after 1988, adopted a pragmatic attitude, realizing that, in return for sell- ing long-term use rights, they could raise convertible currency revenues while improving the local business climate. In this regard, the examples of Hong Kong and Singapore proved instructive. But, as discussed below, the actual introduction of leases proved inconsequential in scope and did little to cre- ate competitively determined market prices. A Review of Exoeriments to Date vii. Beginning largely in 1984, China's cities embarked on land use man- agement changes across a broad front and have few if any peers among cities in post-command economies. These ci'anges are proceeding at a different pan.e in different cities. Southern and ;entral coastal provincial cities as weil as the "Special Economic Zones" are well ahead of their counterparts in the inte- rior of the country and even northern coastal cities. However, in every city visited by the sector team during preparation missions, the absence of a clear vision of the end-objective--to create a regionally, nationally, and/or inter- nationally competitive city--meant that these changes lacked coherence, and the critical need to create a land market, governed by price signals and guided by appropriate planning and regulatory mechanisms, was unfolding too slowly. In retrospect, it was only during the latter half of 1992 that evi- dence emerged of an acceleration in reform efforts, with land leasing being adopted at an unprecedented scale in selected cities. Among the changes undertaken, whose adequacy is discussed below, the following are the moAt significant: * creation of profit-oriented real estate development companies (REDCs), mostly with public sertor equity, with a mandate to develop commercial estates or properti.s, and sell these at prices that include land location premiuias; * promotion of economic development zones, though often far from the city center, to attract foreign investors; * selective use of land use lease mechanisms, giving users certain ownership rights in return for paying land use premiums; * introduction of land use development fees, taxes, and other exac- tions: - ix - * promotion of spot redevelopment of inner-city areas, with occasional recourse to "value added" strategies linked to a change in lard uses * greater seneħtivity by planners and developers alike, prompted by rising implicit or explicit land prices, to the need to increase land use efficiency at the project level, promoting densification previously blocked by adherence to centrally supplied guidelines concerning "norms and standards"; * reintroduction of land and buildinfs registration cadasters, care- fully defining the dimensions, uses, and occupancy rights i.ivolved. Urban Land Management Goals and Overall -,Jtructur1nz: What Should Drive Land Reform? viii. Like the typical factory in the state-owned sector, large areas of China's cities are still obsolete, i.e., saddled with inadequate infrastruc- ture, housing and nonresidential facilities. Two thirds of their output is generated by the industrial sector, utilizing equivalent proportions of the labor force and occupying up to 30 percent of the built-up area of the central city when warehouses are included (with higher amounts in the high-accessibil- ity old-city core areas). Using indicators from comparator cities in Asia and among developed countries, all those Rronortions should be cut in at least half. And, if the core areas of the old industrial cities are to shift from industrial to commercial uses, investments in infrastructure will have to vastly increase the amount spen,; in upgrading the central core area "carrying capacity." Road space, for instance, will have to be increased by a factor of two. This will require a rationalization of land use. ix. Such a transformation would allow major Chinese cities (particularly those wishing to be internationally competitive) to shift to a focused devel- opment strateg7. The economy of the future must rely more heavilj on informa- tion services to generate competitive manufacturing products in selective categories of goods. The successful, international competitive cities of the twenty-first century will have vibrant producers' services sectors, employing up to 40 percent of the local labor force. The structural transformation to such a knowledge-based economy will require an accompanying unprecedented revolution in urban land management practices. The old city, full of facto- ries and warehouses located in inappropriate locations, cannot be modernized by administrative fiat and the moving of 30 or 40 facilities per year. Little will be achieved unless incentives to move are greatly sharpened and occupants of sites having high implicit valt s are allowed to undertake land transac- tions that yield substantial fina -ial rewards and permit new investments by the original occupants, who would retain the bulk of the earnings from any such transaction. In some cities, including Guangzhou and Shanghai, this message has been accepted in principle. Towards a Co.i1rohensive Urban Land Systems Reform Strategy x. To achieve these sector goals, a multifaceted strategy is needed to enAble urban land use rights to be obtained and exchanged on commercial terms. Government must craft a plan that (a) allows for the expansion of land grice reforms, in place of administered allocation practices; (u, develops a legal framewovk that reintroducbs the full range of property tightas with the coor- dinated registration of all property rights related to a single parcel of land, in a manner consistent with best international practices, anc. transpar- ency in both the mechanisms and rules for property dispute iesolution; (a) ir.stitutes a loal PlanninR framework that helps extend serviced land developmunt in an affordable manner, while promoting infill devolopment, upgrading, and redevelopment in core areas of cities; (d) creates land-based revenue generation mechanisms that provide a return to a very valuable state asset, while requiring identifiable beneficiaries of government investments in local infrastru:ture to pay a "fair share" of the costs of conbtruction and maintenance; and (e) provides adequate incentives and auidelines that Rive hiah grioritv to historic preservation objectives and ensure that land-related environmental degradation is first contained and then reversed. xi. The transition, implementing this multifaceted strategy, should begin with all new investments that involve land acquisition. All new indus- trial, commercial, residential and nonprofit activities should be subject to the full impact of this strategy, even if some differa.tiel policies are adopted, varying according to the sector involved. The coveraxe of the pre- vailing. traditional system must be proaressivelv diminished in size according to a time-certain schedule for the introduction of new rules. Every opportu- nity should be taken to accelerate t:ae pace of reform among existing urban land users, by moving some into the new system ahead of others, while ensuring a rapid convergence in outcomes on the basis 3f objective u.finitions of "affordability." xii. Because the required reforms are interrelated, and market forces must be -uided by vlannina and regulation. it is important to follow an inte- grated strategy. This is because what is traded on the market is not land but a bundle of rights and obligations which provide for con1trol over particular Parcels of groverty. including stratified interests in buildings. land use rights. mineral rights. and easements. each of which Yields pos.ibilities for generating income. In addition, by announcing the overall strategy simultane- ously, the impact of the reform can be brought forward in time, particularly if that time horizon is short enough to force all users to internalize quickly the implications of that reform. xiii. Given the multifaceted nature of the reform, Government should con- sider creating a Real Estate Sector Reform Group. Eventually, a new Ministry, able to deal with property markets in a comprehensive fashion, should be established, replacing existing entities with more narrowly focesed mandatas and agendas. First Things First: Prices and Pro2erty Rights xiv. Since rural land converted to urban uses will continue to supply the bulk of land utilized for new construction, one of the first issues to be faced is who will undertake land requisition in suburban areas and on what terms. Here, international experience provides a range of models, from one- to-one negotiations between users or developers and rural occupants, to a largely centralized requisition of land by public authorities. In most coun- tries, the process of rural-to-urban land conversion is bedeviled by the fact that the "announcement" effect of impending urban development, and the intro- - xi - duction of the infrastructure required for inte.asive urbanization, will drive prices to a level that is a significant multiple of its value in agricultural use, because the marginal productivity of land has increased dramatically. It is not unuo.ual for the price of rezoned, serviced land at the urban fringe to exceed the cost of farmland by a factor of 15 to 20 times. This orice differ- ential is not due to a lack of inadequate land supplies in any one category but by the fact that each is a recognizablv different -roduct. carrvinp dif- ferent locational advantages. Property rights. etc. Flooding the market with greater supplies of the "higher-value" land, disregarding the rates of return of such actions, merely to "flatten" the price gradient, would obviously nar- row the differential but is inefficient and not in line with practice followed even in "developer-friendly" market economies. xv. The rising cost of suburban and peri-urban land may encourage agri- cultural land preservation, assuming hard budget constraints prevail. Urban land buyers would face an increasingly expensive input, whose price would help contain demand for land. But the Prospect of significant revenue windfalls also creates ar incentive for rural dwellers to extend negotiations with pro- spective urban users until they can capture a significant Portion of the increased market value, even though they contribute little to the revaloriza- tion taking place. The bargaining process may create supply-side bottlenecks and inequities. This is particularly so since, at present, there is no effec- tive income tax system in place to capture part of any "windfall gains" obtained by the rural communities. And, since hard budget constraints among urban consumers of land are only just emerging, one-on-one land negotiations, without State Land Administration (SLA) mediation, could lead many urban bid- ders to "cave in" and allow rural dwellers to reap large capital gains. Dur- ing the transition to the market economy, some form of local government over- sight is needed to keep the process from generating politically dangerous charges of runaway "speculation." xvi. Under the 1982 constitution, as amended, rural communities own rural land, while the state owns all urban land. The state has the right to acquire land from rural communities and may trade use rights in the market. Local government should consider taking complete control of the suburban land requi- sition process. exRroRriatina rural land for all urban conversion Purposes. including those that promote economic development. in a manner that is trans- parent. in terms of comDensation rules. as well as equitable. Compensation amounts could be subject to appeal before regulatory bodies, but not the req- uisition itself. The actual location and use of the land would be market- driven, and SLA would only be involved in a technical function, attempting to restrain the length of requisition bargaining and moderating the monetary demands of farmers, who have an incentive to seize as much of the future appreciation of land value as possible, as conversion occurs. Occupants could continue to use the land until actual urban development became necessary. When urban construction would begin, compensation would be paid, based on the value established at the announced requisition date, adjusted only to cover inflation. Government would then lease the remaining land, through auction, negotiation, or grant, to urban users, including land development companies who could transfer leases to p3ots of serviced land to smaller private, coop- erative or state-sponsored developers. The medium-run goal should be the creation of an effective income and capital gains tax system, as well as reforms that force all enterprises to operate under hard budget constraints, - xii - limiting and finally aliminating the need to use administrative requisition mechanisms. xvii. Such a strategy would have to be matched by a reform of prices and use rights over existing urban land. Vacant serviced vlots should enter the new system immediately. with new occupants siRning leases reRardless of user category. As much as possible, sites should be leased in a publicly adver- tised and competitive manner, to allow market prices (and property valuation data) to emerge spontaneously. Regulatory mechanisms should be in place to ensui-:e the fairness with which lease bids are evaluated and awards are made. Institutional nonprofit users (hospitals, research institutes, schools and universities, librarios, government offices) should only receive land grant leases after a high-level review of any proposed use to ensure that the oppor- tunity cost of relinquishing the particular site to nonprofit use is justified by the benefits involved and that alternative, lower-value sites have been considered. In addition, all users favored by such a land grant policy should be subjected to minimum floor to land area ratio restrictions, so that the density of the development reflects the implicit, shadow price of the land area. Institutional users wishing to transfer their lease rights to profit- making users would be subject to capital gains taxes or other forms of revenue sharing with the land-owning authority. xviii. Developed land has occupants who must suffer the effects of Rrice changes that are retroactive in character. This will force them. in many instances, to recycle a repriced product rather than simplv worry about how to use 'future inputs" more efficiently. Nonresidential occupants of existing. developed sites should be given the option to lease or rent. with rent rates set hi&h enough to Rrovide a competitive rate of return on the implicit lease value of the Property. Lease and rent levels would be set with regard to market-value data in locations and corridors that are self-evidently choice sites. In those areas, the number of market-value data points could be increased by ensuring that redevelopable sites vacated through relocation would be made available to new users through a properly monitored competitive bidding process. Relocated users would retain the bulk of the capital gains involved, as an incentive to move. Price signals from regularized "black market" transactions should also be allowed to guide the process. Elsewhere, occupied sites should be leased or rented at levels no lower than the prices of comparable suburban, serviced sites, and again be subject to the market test provided by emerging implicit and explicit rents and prices. Residential users of land should be incorporated into a system of rental reform, aimed at full cost recovery, as expeditiously as possible, following the recommenda- tions of the earlier Bank report on housing reforms, and discussed further in para. xxii. xix. The issue arises: how to deal with the political opposition to such a move to a regulated market, assuming regulatory mechanisms are set in place tc dral with issues of transparency and fairness, and the possibility exists of appeals to impartial bodies in cases where corruption of the process is suspect? For users of new sites, affordability should not be a major impedi- ment to reform, since the potential occupant could move to any one of many locations or simply not bid for a new site. Industrial enterprise should include likely land values in feasibility studies. All owners of housing estates and individual units wo.ld be expected to compensate the state through - xiii - the signing of a negotiated lease. Commercial activities should not take place on new leasable space if the activity proves unprofitable when land prices are factored in. Furthermore, since there is no compelling economic reason to maintain state control over most commercial operations, these should be privatized as expeditiously as possible, making market real est:.te transac- tions possible. Institutional, uonprofit users could be taken caL3 of on an exceptional basis, as already discussed. xx. For existing occupants, annual market rents would provide an alter- native to buying a lease. Those rents could be phased in over three to five years. Existing users occupying valuable real estate within the existing built-up area and not willing to pav such rents would be given the opgortunitv to relocate to suburban locations, in return for sianing a lease for the new site. with the added incentive of receiving a significant -ercentaae of any new lease income obtained from the old site. This particular option can be defended as promoting the overall goal of introducing a land market, even if the income obtained is, in a strictly legal sense, unearned, because the original occupant did not have legal rights to dispose of the administratively allocated sites. As noted, this view has now been accepted, in principle, by national and selected local authorities. Much is made of the weakness of the state-owned enterprises in China and for the need for enterprise restructur- ing. Land reform could provide an unexpected tool in this process because, fortuitously, many urban loss-making enterprises are located on valuable sites whose sale could generate resources that would allow the management or its workforce to begin new ventures elsewhere. Because Government has allowed occupants to develop implicit "ownership" rights in the past, it must now adopt some fcrmula to deal with the issue during the transition period. And the dislocations created by a transition to a regulated market economy are too far-reaching to correct by administrative fiat alone; incentives are needed. xxi. This price reform for present occupants of developed sites would be facilitated by several factors. Industrial income tax reform now under way will reduce statutory tax rates for state-owned enterprises from 55 percent to 33 percent. Taxable income will exclude land costs, making them directly deductible or treated as part of depreciation allowances shielded from taxa- tion. With deductibility taken into account, prices or rents can be a multi- ple of that which would otherwise prevail, and thus not affect retained earn- ings significantly. Commercial activity is now increasingly disciplined by market forces and able to avoid price controls and to maximize profits. As one indication of this, evidence already exists of a lively trading of urban use rights. Desirable real estate in favorable locations, regardless of for- mal ownership, is being transferred to other users at very high explicit or implicit prices. xxii. After a decade of rapid increases in real household income, and in the wake of forthcoming housing rent reform with compensating wage adjust- ments, it is not unreasonable to ask residents of existing housina units to pax land rents. Since these rents will be phased in over several years, and since the massive housing construction effort of the past decade can be expected to continue, there should be little hardship involved because reloca- tion to lower-cost sites will be a realistic option. "Grandfather" clauses could allow long-term residents to live out their lives in such areas, under rules and conditions not applicable to others, and revocable if they agree to - xiv - "buy-out agreements" that provide generous relocation incentives. But all other residents should be subject to market rules, including relocation. Additional hardship cases could be dealt with through "safety net" programs that provide housing allowances to the certifiable needy. Government workers, whose pay scales lag behind those of comparably experienced workers in the for-profit sectors, could be included in the "safety net" program. Perhaps, however, it is time to include government wage-scale restructuring into the overall reform agenda; there is little justification, political sensitivities aside, for paying government workers wages that are clearly inequitable (and inefficient) when compared to pay scales prevailing elsewhere in the economy. xxiii. The development of the legal framework and institutions will have to be accelerated in order to clarify rights and obligations of owner and occu- Pant alike, as well as to establish an ordered and Publicl' accessible svstem for recording such rights and obligations. Fortunately, the issue of state ownership per se appears to pose no problems in terms of efficient occupant behavior. Those who obtain leases can be expected to behave in roughly the same manner as landowners in other systems, if the leases are long-term (50+ years), renewable well ahead of the lease expiration date, and provided lease- holders can easily transfer use rights while retaining attractive after-tax capital gains. Developing this framework for a leasehold system should con- tinue to be the responsibility of the central government. xxiv. In addition, the emeraina Dropertv market reauires changina a per- ception that the established conflict-resolution and court system fails to protect propertz rights. It should be presided over by judges who can quickly hear disputes, issue and publish rulings to protect the rights of borrowers and creditors in foreclosure proceedings 1-n accordance with published laws, rules, and regulations. The courts should have enforceable powers of eviction where land users operate under land rental agreements and fail to meet the obligations defined in those agreements. Any political interference with court decision-making should be eliminated. xxv. ProDerty markets also require that all land holdings be registered and, thus, that any potential buyer of a leasehold be able to establish whether or not the title to be transferred is sublect to any obligations to third parties. including those created when land is used as collateral in obtaininz loans. The precise definition of the boundaries of each and every parcel and the nature of interests in each property must be very clearly defined. xxvi. The state should retain expropriation rights within the urban built- up area parallel to those exercised in the periphery. Locai government can designate any property as subject to expropriation if the site is needed for approved public works such as a transportation project requiring additional rights-of-way. "Excess condemnation" (i.e.. additional exproDriation) of high-access nodes created by new transport investments should be encouraged. Similarly. areas designated as priority redevelopment zones should be subiect to expropriation if the occupants. through grivate arrangements. fail within a time certain to assemble sites and provide for the prescribed redevelopment investments. This measure is made necessary by the fact that, particularly in older "downtown" areas, the fr..gmentation of land use may make redevelopment financially unattractive and encourage certain occupants to "hold out" for - xv - remunerative or noneconomic reasons. The example of Singapore's laws, which facilitate redevelopment, can be cited in this regard. Compensation criteria must be carefully spelled out, avoid inclusion of any opaque in-kind rewards, with any compensation package (but not the expropriation itself) subject to appeal through some impartial judicial mechanism. Unfortunately, the ability to challenge administrative resolutions and decisions in a Chinese court of law and the ability to utilize the Chinese court system for the protection of property interests by foreign investors is perceived by investors as untested. Problems encountered include the inability to obtain applicable laws, regula- tions and rules of court procedures. Neglecting this strand of the reform could prove to be a major impediment to fulfilling the agenda defined in this report. Reinventing the Chinese Citys Planning and Regulatinx Development with Prop- erty Markets in Mind xxvii. Any municipal or metropolitan government has a limited capacity to add to trunk infrastructure within a medium-term horizon, say between now and the year 2000. Where that infrastructure is located will determine what options will be available to investors and consumers. The redevelopment of the old core areas of the cities, in particular, cannot be accelerated at a sustainable rate without a redirection of infrastructure investments from peripheral locations to the CBD. Furthermore, the supply of serviced land developed during any given time period should be emploved to its highest and best use, unless welfare, health. safety. historic preservation. and environ- ment considerations., codified in the form of "Performance indicators." dictate otherwise and. in the short to medium term. unless infrastructure caPacitV cannot sustain the scale of investment involved. In this latter case, plan- ners should examine ways to accommodate user-led trends, identified by an adequate monitoring system, by modifying trunk and connective infrastructure plans. xxviii. Planners should foreo the temptation to endorse a "blank sheet" approach. whereby redevelopment is viewed as too messy, and "orderly" areen- field development is slstematically preferred. In core areas of the city, this approach has yielded to ad-hoc approval of spot redevelopment which, cumulatively and over time, will force infrastructure investments to be refo- cused and recentered geographically, regardless of any current strategy of denial. Planners should internalize the implications of limited infrastruc- ture expansion options by restricting the number of new spatial priority zones. Trunk infrastructure investments are both costly and a function of distances involved; compact development is therefore to be promoted, all other things being equal. A metropolis cannot, in a financially sustainable fashion. simultaneously create a large number of satellite cities. industrial and residential suburbs. and major. new special economic development zones 20 to 40 kilometers from the city center. while accommodating redevelogment of the historic core area. A more compact development strategy, promoting rede- velopment, would also be more consistent with strategies that discourage unnecessary conversion of agricultural land to nonagricultural uses. xxix. The potential to redevelop the existing core area deserves reempha- sizing. Even though reuse is dictated by the need to adapt urban areas to accommodate the shift from industrial- to commercial-led development, and from - xvi - welfare- to market-oriented housing solutions, observers standing at the top of the tallest buildings in any city will notice that such redevelopment has been exceptional. In some areas, like Guangzhou, which is at the forefront of market experiments, completed redevelopment often rings the core rather than tackling the center. Similarly, Shanghai's leasehold experiments through mid- 1992 were usually located outside the old (pre-1949) city boundaries. Why? The answer lies in the past absence of forceful public sector leadership. High value-added opportunities have been forgone because the profitability inherent in radevelopment was diminished by (a) limits on use and density optionsa (b) limited commitment to new, facilitating infrastructure invest- ments; (c) associated, wrong-headed citywide dedensification strategies; and (d) a failure to tackle the excessive entitlement demands of existing occu- pants of redevelopable sites. These obstacles to financially feasible rede- velopment can be removed, following suggested options detailed in this report, but they will require political action to bring them about. xxx. There are many legitimate reasons for a local government to have plans governing property markets. Open space and its distribution, the place- ment and scale of public facilities, and the location of public infrastruc- ture, over the medium term and governed by a comprehensive framework, all require macrospatial plans. The preservation of environmentally sensitive areas and historic preservation districts is another responsibility of the public sector. In China. Planners at the macro and micro level already exer- cise enormous urban Rlanning control. They specifv macrosDatial Dlans and related infrastructure and they Provide for detailed district-level land use and building Plans. Unfortunately, master plans in China have been developed as rather abstract, architectural exercises without the use of analytical tools to test alternatives in terms of costs and consequences. And the cumu- lative effects of district-level permits for land development are usually not monitored and reflected in revised macrospatial infrastructure plans. The issue is thus not omission of Rlanninf. but the need for more informed nlan- nina. taking into accourt feedback gathered by monitoring indicators and common-sense economics. xxxi. Mayors and city councils should be presented with costed alternative analyses that document the amount of potential development made possible per unit of infrastructure investment, under various scenarios. These should reflect the shadow value of land. Planners must also internalize the concept of consumer sovereignty, unless there is a clearly articulatod public purpose for not doing so. Thus, if what is happening on the ground is very different from what is projected by the masteL plan, then the objective of monitoring and planning is not to identify departures as delinquent events but to assist governments to revise the scale and location of public investments. xxxii. Risk Assessment. Bold reforms may produce unanticipated side effects. For example, will land reform create a surge in retail prices? Will workers faced by the resulting reuse of land be forced to undertake onerous additional commuting journeys to work? The Bank's previous work on relative price adjustment reform suggests that retail price inflation can be contained if central monetary policy follows its current, conservative course and if new price signals yield additional sup-lies of inputs, moderating price inflation. That conclusion is assumed to be applicable in this case. That is, a gradual decontrol of land prices with adequate development of market institutions - xvii - should minimize the risk of price spirals. As for the commuting-related con- sequences of land use changes led by market reform, the answer lies in the results of costly simulation exercises. Consider, however, that large numbers of workers have been moved to residential locations, including Pudong, rela- tively distant from their CBD-located factories. Reform may well decrease unnecessary commuting. The best advice is to monitor trends and develop reme- dial measures as necessary. Coordinated policies at the central and at the local level would then follow. The Equitv Issue xxxiii. Equity concerns have to be suitably defined, where China's cities and towns are concerned. A "level playing field" must be provided for newcom- ers who engage in activities involving real estate development or its use. Though local, state-sponsored entities have been favored in the past, the future should prove different, if events in late 1992 are taken as a guide. Any remaining formal or de facto barriers to real estate developers from other countries, provinces, counties or from the private/cooperative sector should be reexamined and removed. If the less well-regarded can outperform their rivals, including those with local state-owned credentials, so be it. Out- comes, not ownership composition, should be what matter. xxxiv. Sooner or later, Government must also face up to the inherent ineq- uity involved in the treatment of the increasing number of temporary resi- dents. On an experimental basis, temporary residents who have lived in a city or town for more than a certain period of time should be granted the rights to purchase or develop real estate. The barriers imposed on such residents to buy or lease property are counterproductive and should be reversed, even if Government wishes to maintain strict controls over population movement in general. Sitting on Golden Stools? Land-Based Revenue Mobilization and Reform Xxxv. The Primarv objective of reform should be the establishment of explicit land price sianals to govern land use behavior choices. Land should be priced like any other productive input, at rates set by the market. Then land users will treat land more efficiently. Distributional issues, dealing with who gains from0land price reform, and creates what some might perceive as "windfall profits," are best dealt with by a revised tax system, which is not a major focus of this report. The issue of unfair individual gains. revenue generation and the question of how revenues-are split between locul and cen- tral governments should not obstruct the achievement of that nrimarv obiec- tive. xxxvi. The People's Republic, as a sovereign state, has the status under international law to act as a "legal person." Nevertheless, the state must clearly designate an institution (or lower-level government to act for it) as the titleholder and there must be clear rules involving revenue assignment. xxxvii. The revenue issue would then be reduced to who would receive the benefits from rents, leases, property taxes, and capital gains or transfer taxes. At present, this may be an academic question in most cities, since relatively little revenue has been generated by urban land. In the exurban - xviii - counties of the Pearl River Delta, however, the issue must already be faced. If urban land reform, as defined in this report, is to become a reality, incentives should be in place to encourage those price signals to grow rapidly. Furthermore, the variation in revenues per square meter (i2) within a city, and the variation in average revenues across cities, must be allowed to emerge spontaneously. This argues for local-level incentives and limited central interference. Finally, international experience with urban land reve- nues suegest that, even under different political systems and different levels of economic develoyment. local a3DroRriation is the norm. This leads to the following recommendation: all land-related revenues should be assizned to local governments. xxxviii. Having said that, the fact remains that China's cities are expanding their expenditures at a rapid rate, and using off-budget, nontax sources to finance these to an increasing degree. Whether or not land revenues prove a significant contributor to local revenue generation, the increased reliance on loans means there is an urgent need to identify clear sources of loan repay- ment, linked to the beneficiaries. User charges for public utilities will have to increase significantly during the next decade, since there is little likelihood that central or provincial authorities will provide grants as sub- stitutes. And budgets overseen by local finance bureaus must be made more comprehensive in nature, to allow such documents to be used for management purposes. xxxix. There are other issues to be tackled. In cities like Hong Kong and Seoul, as a matter of deliberate policy, government has set limits on the number of hectares leased per year. Resulting land revenues can provide a significant source of income, when combined with other land-related taxes and fees. However, where such very restrictive policies are pursued, development may be handicapped, and the availability of affordable housing may be compro- mised. Given the overwhelming fear of "unaffordable" reforms, it seems sensi- ble to set aside strategies that promote revenue generation rather than accom- modating user demand for serviced urban land. xl. The issue of real estate development company exactions, for the pur- poses of promoting infrastructure development, is more complicated to resolve. These companies, given managerial autonomy during the last half-decade, often control land resources for which no lease price was paid; occasionally, the requisition cost required to free up the land was also shouldered by the gov- ernment. Furthermore, companies have usually benefited from temporary tax exemptions; in some cases, the companies have been classified as nonprofit institutions. Under these circumstances, and given the apparent profitability of their operations, it is not unreasonable to require that they pay for more than the costs of servicing the land they occupy. The question is how. Here, the rule should be: transparent regulations fairly applied. Exactions should be replaced by land lease arrangements, openly arrived at and supervised by appropriate regulatory bodies. Information as Infrastructure xli. Proof of title is fundamental to the exercise of property rights, including the use of land as loan collateral and the ability to transfer land use rights for profit. The new registration system intended to establish I,I , - xix - proof of title is based on a meticulous process of examination and field work. The separate building registration process is also very thorough, and includes an examination of the structure to ascertain its conformity with existing building controls. This latter anomaly was brought about by the fact that, at the national level, land use rights were, and continue to be, the purview of SLA, while building rights are registered by the Ministry of Construction (MoC). The land and building cadasters are kept separately. There is no com- pelling justification for such practices. xlii. One important reform initiative, the full regularization of transac- tions between individual land users, is likely to be stymied by this fragmen- tation of responsibility. The transfer of uses will be restricted to those holding registered land use certificates, who are generally also the "owners" of any associated structures. But the regulations are conspicuously silent concerning the rights of public housing tenants to rent or sublet the rights to space they now occupy, particularly attractive ground-floor residential space located on commercial streets. This is the major source of "black mar- ket" activity. Rents or sublease values reflect land-related location fac- tors, while what is actually traded is building space. Both MoC and SLA will have to cooperate to solve the problem, by taxing a portion of the monthly or yearly payments in return for a guarantee of "regularization." xliii. Aside from the inefficiency associated with the maintenance of sepa- rate records, the key problem to date is the incomplete coverage of the regis- ters. Progress has been greatest in the newer built-up areas of cities and in the suburban counties under the municipalities' jurisdiction; elsewhere, the process is ad-hoc. Furthermore, in some cases, real estate occupied by public agencies has been ignored. xliv. The cadastral maps and associated topographical base maps represent a source of fundamental physical information. Other important data, including land uses, land values, demographic and economic information, and infrastruc- ture network location, when overlaid onto the fundamental information, can provide a comprehensive and coordinated database that can be used to monitor urban development. The advent of digital database technology now allows stor- age of information in a form suitable for convenient manipulation. Thup, planners, administrators, and those seeking title information, can make pro- jections or display results quickly. Information can be used, for example, to forecast where people will live and work in the future, the problems they will face, the services they will require and can afford, as well as the public revenues that can be expected. These data can help select and appraise investment projects, as well as manage related operations and maintenance (O&M) programs. xlv. Unfortunately, within China, the self-evident advantages of auto- mated Geographic Information Systems (GIS) have led to a set of narrow bureau- cratic responses. The economies of scale and advantages of integrated plan- ning have not emerged, because each agency or bureau has attempted to develop its own independent GIS. Furthermore, the technology is currently in the hands of information engineers who do not place consumer requirements high on their agenda. Finally, GIS cannot really flourish and maximize its potential impact while the urban planning system remains rooted in a traditional static mode that allows for little feedback. As always, urbe- land management reform requires a multifaceted response. Urban Environmental Strategies xlvi. Industrial redevelopment is of critical importance in China's largs urban areas not only in terms of potential industrial productivity gains, but also in terms of urban environmental pollution. Industrial pollution in urban areas has created dangerous conditions--houses next to polluting factories, chemical spills, noise pollution, vibrations, dust, congested alleys and lanes. Many of the polluting firms that are left behind in the urban areas are the least profitable ones and financing their relocation is difficult. One means of moving such firms out is to ensure that they receive adequate compensation for relinquished land. Therefore, urban land market reforms can also create important environmental benefits. xlvii. Initial concerns about land use sprawl and the poisoning of agricul- tural land in peri-urban areas were motivated by satellite images that revealed uses unanticipated by master planners and by vocal Government con- cerns about the adverse environmental impact of townnitip and village enter- prise activity. The transfer of polluting enterprises from central to periph- eral locations also raised doubts about the degree to which relocation merely exported pollution. In general, these concerns have proved somewhat over- stated, but not unwarranted. Aside from insisting that an environmental impact statement accompany all new factory and storage plans, and that enter- prises, iaxistir&g and proposed, have budgets to deal with effluent disposal, rural areas must move quickly to introduce elementary "township" zoning or planning procedures, creating industrial zones, providing for areawide efflu- ent disposal solutions, and tackle the issue of "township" budget and finan- cing sources associated with such measures. In this process, there is a need for municipal authorities to extend the coverage of the controlled "urban planning area" to cover peri-urban areas. curbingt the autonomy of township authorities to dispose of land in an unplanned manner. Summing Up xlviii. The transformation of formerly centralized, command economies is now a worldwide phenomenon, with very few exceptions. In both the rural and urban sector. China is a gioneer and. in assessina economic achievements to date. has few yeers. While generally preserving macroeconomic stability, Government has encouraged localities to experiment on a wide range of changes and then internalized that experience and applied the lessons nationwide. In the final analysis, though any sector report embodies a critical assessment of unfin- ished tasks, note should be taken of overall achievements to date. China's future urban development will be far smoother than that of its post-Socialist counterparts. Government must now begin to measure its progress not by that standard, but by that of the newly industrialized countries and those on the verge of achieving that status. They are the wave of the future. They will be China's key competitors in the export market. With trade liberalization, they will provide the stiffest competition for domestic market shares. Bolder reforms must follow. - xxi - xlix. The main message of the report bears repeating. The key objective of the reform is to create property markets, where land and structures are traded at prices set competitively. This will improve the efficiency with vWhi^ a key input of production is used and promote the emergence of interna- tionally competitive cities. This will require a complementary set of legal, fiscal, planning and institutional reforms. The reform cannot be driven by mere administrative means; incentives are needed to encourage individuals to buy and sell land use rights. Regulatory mechanisms must be set in place to ensure the public of the inherent fairness of the process. Distributional concerns, encouraging authorities to tax away substantial portions of the gains involved, should be moderated by the need to "lubricate' the land market transaction reform. Planners must become more "developer-friendly" and accom- modating to land use demand regarding the location, density, and building use desired. The legal system, now viewed as weak in this area, must be reviewed and strengthened, as necessary, to guarantee property rights and resolve prop- erty disputes. And a macroeconomic environment that sustains rapid growth while maintaining low inflation is also necessary, as the negative experience of other post-command economies in transition makes clear. All these strands hang together. Neglect any one and the reform will fail, if any reasonably short time horizon is envisioned to accomplish the reform. - 1 - I. THE RATIONALE FOR URBAN LAND MANAGEMENT SYSTEMS REFORM A. Creating Competitive Citiest Obiectives of Urban Land Management 1.1 With barriers to trade and economic activity falling dramatically in recent years, and the scope of market forces rapidly expanding, urban land management in China should increasingly seek to establish and preserve well- functioning land markets, as part of a larger strategy to create regionally, nationally, or internationally competitive cities. At least four objectives for urban land management can be set forth; they can easily work at cross- purposes with one another if policymakers pursue any one of them in a single- minded fashion. Trade-offs, bqsed on the analysis of costs and benefits, are essential tools in resolving debates and reaching decisions on urban land policy. 1.2 One obiective is to promote land conservation and control the length of trunk infrastructure investment by Promoting comiact cities. Such cities would contain a substantial proportion of residents and jobs within a "reason- able" distance from the identifiable Central Business District (CBD). The radius involved will clearly be positively correlated with city size. It is, however, not the case that larger cities are, per se, more wasteful in their use of land than smaller urban and rural centers, as China's national urban planners assume, discouraging the growth of metropolitan areas in favor of towns and smaller cities (Box 1.1), Box 1.1: LAND USE PER CAPITA, BY SIZE OF URBAN SETTLEMENT The Center for Policy Studies, within the Ministry of Construction, revealed in 1992 that large cities are far more efficient users of land than smaller towns or nerl-urban townships and villages. Using 1990 data, gross per-capita land consumption per "nonagricul- tura2. resident" rises from 61 m /capita for cities with more than one milliopi persons, to 80 m /capita for smaller cities vith populations exceeding 500,000; and 86 m /capita for those cities with 200,000 to 500,000 residents. The smallest cities consume 103 m2/capita. Bank mission estimates suggest that examining the built-up area of towns and rural settle- ments adds further evidence of how land use/capita increases drastically with decreasing settlement size. Deslignated towns require, on average, 134 m /capita, while rural townships and villages occupy 168 m2/capita within their built-up areas. Source: Lin Zhiqun, "The Development of Land for Urban Development in the 1980s," Beijing, Center for Policy Studies, Ministry of Construction, 1992, processed; and mission estlmates. 1.3 A second obiective stresses the need to expand serviced land devel- onment. and promote redevelopment of existina urban areas. so that land inputs are available at "affordable Drices." or prices perceived to match the "real" marginal value Droduct of land. In the simplest possible terms, there is no obvious reason that the internal rate of return of redeveloping urban land or rezoning and servicing greenfield plots should exceed substantially the inter- nal rate of return of comparable investments. An adequate supply of appropri- ately priced land is a key component of a broader strategy for achieving urban competitiveness. Quite obviously, though, the goals of achieving compact - 2 - development and of maintaining a steady supply of "affordable" land can come into conflict (Box 1.2). Box 1.2, INrERNATIORAL URBAN DEVELOPMENT DENSITIES Seoul, Korea is a model of compact development (especially if the "need" to build diatant satellite c-ies and the related connective trunk infrastructure to accommodate vast greenbelt reservations around the city is temporarily Ignored). C.;mpared to a set of simi- lar market- and vonmarket-economy cities, Seoul manages to accommodate 6 million people within 10 kilometers of the city center, while Paris requires a radius of 14 kilometers to achieve the same objective, and Is outperformed only by Shanghai, which has one of the dens- est *old citles" in the world, where 6 mil1Uon psople live within a 7-kilometer radlus frao the CBD (Figure 1.1)./ Yet, Seoul has achieved that first objective at an extraordinary cost.2/ This is because the city planners have systematically withheld more than half of the total area within Seoul's boundaries from development, while allowing very little land zoned residential to remain vacant (5 percent) and ensuring that new large-scale suburban residential districts achleve relatively high gross floor to land area ratios, averaging 2.3 for walk-up apartments and 3.5 for high-rise apartments (Figure 1.2). By contrast, cities that have treated land as an Input with only nominal value (Warsaw, Poland; St. Petersburg and Moscow, Russian Federation) exhibit a degree of sprawl that can be traced directly to planner-led decisions, and appears to have little to do with the implicit value of land, however roughly calculated. Warsaw's population of 2.5 million requires an urban area 40 kilometers in radius from the CBD, while Paris accommodates that many people within a radius of 5 kilometers, without resorting to Seoul's extreme planning practices. Warsaw's extraordinary sprawl Is due quite simply to a set of regulations and planning practices, backed by massive transport investments, that have forced development along five radial co:- ridors, and encouraged fragmented development, as communities "leapfrog" across greenbelts, and Individual projects make lavish use of land, with gross floor to land area ratios well below 1. In addition, large tracts of land are zoned for agricultural and open space uses, even near the city center (Figure 1.3). l/ Figures can be found in Volume II, Annex 1. 1/ In Korean cities, up to 60 percent of the value of a housing unit is accounted for by land costs, while in the more land-abundant, relatively deregulated US markets, 20 per- cent is the norm. (B. Renaud, "Confronting a Distorted Housing Market: Can Korean Practices Break with the Past?," World Bank, June 1992, processed.) 1.4 A third obiective is a more subiective one: to build a "city beau- tiful," creatina a mix of green space and amenities that are both attractive to "footloose" businesses and skilled workers, and otherwise intrinsically avnealing (Box 1.3). Pushing this objective without regard to the economic costs involved, however, can quickly overwhelm attempts to meet the first two objectives. 1.5 Finally. and often harder to quantify, are the costs and benefits of promotina conservation and preservation. Few would dispute that cities have entire neighborhoods or groups of neighborhoods which have great historic value, and which, by their very fabric of narrow streets, examples of vernacu- lar architecture, and numbers of cultural monuments, are irreplaceable. Cen- tral Paris and Amsterdam, in the Netherlands, along with Venice and Florence, in Italy, are four examples. Neighborhoods in United States cities, like Georgetown (Washington, DC), the North End (Boston, Massachusetts), and the newly restored shophouse areas of Singapore, fit this category, though they were all once viewed by urban planners as prime targets for wholesale demoli- tion and "modernization." Interestingly enough, conservation objectives can also make good economic sense and, as such, can be fully compatible with an -3- lox 1.3i THE REDEVELOPMENT OF BALTIMOREE'S INNER HARBOR Properly conceived, "city-beautiful" objectives can prove extremely beneficial to a city. Many cities have high-access areas formerly devoted to port and warehouse activi- ties outmoded by changes In transport technologies and locational choices. Baltimore, Maryland, United States, transformed 121 ha of "Inner Harbor" land into a vibrant office, comercial and entertalumout center that created a "living room" for the city while attract- ing 25 million tourists annually, where few existed a decade ago. A public-private partner- ship has created a real estate bonanza involving 100 buildings and $1.8 billion in Invest- ments. Conscious of conservation and preservation, blocks of older buildings surrounding the Inner Harbor have been saved from "urban renewal" (i.e., destruction) and have added to the amenity value of the area. urban competitiveness strategy (Boxes 1.4 and 1.5). The locations just cited are immensely desirable as business and/or residential locations, and have very high market values, precisely because they have been preserved and upgraded rather than destroyed. Until very recently, this potential has received very little emphasis in China. In hindsight, who would not deplore the aesthetic and economic loss involved in the insensitive disregard of the historic core of certain Chinese cities where "urban renewal" wiped out much of the character that made them so distinctive only one-half century ago. Cestainly the policymakers of the day would be hard-pressed to defend these measures today. Box 1.4s SINGAPOREs TANJONG PAGAR HISTORIC PRESERVATION PROJECT Singapore's Government commitment to historic preservation and coneervation evolved only during the last decade, and moved from the preservation of individual struc- tures to renovating entiri neighborhoods. The best example of this new approach can be found In Tanjong Pagar, on the western edge of Chinatown. It covers over 200 shophouses on five streets. Units have been renovated with "adaptive reuse" In mind, promoting commercial occupancy (restaurants shops, offices). The buildings and land livolved were compulsorily acquired by the Urban eadevelopment Authority, and then leased. The first shophouses were put up for tender in 1987, and the process continued through 1991. The tender offers stress the fact that, though usually only three stories high, the structures are "rich in architec- ture and history" and "strategically located" to subway stations and nearby commercial and financial activities in Shenton Way. The government invested funds in upgrading Infrastruc- ture "by building an electric substation ...; construction public amenities like water, telecommunication mains and modern sanitation; improving the physical environment by plant- ing trees and creating sidewalks." Tender offers were publicly advertised, and required successful bidders to mconserve (the structures] and to restore and preserve the same according to the architectural design nnd concept of their original construction, within a period of 35 months from the date of tJ1e Authority's acceptance of tender." Lease periods were granted for 99 years, with rights to mortgage and sell or sublease, under conditions spelled out by the Authority. Bids have far e; eeded what the Authority expected, and the restoration effort has proved to be a "value-addt ' -ercise, generating such a high rate of return that comercial developer, have begun to V ! the conservation market, notably in the case of Clarke Quay, which involves the renovat. -- of several hundred warehouses and shophouses on the eastern bank of the Singapore River, and is slated for completion In 1993. Source: Singapore Urban Redevelopment Authorlty. 1.6 Conservation and preservation also covers the natural habitat. In some cases, the issue is clear-cut. Failure to protect the water catchment -4- Box 1.5: AMSTERDAMs REDISCOVERING THE VALUE OF HISTORIC PRESERVATION The Structure Plan published ln 1982 openly endorsed the concept of a compact CitY, with high population and construction densities. The land use policy underlying Amsterdam's development parallels China's chosen path in several respectsl most (75 percent) of the municipality's land is available only on leasehold terms, with general conditions and inflation-adjustel lease rents subject to revision every 50 years. Earlier still, and building on legislation datiag back as far as 1918, 7,000 bu:.ldlngs in the historic city center were slated in 1969 for protection and rehabilitation. As In Singapore, the strategy has proved econom.4 ally viable: 'When existing houses are preserved and restored, there is a less dras- tic cut in the number of dwellings, and rents are usually lower than for new houses. There is a great demand for small, low-rent apartments. Businesses also ?rofit from the preservation and restoration of old neighborhoods, especially the very many small businesses. Not only do these plans provide for maintaining the present population level, and thus the number of potential customers, but low-cost accommodation will also encourage new business developments, so that the whole city may benefit.* The same strategy was seen as providing valuable income from tourism, as Amsterdam became the fourth-largest tourist center in Europe. Source: Amsterdam: Planning and Development, Amsterdam Physical Planning Department, 1983. areas of a city, as Shanghai has discovered, leads to very costly and other- wise avoidable investments in substitute source development. Wetlands in and around urban areas can play a significant role in enabling nature to carry out functions otherwise requiring expensive capital investments. Ecologically fragile hillsides, if opened to development, can lead to man-made disasters (mud slides in the slums of Rio de Janeiro, Brazil) or require extravagant infrastructure retrofitting investments (Bogot&, Colombia). In such extreme cases, market demand or infrastructure costs are not really at issue. Case- by-case decisions need to be made. Decisions to remove potential urban land from development for ecological reasons must be soundly justified and accompa- nied by complementary measures to accommodate demand-driven pressures by pro- viding attractive serviced land alternatives nearby, if the other goals of efficient urban land management are to be met. 1.7 All in all, urban planning raises critical choices. Decisionmakers face the challenge of creating cities that are nationally, and often interna- tionally, competitive, as defined in economic terms. Yet professional urban planners in China, by training and experience, are often ill-equipped to face these choices and challenges, given a historic legacy emphasizing architecto- nic values, abstract planning exercises, and imported Soviet norms which were never exposed to hard-headed cost/benefit review and alternatives' analyses. B. The Urban Sector's Decade of Achievement 1.8 During the last decade, particularly after 1984, China's cities and towns have undergone a series of policy and institutional reforms, as part of a larger, national transition to a regulated market economy. These reforms have been reviewed in a number of World Bank reports, including the 1991 Coun- try Economic Memorandum and a variety of sector-specific studies on resource - 5 - mobilization, financial intermediation and monetary policy, housing, social security, urban labor markets and industrial restructuring.1/ Within the context of shifting macroeconomic objectives, the reforms have sought to decentralize governmental power to lower-level authorities, particularly muni- cipalities; differentiate the functions of government departments from those of business enterprises (granting the latter increasing autonomy in decision- making); increase the scope of market-guided plices and of profits in promot- ing ecoromic restructuring and gruwth; shift resources to the household sector by raiLdly increasing the size of the total wage bill (including wages, bonuses and a variety of monetary subsidies); and expand the scope of property rights for individuals, households, and cooperatives in and around cities and towns. Over the last decade, in response to the implementation of various changes and reforms, GDP has grown at an average real rate of 9 percent per annum. 1.9 These reforms, together with rapid growth, have led to a massive transformation of the urban landscape, particularly outside the "old" city, defined by its 1949 boundaries. Residential and nonresidential construction has flourished. Based on a 1985 end-year census of housing and nonresidential structures and estimates of additions to stock since then, it is clear that half of all the urban structures in China, measured in terms of square meters, were added during the last decade alone (Table 1.1). This can also be docu- mented at the individual city level (Table 1.2). Annual investments in real estate accounted for approximately 10-15 percent of urban GDP, adding approxi- mately 300 million m2 of space in 1990 alone.2/ The overall stock of urban residential and nonresidential space, at 6 billion i29 has a replacement value j/ World Bank, China: Country Economic Memorandum: Planning and Reform in the 1980s, Report No. 10199-CHA, April 15, 1992; China: Revenue Mobili- zation and Tax Policy, Report No. 7605-CHA, 1989; China: Reforming Social Security in a Socialist Ecoromv, Report No. 8074-CHA, June 25, 1990; China: Industrial Policies for an Economv in Transition, Report No. 8312-CHA, June 29, 1990; China: Financial Sector Review: Financial Policies and Institutional Development, Report No. 8415-CHA, June 29, 1990; China: Urban Housing Reform: Issues and Implementation Options, Report No. 9222-CHA, June 26, 1991; China: Reforming the Urban Employ- ment and Wage System, Report No. 10266-CHA, March 1992; and China: Industrial Restructuring: A Tale of Three Cities, Report No. 10479-CHA, April 1992. 2/ As a rule of thumb, total building space additions are roughly twice residential stock additions. The annual investment in urban construc- tion, as a percent of GDP, is thus estimated at at least twice the size of annual investments in urban residential construction, which the Bank's housing sector report estimated at roughly 3.5 percent of GDP; thus, as a percent of GDP, urban building investments exceed 7 percent of GDP. Urban GDP is at least 50 percent of total GDP, when statutory towns are excluded. Actual urban GDP is probably equal to 65-70 percent of total GDP, even if town-level productivity is heavily discounted vis-a-vis city-'evel productivity. Thus, real estate investments as a percent of urban GDP can be set at 10-15 percent. - 6 - Table 1.1: 1985 END-YEAR HOUSING AND NONRESIDENTIAL STRUCTURE CENSUS: VINTAGE OF EXISTING STOCK (million m2) Total Before 1950 19508 1960- 19709 1980-85 1. 323 Cities 2,833.9 274.0 320.1 379.3 853.5 1,006.9 (of which 1,363.3 residential) 1I. 5.270 Statutorg Towns 1,842.8 167.3 104.4 222.8 654.4 693.8 (of which 927.8 residential) Source: Ministry of Construction. Table 1.2: SHANGHAI: CHANGING FLOOR AREA IN CITY DISTRICTS (1980-90) (million mn) Floor/land Land area Building stock area ratio Building stock (km2) 1980 1990 1980 1990 change (2) City proper 748.71/a 91.34 172.56 0.12 0.23 88.9 Huangpu 20.46 8.30 13.39 0.41 0.65 61.4 Nanshi 27.92 7.84 14.23 0.28 0.51 81.6 Luwan 8.05 6.18 8.29 0.77 1.03 34.2 Xuhuei 46.64 11.05 18.81 0.24 0.40 70.3 Changning 28.82 6.38 13.88 0.22 0.48 117.5 Jingan 7.62 6.94 9.08 0.91 1.19 30.9 Putuo 29.88 8.71 16.73 0.29 0.56 92.1 Zabei 27.95 8.49 14.04 0.30 0.50 65.4 Hongkou 23.48 9.62 16.19 0.41 0.69 68.3 Yangpu 59.63 14.63 27.72 0.25 0.46 89.5 Minghang 43.08 n.a. 7.51 0.00 0.17 n.a. Baoshan 425.18 3.21 12.69 0.01 0.03 295.6 la Boundaries as of 1990. Sources: Shanahai Statistical Yearbook 1991, p. 407; and Shanghai Citv ProDs? Real E tate Statistical Material 1980. - 7 - of at least Y 5 trillion, and a depreciated value of at least Y 2 tril- lion.I/ By comparison, national GDP in 1990 was Y 1.8 trillion, while urban GDP totaled at least Y 1.1 trillion. This estimate seems consistent with evidence from other cities where such information is available;A/ in Hong Kong, where land is considered a relatively scarce good, the Government esti- mates that the market value of all property is equivalent to 3.4 times ctrrent GDP (Table 1.3). 1.10 China's remarkable record in urban construction stands in contrast with the relatively small proportion of China's population living in urban areas (26 percent) and the fact that the built-up area of China's cities and statutory towns amounts to only 20,000 km2 (in turn, equivalent to only 2 per- cent of China's documented cultivable agricultural land area). These propor- tions should not, however, mask the fact that China now has the world's larg- est urban population. 301 million people. which is growing at a rapid rate of 100 million persons per decade. Temporary residents swell these numbers by approximatelv 66 million on any given day (Box 1.6). China is undergoing the transition from a rural to an urban society rapidly enough that. by the start of the new century. nolicvmakers will have to plan for the fast-avvroachina day when half of China's population will live in officially desiRned cities and towns. That event, achieved ia advanced countries like the United States 100 years ago, will come sooner than official statistics are likely to docu- ment because, fanning out from each of China's most vibrant cities, in subur- ban areas, and along major interurban highways and water-ays, a vast set of 3/ The replacement value of different types of buildings is based on reports of average building costs for residential and nonresidential structures. The latter systematically exceed the former by about 25 percent. Mission estimates gathered on previous and current work in the housing sector suggest that total invest-ment costs for average residential structures rarely fall below Y 600/m2; the estimated "average" for all space assumes that total investment costs for all building investments average around Y 800/i2. The depreciated value estimates place an average value of Y 320/in, and give heavy implicit weight to the 50 percent of the exist- ing stock added during this decade, particularly after urban growth accelerated dramatically in the period after 1984. These are obviously order-of-magnitude estimates. They do serve to underscore the importance of the real estate sector, no matter what sensitivity analysis may sug- gest as appropriate lower and higher bounds. A/ Ingram, analyzing the fraction of nonagricultural land value in relation to nonagricultural output in the United States, during the 1970s, sug- gests that the value is approximately 0.62. If land values in urban areas represent no more than one quarter of real estate values, overall, then this example suggests that real estate stock as a fraction of local urban GDP equals 2.48 in the United States (G. Ingram, "Land in Perspec- tives Its Role in the Structure of Cities," in M. Cullen and S. Woolery, eds., World Contress on Land Policy, 1980, Lexington, Massachusettet Lexington Press, 1982). Data for the United States, as a whole, as reported periodically by the US Bureau of Economic Analysis, suggest that the ratio of gross value of structures to GDP is approximately 2.74 (David Dowall, personal communication). -8- Table 1.3t HONG KONG: CURRENT VALUE OF PROPERTY AND ITS RELATION TO GDP (in 1992 Hong Kong dollars) District No. of assessments Total rateable value Hong Kong 377,102 67,988,027,660 Kowuloon 206,255 31,340,973,040 New Kowloon 206,053 30,308,790,630 New Territories 409,162 47,957,524,560 Total 1.198.572 177.595.315.890 Notes: (1) Market value of all properties (residential, commercial, indus- trial) (including buildings and land) - Rateable value x 12 - HK$177.595 billion x 12 - HK$2,131 billion (2) GDP at current market price (1991/92) - HK$633 billion (3) Property + land value/GDP - 2,1311633 - 3.37 Source: Rating and Valuation Department, Hong Kong Government. Box 1.6: SELECTED TIME SERIES FOR TEMPORARY RESIDENTS /a IN EIGHT CHINESE CITIES ('000 persons) 1984 1988 1989/b Beijing 300 1,310 1,310 Tianjin 417 (1985) 813 612 Shanghai 750 1,246 1,260 Guangzhou 500 1,170 914 Shenyang 86 (1985) - 328 Wuhan 350 745 Chengdu 270 - 420 Hangzhou 250 (1985) 295 358 la Present for more than three days at the time of the sur'ey. /b Recession year. Temporaries were often involunta^rrly rrpatriated. Source: M. Rutkowski, "China's Floating Population and Labor Market Reforms," World Bank, December 1991, processed. peri-urban settlements is developing. There, township and village enterprises (TVEs), engaged mostly in industrial activity, are already producing over one fifth of the gross value of industrial production, while the state-sponsored urban enterprise sector now accounts for barely half of national output.5/ Despite the existing size of the urban vopulation. China can still benefit from being at a relatively early stage of the rural-urban transition. avoiding mistakes, at the martin. and options not available to other. already larmely urbanized, post-Socialist economies. C. The Challense of RestructurinR the Urban Economy 1.11 If one resorts to a set of stylized facts to describe a large nationally and internationally competitive city in the late twentieth century, one would single out a few key items. First, whether in the developed or developing world, the manufacturing sector's share of output and employment ranges from 20 to 30 percent. Second, employment tends to concentrate in particular subsectors, rather than range across all output categories. Third, these activities operate in a relatively fast-paced environment, as product life spans are progressively shortened, competition increases, and quality and price depend increasingly on an evermore complex, interrelated strategy that -works to improve efficiency at the physical plant level and within executive and financial management operations, while giving greater prominence to mar- keting research. 1.12 In this process, information-intensive inputs have grown enormously (whether provided "in-house" or through the use of external suppliers). In highly urbanized Singapore, where manufacturing accounts for 29 percent of GDP, financial and business services total 30 percent of GDP. And there, as elsewhere among advanced Asian cities, services are the key generators of GDP and employment. In Hong Kong, 66 percent of employment is in the services sector, while 80 percent of Seoul's employment is tertiary in nature. By con- trast, in those cities, manufacturing accounts for 34 percent and 19 percent of all jobs, respectively.6/ 1.13 For a highly urbanized country like the United States, the story is much the same. Manufacturing accounts for 25 percent of employment in a typi- cal large city, government takes up another 5-8 percent, as does construction. The remainder involves transportation and business and personal services. 5/ The spatial distribution of TVEs is not known to Chinese authorities except on the basis of anecdotal evidence. That evidence suggests that it is largely a suburban phenomenon, in terms of land occupied, employ- ment, and value added. This issue should become a high-priority research item among urban and economic development planners in China, who alone can assemble the necessary corroborating evidence. 6/ Information derived from local annual statistical yearbooks issued by each city. - 10 - Business services alone may now account for as much as 40 percent of city employment.7/ 1.14 The driving force behind these figures can best be described by the following quotation: "The developed world at the end of the twentieth century is still in transition from an industrial to an informa- tion economy where information becomes the most important, indeed all-pervasive commodity. This does not mean that manufacturing industry will atrophy or that the demand for goods will diminish; on the contrary, there will almost certainly be massive increases in demand for manufactured products. But it does mean that various complex substitu- tions will take place. One involves the introduction of ever more powerful information technology into the manu- facturing process, hence the substitution of this technol- ogy --based on robotics and automation--for labor and traditional energy-intensive capital. Another is the substitution of knowledge and intelligence for some mate- rial and energy resources, thus achieving the more effi- cient design of products, greater knowledge of their mar- kets, and tighter, more robust planning and operation of their production and distribution processes. These two changes are also facilitating the shift to increased qual- ity, diversity, and 'personalization' of the products involved. The mass markets of the industrial age are evolving into smaller, more specialist niche markets, offering the consumer specialized services and products which in the past were assumed to be totally inconsistent with tradi- tional manufacturing."8I 1.15 By contrast, for all 467 cities of China, taken as a whole, 56 per- cent of municipal GDP is still derived from the secondary sector; and in cities like Shanghai, the number rises to 65 percent. Among major cities surveyed by the mission, the lowest percentages were found among the "south- ern" cities considered the most advanced in reforming their economy. Even ZI Thi, includes engineering, legal, real estate brokerage, telecommunica- tions, finance, accounting, consultants, iniurance, and marketing, as well as nongovernmental quaternary services such as research and develop- ment. 8/ "Epilogue," in J. Brochie, M. Barry, P. Hall, and P. Newton, eds., Cities in the 21st Centuryt New Technologies and Spatial Systems, New York: Halstead Press, 1991. - 11 - there, the proportion contributed by the secondary sector totaled or exceeded 40 percent.2/ 1.16 Furthermore, China's cities typically produce every conceivable manufacturing product, revealing little selectivity or specialization, while operating with plants that are, by almost any measure, generally obsolete in layout, technology, inventory management, quality control, labor efficiency and raw material utilization rates. 1.17 Worse still, as documented below, much of this industry is located in the built-up areas of the city, often within the boundaries of the 1949 city, utilizing 20-30 percent of the land.10/ In Hong Kong, by contrast, industry utilizes only 6 percent of the built-up area, while in Seoul, the proportion is 9 percent. In North America, the norm is 10 percent or less.1l/ 1.18 At a small-area level, the situation is even worse. Shanghai's port authority, for example, controls 229 hectares of prime real estate along the Puxi side of the Huangpu River, sloag the Bund and on either side of that center. This land is typically devoted to low value-added activities, such as warehouses. That amounts to 2.5 percent of the area of the old (pre-1949) city, and obviously a much higher percentage of its land value. If freed up, much of this area could potentially be transformed into a much higher-value mixed-use commercial and entertainment center, following examples set by such cities as Boston and Baltimore in the United States. 1.19 In places like Shanghai, restructuring will involve important land use readiustments. and infrastructural investments to accomyanv them. These investments will be massive in scope within any one sector, and may require significant spatial targeting, directed at the densely populated old core 9/ State Statistical Bureau, ChiT.s St%tistical Yearbook-1991, Beijing, 1992. Ironically, when Shanghai was irxa of the premier urban centers of China, and a contender for Pacific Rim leadership in international trade, in 1952, 42 percent of GDP originated from tertiary activities suci as com- merce, finance, communications, science and technology research and development, education, and "culture." Today that proportion has fallen to 30 percent. (K. Fung, Z. Yan, and Y. Ning, "Shanghai: China's World City," in Y. Yeung and X. Hu, eds., _ina's Coastal Cities: Catalysts of Modernization, Honolulu: University of Hawaii Press, 1992. 10/ According to Lin Zhiqun ("The Development of Urban Construction Land in China During the 1980s," Beijing, Ministry of Construction, 1992, pro- cessed), the percentage of built-up area land devoted to industrial uses in the sector study cities varies from 18 percent in Shenzhen to 27 per- cent in Guangzhou. An additional 3 to 6 percent is taken up by ware- houses. The same phenomenon emerged in Soviet cities. Approximately 30 to 40 percent of the inner areas of Soviet cities are occupied by indus- try (G. Andrusz, Housina and Urban Development in the USSR, London: Macmillan, 1982). II/ Ibid. - 12 - areas of the city. To cite but two examples. First, in the competing Asian cities like Singapore and Hong Kong, there are approximately 45-50 telephones per 100 persons; in London, Paris, and New York, the number of phones per capita is greater than 1. In Shanghai, there are only 6 telephones per 1OQ persons. Second, within Shanghai's inner-ring road area, equivalent to the 1949 city boundaries, only 10 percent of land is used for transportation, whereas the norm in developed county cities is 20 percent or more, and similar proportions are found in such high-density centers as Hong Kong and Singapore. 1.20 Thus, by any reasonable measure of scope involved, the urban eco- nomic transformation under way will require policymakers within China to focus serious attention on urban land management issues. China justly prides itself as one of the few developing countries that has largely avoided "disorderly" development, characterized by the proliferation of unplanned and unanticipated slums and squatter settlements. To maintain that enviable record will become much more difficult during this decade and beyond because the emergence of the market economy is loosening the control exercised by urban and economic plan- ners ovqr the timing and location of real este_e de-elopment. Those p.Lanners are having to learn to live with an unpredictable world, where citywide and, more particularly, small area (district and subdistrict) population and eco- nomic activity cannot be anticipated as easily as in the past. This is par- ticularly the case now, as cities like Shanghai and Beijing have formally granted lower-level authorities the right to approve project proposals, feasi- bility reports, and project contracts independently--up to a certain thresh- old--without clearance from the municipal governments themselves. Efficient and equitable urban development, following environmentally sound practices, will call for new skills, new policies, and new institutions. And China's urban planners will have to learn to live with the challenge of reinventing the Chinese city not merely to effectively compete within and across prov- inces, but also to gain an increasing share of direct foreign investment and world trade in goods and services. Success is thus like a two-edged sword. Having committed themselves to developing a world-class urban economy, closely interconnected with the international market, China's policymakers cannot rest on their laurels. What from the vantage point of 1984 looked like thoroughgo- ing reform is but a first step. D. The Urban and Peri-Urban Environmental Challenge 1.21 Industrial redevelopment problems are enormous and pose serious challenges to planners. The most pressing problem is undoubtedly the old age and dilapidated conditions common to many industrial factories.12/ Land 12/ In the core area of Shanghai, defined as containing 3,000 hectares, there are 5,603 factories occupying approximately 20,000 sites. These sites average 2,145 m2, and total 429 hectares. Other analysts paint a picture which is equally startling by adding in the area occupied by warehouses: "It is estimated that over 30 percent of the built-up area of the central city is occupied by factories and warehouses. In the urban district of Nanshi, factories [and warehouses] occupy 61.6 percent of the total area of 4,175 hectares. Further, because of the shortage of storage space, many workshops use the neighborhood streets for storage ...." (Fung, Yang, and Ning, "Shanghais China's World City," op. cit.). - 13 - reform, providing factories with incentives to move from inappropriate, high- value CBD locations, can help improve environmental conditions while allowing technological renovation or the development of new product lines to take place. Other land users, more capable of using high-access locations, can then occupy the old sites and accelerate the transformation of Chinese cities, allowing land uses and production patterns to more closely resemble that of an internationally competitive city. 1.22 The physical plant of Shanghai's industries is, on average, uncom- petitive by international standards; nearly 50 percent of the physical plant and equipment of factories was built before 1949. Across Shanghai, manufac- turing facilities have expanded incrementally into residential neighborhoods, and as a result, they are highly inefficient. Many of the inner city facto- ries reveal a pattern: a series of houses or commercial shops have been linked together as businesses expanded output. In some cases, clusters of buildings in a residential district have been converted to industrial uses and raw materials and products at various stages of processing and fabrication are shuttled back and forth between these buildings. In more extreme cases, mate- rials are literally hauled and backhauled across cities like Shanghai simply because industries have no access to contiguous sites for their operations. Such practices severely penalize manufacturing productivity. 1.23 Another serious and related problem posed by the inadequate manufac- turing facilities found in most central cities is the fact that cramped quar- ters frequently makes it difficult for factories to update their equipment. Despite the fact that new machines might double or triple process efficiency, low ceilings, small elevators and small or awkward floor-space configurations make upgrading impossible.13/ 1.24 Overall, in terms of potential industrial productivity gains, indus- trial redevelopment is of critical importance in China's large urban areas. But there is another pressing reason for industrial redevelopment--urban envi- ronmental pollution. Although there have been few systematic surveys of industrial pollution, officials in every city visited expressed considerable concern over dangerous conditions--houses next to polluting factories, chemi- cal spills, noise pollution, vibrations, dust, congested alleys and lanes.14/ 1.25 Therefore, one aspect of environmental policy involves moving pol- luting enterprises away from dense urban areas into rural areas. Except for the Ministry of Light Industry, industry ministries and the State Planing Commission appear to support this policy. Rural areas clamor for such firms 13/ Shanghai is merely a representative case. In Guangzhou, most buildings are no more than two or three stories high and, as of 1984, half were built before 1949. Within this low-density environment, extensive indus- trial use of former domestic structures, and a lack of separation of residential and factory structures is the norm (S. Li, "A Comparative Study of the Urban Land Use Patterns in Guangzhou and Hong Kong," op. cit.). 14/ Guangzhou's Construction Commission made particular note of this issue. _ 14 - to augment the local tax base.l5/ For a variety of reasons, firms that move tend to be small and fairly profitable. This leaves the large firms (polluting or not), and the small unprofitable firms, in the cities. The latter are likely to pose severe environmental problems since they are likely also to be scrimping on pollution control. The firms that move may also be presumed to be highly polluting. Whether moving them makes sense depends, in part, on the economic costs of such a move and the net environmental benefits. In a market economy, profit maximizing firms take the price of land, labor, and capital into account (as well as any benefits from being located in a city or outside a city) in their location and output decisions. In this case, it may be presumed that forcing a relocation will reduce profits and social wel- fare. When pollution of the firm is taken into account, this presumption cannot be made, as environmental damages from the firm's activities may be far larger if located in some locations than others, but such damages are not considered in the firm's location decisions. In a planned economy with tight restrictions on firm and labor mobility, it is difficult, if not impossible, to make inferences about whether economic benefits would be reduced after relocation. The presumption is that health impacts would fall after the move because population exposures to the firm's emissions would be lower when the firm is in a rural area, even if emissions were the same. However, as the relocated firms are expected to meet the restrictions in the environmental impact assessment process, total emissions should to be far lower than they were in the central city. Nevertheless, if the firm locates near fragile ecosystems or irrigated areas and has high levels of pc'lution,16/ serious damages could result and could conceivably exceed those avoided in the urban area. 1.26 In rural areas, where implicit and explicit land values are low, the density development would be expected to be low as well, and evidence of "market-driven" trends is limited to measuring the extent to which land use has switched from agricultural to nonagricultural purposes. Though the evi- dence is spotty, the impact on urbanization is undeniable (Tables 1.4 and 1.5). Ironically, the key issue is whether there is too little Planning in such areas and thus how best to introduce approRriate -lanning practices. consistent with market forces. into these areas (Box 1.7). The worst-case scenario has it that in peri-urban areas, agricultural land is unnecessarily squandered by the proliferation of unconnected urbanized sites, scattered across the farming landscape. In addition, concerns exist that limited infra- structure and neglect of the environment is allowing polluting factories to dump untreated effluents and contaminate adjacent agricultural land and the water sources that nourish it. In most instances, land management reform in 15/ World Bank, China: Environmental Strategv Paper, Report No. 9669-CHA (Washington, DC: 1992), Annex 1, p. 9. 16/ The law places very fragile areas off-limits to industrial location. Enterprises are not permitted to be sited on the windward side of a city or in protected (including water resource) areas. A related policy pro- hibits certain types of TVE production activities from even being estab- lished (such as sulfur production, tanning, and dyestuff manufacturing). The study team was not able to ascertain the degree to which these laws and policies are implemented. - 15 - China's cities imDlies greater deregulation: in neri-urban areas. more reaula- tion may be necessary (Box 1.8). coverina an area which is arowing in scale as fast as or faster than that of China's cities.17/ Table 1.4: SHANGHAI: CHANGE IN CULTIVATED LAND IN PERI-URBAN TOWNSHIPS, 1980-88 County (no. of Cultivated land chanae (ha) _ townships) 1980 1988 Reduced land Z Change Baoshan (5) 2,226.67 1,746.67 480.00 -21.56 Jiading (3) 3,046.67 2,593.33 453.34 -14.88 Shanghai (7) 7,420.00 6,586.67 833.33 -11.23 Chuansha (10) 7,666.67 7,106.67 560.00 -7.30 Total (25) 20.360.01 18,033.34 2,326.67 -11.43 Source: State Land Administration. Box 1.7: THE ROLE OF SUBURBAN COUNTY AUTONOMY IN HINDERING HETROPOLITAN GUIDED LAND DEVELOPMENT STRATEGIES In Fuzhou, the urban districts do not contain any farmland. They are, in general, smaller than the actual urban built-up area spilling out beyond its boundaries. As a result, those built-up areas outside the urban districts, located in suburban districts, are very difficult to monitor. Although the urban district will expand to reflect new urban development, such expansion lags far behind reality, and not all, new development is covered. One reason for this is that the suburban district governments, who control all surrounding rural townships, do not want to give up these areas, since they are the most productive 'revenue cows" within the suburban district. As a result, nobody, even local planners, have any comprehensive data regarding urban built-up area and population. 1.27 Recent field work and the results of large-scale surveys undertaken by the Nanjing Research Institute for Environmental Studies suggest that peri- urban pollution is a matter for at least selective concern, particularly in the case of such contaminants as hexavalent chromium, lead, and waste gas emissions (particularly dust emissions). The chief offenders are older enter- prises, established before the widespread introduction of environment contract responsibility systems for enterprises built during and after the Seventh Five-Year Plan Period (1986-90). New enterprises are increasingly subject to environment impact statements, incorporated into project feasibility studies, and scrutinized at various critical project development stages, includiuig 17/ The built-up area of China's cities is expanding at a rate of about 50,000-60,000 hectares per year. Nationwide, rural land converted to nonagricultural construction (industry and related urban uses) is esti- mated to average 100,000 hectares per year, with the presumption that most of this is peri-urban in nature. Most rural industries, for exam- ple, are located within 25 kilometere of cities or along transport arte- ries fanning out from those cities. - 16 - Table 1.5: SHANGHAl: DISTRIBUTION OF REDUCED CULTIVATED LAND, 1980-88 (ha) County (no. of Produc- Nonpro- Fish- townships) tivela ductive/b pond River Road Total Share Baoshan (5) 374.00 82.67 16.00 3.33 2.00 478.00 20.5% Z share 78.2% 17.3% 3.3% 0.77 0.47 100.0% Jiading (3) 295.33 46.67 60.00 8.67 41.33 452.00 19.4% % share 65.3% 10.3% 13.3% 1.9% 9.1% 100.0% Shanghai (7) 583.33 179.33 38.00 0.00 38.67 839.33 36.0% Z share 69.5% 21.4% 4.5% 0.0% 4.6% 100.0% Chuanaha (10) 397.33 107.33 45.23 3.33 6.00 559.32 24.0% % share 71.0% 19.2% 8.1% 0.6% 1.1% 100.0% Total (25) 1.652.12 416.49 159.54 15.36 88.14 2,331.65 100.0% Z share 70.9% 17.9% 6.8% 0.7% 3.8% 100.0% /a Productive: industrial and storage land uses. !b Nonproductive: housing and other nonfarm land uses. Source: State Land Administration. applications for loans from the China Investment Bank and the Agricultural Bank of China. It is too early to assess the degree to which these require- ments are being implemented, and whether supervision is adequate. Both bear monitoring. 1.28 The Nanjing Institute survey estimates that about 70 percent of the output of TVEs is derived from TVIEs, employing 57 million workers and concen- trated in the periphery of large coastal cities.18/ Rural industrial pol- lution is caused largely by 600,000 enterprises, or 8 percent of all TVEs. Heavily polluting enterprises total 366,000 in number, and are responsible for 75 percent of measured discharges. As is typical of the industrial structure of Chinese cities, TVIEs are involved in all of China's 40 designated indus- trial sectors. The main polluting industries include pulp and paper making, dyeing, electroplating, chemical production, food and alcohol processing, smelting, asbestos, refining, coking and building material production (cement, bricks and tiles). Production efficiency and pollution control are serious prob'ems for these industries in the face of strong market competition. While rural industry in the aggregate may be less polluting than its urban counter- part, either in absolute terms or in terms of total output of pollutants per unit of production value, this in large part reflects differences in output composition. Individual TVIEs' waste discharges can be quite serious in rela- 18/ By way of contrast, the entire urban labor force with permanent residency status is only about 140 million in number, and the total TVE labor force amounts to 92 million. - 17 - Box 1.8: PERI-URMAN DEVELOPMENT REGULATIONi IN HONG KONGt A NEW STRATEGY For historical reasons, the Town Planning Ordinance only covered zhe existing and potential urban areas which basically Included the main urban areas and the new towns in the New Territories. Area. outside these areas were not covered by the ordinance. There was little problem in the pro-war period (1939) because most of the population and development was mainly concentrated in the main urban areas of Kowloon and hong Kong Island. However, with the development of new towns in the New Territories, many developments have spilled over to places outside the new towns. Much agricultural land has been converted to open storage uses, for use as car dumps, open-air workshops, container storage and vehicle car parks. They have been carried out in a haphazard and disorderly manner, creating iticrompat- ible land use, and environ'ental and traffic problems. To stop the environment from further rapid deterioration, the . mernment proposed to amend the Town Planning Ordinance in July 1990, while a more comprehensive review of the ordinance was still under way. The Town Planning (Amendment) Bill 1990 was passed in January 1991. It extended the jurisdiction of the Ordinance to cover the whole of Hong Kong, with new types of zonings covering country parks, greenbelts, village-type development and open storage uses. Interim planning control through the use of development permission area plans was used to guide developments in the New Territories while the Outline Zoning Plans, which are statutory plans for development control, are under preparation. The Town Planning Board may designate areas which require immediate planning control as Development Permission Areas (DPAs). The DPA plans have to be exhibited and publicly consulted, the same as the statutory Outline Zone Plans, and will be effective for three years. All developers in DPAs have to obtain planning permission from the Town Planning Board. To give the Town Planning Ordinance enforcement power to control unauthorized development in the DPAs, enforcement notices, reinstatement notices and stop notices, similar to those used In Britain, were introduced. Enforcement notices may be werved to the landowner/occupier/responsible person to reinstate the land to authorized development if planning permission has not br..n obtained upon expiry of the period of the enforcement notice. If the unauthorized development seriously constitutes a herlth or safety hazard, adversely affects the environment or makes It impractical or uneconomic to reinstate the land within a reasonable time period, a stop notice will be served to discon- tinue the unauthorized development Immediately. Any person who fails to comply with the enforcement notice, stop notice, or reinstatement notice is liable to a fine and liaprison- ment. An Appeals Board will be set up to deal with appeals related to objections and refus- als of planning permissions of the DPAs. Source: A. Yeh, "Land Leasing and Urban Planning in Hong Kong," Hong Kong: University of Hong Kong, November 1991, processed. tive and absolute terms. For example, in 1991, TVIEs as a whole discharged only 10.7 percent of the industrial wastewater in China, while accounting for 28 percent of production output value. More revealing, however, is that the TVIE pulp and paper industries alone accounted for 44 percent of the TVIE wastewater discharge. Many such plants do not comply w4tb effluent standards and cause serious local damage. The concentration levels are due to low pro- cess efficiency and poor waste treatment in the factories concerned, despite fines and sanctions. 1.29 The situation with regard to air quality and solid wastes is simi- lar. In 1991, TVIEs produced about 9.4 percent of total waste gas emissions and 11 percent of solid wastes, again levels well below their share of output. Treatment levels for waste gas -.re about 20 percent of urban levels, them- selves often inadequate. The chief hazards are fluorine and sulfur dioxide. The cement industry accounts for the bulk of the dust emissions. 1.30 The sector study team did not have the resources to investigate this topic thoroughly. Other sources, including those from Shanghai's Municipal Bureau of Environmental Protection, confirm progress, while emphasizing haz- - 18 - ards yet to be tackled. On the one hand, sewerage treatment facilities have improved dramatically in selected suburban counties, including Chuansha and Shanghai. Yet the Bureau's Agro-Ecological and Macro-Natural Protection Divi- sion claims that its suburban counties contain 3,211 rural enterprises (out of a total of 16,000) with pollution problems requiring remedial action. Some relocated state-owned factories were also identified as polluters. The full scope of the problems involved remain beyond the scope of this report. The issue does, however, merit close monitoring, as peri-urban development gradu- ally merges into the "field" of a greater metropolis, and can no longer be considered in isolation from urban land use issues, broadly defined. E. The Issue of Industrial Redevelopment 1.31 Urban planners in China's largest cities are starting to focus their attention on industrial relocation and redevelopment. Most planning strate- gies call for the relocation of factories to outlying areas, often located in new towns and smeller villages, which have been targeted for industrial devel- opment. In some cities, these efforts are potentially enormous. In Shanghai, for example, plans to develop Pudong call for the construction of millions of square meters of industrial facilities. In virtually every case, efforts to relocate industrial facilities have been, until very recently, administra- tively driven and constrained by a lack of resources. While this difficulty is to be expected given the scsle of potential redevelopment, the lack of a transparent land market prevents the relocating factory from capturing the income associated with the sale and/or redevelopment of its relinquished site to a more productive use. Financing its modernization at a new, more suitable site is thus made more difficult. Very recently, some coastal cities, like Shanghai, have begun to allow existing occupants to retain most of the income derived from the transfer of land-use rights. 1.32 Chinese planners and industrial developers and factory operators confirm the similarity of the present pattern of industrial relocation and redevelopment across most cities (although activities in Guangzhou are the exception). The usual process uncovered during the sector study preparation mission was as follows. First, based on some rudimentary surveys of indus- trial conditions and urban environmental pollution, factories were targeted for relocation. Those factories targeted for relocation were normally compen- sated for the value of vacated buildings, and equipment and ancillary facili- ties that could not be relocated. In some cases, the party carrying out the redevelopment would also pay for lost wages resulting from the idling of work- ers. Compensation provided for relinguished land was the exception, though estimates of the depreciated value of vacated structures may have been - 19 - inflated to allow for some implicit compensation for land rights relin- quished.19/ 1.33 In some cases, the new user of the relinquished site would pay the actual costs of acquiring a new site. In other cases, government grants and loans were provided to cover the costs of land acquisition. In all cases sur- veyed, the industries paid for the actual construction costs of new facilities themselves, and in most cases, the factories invested heavily in new equipment and accessory investments. 1.34 The limitation of capital markets for industrial debt and the fail- ure of factories to receive adequate compensation for relinquished land together made the financing of industrial relocation difficult. As a result, the pace of industrial relocation so far has been very slow, as the following sections describe. Industrial Relocation in Shanghai 1.35 Shanghai has been hampered in implementing industrial relocation, inter alia, by limited resources. Overall, in the city proper, between 1985 and 1990 (the Seventh Five-Year Plan), 103 factories were relocated from old areas, accounting for less than 0.02 percent of all industrial facilities in the municipality. These factories had a total constructed area of 810,000 i2. The land area of these sites totaled 820,000 m2. In the new areas, these 103 factories were given land totaling 1,270,000 m2. The new constructed area was 760,000 m2. The industries usually require less space because the old space was used in an extremely inefficient manner. 1.36 Even though inner city land could be extremely valuable if converted to commercial shops, office buildings or commodity housing; during the Seventh Five-Year Plan (1986-90), only limited attempts were made to capture the dif- ferential value associated with relocating inner city factories to suburban locations. Instead, the government stitched together a variety of programs to partially fund industrial relocation. Funding for industrial relocation came from five principal sources: soft (i.e., concessional loans from the Environ- mental Bureau (20 percent); the Economic Commission for Industrial Restructur- ing (20 percent); general revenues from the municipal budget (12 percent); factory payments (14 percent); and bank loans (42 percent). 1.37 During the Seventh Five-Year Plan, total relocation costs were Y 1.4 billion. Table 1.6 illustrates the breakdown of the costs for industrial 191 The State Land Administration has monitored several forms of spontaneous transactions involving the transfirauation of administratively allocated land into an active commodity with market prices. It also reports these are taking place on a large scale. First and foremost, buildings and land are "sold" together, and the land price is included in the total building price. Second, buildings are leased, along with the associated land use rights. (Zou Yuchuan, "Speeding Up the Regularization of Land Market Tra :tions and Strengthening the Management of Administratively Allocated Land Use Rights According to the Law," Beijing, State Land Administration, December 20, 1991, processed.) - 20 - relocation in Shanghai. The total Seventh Five-Year Plan costs work out to an average of approximately Y 1,700/mr of relocated old factory land. Of the total, about 39 percent went to pay for land requisition and infrastructure provision--Y 550 million. This works out to Y 433/m2 or Y 289,000/mu, for the 1,270,000 m2 of land acquired for the relocating factories. Table 1.6: SEVENTH FIVE-YEAR PLAN INDUSTRIAL RELOCATION COSTS, SHANGHAI Per m2 of Total cost relocated factory (Y mln) (Y) Land and infrastructure 550 671 Building construction 750 915 Moving costs 20 24 Other costs 80 98 Total Costs 1.400 1.708 1.38 Clearly, a more ambitious industrial relocation program will require substantial resources. While there are no surveys of industrial conditions in the inner-city area, a rough estimate of the potential costs of industrial relocation for Shanghai can be made. If one conservatively estimates that 25 percent of the city's core area factories require relocation, then about 5,000 sites need redevelopment. The land area covered by this redevelopment activity would approach 1,100 hectares. Using cost figures from the Seventh Five-Year Plan, the total cost of such an industrial redevelopment effort is likely to exceed Y 18 billion. Based on estimates of the actual relocation cost during the Seventh Five-Year Plan, the cost of acquiring and servicing new suburban land was about 40 percent of the total relocation cost. Using this proportion, the total land acquisition and infrastructure cost of relo- cating these 5,000 establishments would be approximately Y 7.2 billion. 1.39 Since, under the approach followed until 1992, factories did not receive explicit compensation for the land that they left behind, ainA since the new users did not finance the acquisition and improvement of new indus- trial estates, either the old factories or the government were required to come up with the funds for relocation. While this may have been easy for profitable firms, many firms do not have sufficient cash flow to cover the cost of relocation. Thus, the government has more often than not been called upon to provide subsidies. During the Seventh Five-Year Plan Period, enter- prises provided only 54 percent of the costs of relocation (12 percent in cash and 42 percent in loans). 1.40 The Eighth Five-Year Plan (1991-95) calls for the relocation of 131 factories from the city center. The projected cost of these relocation efforts is estimated at Y 2.5 billion, about Y 19 million per factory. This estimate is approximately 40 percent greater than the per-factory cost incurred during the Seventh Five-Year Plan, and reflects the fact that the - 21 - next round of relocation projects will focus on more problematic cases; there will be more difficulties in relocating such factories. Other problems will emerge as well--good suburban industrial estate sites will become more expen- sive and factories will be forced to move out to more remote locations. Oper- ating at these distant sites will make relocation more costly in terms of worker relocation, if commuting is excluded as an alternative. One possible way around these problems is to capture the differential land value associated with the industrial relocation; this policy is now gaining acceptance in Shanghai. Industrial Relocation in Hangzhou 1.41 In 1984, the city conducted a detailed survey of the land uses in the West Lake area, identifying 86 factories and work units which needed to be relocated. Of the total, 52 enterprises were targeted for relocation because of pollution problems. The remaining 34 factories and shops were identified for relocation because of other redevelopment requirements, such as the expan- sion of infrastructure and the upgrading of old urban areas. The combined constructed area of these 86 businesses totaled 694,000 m2. The new indus- trial areas will have a FAR of 1:0.87, a level more consistent with modern standards. This will require the requisitioning of 800,000 m2 of land for new industrial estates. 1.42 By 1990, 59 factories had been moved. All of the polluting facto- ries had been removed from the West Lake area and all temples previously con- verted to other uses had been returned to religious organizations. With the relocation, factories have had their industrial production facilities revital- ized. Beyond improving environmental quality in and around Hangzhou, reloca- tion has greatly enhanced factory productivity. Most of the ol; 'actory sites have been recycled to uses consistent with the Comprehensive Pla.i. In the West Lake area, all industrial sites vacated will be transformed to open space or tourist-oriented uses. 1.43 Like Shanghai's, Hangzhou's industrial relocation program is expen- sive. A total of 376,000 m2 of new industrial space has been built at a total cost of approximately Y 380 million--an average cost of approximately Y 1,010/ m2. Costs of relocation are higher than projected, and the city expects the actual costs to approach Y 550 million, about 20 percent higher than planned. Another problem which compounds the high costs is the fact that some of the factories needing relocation are not profitable. Of the 86 firms targeted for relocation, 17 do not generate retained earnings which could be used for relo- cation. The government creatively assembled a variety of funding sources. Of the 59 factories which were relocated up to 1990, the funding sources break down as follows: - 22 - Y million Percent Grants from state and province 17.6 4.6 People's Construction Bank loans 54.5 14.3 Local grants (municipal) 47.9 12.6 Enterprises 260.0 68.5 Total 380.0 100.0 1.44 Despite the relatively high costs of industrial relocation in Hangzhou, and the problem of low profitability of targeted industrial enter- prises, industries paid a higher share of the relocation costs (69 percent) than in Shanghai (54 percent). Many enterprises in Hangzhou have received substantial compensation for their old buildings since they were then occupied by other users. Vacated industrial buildings in Hangzhou can be recycled and, as a result, relocating factories receive significant compensation for the structures left behind. Industrial Relocation in Fuzhou 1.45 Like Hangzhou and Guangzhou (see below), Fuzhou has conducted an -^atironmental and land use assessment of its industrial uses located in its inner urban area. As a result, a number of old factories have been targeted for relocation. Over the past seven years, 22 factories have been movod out of the city center. To date, the government has invested Y 50 million to underwrite relocation and the total cost of the industrial relocation to date is estimated at Y 200 million--about Y 9 million per factory. Most of the vacated industrial land has been recycled to accommodate residential uses and commercial activities. 1.46 The process of industrial relocation starts with the factory agree- ing to relinquish its site. When it does, the local SLA Bureau obtains the use right. The new user pays the factory compensation for the old buildings, fixed investments, and lost wages and production and moving costs, but not for land. Instead, compensation is paid for requisitioning a new piece of land for the factory in one of Fuzhou's new suburban industrial estates. At pres- ent, differential land values are not used to determine the level of compensa- tion paid to the old factory. Monies paid are based on the actual costs of finding a new site, compensation for old facilities and equipment and payments to idled workers. Brokers operate in Puzhou, and negotiate with factories to identify redevelopment possibilities. New users typically pay these brokers commissions of 1 percent of the total level of compensation costs. In some cases, where the site was used by heavy industry, the city will itself iden- tify new users, and clean up any pollution. 1.47 The costs associated with acquiring new industrial sites in suburban areas have been rising over the past three to four years. In 1988, the cost to requisition unimproved land in the suburbs averaged between Y 150 and Y 190/i2. In addition, the relocating factories must pay infrastructure fees (Y 150/m2), other fees totaling Y 15 to Y 30/m2, and costs for the construc- tion of new buildings (averaging between Y 400 and Y 6001m2). Given a FAR in the new industrial areas of 1:1.0, the typical costs for new facilities on a - 23 - 10,000 m2 site with a 10,000 m2 building would range from Y 7.9 million to Y 10.8 million. 1.48 In Fuzhou, the net cost to factories for relocation should be the difference between the cost of construction for the new building and the com- pensation received for old buildings. By paying this difference, the factory gets a new and much more efficient physical plant. The typical relocation compensation received by factories in 1991 exceeded 50 percent of the total relocation costs, and in some rare cases, where more than one new users com- peted for the old site, the relocation compensation was even higher. 1.49 Nevertheless, despite the generous compensation received by facto- ries for relocation, planners and brokers claim that it is difficult to per- suade factories to relocate. The process of industrial relocation is long and complex, and most real estate development companies (REDCs) 20/ prefer to engage in greenfield development or in residential redevelopment projects. Despite these difficulties, the city has targeted seven major polluters for relocation from the city center: Fuzhou Barrage Factory; Fuzhou Tire Recap- ping Factory; Fuzhou Forged Steel Factory; Fuzhou Sewing Machine Factory; Fuzhou Roofing Materials Factory; Fuzhou Motor Factory; and Fuzhou Dyeing Factory. 1.50 The present system of financing industrial relocation in Fuzhou (and to a lesser extent in Hangzhou) whereby land requisitioning costs are borne by the new user of the original site, goes a long way towards covering the costs of industrial redevelopment. However, it still leaves the relocating enter- prise with the problem of covering 30 to 50 percent of the costs of reloca- tion. Guangzhou's market-driven approach, discussed below, serves as an alternative worth considering. Industrial Relocation in-Guanfzo 1.51 Guangzhou, like other cities, has conducted environmertal assess- ments of its inner-city factories and identified 14 polluting factories for relocation in 1985-90 and 28 for relocation in 1991-95. Factories targeted for relocation are provided with replacement sites and are compensated for their buildings and immovable facilities. Vacated sites are put up for tender or otherwise allocated and the proceeds are used to finance industrial reloca- tion. 1.52 Enterprises targeted for relocation normally receive financial assistance from the state investment plan. There are four sources of funds to pay for relocation: (i) state funds; (ii) retained earningst (iii) the rele- vant manufacturing or industrial bureau of the municipality; and (iv) bank loans. These channels are similar to those outlined above for Shanghai and Hangzhou. 1.53 As is typical in southern China, industrial redevelopment is, to a relatively large degree, market-driven. In Guangzhou, some industrial users 20/ These for-profit companies were first authorized in 1984, and are dis- cussed in Chapter III. - 24 - in the inner city area have decided to move to suburban areas to acquire more space for expansion. Enterprises may approach the municipality's Construction Commission and request relocation assistance or they may decide to sign coop- erative agreements with one of Guangzhou's numerous REDCs. When industries link up with REDCs, the REDC typically acquires and develops a new site for the industrial enterprise, usually in a suburban district. In exchange for agreeing to relocate, the industry receives a percentage of the profits (usu- ally 50 percent) associated with the redevelopment of the original site, less the costs of the new suburban facility. 1.54 This arrangement is legal only if the factory reports the business arrangement, obtains government approval, and pays a transfer fee to the municipal government. The fee averages about 20 percent of the profit of the project. This fee was imposed two years ago. Through early 1992, it had been levied on 30 to 40 industrial redevelopment projects. Such an approach has considerable merit and has the obvious benefit of being able to finance rede- velopment without the use of massive government subventions. F. Residential Redevelopment Issues 1.55 In market economies, redevelopment (residential or otherwise) occurs only when the revenues generated by building and selling or leasing new space exceeds the cost of site acquisition, clearance and new construction. Where there is government intervention, the government assists in assembling land for redevelopment and provides infrastructure to support and service resi- dences and businesses operating in redeveloped areas. Regardless of the extent of government intervention, current users of the land and property to be redeveloped are compensated for the loss of their property. This compensa- tion is almost always paid in the form of cash compensation, although in some countries, such as Japan and Korea, compensation can be paid in the form of plots of land in the project after the site has been cleared and equipped with new infrastructure services. Rarely is compensation made in terms of in-kind transfers--flats for flats, or shops for shops. 1.56 In China, since the land is owned by the state, it has rarely been transferred for financial consideration, except when long-term leases were made to foreign investors. In such cases, the local government cleared sites and incurred the cost of relocating and resettling former users, with the pro- ceeds from the leases used to cover these costs. 1.57 The situation when inner-city urban land is redeveloped for domestic users has been different. The first and foremost difference is that reloca- tion laws revolve around the principle of in-kind compensation--if a household or business must give up its accommodation, the redeveloper must replace tho space in kind. This concept leads to a very high incidence of on-site reset- tlement of households and businesses. Most cities seem to strive to provide as much on-site relocation as possible; examples abound of virtual 100 percent on-site relocation in Guangzhou, Tianjin, Hangzhou and Shanghai. Other cities, including Beijing, recognize the high costs of in-kind, on-site replacement of housing and have started to develop alternatives aimed at reducing on-site resettlement and increasing the percentage of households willing to move to lower-cost suburban projects. - 25 - 1.58 While reducing the incidence of on-site resettlement works to increase the financial feasibility of redevelopment projects, other government policies also impose financial costs on these projects. In virtually all cities, REDCs charged with carrying out inner-city redevelopment projects have oiten been required to make significant improvements to the base of community facilities provided in old areas (unless the site involved is very small), and to do so without receiving compensation from the district governments to which these facilities are transferred. Such requirements have imposed considerable financial burdens, making it difficult for REDCs to build feasible redevelop- ment projects. 1.59 Another burden most redevelopment projects must shoulder is the pay- ment of fees, taxes and charges. A myriad of exactions is levied on real estate development projects. In Guangzhou, most inner-city real estate devel- opers assume that taxes, fees and charges will comprise at least 20-25 percent of total project costs.2l/ 1.60 The high incidence of in-kind and on-site resettlement, entitle- ments, heavy requirements for the provision of community facilities at no cost to local district governments, and the payment of numerous fees, taxes and charges, makes it difficult for REDCs to undertake redevelopment projects. Unless they are permitted to redevelop cleared sites at substantially higher densities, projects are financially feasible only if the price of commodity housing or commercial space can be aggressively marked up to provide enough revenue to carry the cost of project development, including the provision of replacement housing. In some instances, the associated high markups might not be feasible and the purchasers of commodity housing units pursue other alter- natives. To the extent that these "clawbacks" are due to the fact that the REDCs receive their land outside the market economy. DSavina little or nothing for acquiring rizhts. the issue is best dealt_with by forcing them to switch to leasehold azreements: transparency and fairness is the issue._not the _aX- ment of fees based on the possession of land. G. An Overview of RedeveloRment Prgoects 1.61 This part of the report reviews policies and practices in use in a range of Chinese cities and assesses alternative policy options for improving the financial feasibility urban residential redevelopment. The survey is based on case studies prepared on redevelopment projects in Fuzhou, Guangzhou, Hangzhou, Shanghai and Tianjin. Annex 2 provides detailed information on these projects, which, in spatial terms, have tended to avoid the oldest and densest core areas of the city, except in the case of Fuzhou. 1.62 Over the past several years, missions have gone to Beijing, Fuzhou, Guangzhou, Hangzhou, Shanghai and Tianjin to assess housing and land develop- ment practices and issues. As a result, 11 detailed cases of redevelopment projects have been assembled. These are listed in Table 1.7. Most projects were started between 1985 and 1987 and were completed within the past two to three years. The projects range in size from 0.5 to 29 hectares, and after completion, range in size up to 737,000 m2 of constructed space. There is 21/ This topic is reviewed further in Chapter III. - 26 - considerable variation in terms of the percentage of former residents which were resettled on the site, ranging from 0 to 100 percent. Table 1.7: PROFILE OF REDEVELOPMENT CASE STUDIES Date Site Built area X On-site Project name City started area Old M ~w resettled (ha) (ma) 'ma An Deng Fuzhou 1987 0.5 6,213 0 0 Jin Hua Guangzhou 1986 29.0 335,000 737,000 90 Xiao Fuqing Hangzhou 1985 0.7 6,000 15,417 100 Hu Lang Garden Shanghai 1985 1.7 12,589 84,000 100 Hui Yi Garden Shanghai 1984 1.3 8,630 8,250 0 Jian Guo Shanghai 1985 1.9 26,221 88,400 0 Ordinary Citizen Shanghai 1985 16.5 116,040 325,000 90 Tian He Shanghai - 0.5 13,880 17,200 45 Ying Xiang Shanghai 1985 5.5 33,000 190,000 36 Pingshan Road Tianjin 1985 1.4 8,756 33,967 100 Wujiayao Tianjin 1985 3.4 15,715 61,319 100 Source: Redevelopment project surveys, 1991, 1992. 1.63 Table 1.8 illustrates the project densities before and after rede- velopment. In all but two cases (An Deng, where the site was converted to a park, and Hui Yi Garden, where a high-income project was developed), redevel- opment took place at considerably higher densities. On average, the FAR of the projects increased by an average of 151 percent. The greatest increases took place in Shanghai, where the density of development increased by 567 percent for Hu Lang Garden and 477 percent for Ying Xiang. These project statistics illustrate that developers have been successful in overcoming urban planning policies aimed at reducing inner-city development densities. The higher densities have enabled REDCs to construct and sell commodity housing to finance their projects. Before turning to the actual financial performance of these projects, we need to consider the process and approach of redevelopment. An Overview of Compensation and Relocation Policies 1.64 The most critical determinant of the financial feasibility of rede- velopment projects is how the owners and tenants of demolished housing are compensated. The exact nature of this compensation, in terms of the size and quality of the replacement housing, where such housing is sited and what kinds of temporary accommodation are provided, largely determines the costs of rede- velopment projects. This section outlines relocation policies followed in a variety of Chinese cities through early 1992 (see Annex 3). 1.65 In most instances, but with some exceptions (Box 1.9), the redevel- opment of any site begins with a careful and protracted negotiated assessment of property compensation. While practices vary from city to city, the princi- ples generally are similar, and are dictated by municipal and provincial stat- - 27 - Table 1.8: PRE- AND POST-REDEVELOPMENT FLOOR AREA RATIOS FAR FAR Percent Project City before after increase An Deng Fuzhou 1:1.24 1:0.00 Jin Hua Guangzhou s1:.15 1:2.54 120.9 Xiao Fuqing Hangzhou 1:0.81 1:2.08 156.8 Hu Lang Garden Shanghai 1:0.74 1:4.94 567.6 Hui Yi Garden Shanghai 1:0.66 1:0.63 -4.5 Jian Guo Shanghai 1:1.38 1:4.65 237.0 Ordinary Citizen Shanghai 1:0.70 1:1.97 181.4 Tian He Shanghai 1:2.78 1:3.44 123.7 Ying Xiang Shanghai 1:0.60 1:3.46 476.7 Pingshan Rd. Tianjin 1:0.63 1:2.43 285.7 Wujiayao Tianjin 1:0.46 1:1.80 291.3 Average 1:1.01 1:2.54 151.5 Average excluding An Deng 1:1.11 1:2.79 151.4 Source: Redevelopment project surveys, 1991, 1992. ute. Thw resulting compensation depends on the ownership of the structure to be demolished. In the case of municipally owned properties, the typical pat- tern is for the developer to provide replacement housing. Sometimes the units are located on the site, and in other cases, they are located in a new project area. Box 1.9: THE FUZHOU REDEVELOPMENT MODEL OF RESIDENTIAL NEIGHBORHOODS Fuzhou has tried to overcome some of the hurdles created by conventional housing redevelopment projects, which account for 40-50 percent of total land redeveloped, and thus has Increased the proportion of land redeveloped (as a percent of all new land development) to a relatively high yearly average of over 40 percent. Elsewhere, 25-30 percent is the norm. One option adopted In Fushou is to encourage the redevelopment of sites leased to foreign joint ventures. These areas now account for 30-40 percent of all land redevelop- ment. In addition, in situ "self-help" redevelopment by enterprises or by neighborhood block ls promoted, with coumercial use of ground space permitted as an added incentive. Ten to 20 percent of all redeveloped land now follows this pattern. The conventional system Is thus curtailed, as are the onerous compensation arrangements associated with such planning strategies. Source: Fuzhou State Land Administration Bureau. 1.66 The actual negotiations take place between the REDC and building owners. In the case of private owners, the REDCs negotiate directly with them and payments go to each and every individual owner. Compensation proceeds can be used to purchase a new unit or owners can forgo compensation and, instead, exchange their property ownership right for the right of occupancy in munici- pally owned rental housing. In the case of units owned by enterprises, nego- - 28 - tiations are with the enterprise, not the tenant, and the payments or exchanges take place between the enterprise and thi REDC. After resettlement, the enterprises are free to set new rents. In most instances, new rents increase to partially reflect the higher costs of the units. In cases where the municipality owns the units, the housing authority negotiates with the developer over the level of compensation. Rental rates for the tenants usu- ally remain at the same level on a per ma basis. H. Looking Back: The Government's Rationale for Urban Land Management Reform During the Transition to a Regulated Market Economy 1.67 As in most sectors, the Government's agencies charged with land and real estate management have had no unified vision of issues and options.221 Particularly at the national level, there is no urban development ministry or systems reform commission overseen by a vice-premier that can encourage all relevant actors to agree on what needs to be done and by whom. The problem at the local level is less severe, because "urban construction" is under the supervision of one vice-mayor and because metropolitan areas are administered as single jurisdictions, without competing central cities and independent sub- urbs. Even so, no vice-mayor has the authority to oversee land management, if that rubric is broadly defined to include all the issues and actions required for thoroughgoing systems reform. 1.68 This fragmentation results in a set of partial reform agendas, often working at cross-purposes with one another, and uninformed by the need to overhaul the land management system in a comprehensive manner. What should that agenda consist of, if the overarching objective is to maximize the con- tribution of the real estate sector to the effort to create an internationally competitive network of cities and towns working within a regulated market economy? 1.69 China's overall reform agenda needs to focus on five major objec- tives. First, a framework that encourages rapid, noninflationary growth is needed, along with strong central macroeconomic management institutions to implement it. Second, sectoral policies are required to guide the provision of local and regional infrastructure, so as to lessen the constraints on eco- nomic activity, while promoting the wider use of beneficiary financing. Third, an adequate "safety net," financed by the broadest possible revenue base, must be provided for the unemployed, the sick, the disabled, the retired, and the poor. Fourth, economic transactions should be shaped, to the greatest extent possible, by market signals generated in competitive markets free of barriers to trade. Finally, economic development must take place in a manner consistent with environmental protection. 1.70 These objectives all have implications for urban land management. Macroeconomic considerations require enterprise reforms that force businesses, particularly state-owned enterprises, to operate within hard-budget con- straints, forgoing subsidies such as tax relief and nonpayment of debts. Enterprises must be able to change product lines, the mix of inputs, and loca- tions so as to respond to market demand. Such flexibility would allow exist- 22/ Chapter VI suggests that this situation is now changing. - 29 - ing businesses to relocate from inappropriate locations, in part assisted by the proceeds from the resource value of the older, more valuable sites relin- quished. This would, in turn, allow existing and new Rersonal and nroducer services greater opportunity to expand in the downtown core of China's cities. where such activities are seriously underdeveloped. 1.71 The goal of progressively removing infrastructuro constraints, while devising ways to increase beneficiary contributions to finance the required investments, also has implications for urban land management. The development of urban areas is dependent, in part, on the provision of adequate serviced land and of related trunk infrastructure. The urban planning and budgeting practices followed at the local level have serious implications for economic efficiency as well as consumer welfare. Practices developed for a simpler "command" economy, where local and central government functions were not sharply differentiated, are increasingly out of step with the emerging demand- driven economic system, where local governments are largely self-financed. 1.72 The introduction of "safety net" programs that provide worker hous- ing and social security benefits independent of any given place of employment is a prerequisite for the better use of enterprise assets and the revenues derived from the use of those assets. This, in turn, will allow urban land use rights to business users to be priced and traded like any other commodity, and for "affordability" concerns involved to be set aside as irrelevant. Social protection of family incomes will also enable housing demands to be duly reflected through market signals, rather than treated as a near-free good. With the emergence of explicit market prices for land users, achieving the fourth goal--competitive markets driven by deregulated prices--will be facilitated. 1.73 Environmental considerations require a set of interrelated actions in the field of urban land management. It is important to expedite the relo- cation of polluting factories from residential areas, protect ecologically sensitiva locations, selectively upgrade neighborhoods slated for preservation on historical or cultural grounds, and ensure that peri-urban development does not create land use patterns that threaten unnecessarily the agricultural sector or make the provision of infrastructure excessively costly. 1.74 Overall, as explicit land prices emerge, land users' willingness to pay will depend, in part, on interrelated government decisions on rules affecting property rights, planning restrictions, infrastructure availability, tax laws and regulations, and the availability of up-to-date cadastral infor- mation. 1.75 By this standard. the priorities of kel aovernment agencies to date have been inadeguate. The principal actors in urban land management are SLA, MoC, and the Ministry of Finance (MOF), along with their subordinate agencies at the local level. Clearly involved, as well, are those agencies that super- vise state-owned and collectively owned enterprises in the industrial and services sector, as well as nonprofit institutions (Annex 4). 1.76 MoC, at the central level, and the Construction Commissions, at the local level, has placed highest priority on developing and overseeing the implementation of metropolitan "master plans" that set out the general outline - 30 - of future trunk infrastructure, provide for the pattern of permissible (large- area) land uses, and set limits on the overall expansion of the built-up area. Within this framework, district and subdistrict plans often provide for more exacting use and density controls. The REDCs, as discussed in Chapter III, fall under the ju-isdiction of the Commissions, as does the registration of buildings and the enforcement of building codes. Operating in a system within which land is largely allocatecd administratively, the Construction Commissione and their suporvising Ministry view the actual requisition of a land plot for an activity as only the last step in a process otherwise controlled by them (once SPC and other economic control agencies have approved a given invest- ment). Their top priority, in the reform, has been to continue to do what they have done, but to do so more "scientifically" and to exercise, on occa- sion, more control than in the past. For example, compiling and updating a cadaster of building registrations is identified by these agencies as a reform objective, as is the introduction of GIS to facilitate project preparation, and the eventual provision of zoning or quasi-zoning plans for each part of the city. However, enabling market-led development by redirecting infrastruc- ture investments, providing land users with the right to choose where to locate, letting prices regulate and control land use densities, and facilitat- ing the recycling of land (and limiting pressure on greenfield development) through the creation of redevelopment strategies, are perceived to have lower priority. In sum, these agencies have acted as if conventional land transac- tions, as foL.d in market economies, are likely to emerge only on an excep- tional basis, at least for the foreseeable future. As documented in Chapter VI, though, this "framework" may change dramatically as a result of policy work currently under way. 1.77 SLA, created in 1986 after "exceptionally large" areas of cultivated land were converted to other uses, is the result of a merger of persotnel previously working for construction and agricultural commissions. Their key responsibilities are to monitor ind strictly control land consumption at the project and macro area level, while implementing land requisition procedures within the administratively allocated system. They are also charged with developing and updating land cadasters in urban and rural areas. 1.78 Reflecting the agency's original mandate to control the conversion of rural to urban land, and hence preserve cultivable land; and reflecting, as well, the fact that the staff was recruited largely from specialists in agri- cultural land use, the Administration places high priority on control of "wasteful" land use in rural areas, and the preservation and reclamation of cultivable land, at the project and citywide level, through administratively determined norms and quotas, enforced and monitored at different levels of government (depending on the amount of land requisition involved). Various fees and charges have been introduced to finance land reclamation and deter "excessive" requisition demands. This approach has little "market" content, however, and reflects the concerns of planners in the late 1950e, who viewed "wasteful" land requisition as an administratively created problem, requiring administrative solutions. Until 1992, the Administration measured successful performance principally by the extent to which the enforesment of rules mini- mized the loss of cultivable land. 1.79 SLA also monitors and attempts to control urban land-use conver- sions, though these usually involve a change of building use or ownershilp, - 31 - rather than vacant land (thus placing SLA bureaus in conflict with their Con- struction Commission counterparts). While acknowledging the need to supervise the nascent market in the leasing of land use rights, they have focused more on technical changes (how to prepare lease documents, and how to develop the valuation procedures needed to determine "reservation" prices for leases) than on developing strategies that would promote the rapid emergence of land mar- kets with explicit prices and clear property rights (including the trading of land use rights among existing occupants of administratively allocated land). This too is changing. By 1992, a commitment to promote land leasing as a maior component of reform became clear both at the central level and in selected cities. 1.80 MOF has had two "reform" objectives: to minimize the loss of tax revenues that may result from the introduction of explicit land prices, as these create exemptions that impinge on business enterprise taxable income; and, where land revenues do become explicit, to ensure that these are moni- tored and, wherever possible, shared with the central government. The emerg- ence of land markets, per se, has not been viewed as a high priority except to the extent that reforms enlarge the state's revenue sources. 1.81 As subsequent chapters make clear, in the past these agencies sup- ported a number of "reforms" in cities across China that were disarticulated and did little to advance systems reform as defined in this report. More often than not, "reform" in practice meant greater control over land-use activities, and the refinement of mechanistic tools and techniques, with the hope of im;proving the operation of the administrative allocation system inher- ited from the past. A significant reevaluation of objectives and policies is now taking place, for reasons to be discussed further.23/ 1.82 Why does the "vision" of central agencies really matter? Within the present system, the Center passes general laws or executive regulations that often impose conditions affecting reform. Cities must prepare Master Plans and have these approved by the State Council; they are required to follow land use tax schemes that igfiore local conditions and bar experimentation; land requisition and even investment projects are often subject to central approval; and, until recently, trading of land use rights was forbidden and could not, therefore, allow for the emergence of land markets. Norms and guidelines issued by central urban planners have intimidated local authori- ties, who have often treated these as legally binding. The fact that local authorities actually prepare the detailed implementation regulations, and the fact that these localities have often found creative ways to undermine central supervision, means that some haappropriate practices mandated or suggested by the Center have been modified to fit local "realities." But these local adap- 23/ SLA has produced an internal report (dated October 1991) identifying the need to administer the urban land management process in a holistic fashion. The State Council (Cabinet) Policy Research Office has also produced a new report on land reform, whose contents are confidential. MoC expects to produce a similar type of report by June 1992, under the leadership of the Center for Policy Research. Other key players, includ- ing MOF, have so far failed to take active policy research initiatives. - 32 - tations are often illegal and may subject local authorities to disciplinary action. This is hardly a healthy state of affairs. 1.83 What should the Center's role be? Simply put, to promote regulated autonomy. Issues such as who really holds title to the state land, and how the roles of MoC a-id SLA are to be defined, require central direction. The fiscal framework in place must be clarified, as must regulations involving the environment. Beyond that, the Center should remove obstacles to market reform and deregulate urban planning on a selective basis (e.g., eliminate the requirement that cities produce static Master Plans subject to State Council approval; concentrate on disseminating best practices gathered from domestic and international experience, with a competitive local economy as a key objec- tive; while strengthening township planning regulations in peri-urban areas, now subject to limited supervision). 1.84 Subsequent chapters first review the theory and international expe- rience on the functioning of urban land markets. They then discuss in more detail what has been accomplished to date under the broad rubric of urban land management reform, what reassessments t%ie Chinese authorities have themselves begun to make, and what remains to be done. The range of reforms involved is large and interrelated. Prices and property rights must be redefined, laws must be clarified, planning regulations must be reviewed, and the whole pro- cess of urban land use planning and legal and planning institutions must be extensively reformed. - 33 - II. URBAN LAND PRICE BEHAVIOR IN A MARKET ECONOMY: AN INTRODUCTION A. Economic Model of Land Price Behavior 2.1 This section outlines a relatively simple economic "model" of how land prices in different locations are determined by market forces--a model on which general consensus among economists would exist. This model would be applicable to a regulated market economy in China. The second section of the chapter reviews available evidence on the actual behavior of land prices in a range of economies around the world. 2.2 In simple terms, the market price of any productive assets--includ- ing land--will be determined by the expected future flow of income, or returns, to be yielded by the asset. The technique of discounting, using an appropriate interest rate, is applied to convert the future rever.ue stream into the "present value" of the asset.l/ 2.3 Postponing, for the moment, the issue of the spatial pattern of land prices, many factors affect the price of land. Whoever invests in land does so with alternative investment opportunities in mind. This means that land prices respond to the paucity or abundance of attractive alternative invest- ments, the tax treatment of all such investment opportunities, and expecta- tions about the yields of such alternatives over a long period of time. If land supplies available for urban development are severely restricted, rela- tive to expected demand, or if financial assets yield relatively unattractive rates of return, then the expected annual yields from land may be relatively enormous and prices will be commensurately large. If, on the other hand, developable land supplies are viewed as relatively abundant and if alternate investment, after taxes, are viewed as competitive, then prices and their rate of increase will be more moderate in scope. 2.4 Tokyo and Seoul provide excellent examples of urban land price trends influenced by factors involving a number of apparently unrelated deci- sions. Yukio Nogushi compared the price of inner-city sites in London and Tokyo and found that, per m2, Tokyo's prices were 40 times higher than 1/ The price of land, then, is the sum of actual or unrealized earnings from that land, per time period, discounted in each case by an appropriate discount factor that translates these into "present worth or value." Expressed as a formula, (l)Po=e Rt t=0 (1+1) where Po - present price, Rt - earnings in any given year, t - the number of years involved (here expressed as o - the present and - - infinity) i - discount or interest rate considered appropriate. - 34 - London's.2/ To a lesser extent, the same proved true for commercial proper- ties. However, annual per m2 rents differed by only a factor of two. What appeared to be at work was a pattern of annual rental increases that rose in step with the (differential) growth of the two economies, while land prices in Tokyo seemed very susceptible to "speculative bubbles" that allowed the ratio of prices to rents (i.e., the flow of expected future rents and the expected level of interest rates utilized to discount such rents into "present values') to deviate dramatically between the two cities, particularly during the period 1986-89, when Japan's interest rates fell sharply. Japanese bank practices, willingness to lend heavily against land used as collateral valued at ever- rising levels, added fuel to the speculative frenzy. This exacerbated the impact of other local constraints operating in Tokyo, including the relatively large amount of land within the city subject to explicit (planning regula- tions) or implicit (tax regulations involving land) penalties discouraging rural-urban conversion. Land prices grew at unprecedented rates. Land price trends in Seoul reflect the complementary impact of local planning practices, restricting serviced land development described above, and financial sector policies that provided relatively low after-tax rates of return for those holding financial assets, thus inducing investors to readjust their wealth portfolio toward real estate acquisition.3/ As a result, Seoul's average annual land price increases far exceeded such predictors as the growth rate of nominal GDP in the municipality. 2.5 Certain underlying facts cannot be denied, however. Land price increases are a normal and expected feature of urban development. Using data from land markets in developed economies, where constre4nts are relatively small and alternative investment opportunities are relatively ample, one can confidently assert that land values will grow, in nominal terms, at least as fast as economic development pressures dictate. If an urban economy expands at 10 percent per year, then urban land prices will almost inevitably respond to that by growing by at least that rate. By contrast, in a stagnant economy, expanding at only 1 or 2 percent per year, or experiencing a recession where output is actually declining, then, all other things being equal, land prices will mimic the underlying economic development "fundamentals" and increase more slowly. 2.6 Similarly, if an economy undergoes a fundamental set of "shocks," including a massive increase in remittances from abroad, a drastic change in the rate of return on financial instruments, or a sudden change in the "rules of the game" under which land is made available for development, then inves- tors will react rapidly, creating land "booms or busts" that would only sur- prise urban policymakers ignorant of the interconnected nature of the urban economy. The worst possible response to such a "crisis" is to view the evalu- ation of land prices as an isolated "land manazement" issue. when. in fact. a far wider aaenda is at stake. 2/ Yukio Nogushi, "Land Problems in Japan," Hitotsubashi Journal of Econom- iCs, Vol. 31, No. 2, December 1990, pp. 73-86. 31 B. Renaud, ComDounding Financial Repression with Rigid Urban Reaulationst Lessons of the Korea HousinR Market, World Bank, Report No. INU21, June 1988. - 35 - 2.7 For the sake of simplicity, land prices can be viewed as having two basic components: current use value and potential reuse value (capital gains). Owners of a nonproductive resource, such as vacant land, must be compensated for not developing (or transferring for development) land in the present period. That compensation comes in the form of as-yet-unrealized "capital gains" that emerge in monetary form when the land is sold. Developed land yields explicit or implicit use rents. In between, land can appreciate because realizable rents And future appreciation through resale will enter into value calculations. 2.8 From a spatial perspective, capital gains drive land prices at the urban periphery, while the s.ream of rents is the key determinant of land in developed areas. This leads one directly to the need to differentiate between the land values across space as well as the value changes at different dis- tances from the CBD. 2.9 Implicit in the discussion that follows is the assumption that, unlike sectors such as agriculture, urban development follows fairly standard spatial patterns. Cities around the world are basically alike, under market- type incentive structures. The reason for this phenomenon has never been fully explained, but one plausible hypothesis can be labeled as the "Ingram paradigm."4/ The first assumption ii that the prices of tradeable goods produced in any city reflect world prices, and that the wages paid workers producing such goods are a function of both those uniform prices and of the technology employed, which can vary greatly. Second, the wages of workers producing nontradeable goods are a function of prevailing wages in the trade- able goods sector; the higher the latter, the higher the former. Third, the prices of nontraded goods and services is a function of the wages of workers of nontraded goods and services, with very little leeway provided in terms of technology used, which thus becomes a minor explanatory variable. Fourth, urban infrastructure technology (particularly for large, densely populated cities) has the characteristics of nontraded goods, and is similar across cities around the world. Where real estate and urban infrastructure are involved, the ratio of output prices, input prices and incomes are roughly the same around the world, and vary only in absolute levels. Thus, the behavioral responses expected of producers and consumers are roughly the same around the world, since relative price signals and incomes are similar. Chinese cities look very much like American cities, while American farms and Chinese farms have spatial and technology characteristics that are polar opposites. One can thus generalize about cities and label certain types of outcomes as policy- distorted much more easily when discussing city st tures than when reviewing other types of spatial behavior. 2.10 The easiest way to illustrate this phenomenon is to assume that, within cities, land values and activity densities are radially symmetrical around the city core. Simply put, this allows the at,alyst to collapse a three-dimensional story into a two-dimensional graph that features prices on the vertical axis and distance from the CBD on the horizontal axis. Experi- A/ Though never published, Gregory Ingram developed a hypothesis several years ago, and graciously shared his assumptions with the sector study team. - 36 - ence suggests that the best mathematical expression of land price behavior, under those assumptions, is a negative exponential function; which assumes a constant percentage decline in price per unit of distance from the CBD center: (2) V,=Vo e-^x where V;, level of activity or price at distance "x" from the CBD, Vo - level at the city center, b - gradient value representing the proportional change in V for every one-unit change in "x". 2.:11 Using this formula, one can demonstrate a number of empirical facts: (a) The gradient can typically be described as a concave curve peaking at the CBD. (b) The gradient flattens out over time, even if the overall absolute average value of prices is increasing. (c) That flattening process reflects both the expansion of the city and the relatively more rapid appreciation of land prices in the periph- ery versus the CBD. (d) The gradient declines more slowly in big cities than in smaller ones, implying that the "urban field" of a smaller city ends more abruptly, while that of a metropolitan area obviously covers a larger radius. (e) The gradient of "activity" (population, employment) is closely linked to the land price gradient, though land value gradients are consistently less steep than "activity" gradients, implying that the user response to price increases (in terms of land use density) is less than 1:1. In other words, the "elasticity of substitution between land and structures is less than unitary."5/ (f) Further empirical work suggests that "activity" response to land price gradients is not instantaneous, and that the lags in adjust- ment are large enough (data from the United States and Japan suggest that only about 10 percent of any disequilibrium in the population 5/ Estimates of such elasticity of substitution have been calculated for various cities. In Seoul, within the housing sector, the elasticities range from 0.5 to 0.8 (L. Hannah, K. Kim, and E. Mills, "Land Use Con- trols and Housing Prices in Korea," World Bank, February 1991, pro- cessed). For Bogot&, a similar rate, around 0.7, was documented in D. Dowall and T. Treffeisen, "Spatial Transformation in Cities in the Developing World: Multinucleation and Land-Capital Substitution in Bogot&, Colombia," Rezional Science and Urban Economics, Vol. 21, 1991. In the United States, values of 0.8 are commonly cited as appropriate "stylized facts." - 37 - gradient is corrected each year)6/ that one can expect investment behavior, resulting in job and residential locations/relocations, with use density consequences, to continue for several years, reRardless of any other changes in underlvinz economic or Policy variables. Population and employment locations and densities are always "catching up" with a "moving target" equilibrium land price gradient that leads the way. The rate of adjustment can be acceler- ated or delayed by policy decisions, often made by decisionmakers in the public sector who have little understanding or interest in real estate development. 2.12 Work on a number of large cities across Asia and elsewhere found the difference in land prices between average prime commercial land in the CBD core and exurban land not yet been considered ready for urban development to be roughly 100:1, though CBD prices show enough dispersion that the sample chosen to calculate "the" prime average price will influence the ratio.7/ In cases whe:e impediments to land development are relatively few, exurban land not yet deemed for inclusion into the urban development stream can be assigned an index number of 10. The same land, once considered developable for urban purposes, would have an index number of 30; and, once zoned for officially sanctioned development, the number can reach 60 to 70, while the provision of services will double or triple the figure to an index number of 150-200. By contrast, the average value of prime commercial space would hover around an index number of 800 to 1,000. 2.13 By implication, the average annual increase in land prices reflects vastly different growth patterns in the rate of realized or unrealized land values at different distances from CBD (as well as dispersion in increases at any given radius from the CBD). Much of policymakers' concern about urban land prices typically lies less in what is occurring in and around the CBD than in what is occurring in the suburban areas undergoing the transition from rural to low- and medium-level density development. Failure to understand the fundamentals behind land price increases can lead to misguided land price control interventions (Box 2.1). 6/ E. Mills and K. Ohta, "Urbanization and Urban Problems," in H. Thomas and H. Rosavsky, eds., Asia's New Giant--How the Japanese Economy Works, Washington, DC: Brookings Institution, 1976. The adjustment rate in China, given existing policy and institutional constraints, is probably much slower. 7/ The actual ratios cited are based on field work by David Dowall and col- leagues in various Asian countries. However, in cities like Tokyo, sub- ject to a mass of disincentives to services land development, the ratio between land around Shinjuku Station and land 50 to 60 kilometers away is 350:1. (G. Tolley, "Urban Land Prices and Land Use in Market Economies," 1992 processed.) - 38 - Box 2.1: THE RELATIVE ADVANTAGES OF CBD AND SUBURBAN LOCATICOS The CBD has one enduring advantage, traced back to the time when urban areas emerged as major centers of economic power. Central locations appeal to economic activities that need face-to-face communications with customers or suppliers, given the state of commu- nications and transport technology in place. In developed countries, these centralized activities are now generally restricted to executive functions, dealing with nonroutine decisions that often require rapid updating of information based on instant access to exter- nal expert professional advice. These are largely service sector activities and include certain functions related to government, law, accounting, insurance, banking, trading, advertising, special-interest associations, consultancy, and related support services (hotels, restaurants, entertainment). Some specialized types of manufacturing activity can also benefit from core locations because of the need to have access to "external economies" that are reaped by proximity to other types of nearby firms. These include firma producing unstandardized goods, continually changing products, uslg little space, and able to avoid significant commitments to fixed capital investments. These enterprises share an overriding characteristics instability and uncertainty associated with swiftly changing product lines, driven by highly unpredictable demand. High-fashion clothing and publishing are but two examples. In each case, these activities must be able to tap a wide pool of nearby skills and facilities on very short notice, and do so in locations where sufficient demand exists to make such ancillary suppliers available at relatively low cost. Given swiftly changing product lines, such businesses cannot afford to invest in self-contained, large plants filled with extensive equipment. They do not require large amounts of inexpensive land. Instead, these management-litensive operations rely heavily on outside suppliers for goods or services, supplies that typically demand face-to-face communication to specify the exact nature of the specified inputs required. Letters, telexes, facsimiles, or telephone calls ulone caunot, in these cases, adequately satisfy the requirements involved ln these buyer- seller relationships. Physical proximity, or clustering, is an economic necessity. Why then does decentralization proceed across the world's cities? Because, for many activities, the elements of uncertainty and the need for frequent and speedy communica- tion, as defined above, have declined or have simply become lnapplicable. The revolntion in infrastructure technology and the massive investments that have accompanied it, primarily in transportation and communications, have rendered CBD locations less relevant, releasing its land for the new uses created by the growing producer services sector. Suburban locations, in most instances, are vastly less expensive places to locate Industrial plants and residences, along with many personal service activities. In most developed countries, even exurban locations in metropolitan areas or relatively isolated regional centers (Omaha, Nebraska in the field of telemarketing, credit card processing, and hotel reservations) can depend on the nation's accumulated stock of regional trunk infra- structure investments to facilitate rapid development. In such suburban, exurban and regionally isolated environments, the population's choices concerning preferred residential locations coincides with that of employers. Suburban and exurban locations allow consumers access to relatively inexpensive residential amenities, including housing space, yards, and recreational opportunities that cannot be matched by close-in locations. The population that remains clustered in and around the CBD has become increasingly idiosyncratic: "foot- loose" households that place great emphasis on the core-area man-made recreational opportu- nities and the ethnic diversity available in and around the CBD, along with others who weigh the spatial Implications of policy distortions (rent control in central areas; planner restrictions on low-income, single-family or apartment dwellings in the suburbs) and decide that suburbanization is not a viable option. B. Land Price Behavior Across Socialist and Market Economies: Evidence from Larae Cities 2.14 Where dramatic deviations from the above stylized facts are observed, they can generally be linked very closely to the impact of massive planner-led distortions on land markets. For example, de facto public control - 39 - over recent land development,8/ and the simultaneous imposition of a strictly controlled wide greenbelt around Seoul have raised developable subur- ban land prices to levels not dissimilar from those in core areas. The price density gradient is, essentially, flat and thus incentives to redevelop the CBD, any other factors aside, are dramatically diminished by the inability of original occupants to acquire the rights to consume more land, per unit of expenditure, in the periphery (Figures 2.1 and 2.2). 2.15 The juxtaposition, in turn, of the population density gradients of Paris, four Chinese cities and Moscow, measured from the CBD, reveals that Paris' pattern is very much in accord with the stylized facts (Figures 2.4 and 2.4). St. Petersburg and Moscow, by contrast, have positive (and counterin- tuitive) density gradients which reflect a land development policy that has ignored inherent land values in determining the location of activity over several decades (Figures 2.5-2.8). In particular, note should be taken that St. Petersburg, Moscow, and Warsaw deliberately reserved large amounts of prime land, close to the CBD, for low-value nonresidential uses (primarily industrial and warehouse), invested enormous amounts of money in trunk infra- structure linking the core to the distant suburbs, and then drove large num- bers of core-area residents to peripheral locations, where combinations of walk-up apartments and high-rise structures were erected in configurations exactly like those found in close-in residential neighborhoods outside the historic core. The consequences of this were not merely aesthetic. This deliberate disregard for consumer preferences, and for the necessary trade- off a between loss of accessibility and improved amenities, including greater residential space at distant locations, meant that the "market value" of the more distant communities probably does not cover their investment costs. Neflect of elementary economics has created a potentially large stock of Rrop- erties yielding neRative value addedi Box 2.2 suggests that even some Western economies suffer from excessive regulation and uninformed urban planning. 2.16 Perhaps surprisingly, Figure 2.3 also illustrates the fact that some Chinese cities, in spite of their exposure to 40 years of "Socialist" urban construction policies, resemble market-economy outcomes in some respects, and have not merely reproduced the anomalous patterns found in the Russian Federa- tion or Poland. The reasons for these divergences are explored further in Chapters III and IV. 2.17 Interestingly enough, during the last five years, researchers in command economies undergoing the transition to a regulated market have been able to mimic the stylized facts referred to above by the use of "scientific" formulas that give varying weights to accessibility, amenities, and infra- structure availability. They began by noticing that urban land price gradi- ents in Chinese cities, around 1950, corresponded exactly with the predictions of the simple model discussed earlier. In Beijing, for example, land prices per m2 in 1950, averaged across 18 land classes, varying from most desirable and accessible to those outside the area of immediate urbanization pressure, 8/ Renaud reports that in 1989 only 22 percent of residential development in Seoul is controlled by the private sector (B. Renaud, "Confronting a Distorted Housing Market: Can Korean Policies Break With the PasLS" op. cit.). - 40 - Box 2.2: STOCKHOLM'S HOUSING RENTAL PRICE GRADIENT Stockholm operates largely within a municipally controlled land lease system. Land for residential development is leased largely to nonprofit municipal corporations at nominal rates, who build rental units. The Municipality can force housing development to occur in areas which would be rejected as undesirable If "market" rules applied and develop- ers had a greater choice of sites. Construction and rental or homebuyer mortgage finance is heavily subsidized. However. the building materials industrv is oraanized along oligopolis- tic lines and is unregulated, charging "market" prices for its Droducts. Rents are set yearly so that the average rent per m' for the pooled stock of any one development company cowers Its costa and the company gains no "profit." Clearly, different companies have hous- ing stocks of different vintage. Newer companies, whose stock is relatively new and, almost always, built in outlying areas, have much higher average costs than older companies, whose stock is largely concentrated in the core of the city. The rent-setting rules, set on a company-by-company basis, yields a perverse result: apartments that are nearly identical ln characteristics, except location, carry very different rents. A unit in central Stockholm is cheaper than one built (more recently) in the suburban areas. The price gradient thus behaves in a perverse way, contradicting the presumptions of accepted land use models created with market economies in mind. As in the Post-Socialist economies, deregulation of prices would vastly increase the value of older, core-centered housing and yield many subur- ban development unmarketable except at deep discounts that might not even recover investment costs. yielded a predictable land gradient and a ratio of 100:1. Data from Guangzhou and Tianjin provided roughly similar orders of magnitude.9/ 2.18 In China, SLA published "Procedures for Land Classification in Cities and Towns" (1989), to direct the classification of land values across China's urban areas. Each land parcel is classified by aggregating a large set of land-value-determining elements, weighted through a variety of methods ranging from the intuitive to the "scientific," using regression analysis. The key elements included various obvious factors such as transport condi- tions, infrastructure services, environmental conditions, commercial develop- ment, and population density. As summarized by the urban land management report prepared by the Chinese Academy of Social Sciences, results from cities like Shanghai, Beijing, Tianjin, Nanjing, and Ningbo all exhibit the negative land price gradient pattern predicted by the theoretical discussion and market signals summarized above.10/ St. Petersburg also performed a similar exer- cise in 1992 and, again, the synthetic land price gradient is consistent with outcomes found in market economies, though few would agree that the implicit extremely high absolute values would be "validated" by market transactions negotiated freely among individuals and/or enterprises facing current hard budget constraints.11/ 2.19 Unfortunately, these exercises, while useful introductions to the subject, are misguided if they assume that planners can estimate absolute land 9/ State Land Administration. 10/ Institute of Finance and Trade Economics, Chinese Academy of Social Sci- ences, Urban Land Use and Nanaeement in China, draft final report (pro- cessed), Beijing, June 1991. ll/ Alan Bertaud, personal communication. - 41 - prices. varcel-bv-2arcel. by the use of formulas and if these exercises are not constantly updated to take into account real transaction data (Box 2.3). Land prices, regulatory and policy impediments aside, can only emerge sponta- neously within market economy frameworks, as individual sellers and buyers bargain with one another in much the same way that other commodity markets operate. Policymakers in post-command economies are spending an inordinate amount of time trying to outguess market outcomes. Instead, as Chapter III suggests, they should be focusing their attention on facilitating the sponta- neous emergence of market prices, and enabling existing users of property to release plots, where potential gains to existing users coincide with the expected gains to new occupants. This conclusion appears to be gaining accep- tance in China's coastal cities, after the rapid increase in land-lease activ- ity proved the limited utility of standard land prices as a predictor of "real world" outcomes. The top priority of systemic reform in urban land markets lies in price and property rights reform. There are no shortcuts. Box 2.3: STANDARD LAND PRICES IN CHINESE CITIES In the case of China, the so-called standard land prices are set across five to eight zones, any within each zone, prices are set at different levels depending on current use. A commercial enterprise set next door to a residential user may carry a differential theoretical price burden which is In the neighborhood of 25:1. Unless these "pricese are used as very rough guidelines, sensitive to the fact that land prices (in a market economy) can drop precipitously from a high-access road to a more isolated land site one block away, Ozone" considerations aside, they will add little information and provide planners and regu- lators with bogus specificity with respect to lane- transactions. These exercises are acceptable as academic experlments. They may even alert urban planners that in high-access zones and locations *here are many users whose relocation would be desirable. - 42 - 111. FIRST THINGS FIRST: PRICES AND PROPERTY RIGHTS A. Do Urban Land Prices Exist in China? 3.1 The prevailing model of land allocation in urban China evolved dur- ing the 1950., following the establishment of the People's Republic in 1949. The most simplistic descriptions of the process of land allocation have implied that urban land became a free good, and that this fact is still true today, complicating any move from a command economy to a regulated market economy, within which land would become a tradeable commodity like any other. The truth is that, while the process followed prior to the introduction of urban economic reform in 1984 may be interpreted as confirming such a hypothe- sis, the reality today is different and land in urban areas increasingly has a "value" ranging from plots with prices revealed by "black market" property transactions; to those that have prices that are more clearly sanctioned (land lease contracts); and those that can be described as quasi-market prices in nature (land requisition from rural communities; compensation paid to those land users involved in reuse and redevelopment projects; and the value of land assets transferred by bankrupt enterprises in the process of merger with prof- itable enterprises). Today, real estate market researchers can draw rough actual land value maps for any city in China, on the basis of the indicators listed above. At the margin, market forces are emerging, even if the scope of the market is limited to particular cities or areas of cities, and especially if stringent definitions of "market prices" (one-on-one deregulated transac- tions) are used as the standard of evaluation. 3.2 Furthermore, along with these price signals, there is a real estate market constituency in place, ready to promote change. In China's cities, there is an active real estate development market emerging, with a small but increasing proportion of property developed on a for-profit basis by domestic and foreign-funded business enterprises; and a growing number of occupants ready to acquire that property on commercial terms. The key issue. then. is not how to invent prices for a free good but how to introduce mechanisms that expand the scone of exDlicit market Drices. and orovide for the Dro2erty rights that must accompanv them, allowing these to emerze as expeditiouslv as Dossible. B. The Abolition of Urban Land Prices and Propertv Rights: A Retrospective Look 3.3 The 1982 Constitution, written prior to the liberalization of the urban economy, nationalized all urban land, leaving rural land in the hands of the local "collectives" that emerged in the aftermath of the collapse of the "commune' system. In fact, private urban property rights, except for those over pre-1949 individually owned housing stock, ceased after 1953, in the political upheaval that coincided with the Korean War. At that stage, private land was first converted into a joint state-private ownership system and then nationalized by the end of 1956. 3.4 Private ownership of rural land, including that in the periphery of the preexisting built-up areas of the cities, was also terminated after 1949. In 1950 the Chinese Agrarian (Land) Reform Law formally confiscated land held - 43 - by "the landlord class" and vested property in the hands of the cultivating farmers, who were promised title deeds and the right to manage, buy, sell or even rent out land freely, subject only to Government's right to requisition land for urban uses. These rights were reaffirmed in the 1954 Constitution. However, subsequent campaigns led progressively to the "cooperative" movement, when private rural ownership rights were restricted to small plots surrounding farmer housing. This was followed, in 1957, by the creation of People's Com- munes, which effectively eliminated private property in the countryside, including rights over housing and the surrounding small private plots. 3.5 The state continued to claim formal title only to urban and indus- trial land, but developed procedures for expropriating rural land, as needed for urban development, with compensation paid to former occupants. These procedures, issued by the central Government Administrative Council in 1953 and 1954, were confined to basic guidelines that called for state construction units I/ to acquire land only as needed, and to avoid encroaching on land suitable for agricultural purposes unless necessary. Other than requiring nominal compensation for land taken, the new users were relieved of the obli- gation to pay fees for using the newly acquired land. And even the funds used to pay for expropriation were covered out of the state budget, freeing benefi- ciaries of the constraints imposed by "hard budgets." Actual implementation of these procedures was divided between local authorities and the central industrial ministries, acting individually under the weak supervision of the Government Administrative Council and the People's Committees (governments) at the local level. 3.6 Following Soviet practice, the relevant supervising economic plan- ning authorities first approved the investment plans of subordinate enter- prises and institutions, and then authorized local land administrative bureaus to acquire land, turning it over to the user free of charge and for perpetu- ity, but with no rights to exercise the unauthorized transfer of such land to third parties. 3.7 Examples quickly proliferated of waste and inefficiency, as users demanded and frequently received far more land than justified by their invest- ment plans. In addition, driven by the desire to create more attractive and "modern socialist cities," local authorities set aside large areas of land for open space, recreational facilities, and large public buildings, and did so j/ In effect, this covered most urban expansion, since private development was quickly limited in scope, and ultimately eliminated. - 44 - well ahead of any conceivable mid-term implementation plan.2/ The problems were first evident among the so-called 18 key cities I/ designated for intensive industrial development during the First Five-Year Plan (1952-57). Many of these were cities with little prior industrial development history. Their built-up areas expanded at often enormous rates, with Zhengzhou, for example, increasing in size ninefold between 1949 and 1957. By the mid-1950s, the central government decided to utilize more fully the existing industrial capacity of older industrial cities, and suburban land acquisition sprees spread to cities like Guangzhou and even Shanghai. Table 3.1 illustrates selective instances of built-up area expansion during this period. 3.8 At the microlevel, enterprises and institutions often squandered land, driven by a desire to build up future reserves; guided by Soviet norms and standards that promoted low floor/land area ratios (FAR); encouraged to establish buffer zones between factories and other uses; allowed to set up "security zones" around their plants; and permitted to build self-contained communities that ignored the potential offered by existing nearby city public and commercial facilities. 3.9 The distortions permitted and encouraged by this approach soon began to receive widespread publicity. Typically, these individual case studies reported that only 20 percent or less of the land area requisitioned was actu- ally occupied by structures. A 1956 central government assessment of land requisition in selected cities, including Beijing, Changsha, Hangzhou, and Chengdu, reported that 40 percent of the requisitioned areas were taken with- out justification, even by the nonmarket standards of that date. Central authorities responded in 1958 with new legislation amending earlier land 2/ "During the First Five-Year Plan, model plans of large Soviet cities were universally adopted for the planning of all Chinese cities, despite wide variations between Russian and Chinese urban centers with regard to size, type, geographical location, terrain characteristics and cultural attri- butes. Thus, even the plan for a small workers' town of less than 20,000 inhabitants would include all the urban features and amenities of a large city. These included large municipal public squares with wide central boulevards, east-west axis roads, civic centers for municipal government buildings, district ;3rks, and several sports stadiums. Also, public squares, a typical feature in Soviet cities, became a basic land use characteristic in many plans for Chinese cities. These included munici- pal squares, district central squares, traffic squares and other squares of special types. All of them occupied large amounts of land. For exam- ple, the area of the municipal central squares in the urban plans of Lanchow, Loyang and Harbin exceeded the 9 hectare Tiananmen Square in Beijing. In some cities, the district central squares were planned to be larger than Red Square in Moscow!" (Source: K. Fung, "The Spatial Development of Shanghai," in C. Howe, ed., Shanghai: Revolution and Development in An Asian Metropolis, New York: Cambridge University Press, 1982.) 3/ Beijing, Chengdu, Luoyang, Zhengzhou, Wuhan, Xi'an, Zuzhou, Shenyang, Anshan, Datong, Shijiazhuang, Qiqihar, Harbin, Jilin, Changchun, Lanzhou, Taiyuan, and Baotou. - 45 - Table 3.1: CHANGE IN THE SIZE OF BUILT-UP AREAS IN SELECTED CITIES: 1949-57 (km2) 1949 1957 Z growth Beijing 109 221 103 Xi'an 13 65 392 Zhengzhou 5 52 900 Hefei 5 57 1,007 Wuhan 34 (1953) 130 (1959) 382 Jinan 23 37 61 Shanghai 80 116 45 Tianjin 61 97 59 Nanjing 42 54 29 Fuzhou 11 19 (1960) 73 Source: State Land Administration. administration regulations, increasing administrative controls over expropria- tions and individual project requirements. Among the changes introduced, the most significant were the following: (a) the enterprises' request for land (size, locetion) had to be approved by the same institution responsible for evaluating and approving the investment projects; (b) local governments were given greater discretion in both controlling land allocation and supervising the use of land after expropriation; and (c) the amount of land that any given level of government could allocate to a project was restricted by physical quotas, so that large land allocations would need central government approval. 3.10 Over the next two decades, particularly through 1975, the expansion of cities built along Soviet lines slowed dramatically, a factor that should be kept in mind when comparing the characteristic features of Socialist cities across countries (outside China, massive expansion continued). Tables 3.2 and 3.3 provide data for one so-called key city, Beijing, and one traditional coastal center, Guangzhou. Land expropriation in Beijing, which had totaled 205 km2 from 1952 to 1962, fell to only 35 km2 between 1962 and 1980. In Guangzhou, land requisition between 1951 and 1960 totaled 8,119 ha, but between 1961 and 1974, dropped to 2,343 ha. Lack of systematic data from other cities prevents generalization, but anecdotal evidence suggests that trend was national in scope. Recently available data provided by SLA summa- rize the best estimates of cultivable land losses since 1957, and include not only strictly urban activities, but also all other sources of demand, includ- ing the growth of the built-up areas of rural communities, the requirements of stand-alone mining districts, and the land taken for interregional infrastruc- ture investments. The data are accompanied by no systematic explanation. They support the hypotheses of the report, however: (a) relatively unfet- tered, land-consuming development during the first decade of Socialist trans- formation; (b) a systematic antidevelopmental strategy during the 1960s and 1970s; (c) a return to unfetteeed growth in the late 1970s and early 1980s; and (d) expansion more mindful of the implicit and explicit cost of land in - 46 - Table 3.2s FARMLAND EXPROPRIATION IN BEIJING: 1949-85 Expropriated Increased Ratio of devel- farmland urban built- oped land versus in suburbs up /a area acquired land om=2) (0=2) (2) 1949-52 42.11 17.60 41.9 1952-57 106.52 53.80 50.5 1958-62 98.47 64.90 65.9 1962-65 5.35 9.60 178.9 1966-70 8.78 16.60 188.7 1971-75 10.56 17.60 167.0 1976-80 10.03 18.80 187.3 1981-85 23.30 24.60 105.7 Total 305.10 223.50 73.3 /a Developed. 'Sourcest Beijing Academy of Social Sciences, Urban Institute, "A Study on the Urbanization Process in Beijing Near Suburbs," 1989; and Yu Xuewen, 1986, "Analysis of Urban Construction Land Development in Beijing by Applying Remote-Sensing Technique," City Plannin2 Review, 1986-3: 9-14. Table 3.3: URBAN LAND REQUISITION IN GUANGZHOU (1951-90) Total Of which: Percent of Year land area arable land arable land (ha) (ha) (Z) 1951-55 2,240.12 642.30 28.67 1956-60 5,879.05 3,652.79 62.13 1061-65 832.21 413.57 49.69 1966-70 860.20 234.40 27.24 1971-75 1,054.33 379.59 36.00 1976-80 1,752.90 950.59 754.22 1981-85 2,937.27 953.97 32.47 1986-90 2,645.39 1,808.81 68.37 Total 18.201.48 9,036.02 49.60 Source: C,uangzhou Real Estate Bureau. - 47 - the period after 1984, driven by rising costs, profit-maxlmizing behavior, and the reimposition of more effective administrative controls.-'/ 3.11 To aome degree, the decline in land requisition after 1960 can undoubtedly be ascribed to the refinement of bureaucratic procedures; however, the wholesale relocation of urban land use planning personnel to farms and factories during the Cultural Revolution period (1966-76) makes this unlikely as a primary cause.5/ Ins&ead, the trends are best explained by the drastic decline in "capital construction" during the period in question, combined with an assignment of roughly half those fixed asset investments to the development of "Third Line" factory centers in very remote regions of China, meant to preserve core industrial capacity in case of a nuclear attack aimed at the preexisting industrial centers.6/ 3.12 The impact created by the period of Soviet-style planning methods was evident beyond greenfield areas. In the name of renewal, some cities, particularly Beijing, were stripped of significant portions of their inherited architectural heritage (though most cities were spared, more through neglect or lack of resources than by c-nscious intent, this force-fed redevelopment). Tiziano Terzani reports that almost half of the surface of 1949 Beijing was transformed, as traditional vialos (arches in marble and painted wood built across streets) and the inner and outer great city wall- were systematically destroyed during the 19509 and 19608.7/ The princely palaces (wang fu) were 41 According to SLA sources, net logs after reclamation fell from 15.1 mil- lion mu in 1985 to 1.0 million mu in 1990. Whereas between 1979 and 1985, 234 million mu of arable land (after reclamation) were taken out of production, only 34.8 million mu suffered the same fate in the period 1985-90. One mu corresponds to 667 e. t/ Hong Kong analysts, in a recent comprehensive report on urban land man- agement practices su6gest that during the period covering 1966-76, "... urban land planning and land management organs were rescinded and the urban planners and property managers were either transferred to a lower level or compelled to do manual labor in the countryside or in a factory. Laws and regulations ... (concerning] urban planning and management were criticized as being creatures of the bourgeoisie" (J. Ratcliffe, S. Tsui, and L. Yu, Land Manateient in the People's Republic of Chinas A Research Report on the Present System of Land Manaaement in China, Hong Kong: Hong Kong Polytechnic, April 1990). 6/ Other assessments of the extent of Third Line investments are even more dramatic. One recent report suggests that from 1960 to the mid-1970s, 73 percent of state investment in capital construction went to these remote locations, resulting in the establishment of 29,030 state enter- prises employing one third of the state-owned workers once the program was terminated (Chreod Development Consultants, Towards Land Nargaaement Reform in Shanahai: A Case Study of the Gentral Core, Shanghai: Shanghai Urban Planning and Design Institute, March 1991, processed draft). V/ T. Terzani, The Forbidden Door, Hong Kong: Asia 2000 Ltd., 1985. - 48 - demolished or drastically altered to fit the requirements of new occupants. Temples became factories or were simply demolished to make way for roads, stadiums, and new structures. During the Great Leap Forward period alone (1958-60), 1,400 factories were opened in the city center. Only the inherited stock of vernacular architecture--courtyard houses, lined up along narrow residential lanes (hutongs)--was left. Today, individual "monuments" deemed worth preserving number only 100-200, though hutona housing rehabilitation is now being pursued vigorously. What remains of thc old city is the Forbidden City, the hutongs, and a few temples. Undoubtedly, as previous Bank reports on urban development have acknowledged, Beijing is an "orderly" city, but it has lost much of its distinctive character. There is a cautionary lesson here for cities like Shanghai, which earlier lacked the resources or the opportu- nity to carry out similarly drastic "renewal." Urban land management reforms need not merely create the new; conservation also has its place, and may well contribute to the economic revitalization of a city center, by providing amen- ities that, directly or indirectly, enhance the competitiveness of the city as a center of diversified economic activity. It does so by attracting scarce international and national physical and human capital that values the overall urban fabric as a key "amenity" that supplements monetary compensation. 3.13 Urban development increased in scale after 1975, but the accelerated economic transformation of the urban sector dates back only to 1984. This point is critical to understanding why China's cities and towns, today, are better placed than their post-Socialist counterparts in adapting urban market management techniques and producing land use outcomes more compatible with a market economy. A core hypothesis of this report, difficult to prove on the basis of the rapid assessment techniques used in preparing sector reports, is that China's land use Planners have had relatively limited opportunities to repVcate the practices of the 1950s and that, within the time span of only one decade. they were soon face to face with an embryonic market economy that forced them, however imverfectly. to refocus their efforts and adopt new poli- cies and tools. In addition. the imperative to save arable land reasserted itself. All this had to be done by the remnants of the original core of urban planners predating the Cultural Revolution and their young apprentices. 3.14 Furthermore, and even harder to document, it appears that, striking exceptions aside, Chinese planners. prior to the internalization of market signals. did not have the investment resources to invest in the regional transPort networkn necessarv to allow the low-density "Socialist" city to expand as oriRinallv planned. The bicycle and the bicycle-bus commutina reauirements of the workinz population acted as a land value proxy. forcint suburban development to be more compact than in other Socialist economies. where extensive highways and metropolitan subway slstems allowed distant Socialist suburbs to emerge. Furthermore. tecause the cities examined in this report were systematically denied the rescurces to fund massive industrial investments in and around old citiep. tle Soviet syndrome, where valuable land close to the CBD was heavily preempted for new industrial estates. was never fully imulemented. Finally, zones 5-10 km2 from the city center, containing lower-density "old-city" developments and later developed as "Socialist" sub- urbs (with a mix of industrial uses and low-density walk-up apartments) are now targets of redevelopment, densification, and land use change. Though the rate of change is relatively slow, and city-specific in its pace, this will - 49 - bring the population density profile that radiates out from the CBD of Chinese cities more in line with what is expected in market economies. 3.15 At first, instead of promoting land market reform, the central authorities reaffirmed their hostility to land pricing and individual property rights, in favor of administrative solutions. Article 10 of the 1982 Consti- tution, which banned the sale, lease, or "unlawful" transfer of urban land, was followed by a 1983 State Council (Cabinet) Circular on Preventing the Sale, Buying, and Leasing of Land. The 1986 State Land Administration Law's opening paragraph was emphatic in placing the Government monitoring and con- trol of land use ahead of any other objective. It was only in 1988 that the Constitution and the Land Administration Law formally recognized the right to transfer state-owned land to users who would pay for the rights to use such land, becoming leaseholders. However, one major benefit that emerged from the political disorder that characterized the decades of the 1960. and 1970. was that China's leaders effectively devolved administrative decision-making power over urban development to local authorities. With the endorsement of the 1978 Third Plenum of the Twelfth Party Central Committee, local experimentation with unorthodox reforms became poseible, even in the absence of a national set of laws and regulations cover:'ng the practices involved. It proved instrumen- tal in accelerating the pace of policy and institutional refc-m in the urban sector, particularly after urban reforms were explicitly endorsed in and after 1984.8/ This one clear and positive outcome of two decades of intermittent decentralization meant that local autonomy became finally embedded in the Chinese political system. 3.16 Local authorities, particularly among the more developed and outward-looking coastal centers, took advantage of this leeway, and experi- ments in urban land revenue generation and enhanced property rights began to take place. Having no comprehensive framework within which to develop a market-oriented strategy, however, these experiments proved, in retrospect, to be largely fragmentary. That comprehensive framework, along with enabling legislation and regulation, is one of the functions the Center could have per- formed, but chose not to exercise. The conclusion, then, is that the momentum for reform is building, but that, by the criteria used in this report, much remains to be done. 3.17 China's recent urban land management reform has been shaped by fac- tors absent until now in command economies in transition. First, as noted, the massive urban investments led by Soviet-style planning in China's cities were both spatially and temporally contained and were not able to transform most cities as rapidly as was the case in other Socialist states. Moreover, during the last decade, foreign and overseas Chinese investors became impor- tant actors in the real estate scene, both downtown and in the suburbs and exurbs. They brought with them a new perspective: that "capitalist" tools could be put to work in reforming a command economy without necessarily threatening the prevailing political order. Greenfield development became more expensive, as land requisition from rural "collectives" was often trans- formed into a set of negotiable transactions among increasingly equal parties. 8/ These reforms are documented in China: Country Economic Memorandum: Planning and Reform in the 1990s, op. cit. - so - The key ingredients of market-led enterprise behavior began to emerge from a variety of sources. (hie quarter of all industrial production is now produced by the township and village enterprises operating largely within the urban field, having no access to preferential policies in the acquisition of inputs, and selling output in a relatively competitive environment that allows for bankruptcy when economic conditions justify such outcomes. Collective and private enterprises have grown in scope, facing largely similar market- determined constraints. The state-owned enterprises, which raise the stiffest challenge to the introduction of market forces, now control barely over one half of industrial production, and even they are progressively being exposed to market discipline. China's urban economy was the first among the economies in transition where market prices became dominant and a buyer's market emerged, so that producers could no longer ignore the consequences of neglect- ing consumer demand. Today, for example, only one third of agricultural out- put is purchased at fixed prices. Nearly 70 percent of all consumer goods have been deregulated and price controls have been removed from 58 percent of industrial raw materials and producer goods. Producers and managers now have full authority over about two thirds of the country's price decisions. In some provinces (Guangdong, Fujian) and cities (the Special Economic Zones and Wenzhou), inputs and outputs are traded almost entirely within a market sys- tem. 3.18 Furthermore, central authorities overseeing the unified state budget have been quick to identify price subsidies, forgone tax income and otherwise unnecessary infusions into loss-making enterprises;9/ measure the inventory of unsold goods lying in warehouses, victims of bad marketing decisions;10/ track the enormous amount of unpaid interenterprise debts that have accumu- lated to date;1j/ and realize that the banking sector's likely unrecover- able loans threatened the viability of the country's financial intermediaries. The year 1992 may mark the time when state-owned enterprises, many of which are overstaffed and poorly managed, will be forced to take restructuring seri- ously. The mechanisms used--mergers between firms with differing levels of profitability; managerial demotions and dismissals; de facto layoffs of redun- dant workers (who are sent home, receiving only one third of expected monetary compensation); and the creation of subsidiary service enterprises to absorb unneeded manufacturing labor--may not be "first best" solutions from the view- point of market-system economists. Given Chinese conditions, however, these events, if not reversed, mean the "rules of the game" are chanaing dramatic- ally. reinforcina trentds anticiRated over the last several years. 9/ Tienty percent of the 1991 unified central-local budget, or Y 88 billion, was devoted to loss-making enterprises, excluding unrecorded tax exemp- tions and portions of budgetary expenditures under the rubric of "enter- prise development funds" that were used to bail out loss-making enter- prises. 10/ By June 1991, Y 136 billion worth of goods were stockpiled by state-owned enterprises, adding to commercial inventories equal to three times that amount. II/ Intercorporate arrears totaled approximately Y 200 billion at the end of 1991. - 51 - 3.19 Managers and the supervisory governmental bodies that act as proxy "boards of directors" have now no choice but to behave increasingly as hard- nosed businessmen whose investment decisions--including those involving loca- tion and relocation--must respond, more and more, to the same incentives that drive their counterparts in market economies. Urban planners cannot but adapt to this environment. And, from the real estate market's perspective, this situation was immeasurably assisted by the 1984 State Council decision to issue "Temporary Regulations Involving Systemic Reform of the Administration of the Construction Industry," which directed that "urban real estate develop- ment companies (REDCs) should be formed to carry out the comprehensive devel- opment of urban land and housinb" and that these companies could transfer serviced land, housing and related facilities on a for-profit basis, acting as business enterprises largely responsible for their profits and losses, even if their initial equity was "public" and thus "ownership" rights were vested in municipal, or other governmental units. 3.20 Chapter IV will document the fact that Chinese urban planners are clearly (if selectively) recponding to merket demand, even if this means radi- cal departures from past practices. That being the case, the first priority is to expand the scope of market forces in the land allocation process, and expect that urban planners will play "catch up" and do so at an increasingly rapid rate. Once convinced, Chinese policymakers have few rivals in the implementation of new approaches. Technical assistance needs aside, the key problem facing uroan economic development is enhancing the underlying incen- tives being set in place. Nevertheless, one should not underestimate the difficulty facing Chinese planners. If, as this report declares, the redevel- opment of deneely populated areas and the associated massive relocation of factories and warehouses is the key priority in urban land management, then the task will not be achieved in the short or medium term. C. The Search for a Third Way: Evaluatina Price and Propertv Reforms to Date 12/ 3.21 The concept of land 4s a commodity resurfaced, oddly enough, as early as 1980, at a National Conference on City Planning Work, which proposed the collection of locally retained land use fees as a source of revenue for urban '.nfrastructure development. In 1980, the State Council endorsed the collection of land use fees from foreign-funded v -ures. Urban land markets began to be reinvented. 3.22 Experiments to date have been modest in scol.- f cAe uses a simple and absolute indicator of systemic reform, such as the e.K c nt of land use rights that have been sold off as fixed-term leaseholds th. eive property owners the right to mortgage and transfer their lots. Little ' this was leased on the basis of competitive bidding, and most "prices" . - merely meant to recover serviced land development costs. Typically in t&. cities examined, land leased through competitive bids aqualed an area only two to five times the size of Beijing's Tiananmen Square. As of August 1991, a total area of only 2,500 hectares of state-owned land had been leased, covering 12/ Selective updates, reflecting developments after the preparation mis- sions, are reported in Chapter VI. - 52 - 1,071 tracts, and acquired for Y 2.47 billion (1988-August 1991). This amounts to approximately 1 percent of all annual built-up land expansion in urban areas. MOF received only Y 20 million of this amount. Through the end of 1991, 645 hectares had been leased in Shenzhen, yet all but 3.5 Percent of the land area involved was made available through private treaty or negotia- tion at nominal averaae Drices. or exempted from payment altoaether. Land actually sold by bidding and auction amounted to 15 hectares between 1987 and 1989, while an additional 11 hectares were subject to bidding in 1990; and 9 hectares were let through bids in 1991. Auctions were discontinued after 1968, out of a declared fear that participants in auctions might bid "irre- sponsibly" and "distort" market values to the extent that they operated under soft budget constraints. Some sense of the revenue gained from land lease sales can be gained from an examination of Table 3.4, while Table 3.5 provides evidence of current market prices and their range (from Y 620/mn to Y 3,012/ m2) around the average of Y 1,213/m2 or $220/m2. Table 3.4: LAND USE RIGHTS TRANSFER IN SHENZHEN: 1987-91 Total 1979-87 1987 1988 1989 1990 1991 1987-91 Number of lots 5 110 83 123 324 645 Total land (ha) 8,125.0La 15.73 225.84 199.78 204.00 308.60 953.95 Of which: Negotiated 10.24 221.87 194.57 193.37 300.00 920.05 65.1X 98.22 97.4Z 94.82 97.22 96.42 Bidding 4.64 3.08 5.21 10.67 8.60 32.19 29.5X 1.42 2.62 5.22 2.82 3.42 Auction 0.86 0.89 0.00 0.00 0.00 1.75 5.52 0.42 0.02 0.02 0.02 0.22 Total revenue (Y mln) 31.15 218.27 187.85 - - 437.26 Of whicht Negotiated 12.84 97.21 130.27 - - 240.32 41.22 44.52 69.32 55.02 Bidding 17.06 82.06 57.58 - - 156.70 54.82 37.62 30.72 35.82 Auction 5.25 39.00 0.00 - - 44.25 16.92 17.92 0.02 10.17 Unit px ce (Y/m2) 198.01 96.65 94.03 of which: Negotiated 125.39 43.81 66.95 Blding 367.65 2,666.16 1,106.12 Auction 610.47 4,366.32 /a By administrative allocation. Sourcest (1) Between 1980 and 1989: Chen Rueirong, 1990, "Views on Land Lease and Real Estate Market in Shenzhen," China and Foreign Real Estate Times, 1990, No. 3, p. 12. (2) For 1990: Shenzhen Real Estate Yearbook, 1991, p. 27. (3) For 1991: China Real Estate Information, 1992, No. 2, p. 2. 3.23 Shenzhen's record in land leasing has often been exaggerated; how- ever, local authorities have introduced one transition measure worth replicat- ing across China: while annual ground rents for leaseholders are symbolic and thus very low in value terms, holders of administrative land must DaY annual rents eauivalent to the expected annual return on the value of the land, had it been leased. The strategy is meant to encourage users operating under the old system to switch to leasehold arrangements, benefiting from the rights that accrue to such "ownership." Tables 3.6 and 3.7 illustrate the signifi- cant penalties involved in continuing to hold land under the administratively - 53 - Table 3.5: LAZID LEASE SALES IN SHENZHEN BY BIDDING: 1991 Total Land floor Total unit ID Location Land space revenue price FAR (i2) (mW) (Y mln) (Ylm2) 1 B108-9 13,176 18,200 8.17 620.10 1.4 2 B119-13 5,146 '.8,000 15.50 3,012.30 5.4 3 B119-14 7,274 12,600 19.00 2,612.04 1.7 4 H312-17 5,000 10,000 15.00 3,000.00 2.0 5 B207-5 9,337 14,600 26.70 2,859.68 1.6 6 B119-24 3,580 14,800 18.00 5,028.49 4.1 7 B207-4 11,042 15,800 28.00 2,535.70 1.4 8 B119-25 3,580 14,800 18.00 5,028.49 4.1 9 B119-23 3,976 14,800 20.00 5,030.43 3.7 10 B119-26 3,976 14,800 20.00 5,030.43 3.7 11 B207-6 11,859 35,600 38.45 3,242.51 3.0 12 B207-9 7,694 9,235 19.63 2,551.24 1.2 Total 85.639 203.235 246.45 1,212.65 2.4 Source: Chen Guangyan, 1992, "1991 Land and Real Estate Market in Shenzhen," China and Foreiutn Real Estate Times, 1992, No. 2. Table 3.6; LAND CHARGES COLLECTED FROM USERS WHO LEASE LAND Types of land use Y/mi/year Various commercial 3.0 Residential (excluding villas) 2.0 Industrial 1.0 Others 0.5 Source: Shenzhen Land Bureau. allocated procedures, and suggest potential incentives involved in encouraging all land users to enter into a "market economy" system. 3.24 Shanghai is also one of the cities experimenting with land leasing, though so far it has confined itself to foreign-investor deals. By mid-1992, 16 parcels had been leased, for a total area of 24 hectares, sold for $106 million equivalent, or an average price of $448/m2 (Table 3.8). The land area involved is but a fraction of the :and requisitioned during roughly the same period (1988-91), and equivalent to only 2.5 percent of the total land requi- sition completed within the city proper, or 675 hectares. - 54 - Table 3.7: LAND CHARGE COLLECTED FROM USERS OF ADMINISTRATIVELY ALLOCATED LAND (m2/year) Land Land use gattern grade Commercial Residential Industrial Public square Others 1 32.0 12.0 10.0 2.6 2.0 2 22.0 10.0 8.0 1.5 1.4 3 15.0 7.0 5.0 1.0 0.7 4 10.0 5.0 4.0 0.5 0.5 5 8.0 3.0 3.0 0.3 0.3 Source: Shenzhen Land Bureau. Table 3.S SHANGAI LAND LESES, 1986 - Jun. 1992 Total unit Price per DistanCO ID Location Date Use IUAR Mtbod Length Land revene price floorWpaco to center I(year) (UP) Ulum) T$Ia?) (Sir?) (km) I H0-26 080.08-G8 CC 5.0 Ridding 50 12,973 26.05 2,162.18 432.44 7.1 -2 n-28-31a 01-19-69 CQI 7.2 Bidding 50 3 614 6.26 2.291.09 316.21 7.0 3 cl -4a7-10 03-2-90 ID 2:7 negotiaton 0 4272 10.S6 254.66 94.32 9.2 4 T.o / 10-23-90 se 0.5 Negotiation 70 16 140 0.89 55.00 110.00 14.5 5 14-245 10-30-90 HOT 0 21,992 13181 62792 1395.39 70 6 Cubei-24 07-14-91 SES 1.4 81dd 70 553400 15.00 270.76 193.40 7.5 7 Pudong-1 11-04-91 IND 2.5 Negotiation 50 5.270 0.53 100.00 60.00 5.6 6 Pudong-2 11-07-91 Cti 6.5 Negotiation 30 86116 7.30 900.00 138.46 7.5 9 Guboi-22 11-07-91 C5A Negotiation 30 6,003 4.79 797.93 - 4.6 10 Depu-X. 01-25-92 iES 4.2 Negotiation 70 19 790 1.00 50.53 12.03 3.3 11 Pd-136 03.09-92 C01 6.0 Negotiation 50 3,555 0.96 275.11 5.865 4.6 12 Livang 03-10-92 BBS 3.9 Negotiation 70 2,025 0.71 352.00 90.26 3.5 13 Nuahan 03-23-92 Res 1.9 Negotiation 70 8.524 1.15 134.91 72.93 3.6 14 Beiji-71 04-11-92 CON 5.0 Negotiation 30 23 800 4.60 193.28 38.66 3.8 U Changei 04-21-92 US 2.I Negotiation 70 3,228 0.65 201.36 95.89 3.5 6 Jtingin CM Conversiou 3,699 7.50 2,027.56 total 2368653 106.12 448.04 (T van) (Ti.') 17 Pudon5-1˘ 06-22-91 Negotiation 70 1,510,000 670.00 445.71 5.4 18 PudoNg-ig E5 06-22-91 Negoti tion 70 4 000,000 240.00 60.00 12.1 19 Pudong-gq c 06-22-91 Negotiation 70 4.000.000 240.00 60.00 14.6 Total 9.510.000 1.150.00 120.93 I Unserviced land. These negotiated leases ivolvod the conversion of ehisting joint-venture agreemets, wbhre the Chinese partners had origi- ny contributed land ao equit /c Parcels 10-15 all require resettlont and related expenditures. This land was aecSgd to roal estate development compuise and the lease prico can boet be vied as goerunmt equity inveatmant. Sourcoe Shangbai Lad Bureau. 3.25 In Fuzhou, 17 hactares have been leased to date, with lO hectares leased in 1991 alone. The sales have been made almost entirely to foreign- funded ventures (90 percent). Like Shanghai, they involve primarily (91 per- cent) negotiated "private treaty" contracts, rather than auctions or tender- ing; however, unlike Shanghai, where leased tracts tend to be 7 km or more from the CBD, the land involved in Fuzhou is almost all within the "old city" (85 percent). Prices differ according to use. For example, in 1991 the eight cases of leases for commercial use were sold for an average of Y 829/m', while the nine residential tracts sold for only Y 436/mt. 3.26 In other cities visited by the mission, there was little evidence of leasing activity. Chengdu had none to report; Hangzhou had leased two tracts in the city proper to domestic investors, covering 0.66 ha and sold for -55 - Y 3,080/ma in one case and Y 2,600/ma in the other.13/ Even Guangzhou ha. followed a fairly cautious policy in leasing land. Between 1988 and 1991, 20 sites, covering 14 ha, were leased, primarily to foreigners, in the autonomous Economic and Technological District, 25 km from the city center. Land lease sales in the rest of the city began officially in 1990, and totaled 26 ha, with an additional 6 ha leased since then (Table 3.9). Prior to 1990, a few housing development sites were leased to domestic REDCs. The key transaction involved Huadiwan, and covered 107 ha. It sold for Y 304/m2, reflecting its proposed use and the lack of infrastructure for servicing the site. Little other activity took place prior to 1990. Table 3.9: LAND LEASE TRANSFER IN GUANGZHOU: 1990/91 Number of Total Total Unit Land area leases land area revenue price per lease (ha) (Y'000) (Ynl?) (ha) Total 51 25.87 209,610 810.24 0.51 Of which- For foreign 6 33,620 For 1990 14 3.00 4,122 137.50 0.21 1 0.23 600 262.70 2 0.48 1,030 215.98 3 0.39 342 87.38 4 0.03 n.a. 5 0.05 230 469.39 6 n.a. 220 7 1.06 400 37.59 8 n.a. 250 9 n.a. 10 0.48 300 62.97 11 n.a. 100 12 0.08 200 240.38 13 0.07 100 149.48 14 0.14 350 257.35 For 1991 37 22.87 205,488 898.42 0.62 Source: Guangzhou Construction Commission. Sales made within the semiautono- mous Economic and Technological District are excluded. 13/ In the fast-paced environment of 1992, the "shelf life" of all aggregate data and conclusions has been shortened. Hangzhou transferred 200 ha to a Hong Kong concern in May 1992, at a location 4 km from the West Lake, for 50 years, at a total cost of $i11.5 million. - 56 - 3.27 The process of change, providing building blocks for the comprehen- sive reform process prescribed in this report, can best be described as involving the introduction, recognition and eventual legalization of a variety of land pricing mechanisms that had been largely dormant between the mid-1950s and the early 1980.. These mechanisms include the following: (a) the emer- gence of de facto land market transactions in the urban periphery, both in the requisition of rural land, and the creation of suburban township and village enterprises, operating within a market system; (b) the realization, on the part of local governments in particular, that urban land development can gen- erate resources for financing urban infrastructure development, in the form of fees, taxes, rents, and leases; (c) the recognition that unregulated transac- tions in central (core) area land haze become a significant element in shaping land use changes, followed by a decision to regularize such exchanges rather than merely trying to suppress them; and (d) the acceptance of the fact that redevelopment of older city centers, to accommodate the surging demand of the producer services sector and of consumers willing to buy or rent housing at commercial prices, provides a unique opportunity to reap benefits from land with exceptional access characteristics. In this process, the REDCs, autho- rized in 1984, began to introduce and popularize the concept of commercial transactions in real estate, particularly in the form of "commodity" housing sold on a for-profit basis to both enterprises, nonprofit institutions, and individuals. Today, in the case study cities and nationwide, 40-80 percent of all new housing units completed are sold under "regulated" market conditions by REDCs. Such "commodity housing" has, in varying degrees, become a perma- nent feature of the urban economy (Table 3.10). And. because commodity hous- ine carries an implicit land price linked to location, such housing develoR- ments have had a far areater impact in introducing market prices that the more gublicized land leases.14/ 3.28 Land requisition fees, as noted, are paid to farmers and their com- munities, with additional fees accruing to governmeatal authorities, by those urban land users who encroach on legally defined agricultural land. Prior to the reform period, taking 1958 as a point of reference, these fees were nomi- nal in nature and covered only four categories of compensation: land loss; crop loss; residential displacement allowances; and payment for "improvements" already existing on site. By the 1990s, each of these had been inflated in value, as villagers grew more adept as negotiators, and as market forces increased the value of each of the various categories involved. In addition, new items appeared on the compensauion list. Three of these reflected Govern- ment's desire to conserve agricultural land: a loss of cultivated land fee, a fee to recreate lost cultivated land elsewhere, and a fee to replace vegetable plots nearby. Other fees reflected the fact that the "farm environment" had evolved and new rights and entitlements had emerged: displaced rural indus- trial facilities had to be compensated; additional fees were required to cover 14/ Nationally, 32 percent of all commodity housing space is sold to individ- uals. In Guangzhou, the level has reached 44 percent, while in Shenzhen the level totals 38 percent. In other large cities, having made less progress in introducing regul.ated market practices, the level typically falls to 10-15 percent. Among such cities are Shanghai, Tianjin, and Chengdu. (Sourcee Guanezhou Real Estate, 1990, No. 8, p. 17; China Real Estate, 1991, No. 2, p. 72.) - 57 - Table 3.10: COMMODITY HOUSING SALE IN SELECTED CHINESE CITIES: 1989 Total new Commodity hous- Sold housinf inm completed Amount sold to individuals ('000 m '000 ma z '000o '000 i X Beijing 5,541.0 2,434.0 43.9 1,173.0 48.2 15.0 1.3 Shanghai 3,710.0 750.0 20.2 462.0 61.6 58.0 12.6 Tianjin 3,093.0 750.1 24.3 506.6 67.5 56.0 11.1 Guangzhou 2,990.0 1,931.0 64.6 927.7 48.0 409.1 44.1 Shenzhen 1,700.0 1,091.1 64.2 532.3 48.8 203.7 38.3 Hangzhou 624.7 410.6 65.7 174.2 42.4 n.a. n.ae Chengdu 744.0 301.3 40.5 232.8 77.3 25.2 10.8 National /a 109,557.6 37,840.3 34.5 24,913.8 65.8 8,054.9 32.3 /a The level of marker-led development is markedly higher in the smaller cities of China, where the concept of state-supplied housing, linked to a massive "state-owned"-dominated local employment, was never implemented. Sources: Guanazhou Real Estate, 1980, No. 8, p. 17; China Real Estate, 1991, No. 2, p. 72. the subsidies implicit in turning the legally defined agricultural residents into urban residents (grain price subsidy fees, public housing fees, and, often, new urban employment); and fees were required to cover farm-related "installations" that were more broadly defined than under the prior set of regulations (Annex 5). Box.3.1: GUANG2HOU: A CASE STUDY IN RISING LAND REQUISITION COSTS 7she Jianguan Housing Estate, located in Haizhu district, has, during different phases of Itousing development, faced very different land requisition costs: 1979: Y 30,000 per mu, or Y 44 per m2 1983: Y 37,500 per mu, or Y 56 per m 1985: Y 55,000 per mu, or Y 82 per m2 1987: Y 105,000 per mu, or Y 158 per m2 1991: Y 300,000 per mu, or Y 450 per m2 Within 10 years, land requisition costs for this project have increased tenfold; this is typical of general trends In Guangzhou. One reason behind such rapid increases of land requisition costs is the rising expectations of farmers. Since Guangzhou did not adopt a unified land requisition approach until July 1992, most land requisitions were carried out between individual development companies or land users and local farmers. According to the City Real Estate Bureau, as much as 80 percent of land requisition prior to mid-1992 was done without the involvement of city authorities. Source: Guangzhou City Construction Development Corporation. - 58 - 3.29 Data collected by the Chinese Academy of Social Sciences 15/ as well as mission estimates, indicate that between 1985 and 1990 suburban land requisition prices for comparable land doubled, and now averages from Y 100/X! to Y 200/rm for a range of large coastal cities 16/ (Box 3.1). By way of comparison, the average costs of buildings alone averaged Y 413/rn' in 1990, for state-owned investments. In China, urban building costs do not vary dra- matically on a city-by-city basis; a *25 percent adjustment to the national average should suffice to cover differences among most cities. Thus, land requisition costs have become significant, since the floor/land area ratio ..n suburban locations is unlikely to exceed 1.0. By comparison, at the beginning of the reform process in 1979, suburban requisition costs were relatively minor, amounting to Y 10-15/mr in the suburbs of a metropolitan area, while construction costs already amounted to an average of Y 125/man. Furthermore, as noted, the former price was treated as a free good by the user, covered by budgetary funds. Today, the user is most likely to use its own resources (retained earnings and repayable bank loans) to finance the same transaction. 3.30 The original requisition fees left little for the government; even today, the fees that go to the public sector--25-35 percent--are earmarked for agricultural land reclamation. Urban authorities, therefore, have sought other avenues to extract revenue from land, thereby creating a set of "prices" and "rents" which may or may not bear much linkage to market prices, but do validate the notion of land as a commodity. 3.31 As noted, the above-cited ;980 National Conference inspired various cities to launch land fee collection experiments. These fees were meant to proxy land rents, appropriately adjusted to reflect political pressures from users complaining about "affordability." Fees came into effect irL Shenzhen (Special Economic Zone) in 1982, in Guangzhou in 1984, and 100 cities there- after. Fees for domestic concerns were set uniformly at much lower rates than those applied where foreign capital was involved. Table 3.11 illustrates the range of fees for foreign-funded and domestic firms in Guangzhou, and is broadly representative of the orders-of-magnitude involved. The fees applica- ble to new local users were lower, in nominal terms, than rent levals prevail- ing in the mid-1950s, before pricing was abolished. They acted, more than anything else, to reassert the public ownership rights over urban land, and had little to do with systemic urban land management reform. 3.32 The next sten in making land values explicit, or, alternatively, generating revenues form land, came in the form of land development exactions in greenfield and redevelopment locations. The fcrmer exactions were made possible by the emergence of real estate development companies, described in para. 3.19, often competing for business to remain profitable. The latter exactions, involving massive entitlement payments to existing occupants of redevelopment projects, greatly exceeded the former (Annex 6). Furthermore, 15/ Irban Land Use and Management in China, op. cit. 16/ Shenzhen exproprieted all the land needed for its foreseeable future (say, 15* years) by 1980 and set compensation terms at that time. Requi- sition prices for prime land are only one tenth of those cited for other cities. - 59 - Table 3.11: 1984 GUANGZHOU LAND USE FEES (Y/m2/year) A. Applicable to New Land Allocations for Domestic Enterprises __________________ Land Rrade I II III IV V VI VI1 Fee 4.0 3.5 3.0 2.5 2.0 1.0 0.5 B. Applicable to All Foreign-Funded Ventures _____________ Land arade I II III IV Commercial 36-56 28-44 12-28 8-20 Residential 18-25 14-18 10-14 5-10 Industrial 8-12 6-10 4-8 2-6 Other 28-40 20-28 9.6-16 4.8-9.6 C. ADglicable to Special 3conomic Develogment District Commercial and storage 12 Housing 8 Industry and transport 2 Science-related I Other 9 Source: Guangzhou State Land Administration Bureau. from a market economy's perspective, the redevelopment exactions were directed at compensating occupants; greenfield fees were more clearly meant to finance off-site infrastructure through nontax mechanisms that would not be subject to revenue-sharing arrangements with higher-level governmental authorities. All collected revenues would be retained and spent locally. This theme. wherebv land revenue generation and the emergence of market prices is obstructed or made oDsque by disbutes over distributional (i.e.. tax) obiectives and bv interaovernmental revenue-sharing requirements. recurs throuahout all of the subsequent discussions of related Drices. taxes and fees. This can only impede the emergence of explicit land Prices. and transparent land markets. which this report takes as the key objective of any substantive systemic reform. The emergence of explicit land Prices ib too imPortant to the Chinese urban modernization effort to be stalled by disDutes involvint the distribu- tion of taxation recel2ts. 3.33 In greenfield locations, one would expect development projects to cover land, building and on-site infrastructure, with some debate remaining over responsibility to freely provide "public facilities," particularly when these would normally be a private responsibility in a regulated market economy (e.g., retail space operated by local authorities). Greater controversy sur- rounds the degree to which off-site infrastructure (including thuse labeled as - 60 - citywide in scope) should be paid for by the investors of a particular estate or subdivision. North American and Western European practise, for example, provides precedents for some of these exactions (Box 3.2). Unfortunately, not all such practices are automatically fair or efficient, merely because they have emerged in developed market economies. Retrospective assessments of "Western" practices are often critical, and demonstrate that they should be examined on a case-by-case basis. Precedents do not provide practitioners with "best practices," particularly when those tools are utilized in countries wealthy enough to afford to make mistakes. China's urban planners, operating in a low-income environment, have less leeway than that available to urban land management planners elsewhere. Box 3.2: THE FaENCH EXPERIENCE WITH LAND FEES AND EXACTICNS France has, over the last two decades, experienced a process of administrative and municipal finance decentralization that led to the experimentation with various means to recover off-site infrastructure costs. Vincent Renard has concisely summarized the process of searching for an equitable and transparent solution: "In France, after two decades of wavering between a systematic fiscal device (a local development tax calculated as a percentage of a preset lump-sum price of the to-be-built floor area) and case-by-case negoti- ated agreements between developers and municipalities, such as it Is practiced in comprehensive development areas, a now tool between them has been introduced in 1985 as 'special exaction areas'.... The municipal council delineates the geographical area within which developers ... will have to pay all or part of the costs of the public infrastructure required by present and future residents.... The amount ... is not preset; the statute only says that the amount cannot exceed 100 percent of the real cost of infrastructure. The municipality must establish the nature of these public facilities ..., their cost, time schedule for building; as well as the developer's ehare (per m2 of floor area to be built) of the total cost and how it will be apportioned among the various building types. The exactions ccn be ln-kind or monetary." Renard concludes that this experiment is too recent to evaluate, but that it may provide an incentive to force private landowners to lower their "offer ptice." In China, private landowners do not exist and It is not clear If the rural land occupants will respond to this device, reducing their "reservation price" for land in negotiations with the local land bureau In the eame manner. Renard also offers evidence that local municipalities are adopting in-kind "linkage" payments, requiring developers to prov;.de some low-cost housing, that was also introduced in San Francisco and Boston, In the United States, and Guangzhou, in China. Source: V. Renard, "Affecting Land Prices Through Taxation: Perspective From the French Experience," Cambridge, Massachusetts, International Conference on Property Taxa- tion and Its Interaction with Land Policy, Lincoln Institute, 1991. 3.34 The Doerits of the debate set aside, what are China's municipal authorities collectii, in terms of off-site infrastructure "exactions," largely from REDCs, be they In terms of money or in-kind contributions? In cities like Guangzhou, such exactions were reported to have reached the equiv- alent of Y 250-300/m2 as early as 1985, when sites were opened to "competitive bids." Today, exactions average Y 200-500/m%, measured in terms of square meters of "commercial" building area constructed. These exavtions are often provided as "in-kind" benefits (i.e., physical facilities), but overall begin to resemble prices paid for leases, though without the property rights that - 61 - accompany them.17/ Between 1984 and 1988, approximately half of total urban infrastructure investment was financed through these "off-off-budget" sources, paid largely by REDCs, according to city officials. Thereafter, the majority of such expenditures were covered by "off-off-budget" fees. 3.35 In Shenzhen, practices are even more dramatic. Following natiornal policy guidelines to develop "standard land p-ces" as a basis for negotiating land leases or determining the likely accuracy of property exchange prices which may yield capital gains or other such transfer fees, Shenzhen estab- lished a scheme that divided the city into five zones (or classes) within which different land uses are subject to presumed lease values that vary dra- matically depending on intended use (Table 3.12). Here, as elsewhere, one Table 3.12: STANDARD LAND PRICE FOR NEGOTIATED LAND IN ZHENZHEN, 1991 (Y/m2) Land use Land class 1 2 3 4 5 Class I 1,200 300 180 130 50 Class II 700 280 160 110 40 Claso III 00 220 140 90 30 Class IV 250 180 120 70 20 Class V 190 160 100 so 10 Note: Fixve types of land uses are: (1) commercial, retail, financial, office land use; (2) all types of housing except villas; (3) various industrial, storage and transport facilities; (4) construction sites and other open production sites; (5) planting, and other types of open fields. Source: China Real Estate Information, 1992, No. 14, p. 3. must question why land values in prime locations (Class I) should be assumed to vary by a factor of 24:1. As noted elsewhere, however, the citywide ratio of highest values (Zone 1, commercial) to lowest values (Zone 5, agricultural) is plausible, and is set at 120:1. All in all, this is an example of anachro- nistic changes that do little to advance the emergence of a regulated market economy. More significant, however, was Shenzhen's simultaneous decision to require all new land users to pay high off-site infrastructure charges even if the user received the land as a free good. Though the exactions are only significant in some cases (notably land use types 1 and 2 in zones I through III), Shenzhen has at least taken an important step in the direction of reform 171 As of early 1992, Guangzhou has only one reported major local lease (107 hectares in Huadiwan), which was awarded to the highest bidder, the Guangzhou Credit and Trust Company, for Y 304/i2, excluding land requisi- tion costs. -62 - by monetizing the off-site exactions and making the standardM to be used transparent and not negotiable (Table 3.13). Table 3.13: OFF-SITE INFRASTRUCTURE FEE SCHEDULE IN SHENZHEN, 1991 (Y/m2) Land use Land class 1 2 3 4 5 Class I 1,800 460 270 160 100 Class II 1,050 380 240 150 80 Class III 510 340 210 140 70 Class IV 330 290 180 130 60 Class V 230 220 150 120 50 Source: China Real Estate Information, 1992, N 14, p. 3. 3.36 If one combines the guidelines suggested by the standard land prices and the requirements of the off-site charges, the resultant "price" of land in certain areas of Shenzhen, particularly in zones I and II, and land use cate- gories 1 and 2, begin to resemble outcomes expected from the leasing of land through competitive bidding or tendering (Table 3.14). Though this process is administratively driven, it obviously reflects "feedback" from market transac- tions and, in an imperfect way, promotes systemic reform of urban land manage- ment. Table 3.14: COMBINED LAND PRICE AND OFF-SITE FEE: SHENZHEN, 1991 (Y/m) Land use Land class 1 2 3 4 5 Class I 3,000 760 450 290 150 Class II 1,750 660 400 260 120 Class III 9x0 560 350 230 100 Class IV 580 470 300 200 80 Class V 420 320 250 180 60 Source: China Real Estate Information, 1992, No. 14, p. 3. 3.37 Elsewhere in China, the evidence is that more moderate exuctions are the norm, if local official estimates are deemed reliable. Officials in 14 cities,18/ during discussions with a Bank mission reviewing housing reform proposals potentially eligible for Bank financing, revealed the existence of 18/ Including "outliers," particularly Guangzhou. - 63 - exactions for off-site investments ranging from Y 100/m2 to Y 500/m2, where greenfield development was involved. These exactions are calculated on the basis of per square meter of "commodity" housing, and their recalculation, in terms of contributions per square meter of land depends on the average floor area to land area ratio. Casual empiricism suggests that a 1:1 ratio can be used for purposes of this discussion. 3.38 It is instructive that no outsider, not even the central authori- ties, can say with certainty how much any one city spends on infrastructure development and maintenance. As noted earlier, the sources of funds involved prove particularly difficult to unravel. One recent Bank report estimated Shanghai's urban infrastructure expenditures average 3-4 percent of local GDP; information supplied separately to the Bank by Shanghai's Finance Bureau director suggests levels on the order of 9-10 percent of local GDP, a total which equals or exceeds the locally retained revenues of the municipality, which one would normally label as "the budget."19/ 3.39 It is also clear that, city by city, practices have varied greatly, including the degree to which land-related exactions contribute to the de facto infrastructure budget. In some cities, including Guangzhou, Wenzhou, and Shenzhen, budget revenue financing of infrastructure is being phased out in favor of off-budget earmarked sources. In others, like Hangzhou, the offi- cial budget (including the so-called official extrabudgetary public utility surcharges) may still finance half of all infrastructure investments (even here, however, as in other cities visited9 detailed information on the infra- structure budget was not made available to the sector study team). Finally, there are cities, like Chengdu, which claim to frown upon the use of "off-off- budget" financing to any extensive degree. The full story is thus complex and not likely to be clarified any time soon. 3.40 Compensation paid to occupants of sites redeveloped in recent years in cities like Beijing and Shanghai averaged Y 1,000/m2 to Y 2,000/m2 for cen- trally located projects. In addition, projects in those areas are saddled with additional infrastructure fees and charges similar to those reported earlier for greenfield sites if the scale of a redevelopment is larae. This encourages REDCs engaged in redevelopment to confine the scale of projects to small sites, where off-site infrastructure fees are relatively low.201 3.41 The very year (1988) that land leases were authorized, MO0 conceived of a land market "reform" strategy that could be extended across all of China's cities, under the assumption that leasing was not likely to evolve as a major force in the spatial restructuring of cities during the foreseeable future. The strategy chosen was to use one tool, and to introduce land use taxes, with very low maximum and minimum rates mandated from the center, dependent on city size, and with revenues to be shared on a 50-50 percent basis between the collectors and the center. The taxes were meant to accom- plish a variety of rather ambitious objectives. They were supposed to raise 19/ China: Industrial Restructurina: A Tale of Three Cities, op. cit. 7j1 In Shanghai, Y 95/m2 of all newly constructed space; in Fuzhou, Y 43/r2 of newly built structures. - 64 - revenues from land, giving land official recognition as a semicommodity whose use required the payment of a "rent." Thev were intended to provide users with incentives to return underutilized land to city authorities without any compensation, and encourage spatial restructuring, as enterprises which found themselves in "high-rent" districts would voluntarily, and without any assumed additional compensation, relocate to "low-rent" districts. 3.42 The land tax has proven to be a failure in achieving its ambitious objectives, mostly because central authorities had little understanding of the implicit value of land and failed to realize that, given international experi- ence, land taxes were unlikely to ever reach levels required to match the impact of explicit prices. They lacked any coordinated intergovernmental strategy to create real land markets based on explicit prices.21/ They also gravely underestimated the reluctance of local governments to undertake reforms that resulted in substantial revenue transfers to central authorities, and reduced local autonomy by abolishing the original land use fee experiments except for isolated cases involving foreign-funded land users. And the mount- ing evidence from city-specific studies, begun as early as 1985 in Shanghai, and mandated by the State Land Administration for all cities in 1989, that were meant to reveal implicit land values, together with less systematic stud- ies of regulated and unregulated land use rights transfers negotiated between individual users, was ignored and never internalized. 3.43 Though land taxes were act .lly imple..nted nationwide, the absolute value of the taxes and the range of permissible va ues across any one city proved very low if the taxeo were meant to act as quasi-land rents. The high- est annual tax allowable, at the center of China's largest cities, was set as Y 10/ma/year. Given the revenue-sharing requirements attached to the tax, cities, in issuing implementing regulations, restricted the range and the spa- tial scope of the tax to levels well below those anticipated by MOF. Total revenues for i990 and 1991 equaled Y 6.2 billion nationwide, with OF receiv- ing Y 3.1 billion. 3.44 The tax proved to have little impact as a revenue mobilization device and certainly did not affect the location decisions of present or future land users, caused land to be returned to the local authorities or provided incentives to relocation. The tax was rightly viewed locally as a nuisance tax whose only (albeit very important) contribution to reform was to accelerate the work in creating fiscal cadasters that would provide a rough mapping of occupant land area boundaries. From the viewpoint of this report, with its focus on comprehensive systemic reform, the tax initiative proved that "reform" measures can often be labeled as merely "changes" whose reform content is minimal. Now that this deficiency has been illustrated with this one case study, the central authorities, including MO0, may realize that urban land management reform cannot be successfully order A by the center in isola- tion of local ma.ket forces; that revenue-sharing objectives can obstruct land reform; and that systemic reform requirements involve tapping into the vast 21/ For a brief description of related international experience, see para. 3.96. - 65 - reservoir of knowledge on land market behavior, available at the local level and overseas and documented throughout the report.22/ 3.45 While the land tax initiative demonstrates how "reform" measures can be ineffective, more recent initiatives, driven by SLA and by agencies promot- ing serious enterprise reform, show promise of greater progress in the future. The Government has committed itself to regularizing (and perhaps facilitating) selective unregulated real estate transactions, accelerating not only the emergence of land markets, but also accommodating the restructuring of local economies, viewed by outside observers as chronically short of consumer and producer services, particularly in central area locations. 3.46 Evidence of unregulated transactions is fragmentary, but suggestive. Hangzhou's SLA bureau officials suggest that the transfer of ground-level housing space to commercial users varies from Y 2,000/m/year in prime loca- tions (Jiefang Road), to Y 600-800/m2/year for slightly "off-center" loca- tions, Y 300-400/m2/year around the city railroad station, falling off to Y 100/m2 or less a year in periphery areas with few locational advantages for commercial activity. The same authorities estimate that there are a minimum of 10,000 such cases in the city at present. Chengdu, which has allegedly not entered the process of market reform until recently, reports at least 7,000 similar cases of black market activity. Among the more desirable locations (Sichuan Hualong Department Store), annual rents of Y 2,300/mr were reported; while in less central sites (Galaxy Department Store), rents reached Y 470/m3/ year. 3.47 SLA bureau r*,resentatives in Fuzhou were also able to draw land value maps based on 2gulated rents. Peripheral locations rent for Y 15-30/ m2/month, while prime locations, including those around the intersection of Dongda and Baiqi Roads, rent for over Y 200/m2/month. These, again, reflect the conversion of ground-floor housing space to commercial uses, and appear comparable in magnitude to those described for Hangzhou. 3.48 Extensive studies of unregulated transactions were also conducted by the Chinese Academy of Social Sciences.23/ Based on end-of-1988 data, the study team reported data for two districts in Beijing, the Da-She Lan (DSL) district near Quiamen, a busy commercial center; and the Zhong-Guan-Chun (ZOC) district, an emerging high-technology center. In each case, rents as high as Y 300/mW/month were reported, with averages of around Y 70/m'/month in the first district and approximately Y 80/m'Imonth in the second. These numbers reilect largely the subletting of enterprise structures to other enterprises, exclusive of "Joint venture" arrangement, where profit sharing is also involved. In the small number of cases where enterprise land use rights were rented, with no structure transferred, rents often proved to be higher than those reported above, with average rents of Y 426/m /month in the ZGC, based on deals negotiated in 1985/86. 22/ These contentions are amply documented in R. Bahl ano J. Zhang, Taxing Urban Land in China, World Bank, Report No. INU39, March 1989. 23/ Urban Land Use and Manatement in China, op. cit. - 66 - 3.49 For reasons that are not entirely clear, but that may reflect the relatively vigorous anti-"black market" activity of the Beijing Construction Commission, backed by MoC, public housing "black market" rents in the DSL district proved distinctly lower than those reported above, averaging Y 20/m2/ month, though these should be compared to allowable rents of roughly 10 fen/ m3/month. 3.50 Other measures could potentially help reveal land values. Enter- prise reform efforts at central and local levels have approached bankruptcy options with great trepidation. The prospect of significant numbers of work- ers rendered unemployed in large cities is viewed as sensitive and a potential source of social instability. Mergers of loss-making enterprises with profit- able enterprises offer a possibility of advancing both enterprise reform and urban land management reform if the loss-making enterprises' assets, including land occupied, can be converted to new uses more conducive to market demand, and if the dominant partner is highly motivated to maximize profits after the merger. 3.51 The cumulative land assets of loss-making enterprises can provide significant resources to promote reform and the redundant labor can be trans- ferred to higher-value uses if three assumptions hold true: (a) that signifi- cant numbers of potentially redundant urban workers can be retrained and sblfted to service sector activities; (b) that urban labor supplies will grow more slowly during the next decade, in response to the effects of past birth- control initiatives and continued restraints on migration; and (c) that China's cities will be restructured while experiencing rapid economic growth. 3.52 Once again, such partial information is no substitute for in-depth local research. This repnct cannot assess the magnitude of potential gains to urban land reform made possible by enterprise reform mergers. Anecdotal evi- dence, clearly weighted to emphasize success stories, suggests that the poten- tial is great. If mechanisms promoting the relocation of viable industrial enterprises away from inappropriate central locations are actively encouraged, the ,utcome, in terms of urban land market reform, is likely to be even brighter. D. Implementina Comprehensive Reforms in Urban Land Price and Pronertv Riahts Reforms: A Suggested Strategy 3.53 Reform to date has been piecemeal, with incremental changes too often masquerading as "systems reform." Piecemeal changes obviously promote the ultimate objective, i.e., the emergence of land traded as a commodity, subject to regulatory, legal and fiscal constraints. They should not be belittled. The point is that postponing broad-based reform handicaps China's cities *seking to be internationally competitive centers within this decade. A broader strategy is needed, one that is bold in scope and introduced rapidly enough to give China's land markets the opportunity to contribute fully to efficient, equitable, and environmentally sensitive economic development by the year 2000. The report cannot flesh out the details of the transition to a regulated market economy; only the general outlines of such a transition can be described and justified. - 67 - 3.54 International experience provides essentially two avenues to land price and ownership reform: the freehold system and the leasehold system. Most countries adhere to the first approach, but significant exceptions exist worldwide, including places such as Hong Kong, Singapore, Amsterdam, Stockholm, the inner city of London, and Honolulu, among others. China has chosen to preserve state ownership of urban land, while permitting user rights to be leased out, giving leaseholders most of the significant rights of free- hold ownership. Reiewing evidence fro' other cities. one can araue that the leasehold method is not an obstacle to market-led reform. and that there is no need to reopen the debate and Dresent freehold strategies as a serious option (Box 3.3). Box 3.3: WHAT DO LEASES ENTAIL? A lease involves an agreement between a landowner, in which the owner agrees to permit the lessee to use the land. A sublease is an agreement by which a lessee permits a third party to exercise the land-use rights granted to the lessee av the landowner. Leases contain numerous provisions, end form the basis for the agreements reached on prices paid, whether by negotiation, competitive tender, or auction. Leases typically stipulate: * length of lease; * land uses permitted; * building restrictions; * design, arrangement, and height of structures; * provision for private streets, roads, and lanes, along with vehicular access priv- ileges; * green space requirements; * anti-nuisance clauses, and other Operformance indicators; - § lease modification provisions, including associated premiums to be paid and new restrictions imposed; * lessor rights to repossess property for noncompliance reasons; * Government rights to expropriate land for public purposes; * timing limits for construction of all or part of the investment to be made by the lessee; * rights and terms associated with mortgaging, transferring, or granting the leased land to a third party; * terms and premiums associated with the renewal of a lease at or prior to the for- mal end of the lease; and * dispute resolution procedures. 3.55 Leasehold arrangements are not without possible drawbacks. They presuppose a government bureaucracy competent and honest enough to manage the conveyance of land to individual lesseer in a manner consistent with the over- all objectives of urban land management systems reform. They assume that the length of the leasehold, and the arrangements surrounding leaschold renewal, are handled with regard to possible detrimental effects, particularly short- - 68 - sighted behavior on the part of users as the end of a leasehold approaches. They take for granted that the conflict between changing urban land management objectives, including those that affect plot-specific profitability, can be accommodated by leasehold documents that may contain use rights and conditions covering 50 to 99 years. And they downplay the possible detrimental effects of a rent-seeking strategy open to a local monopolist who has effective con- trol over a key urban resource, and is thue open to the temptation to use mechanisms that restrict land supplies for development or redevelopment in order to achieve short-term revenue-maximizing objectives. 3.56 None of these potential barriers proves to be sufficiently serious to avoid endorsing a leasehold strategy. Each potential problem has a reme- dial antidote. This report recommends that lease terms and renewal conditions be drawn up to minimize shortsighted user behavior; that the process of award- ing leases be subject to maximum public scrutiny, and follow transparent and elementary rules maximizing fairness; that lease documents permit the planning regulations incorporated into them be subject to changes after a certain time period, a period long enough not to disrupt the assumed profitability of the original development planned, yet short enough to allow the "rules of the game" to be altered when and if longer-term redevelopment and conversion to alternate uses is at issue (Box 3.4). Finally, local authorities must be held accountable for balancing all objectives of an urban land management strategy, forgoing the temptation to maximize land revenues when the costs of short- changing other objectives is self-evident and well-documented.24/ Individ- ual Chinese cities which chose to follow an unduly restrictive policy on land supply would suffer, over time, in efforts to compote with other localities. There is always potential for "government failure" as well as "market fail- ure," but there are mechanisms for minimizing both that are well-tested by international experience. 3.57 What then is the recommended implementation strategy? It takes as its model the previously issued urban housing sector reform report, wh -ch suggested two guiding principles: (M) that new investments enter the new system, to the maximum extent possible, thus diminishing the impact of past practices in spatial terms; and (ii) that the transformation of the prevsil- ing, nonmarket system be comprehensive in nature, announced at one time, and provided with an implementation time schedule that is both responsive to the disruptive effects of instant change and yet forces all users operating under the traditional system to begin making transition plans to allow users to enter the market economy within a certain period, preferably witlLn five years. 24! In Seoul, based on project-specific data, serviced land development yields internal rates of return for parcels sold on the market in excess of 70 percent, even though the Government's borrowing costs at the time equaled 10 percent. Internal rates of return simply allow benefits and costs to be "discounted" back to the time of land acquisition, even though cost and sales streams occur over various years. Costs and reve- nues are thus expressed in "comparable" terms, excludin the element of time. (Hannah, Kim, Hills$ "Land Use Controls and HousL..& Prices in Korea," op. cit.) - 69 - Box 3.4: HONG KONG EXPERIENCE WITH LEASEHOLD REGULATION Hong Kong's experiences, past and present, provide the bases for these recommenda- tions. For example, nominally nonrenewable leases not required by the government for other purposes are renewed, at a premium representing the market value of the land. Land lessees can apply for a renewal of the lease withln the last 20 years of the lease, thus providing land users with ircentives to continue to I-vest and redevelop the property. If the lease is due to expire iu 10 years, a new lease can be negotiated covering that period plus the normal time span of a new lease (75 years), or a total of 85 years. Under the recommendations provided in the "Report of the Special Comittee on Com- pensation and Betterment" (March 1992), leases previously written to contain very specific planning conditions would now be subject to a time limit, long enough to allow development "conditions" to remain in place for a specified period of time that would allow the original development to take place as planned. Thereafter, however, new development controls and conditions, including those that would restrict the redevelopment potential of the sites, would be automatic, as overall city zc -in regulations changed, without the need to compen- sate the lessee for otherwise unexpected losses due to "down-zoning." The issue of compen- sation for "breach of contract" is thus mitigated. 3.58 The previous Bank report on property markets focused on housing, issued clear warnings that thoroughgoing reform would affect an interconnected set of issues and interests, including the monetizing of "in-kind" benefits now made available to urban workers, the inability of many loss-making enter- prises to survive the reform, and the fiscal implications of any set of funda- mental changes. Land reform will raise similar issues. The maxim stated in the earlier report remains true in this case: it is far easier to abolish markets and market signals than to reestablish them. Reform must, neverthe- less, proceed. 3.59 One of the most difficult issues to be resolved concerns the mecha- nisms and institutional responsibilities for requisitioning suburban land to be converted into urban land. Two possible models can be identified: one in which a single agency, at the local level, would undertake this responsibility on a comprehensive basis for all the new land required by a partict!lar city; the second under which individual purchasers would directly deal with poten- tial suppliers of land. On the face of it, the latter approach appears more consistent with typical practice in market economies. The report concludes, hoJever, that in current Chinese circumstances, it faces two very serious obstacles. First, at this stage, many potential buyers still operate under soft budget constraints, and may well offer prices that "overshoot" likely market values. Second, rural sellers as yet pay virtually no taxes on any "windfall gains" that accompany the conversion of agricultural land to zoned and serviced subdivisions. They are thus in a position to hold out for a high percentage of such "windfall gains" that far exceed compensation for land and structures 'is presently zoned (i.e., for rural use). As such, a uniform pub- lic land acquisition approach, implemented under appropriate controls, may be considered as a short-term "second-best" substitute for a market where hard budget constraints prevail for end-users and for an explicit tax system to mitigate distributional concerns. Over time, as hard budget constraints take hold and rural communities must share any part of the "windfall gains" with the government, this second model of land can become the norm. - 70 - 3.60 The report therefore recommends that, for the immediate future, local branches of SLA should be responsible for the requisition of all subur- ban land to be converted to urban use, following a plan that allows for market-led urban demand in terms of locations requisitioned and hectares acquired during the year. Given the well-known fact that anticipated urban- ization greatly increases the implicit price of rural land, requisition should take place somewhat ahead of anticipated use; occupants could continue to use the land until actual urban development became necessary. Compensation rules should be reviewed for transparency and equity, with the aim of reducing the degree of leeway available for bargaining, and reducing all compensation to cash terms. There should be no explicit or implicit obligation for the state to provide "welfare" public housing, guarantee "urban" jobs for the displaced population, or provide for price subsid.es covering the "newly urbanized" dis- placed population, involved in the prucess of land requisition. Government might instead explicitly set aside a portion of the land area requisitioned, giving former users the right to "first refusal," exercising the option to lease that land before it becomes available for leasing to outsiders. Govern- ment would then lease the remaining land, through auction, tender, private treaty (negotiation), or grant, to urban users; land development comDanies could service the land and then either develop the sites for sale (land lease included) or transfer the serviced Rlots directly to new leaseholdere. inctud- in.yF mal1 r REDCs. 3.61 All greenfield development would thus enter the reform system iuie- diately, and all new users would hold leases. As far as possible, sites would be leased in a well-publicized and competitive manner, to both state and nonstate-owned enterprises, including those belonging to other cities or prov- inces. This would allow market prices, and property valuation data, to emerge spontaneously. Auctions of land should be encouraged, unless the particular priorities associated with a dite require local authorities to reserve the right to choose a "preferred use plan," as defined in relevant tender docu- ments, even if it is not accompanied by the highest bid. Leaseholds, other than those awarded through auctions, should always be assigned following a competitive process based on the issuance of tender documents easily accessi- ble to any interested potential investor. The evaluation of bids should be conducted in an open and transparent manner. "Private treaty" leases based on negotiations with only one party should be discouraged, because these provide no obvious benefits to the landowner, while opening up too many opportunities for the exploitation of auanxi, influence gained through the possession of good "insider connections." 3.62 Institutional users who are nonprofit in nature should only receive land grant leases where ilttle if any money is paid out, after a well- publicized, high-level local government review, to ensure that the opportunity cost of the relinquished site is justified by the benefits involved, and that alternative, lower-value sites have been considered and found to be unaccept- able. This goal may not be achieved in the short run, yet it is imperative to press forward with its implementation. In addition, all users favored by such grants should be subject to strict minimum floor/land area ratios, so that the density of development reflects the implicit, shadow price of the land areas. Any transfer of grant land to profit-making users should be subject to close scrutiny and taxation. - 71 - 3.63 Occupants of existing, developed sites should be given the option to lease or rent, with rent levels set high enough to provide a competitive rate of return on the implicit price (lease value) of the property. All redevelop- ment sites would be transferred to the leasehold system outlined earlier for greenfield sites, and auctioned off or subjected to competitive and well- publicized tenders, except where land grants are deemed to be absolutely necessary. 3.64 Redevelopment has proceeded at a relatively slow pace in China's cities. Within Shanghai's old city (9,000 ha), actual and planned redeveJop- ment covers 50-100 ha per annum (1985-95). This should not be blamed simply on urban planning and regulatory practices. The scale of potential redevelop- ment involved is enormous, and the density of old core areas, where residen- tial and nonresidential uses are intermingled without regard for underlying land values or environmental considerations, makes the task very difficult, even if those planners had at their disposal all the tools, policies, and institutions recommended in this report. Successful redevelopment requires, inter alia, a citywide (but core-oriented) strategy that provides for swift land acquisition; transparent compensation terms for existing occupants; and a set of planning regulations that balance minimal restrictions meant to ensure public policy objectives while maximizing the returns on invested capital; along with an "accommodating" infrastructure-upgrading strategy. In summary, redevelopment strategies should evolve from a dialogue between physical plan- ners, existing occupants, and promoters of a more competitive economy. Once targeted, following widely publicized hearings that provide for individual and enterprise inputs, land should be expropriated unless local occupants present plans that are consistent with citywide objectives and can be implemented in a voluntary but swift manner. The comparative experience of Singapore and Hong Kong provide polar opposite approaches to large-scale redevelopment; the first being more compulsory in nature; the second resulting in such limited outcomes that the head of the local redevelopment authority himself declared the model inoperative (Annex 7). 3.65 Compensation arrangements should be defined in cash terms alone, avoiding the plethora of in-kind or itemized benefits that bedevil land requi- sition in China's suburbs. As previously recommended in the Bank's housing report, existing occupants should be free to "buy into" the properties built in redeveloped sites; they would not possess significant entitlement rights, particularly while housing reform plans are still so modest in scope that prime locations could be held hostage to prior residents demanding reoccupancy rights for real estate on nonmarket terms, with housing treated as a near-free good. As long as housing costs are not viewed as an individual household's responsibility, "welfare" housing should be accommodated in suburban loca- tions, where the implicit value of land is but a fraction of prime land rede- veloped in the core. 3.66 The formula recommended for nonresidential land transfers, where the original occupants received land through administrative allocation mechanisms (assumed to involve little "price discipline" in terms of a profit-maximizing model), could apply here. Unless the local community could provide a viable alternative, the redevelopment project would be placed on the market for bids, based on tender documents. The original occupants would then receive the bulk of the proceeds, distributed according to space previously occupied, with the - 72 - Government retaining the residual. The locational choices of the displaced occupants would thereafter be driven by budget-constrained preferences. All this assumes that the planning norms and regulations governing redevelopment would be "developer-friendly" in nature, allowing a range of residential and nonresidential "solutions" including, in the extreme, the right of new occu- pants to lease relatively small amounts of prime space, while benefiting from the exceptional amenities provided by close access to the CBD. 3.67 Redevelopmenit aside, rents and lease values for occupants of exist- ing sites, who generally received their occupancy rights as a free good, must be established. The process could begin in locations and corridors that are self-evidently choice sites, either from a commercial point of view or from an assessment of their attractiveness as high-value residential sites. But, in short order, all the built-up area should enter the new system. 3.68 Based on emerging valuation data, present occupants of commercial and nonresidential space, generally, would face several choices. They could sign time-bound rental agreements that would subject them to "market-level" rents within, say, 3-5 years, but would be initially set at lower levels, particularly for noncommercial users. At that point, the occupants could opt to sign new rental agreements, based on annually adjusted market rents; par- ticipate in a competitive lease process, with the right to limited preferen- tial treatment (e.g., only 75-80 percent of the bid would have to be paid in cash); or relocate to new sites, in return for a significant share of any revenue generated from the (automatic) market leasing of the abandoned site.25/ 3.69 Alternatively, existing occupants could lease the site immediately, if they were willing to allow that site first to be placed on the market, and offered to alternate users, again in a well-publicized and competitive manner. As above, the winning bid, if it involved the existing occupant, would require that only a portion of the bid price be paid in cash. If the lease was awarded to a new occupant, then the original occupant would relocate and lease a new site. Here, as elsewhere, it bears reemphasizing that government reve- nue generation from land, though one objective of urban land management reform, should not be allowed to obstruct the emergence of explicit land prices. Thus, local authorities should err, if anything, on the side of gen- erosity, giving the bulk of the lease revenue generated through relocation to 25/ Bank recommendations on a rent and price reform in the housing market are covered in great detail in the previously released report, China: Imgie mentation Options for Urban HousinR Reform, op. cit. It recommends rents be increased, gradually, to cover full costs, using formulas to calzulate such costs, and implemented by fiat. The deregulation of existing public housing stock rents is thus seen as an issue involving a different pro- cess and time frame than that recommended for nonresidential users. - 73 - prior occupants.26/ Central or provincial authorities should, for exactly the same reasons, avoid injecting intergovernmental revenue-sharing consider- ations that would simply delay the reform process or drive it "underground" (Box 3.5). 3.70 This reform plan is audacious but also realistic. To begin vith, actual or proposed experiments along these lines are emerging in 1992. In Guangzhou, all new urban land is now leased or granted on a fixed-term basis. In experimental cities, like Jingzhou (Liaoning Province), all existing users wivl be allowed to pay a conversion tax that depends on the existing land use category involved;27/ thereafter the user becomes a leaseholder free to trade rights, mortgage property, etc. For all new urban land, land requisi- tion will be "unified" and performed by the city, which will service the land and then lease the land at a price that covers: (a) land requisition costs; (b) infrastructure development costs; and (c) a cost-plus markup ranging from zero percent for nonprofit institutions to 10-20 percent for industrial users, 20-30 percent for residential users, and 30-40 percent for commercial users. One disappointing feature of the model is that it precludes competitive bid- ding and views leasing as a mere "cost-recovery" mechanism driver. by arbitrary cost Plus formulas, rather than as a means to arrive at explicit market prices. This view obviously has widespread approval amone local and central Dolicvmakers. since most leases are understood as noncomvetitive "neRotiated" deals, where establishing market prices is obviously not facilitated. 3.71 Shanghai, as part of its Pudong development plan, is considering a series of initiatives, which would subsequently be introduced into Puxi, within a "suitable" period of time. All new land developed for commercial use would be leased. All commodity housing, direct foreign investment or joint venture, would be built on leased land or land requiring quasi-market "land use charges" bearing a relationship to likely leasehold values. New domestic users of industrial land would, however, be exempt from the reform, and would continue to pay only nominal land use taxes, while existing users of land on both sides of the river would find themselves obliged to pay a special "land 26/ The transformation of land use rights involved has few if &ny precedents. The sector team cannot confidently cite what the capital gains or trans- fer tax should be. On the one hand, Government should share in the pro- ceeds of the transactions; on the other hand, "confiscatory" taxes cou'd dimply halt the acceleration of legal transactions, which is seen in this report as a key element of reform. The obvious answer is to utilize the Chinese genius for experimentation. Cities should be given latitude to set the share captured by the Government, on a temporary, experimental basis. It would quickly become obvious, summing up the experience from a large number of cities, what tax rates are appropriate, i.e., rates that reserve a "fair share" of the proceeds for the government while promoting market-validated land transfers. Obviously, very high taxes, say 50 per- cent or above, will stymie the process. But the fine-tuning will have to be done by the Chinese, without the benefit of "best practices" drawn from international experience. 27/ Would a moderate transfer tax make more sense based on t4e time transac- tions occur? - 74 - fBox 3.5: CAN INDUSTRIAL REDEVELOPMENT BE FINANCED THROUGH DflFERENT: RENT? One question to ask is whether the differential in land value is sufficiently large to finance industrial relocation. If one draws on economic principles, then in cases which do not involve large "externalities" (e.g., pollution amelior4tion), It could be con- cluded that--if the transaction does not result In a financial surplus--ths overall economic benefit from the move must be suspect (in market economies, a firm will only move If it anticipates net savings). To look at empirical evidence, consider Fuzhou and Shanghai. In Fuzhou, the currept cost of land acquisition In suburban Industrial estates runs between Y 225 and Y 300/rn. Total land cos2ts, including in'rastructure lmprovem?nts and taxes, cur- rently range from Y 390 to Y 480/m . If we add to this Y 400 to Y 6001/r for building con- struction costs, then the total costs of new facilities is between f 790 and Y 1,080/m2 (assuming a FAR of 1:1). In the inner city of Fuzhou, where there is great development potential and where planning controls permit the construction of buildings with FARs between 11.8 and 1:2.5, the Fuzhou Construction Commission has recently tendered four sites to REDCs for the con- struction of commodity hoyasing and commercial projects. Tender prices for these sites range from Y 1,950 to Y 2,850/m . Thus, the differential land values between the city center and the suburban industrial areas are considersble, with the ratio of serviced inner city to suburban industrial land ranging f.om 4.1:1.0 to 7.5sl.0. If enterprises located in the central area of Fuzhou were permitted to sell their parcels or enter into cooperative agree- ments with REDCa, they could easily finance their industrial relocation. In fact, the Fuzhou Municipality could levy a tax on an enterprise's land sales profit and still leave enough after-tax profit to fully fund a typical industrial relocation project. Thus, it is possible to permit and tax land transfers and still provide the resources to funding industrial relocation and redevelopment. Another added benefit of per- mitting such market transfers is that it introduces market discipline to inner city redevel- opment. The REDCs would be forced to more carefully consider market demand for redevelop- ment before purchasing land for projects. In Shanghai, the potential benefits of permitting factorles to sell their land to REDCs could be substantial. As outlined above, under the Seventh Five-Year Plan, the typi- cal industrial relocation project cost Y 1,700/rn of vacated land area (in terms of the site relinmuished by the enterprise). It Is thus the actual break-even price that an enterprise would aeed to receive in order to pay for the relocation. Recent sales prices of parcels sold to overseas developers for industrial and commercial activities in the city center exceed Y 2,COO/ma. In fact, the Shanghai Municipal Government's recent proposal to the World Bank for the financing of nine pilot redevelopment projects in the city center esti- mates that the cleared sites could be leased for Y 10,000/m- Thus, while the actual market value for specific industrial sites in the inner area of Shanghai will vary according to location, size, and configuration and development potential, many sites In the city center can counavd prices that exceed relocation costs. Limited evidence from Fuzhou and Shanghai sungests that Inner city industrial redevelopment could be fully financed if industrial factories were permitted to sell their *lots to REDCs through competitive bidding or tender procedures. Government could tax a portion of the proceeds and still leave enough revenues for the enterprises to fulls finance their relocation. Given the enormous reguirements of industrial relocation and modernlza- tion. this 2ronosal warrants serious consideration. use charge" which may or may not approach the levels required of new land users. The proposal would be accompanied by other, complementary reforms. It assumes all corporate income taxes will fall from 55 percent to 33 percent within 2-3 years, beginning immediately. The Shanghai Housing Development Company, which now builds the bulk of noncommodity housing, would set up a land acquisition and land development subsidiary which would sell serviced land to the Government, a.id recover all costs incurred. This land would then be leased. - 75 - 3.72 If implemented, these experiments will represent major new initia- tives. They require concurrent reforms in planning, regulations, laws, and institutions, and they require the accelerated development of a central con- sensus. For example, will MOF, which explicitly abolished lozally set land use fees in i988, agree to the quasi-rent land charges proposed by Shanghai? This is not the time to keep key agencies isolated from the cutting edge of reform studies and proposals. 3.73 There are other reasons to believe the time is ripe for reform. After a decade of rapid increases in real household income, and in the wake of a rent reform program offset by compensating wage adjustments, it is not unrea- sonable to ask owners of residential structures to pay land rents or acquire land leases (particularly since reforms would begin with the limited pool of new housing occupants, who would rent or buy the annual increments to the urban housing stock, which average about 5 percent of the existing stock). By the year 2000, the World Bank Housing Report recommended that occupants of the existing (pre-1992) stock be expected to shoulder cost-recovery rents, ful- filling an objective essertially in line with current Government goals. Since these rents would be phased in gradually; since urban households can be expected to continue to experience rapid increases in real income;28/ and since massive housing construction in suburban locations is the norm; afford- able alternatives should be available for those unwilling, for example, to pay for the right to occupy sites with high amenity values (like the CBD). Hard- ship cases could be dealt with through "safety net" programs that subsidize the incomes of the certifiably needy. Government and other workers associated with nonprofit institutions could be included in these "safety net" programs or, better still, be allowed to earn annual incomes in line with their skills and responsibilities. 3.74 The thornier issue is what entitlement rights should be due to existing residents and users, as explicit or implicit land values rise (even if the areas involved are protected from the pressures of redevelopment, through historic preservation legislation). The best approach might be that resident entitlements, even in conservation areas, be limited and subject to market-value tests. No household or business would have the right to live or operate in a particular location at the expense of the community at large. The systematic preservation of a neighborhood can be defended, for reasons that tratnscend economics (even though market forces often validate these choices, after the fact). "Grandfather" clauses could allow long-term resi- dents to live out their lives in such areas, under rules and conditions unap- plicable to others, and revocable if they agree to "buy out" agreements that provide relocation incentives. But all other residents should be subject to market rules, including relocation. Market economies, at the local level, provide many examples of the inequities involved in trying to repress market transactions in neighborhoods slated for redevelopment or preservation. The public sector gains little from obstructing market forces in this context. 28/ Almost uniformly, the 14 cities that submitted proposals for considera- tion in recent discussions on a proposed World Bank Group Housing and Social Security Reform Project assumed household income, in real terms, will grow at 8-9 percent per annum during the 1990s. - 76 - And residents with no claims to entitlement rights should be encouraged to relocate to areas that their budgets dictate as feasible.29/ Recommendations for Improving the Feasibility of Residential Redevelopment Prolects 3.75 In many redevelopment projects across urban China, new commodity housing and commercial space is priced at two to four times its economic cost because redevelopment policies require this new space to fully cover the costs of all replacement housing. This cost differential is inefficient and imposes severe price distortions on the housing market. This section considers sev- eral alternative methods for reducing the break-even price of commodity hous- ing in redevelopment projects. There are essentially two options for reducing redevelopment costs and lowering break-even commodity prices. One approach maintains the existing allocation system and tries to lower the "shadow costs" of the allocations; the other method monetizes the transactions between the existing residents and the redeveloper. The options can be stated as follows: la) To the extent possible, increase the density of new development; limit the ratio of replacement to existing housing provided to existing occupants; limit the provision of public facility space to only what is required to service the project population, taking into account the capacity of adjacent facilities and services; and increase the off-site replacement of housing when off-site construc- tion costs are low and replacement standards are generous. (b) Monetize compensation for demolished housing, providing cash compen- sation for the (depreciated) value of the housing of redevelopment area residents; all housing constructed on the site is then sold as commodity housing and priced accordingly. 3.76 The various figures presented in Annex 8 describe the logic of option (a). To assess option (b) on monetizing redevelopment compensation, Table 3.15 illustrates various break-even commodity housing prices, assuming existing households in redevelopment projects are paid cash compensation for the loss of their units instead of in-kind compensation. Under such circum- stances, all new constructed space (less any public facilities) would be available for sale on the commodity market. As Table 3.15 re7eals, even if substantial cash compensation were offered for old units, break-even prices would still be lower than in comparable cases providing in-kind compensation. 3.77 However, if low rates of cash compensation were paid, existing own- ers of housing would need to find substantial sums for new units built in the redevelopment project area. Depending on the amount of compensation offered to existing residents, the commodity purchase price of new, on-site units would range from Y 60,200 to Y 72,650 (assuming that replacement units are 50 e2). In the case of very low cash compensation, existing households would receive only Y 3,500, less than 6 percent of the price of new on-site units. 29/ Annex 6 documents resettlement regulations emerging since 1991. These new rules have reduced entitlement rights. - 77 - Table 3.15: CASH COMPENSATION AKD 3N-SITE, NEAR-SUBURBAN AND FAR-SUBURBAN COMMODITY HOUSING PRICES (Yuan) Cash compen- Total Commodity grice of replacement housin_ sation per compensation On-site Near suburban Far suburban building (ma) per unit per unit per unit per unit 100 3,500 60,200 50,000 32,500 200 7,000 62,300 50,000 32,500 300 10,500 64,350 50,000 32,500 400 14,000 66,400 50,000 32,500 500 17,500 68,500 50,000 32,500 600 21,000 70,550 50,000 32,500 700 24,500 72,650 50,000 32,500 Note: On-site prices are based on simulations of break-even prices. Near- suburban prices are estimated at Y 1,000/m2. Far-suburban prices are estimated at Y 650/m2. Cash compensation ranges reflect the depreci- ated value of the building, with no compensation for lost land use rights. Source: Redevelopment Project Survey, 1991, 1992. If existing residents received Y 700/m2 in compensation, they would have Y 24,500, about one third of the purchase price of a new on-site unit. 3.78 If households were willing to relocate to other sites, however, they could purchase new flats at lower costs. By relocating to "near" suburban areas, typical flats could be purchased at about Y 1,000/m2, or Y 50,000. The lowest level of cash compensation would provide 7 percent of the purchase price. At the highest level, it would provide 49 percent. If the household was willing to relocate to "far" suburban projects, it would typically pay Y 650/r2 or Y 32,500 per unit. In such cases, the lowest level of cash com- pensation would provide 11 percent of the commodity purchase price; the high- est level of compensation would cover 75 percent of the purchase of a new unit. 3.79 Without a functioning housing finance system, households or owners (enterprises and government units) would be forced to pay cash for new units. If, however, there was long-term mortgage finance available, households or owners could borrow money on a mortgage basis. Table 3.16 illustrates hypo- thetical mortgage loans and per m2 monthly repayments for the repurchase of on-site, near-suburban, and far-suburban housing, assuming that loans are based on the purchase price less cash compensation, and they are for 20 years at an annual interest rate of 10 percent. Two important results are reflected in Table 3.16. First, purchasing on-site replacement housing is expensive, regardless of the level of compensation received. Second, if households moved to far-suburban locations and received Y 500-700/r2, they could actually afford to purchase units and service mortgage debt. - 78 - Table 3.16. PURCHASE La AND FINANCING Lb OF ON-SITE, NEAR-SUBURBAN AND FAR-SUBURBAN REPLACEMENT HOUSING FOR VARIOUS LEVELS OF CASH COMPENSATION (Yuan) Cash compen- On-site replacement Near-suburban Far-euburban sation per Loan Payment/ Loan Payment/ Loan Payment/ building (m2) amount month/r2 amount month/mr2 amount month/rn 100 56,700 10.94 46,500 8.98 29,000 5.60 200 55,300 10.67 43,000 8.30 25,500 4.92 300 53,850 10.39 39,500 7.62 22,000 4.25 400 52,400 10.11 36,000 6.95 18,500 3.57 500 51,000 9.84 32,500 6.27 15,000 2.89 600 49,550 9.56 29,000 5.60 11,500 2.22 700 48,150 9.29 25,500 4.92 8,000 1.54 /a Assumes that loan amount is the commodity purchase price less total cash compensation paid for old dwelling. Replacement units are 50 n2; original units are 35 i2. !b Assumes a 20-year fully amortizing mortgage at an annual interest rate of 10 percent. Source: Mission estimates. 3.80 Redevelopment projects would be far less complex and difficult to finance if sitting tenants and owners were provided with cash contributions for demolished units. Such an approach is far superior to the widespread practice of providing in-kind contributions. It would increase the transpar- ency of redevelopment finance and reveal true resource costs to redevelopment agencies and governments, as well as to the purchasers of commodity housing. 3.81 The payment of cash compensation for the loss of a dwelling unit, excluding compensation for lost land use rights, we-uld even in the best of circumstances still leave the household or owner with a substantial gap. As Table 3.16 reveals, the differential between the cash compensation and the purchase price of replacement housing ranges from Y 48,150 to Y 56,700 for on-site replacement, from Y 25,500 to Y 46,500 for near-suburban replacement, and from Y 8,000 to Y 29,000 for far-suburban replacement housing. The gap reflects the fact that it is current practice in Chinese cities not to pay compensation for land, since it is assumed that urban land is under state ownership. If redevelopment compensation was paid for land as well as build- ings, the gap between cash compensation and the commodity price of replacement housing could be closed. The last section of this chapter considers the financial implications of paying compensation for land. Broadenina the Base of Compensation--Paying for Land 3.82 If all redevelopment housing was sold on the commodity market, and if the pricing of such commodity housing was based on demand as well as supply factors, the prices of commodity housing would be based on location, access to - 79 - services and the quality of construction, not merely on the cost of the proj- ect. In such cases, dev_.-opers would purchase sites or carry out redevelop- ment projects predicated on the potential revenues generated from the redevel- opment, construction and sales of commodity housing or commercial space on a site. When commercial developers in market economies bid for sites, they typically use a technique called the "backdoor approach." This method is based on projecting the total sales revenues from building and selling commo- dity space in the potential project, subtracting all costs of construction and development (including fees, charges, interest, materials, provision of public facilities), developer overhead, and profit. The remaining amount, the "residual," is the amount the developer could pay for the land and still main- tain his target profit level. 3.83 To illustrate this approach, Table 3.17 presents estimated residual land values, asauming various commodity housing prices. The estimates are based on the baseline case presented above, and assume that all existing resi- dents receive cash compensation of Y 400/mh of constructed area for the loss of their old dwelling unit. As the table illustrates, the land residual value can be substantial, particularly with greater FAR and higher conmodity housing prices. Table 3.17: RESIDUAL LAND VALUES FOR VARIOV7 CONMODITY HOUSING PRICES, BY FAR C*ommodity Yuan land value per m2 of site area housing price FAR 1.5 FAR 2.0 FAR 2.5 (Yuan) 1,400 22 136 250 1,600 307 516 725 1,800 592 896 1,200 2,000 877 1,276 1,675 2,200 1,162 1,656 2,150 2,400 1,447 2,036 2,625 2,600 1,732 2,416 3,100 Source: Simulation of base-case redevelopment project, 1992. 3.84 If existing residents and owners were fully compensated for their units, including land value, they would receive substantial sums. If house- holds elect to relocate to suburban projects, they would receive most, if not all, of the funds needed to purchase replacement unite. As Table 3.18 shows, compensation for land can be more significant than for the structure and pro- vide substantial resources for the purchase new units on the commodity market. 3.85 If cash compensation were paid for land and buildings, redevelopment projects could be implemented through competitive bidding for redevelopment project sites; REDCs wanting to redevelop areas would bid for the purchase of the site. The bid-purchases would reflect the REDCs' estimates of the con- struction costs, sales revenues and compensation for demolished buildings and land. The highest bidders would gain redevelopment rights. Existing owners - 80 - Table 3.18: ESTIMATES OF POTENTIAL LAND AhD BUILDING COMPENSATION (Yuan) Commodity Building Land Total housing compensation compensation coapensation price per unit per unit per unit 1,400 14,000 5,953 19,953 1,600 14,000 22,578 36,578 1,800 14,000 39,203 53,202 2,000 14,000 55,828 69,828 2,200 14,000 72,452 86,452 2,400 14,000 89,078 103,078 2,600 14,000 105,703 119,703 Source: Simulation of base-case redevelopment project, 1992. of buildings on the site would receive compensation for the depreciated value of their structure and a pro-rata share of the land value of the site. The pro-rata share should be based on the land value bid, divided by the original constructed area of the site. Occupants of the site would receive a payment based on their preredevelopment ownership of constructed space. 3.86 Local government could tax away some of the land-related compensa- tion. The tax rate should be set to ensure that existing owners were left with enough proceeds to purchase comparable replacement housing. This would be set by limiting total compensation to the lesser of the cash compensation for the structure and the pro-rata value of the land or the prevailing commo- dity housing price (per square meter times the amount of space to be pur- chased) for comparably located commodity housing. 3.87 Cash compensation for buildings and land given up for redevelopment would greatly enhance the efficiency and transparency of redevelopment finan- cing. By broadening compensation payments to include land, affected house- holds and owners would receive substantial funds to apply toward the purchase of replacement housing on the commodity market. Note, however, should be taken that land reform and housing reform are inseparable. Commodity housing options (location, unit size) available to individuals are still limited in many cities. Mere improvement in compensation rules, without greater choice of alternative commodity housing, might render the strategy unsustaina- ble.30/ 3.88 Will business enterprises find this price and property rights reform affordable? Here, one should differentiate between enterprises that do (or should) already face hard-budget constraints and others (largely state-owned industrial enterprises), and then only temporarily. If the Bank's interpreta- tion of Government's reform objectives is correct, standard corporate income tax rates will drop from 55 percent to 33 percent, all enterprises will be 30/ For regulations in other Asian cities, see Annex 7. - 81 - responsible for their profits and losses by the year 2000, and face bankruptcy or marger as the only real options to overcome poor performance. Enterprises subject to price controls wcald, by then, be a vestige of the past. Most prices would be Pet by the m'_ket, and even regulated utilities within mono- poly markets would charge regulated prices yielding cost-recovery rates, including an "adequate" rate of return on investment. 3.89 Even now, commercial enterprises have no reason to operate under a protective "state-owned" umbrella. The political imperative to keep the "ccm- manding heights" of the industrial sector under state ownership ard management no longer covers the urban services sector. There is little reason why com- mercial activity should be run by other than leaseholders with anything but profit maximization ir mind. Certain enterprises, including large department stores, operating outside the market economy, should be transferred to the market system, using the various mechanisms already employed to effectively privatize their smaller-scale counterparts. The "problem" industrial enter- prises owned by the state and today viewed as not viable will, in time, be stripped of nonmarket constraints that prevent profit maximization, and .be forced to "sink or swim." These erterprises will either cease to exist or be able to shoulder the burdens of operating within a market-driven, competitive economy. 3.90 Beyond this, and anticipating the discussion of the net increases in government revenues that can be expected once the new system is fully in place, note should be taken of the implications of the tax code on affordabil- ity. Depending on the assumptions made about the annual tax liabilities of a given enterprise, which, in China, can vary dramatically--for reasons dealt with in other Bank studies--at most one third of any land rent will represent net additional revenue, the rest being offset in the process of estimaating taxable income because these will be treated as pretaxable deductible costs. Similarly, land lease costs would not simply become an added burden to the enterprise, or net income to the Government that first conveys the land; these costs will be treated as assets subject to depreciation allowances that reduce taxable income. Under current rules, allowing loan repayments as well as interest to be treated as tax-deductible, under most circumstances. Enter- prise tax liabilities could be further decreased to the extent that land leases are financed by bank loans.31/ At the same time, of course, the net revenue gains to the Government would be correspondingly reduced. 3.91 Particularly for industrial enterprises located in suburban loca- tions, the new arrangements will have their greatest impact in making land prices explicit, and not in greatly affecting prices, enterprise retained earnings or generating massive net increases in Government revenue. 3.92 Obviously, there are many factories that are inappropriately located. The work by Bahl and Zhang,32/ regressing profits against loca- tion, controlling for other variables in two cities; and less sophisticated 31/ The tax deductibility of loan principal is currently being phased out, as the above-cited tax reform is introduced. 32/ R. Bahl and J. Zhang, Taxint Urban Land in China, op. cit. - 82 - studies cor:elating gross profits par m2 of land use at different locations (e.&., Shanghai's 1985 study)s;3/ document the fact that "centrality" adds littlc to most inductrial activities and that, by implication, the emergence of expħicit land prices could torce relocation. This is precisely one of the intents of urban land reform, one which can be achieved in a gradual manner by pnasii.g in the full impact of market prices, and one which can bring residual benefits to the relocating firms, which would share in the monetized valte- added created by the old site's occupancy by a new user paying market prices. This reform is likely to take place in an environment of rapid economic growth and low rates of inflation, with enormous demands for additiovs to the urban services sector. Both offer favorable opportunities for restructuring in urban areas that are currently danied to virtually all other economies in transition to a regulated market economy, where high inflation and negative growth rates are the norm. E. Land Rev,nues: Will They Prove Significant in Financing City Expenditures? 3.93 Cities and towns will require substantial investments in serviced land development and in infrastructure to recycle land uses in the existing core areas and accommodate greenfield expansion. Because redeveloped land currently provides for only one quarter of that required for new construction, the built-up area of cities and town is expanding at a rate of 6 percent a year, enough to double the developed urban area every 15 years. The true burden of associated infrastructure investments is not easy to calculate because citias resort to elaborate "off-off-budget" mechanisms to finance infrastructure construction. This practice is due partly to a reasonable desire to make direct beneficiaries shoulder a greater share of the costs than in the past, and partly to the fact that China's taxation system allows cities and towns to retain all such reveilues locally, while having to share the pro- ceeds of most conventional revenues with provincial and central government authorities. This creates enormous incentives to shift expenditure sources away from those subject to intergovernmenta.. revenue-sharing (Box 3.6). Mis- sion estimates suggest that the actual average expenditure on urban infra- structure and maintenance is close to 5-6 percent of urban GDP, and may well be higher than that. 3.94 The issue of municipal finance clearly goes beyond one of merely accounting for the ever-widening sources of infrastructure financing. Cities need to consider increased user charges for municipal 3ervices, the monetiza- tion of land values, and greater reliance on long-term financing from capital markets and banks. They must also, however, pay far more attention to expen- diture planning and budget programming, as well as financial management and accounting. This means, inter alia: (a) preparing medium-term consolidated revenue and expenditure forecasts for the purposes of evaluating the feasibil- ity of development options; (b) developing "rolling" three-year investment programs, including operations and maintenance (O&M) implications, based on transparent criteria for determining preferred capital and recurrent expendi- tures; (c) reviewing fiscal subsidies, their objectives, and their incidence, knowing that every yuan in subsidies means one yuan less to promote municipal 33/ Ibid. - 83 - Box 3.6: URBAN CONSTRUCTION FINANCING IN SHANGHAI Before 1980, urban construction was financed by city budget allocations as part of the fixed capital Investment program. Most of the completed urban iafrastructure and ser- vices were provide: free of charge. Certain public uti>Itios utilized user charges, but there were large gapq between revenuep and costs. During the 1980s, demand fvr additional urban infrastructure and services increased dramaically. In order to finance such increases, there was a shift from local budget allocations and greater reliance on drawing upon the resources of enterprises and *society." There are now basically thrne arrangements for obtaining additional urban con- struction funds: (a) to tie nonprofit urban ':nstruction projecte to profit-oriented industrial proj- ects, and implement them at the same time. Once the projects are completed, the profits from the Industrial projects are used to pay back loans used for urban construct projects. Such loans are treated as tax-deductible items. For exam- ple, the Nanpu Bridge, Shanghai's No. 1 Subway Line, and the Shanghai Sewage Proj- ect are all financed through this mechanism; (b) to have enterprises include in their costs items that cover urban infrastructure investments In Industrial districts or development zones. Such costs are tax deductible. The Taopu industrial district and Caohejin development zone have been financed through this mechanism. The urban infrastructure of the Hongqlao and Minghang development zones have been financed through government Itterest-free loans. They will be paid back by enterprises within the zones over the nexc few years; (c) to collect various user charges or fees, including off-site infrastructure fees, bridge and tunnel tolls, sewerage fees, land use fees, excess water consumption fees, and more conventional utility charges. In April 1988, to reinforce the management of the urban infrastructure financial resources, a "Shanghai Urban Construction Fund" (SUCF) was set up. Its main responsibility 'is to collect, use and manage those funds. The SUCF is under the control of the Construction Commission. It has its own supervision board and management board. The supervision board consists of a deputy mayor, and the directors of the Finance and Audit Bureaus. The management board consists of mana- gers from the Construction Commission, Planning Commission, Finance Bureau, the People's Construction Bank of China (PCBC) and other related agencies. The expenditures of the SUCF are first approved by the Planning Comnission. Then the city Finance Bureau determines the budget for urban maintenance and investment, which SUCF uses to Implement Its proposals. The SUCF budget is not comprehensive; it excludes the official urban construction livestment budget and "off-off-budget" spending. Relying solely on the SUCF data, however, it appears that the local financial budget funds cover no more than 40 percent of reported expenditures, and is falling. Source: Shanghai Construction Commiasion. development; (d) reviewing possible borrowing and debt options and management strategies, and develop relevant debt service risk analysis tools; and (e) exploring ways of strengthening tax and fee administration, while reexam- ining the incidence, efficiency, equity, and transparency of such taxes and fees. 3.95 A cross-section of land-revenue time series suggests that such income could become an important source of urban construction financing (Tables 3.19-3.24). At present, however, much of the revenue is earmarked for expanding the supply of land available for cultivable uses, or is subject to - 84 - intergovernmental revenue-sharing rules. Furthermcre, any judgment depends on the rate at which land lease sales expand and what proportion of that mnoney is held back by local governments. Land leasing is clearly growing in scooe, but potential revenues have been kept down because local authorities fear the impact of market prices for land on new occupants. In addition, the incen- tives required to attract present land occupants into the market dictate that the latter be allowed to retain a significant share of lease revenues for the foreseeable future. Given that the central government is also entitled to roughly one third of gross revenues, it is unlikely that local governments can count on land revenues as a providential source of m:ney with which to expand infrastructure co.istruction and maintenance budgets in the ?ihort or medium term. 3.96 Today, local governments are trying to finance serviced laen rel- opment and major infrastructure projects by adding a plethora of fees &itd charges to budget and nationally sanctioned off-budget sourcea. While a move to greater beneficiary financing is necessavy and defensible on both effi- ciency and equity grounds, the proliferation of fees and charges appears to be a hit-and-riss affair. This topic should be studied carefully, to determine both the best sources and the likely size of local government land revenues. International experience already suggests, however, that expectations should be held in check. Hong Kong is often cited as a case where land revenues, driven by the sale of leasehold rtghts, can substantially alter the budget constraints experienced at the loc.t level. One or two years excepted, this is simply not borne out by the data (i.ble 3.25). Worldwide, land lease reve- nues aside, the experience with land tax instruments has proved uniformly dis- appointing. In some market economies, it is possi.-e to collect the equiva- lent of 1-2 percent of real estate market value in the form of property taxes. This is the case in the United States. If land is considered to equal an average of 20-25 percent of the property, then the land portion of the tax typically contributes 2-7 percent of all local revenues in cities like Boston, New York, Philadelphia, Atlanta, Baltimore, and Los Angeles. More likely, China will learn more from the experience of developing countries, where the property tax yield is only one tenth to one fifth of that level, and is even less of a major revenue generator. 3.97 Recent World Bank work on property taxes, moreover, carries a set of cautionary warnings, which, by extension, can apply to any land tax instru- ments:34/ * Property taxes can be extremely unpopular as they often represent one of the few instances where a large percentage of households confront a tax bill. In developing countries, revenues from house- holds are usually derived from indirect taxes, whose incidence is largely hidden in the form of higher prices. Where payroll and income taxes exist, their coverage is often limited (and evasion is minimized through withholding at the source). 34/ W. Dillinger, Urban ProRerty Tax Reform: Guidelines and Recommendations, World Bank, April 1991; R. Bahl and J. Zhang, Taxkng Urban Land in China, World Bank, Report No. INU-39, March 1989. - 85 - Table 3.19: LAND REVE Ui IN SHANCRAI: 1987-91 (Y million) 1987 1988 1989 1990 1991 Total Cultivated land tax La 37 89 57 39 88 310 Vegetable land fund 103 106 65 25 40 339 Land use fee for joint ventures 6 11 9 11 17 54 Land lease sale /b - 104 31 134 149 418 Urban land use tax /c - - 236 206 230 672 Total Ld 146 310 398 413 523 1.793 L The cultivated land tax ranges from Y 3 to Y 10/m2, based on different per capita land use and economic conditions in each township. The revenue is shared with central government, which receives 30 percent of the proceeds. The tax rate is cut in half for farmer housing. For land taken for rural highway use, the rate is Y 2/i2. Total charged land area for each year is as follows: (hectares) 1987 1988 1989 1990 1991 Total 23.67 20.53 12.13 11.67 n.a. 68.00 lb The "state" (MOF) shares 32 porcent of the gross revenue. /c The tax is shared 50/50 -- th the state. In 1990, however, unpa-." taxes by the Baoshan Steel Company (Y 24 million) and Shanghai No. 1 and No. 5 Factories (Y 6 million) were not included. /a By comparison; Shanghai collects approximately Y 500 million each year through its urban maintenance and construction tax. Source: Shanghai Municipal Government. Table 3.20: LAND REVENUE RETAINED BY SHANGHAI: 1987-90 (Y million) 1987 1988 1989 1990 Total Cultivated land tax 26 62 40 27 155 Vegetable land fund i03 106 65 25 299 Land use fee for joint ventures 6 .1 9 11 37 Land lease sale - 71 21 91 183 Urban land use tax - - 118 103 221 Total 135 250 253 2a 895 Source: Shanghai Municipal Government. - 86 - Table 3,21: LAND REVENUE IN FJZHOU: 1986-91 (Y m4llion) 1986 1987 1988 1989 1990 1991 Total Cultivated land tax la - - - 2.33 1.62 1.97 5.93 Vegetable fund 1.80 2.21 0.42 2.15 2.67 4.70 13.95 Fishpond fund 0.51 0.16 0.02 0.57 0.98 1.15 3.39 Off-site fee /b 1.06 9.70 11..00 6.75 4.90 11.46 44.87 Education fee - - - 8.11 10.50 18.61 Location price Ld - - - 1.62 - - 1.62 Land value-added fee is - - - 6.45 9.10 1.60 17.15 Land use tax /f - - 0.16 6.63 10.36 13.55 30.70 Lease sale Lg - - - 10.88 4.64 15.52 Total /h 3.37 12.07 11.60 26.50 48.62 49.58 1A1,73 ja The tax rate ranges from Y 2 to Y 10/rn2, based on different per-capita land area ard different economic conditions in each county. Tax rate Per caRita arable land Paddg field Arable land Less than 0.5 mu (33.5 m2) Y 10/ml Y 8/rn 0.5-1.0 mu (33.5-66.7 F2) Y 8/ma Y 6/ma 1.0-1.5 mu (66.7-100 mr Y 6/12 Y 4/mr More than 1.5 mu (100 m) Y 4/12 Y 2/12 jb The off-site infrastructure fee was implemented in 1985. Based on differ- ent classes (hot spring protection zone and three land classes) and dif- ferent land uses, the fee ranges from Y 45/rn to Y 7.5/m2. /c The urban education construction fee, implemented in 1990k is designed to improve school conditions for Fuzhou. The rate is Y 12/rn of floor space in the city proper, Y 10/mr for county towns, and Y 8/m2 for designated towns. All commercial real estate projects have to pay Y 15/mr. !d The lor-ation price differential fee is collected from all commodity hous- ing development that is not on leased land. About Y 100/rn of floor space is collected. Le The land value appreciation fee is based on the percent of sale price as the comprehensive building cost. If the sale price is below 100 percent of building cost, 15 percent of value appreciation fee will be collected. For the portion between 100 percent to 200 percent of building cost, 30 percent value appreciation will be collected. For the portion between 200 percent to 300 percent, 40 percent value appreciation fee will be collected. For the portion above 300 percent, 50 percent value apprecia- tion fee will be collected. If The land use tax was impler-ited in 1988. Unlike many other large cities, where land use taxes are si.-zed between the central government and munici- pal government, the land use tax in Fuzhou is shared between Fujian Prov- ince and the central government. LS The .and lease revenue is shared with the central government. Land lease sale revenue figures only include the amount retained by the city, not total number. Lh By com,arison, Fuzhou collected approximately Y 28.71 million each year througi its urban maintenance and construction tax (1989 figure). Source: Fuzhou Municipal Gowernment. 87 - Table 3.22: LAND REVENUE IN GUANGZHOUt 1984-91 (Y million) 1984 1985 1986 1987 1988 1989 1990 1991 Total Cultivated land tax - - - 20.7 36.6 25.2 10.9 n.a. 93.4 Vegetable land fund L_ - - - n.a. n.a. n.a. n.a. n.a. n.a. Land use fees /b 1.5 13.5 14.5 20.0 20.2 4.5 5.0 5.4 84.6 (1) 1.5 12.0 12.3 !7.0 17.8 - - - 60.6 (2) - 1.5 2.2 3.0 2.4 4.5 5.0 5.4 24.0 Land lease sale /C - - - - 280.0 109.7 129.7 205.5 724.9 Land use tax /d - - - - - 69.1 66.2 63.9 199.2 Total la _1.5 13.5 14.5 40.7 336. 208.5 211.8 274.8 1.102.! La No data are available. Lb (1) Fees were collected from aomestic and foreign land users from 1984 to 1988. (2) Fees are still collected from joint ventures. Ls The state shares 32 percent of gross revenue. According to city offi- cials, total revenues were over Y 300 million (120.76 ha), plus UK$155 million, as well as 60,000 m2 housing during the period of 1986 to 1990. id It is shared 50/50 with the state. The potential taxable area is 58 kin2, and the area actually covered by the tax, arter 'exemptions," totaled 30 km1. A tax revenue of Y 63 million is an estimate. ie By way of comparison, Guangzhou collects about Y 200 million a year through its urban maintenance and construction tax. Sources Guangzhou Municipal Government. - 88 - Table 3.23: LAND REVENUE IN CHENGDU: 1987-91 (Y million) 1987 1988 1989 1990 1991 Total Cultivated land tax (04/87) La 82.63 25.05 21.24 30.48 30.00 189.40 Vegetable land fund /b n.a. n.a. n.a. n.a. n.a. n.a. Urban land use tax (11/&d) Le - - 18.42 20.50 18.00 56.92 Land use fee for joint ventures (1991) /d _ _ _ _ Land lease revenues /a - - - - - - Total lf 82.63 25.05 39.66 50.98 48.00 246.32 /a The cultivated land tax ranges from Y 1 to Y 10/m2, based on different per-capita land area and different economic conditions in each township. The revenue is shared among four levels of governments: Central government 30Z Provincial government 102 City government 10% County/district government 50% The amount of cultivated land involved are as follow: 1987 1988 1989 1990 1991 Total Average area (ha) 533.13 335.87 49G.73 341.07 n.a. 1,706.80 Average price (Y/m2) 15.50 7.46 4.28 8.94 n.a. 9.34 Lb The rate for the vegetable fund is set at Y 10,000 per mu, or at Y 15/rn2. No data are available on this item. /c The tax is shared 50/50 with the state. /d The land use fee is set at Y 0.5-19/m2. It started in 1991. Estimated annual collections would total about Y 1.87 million per year. So far, Y 0.31 million have been collected, because widespread "exemptions" have been granted. le No land lease sales have been made in Chengdu. /f By way of comparison, Chengdu collects approximately Y 100 million through its urban maintenance and construction tax. Source: Chengdu Municipal Government. - 89 - Table 3.24: LAND-RELATED REVENUE IN HANGZHOU: 1988-90 (Y million) 1988 1989 1990 Total Land reclamation fee Ia 3.10 2.71 1.84 7.65 Vegetable land fund lb 2.50 3.17 1.61 7.28 Land use tax /c - 18.93 21.20 40.13 Cultivated land tax /d 5.39 5.39 3.55 14.33 Total 10.99 30.20 28.20 69.39 /a The land reclamation fee was implemented in 1982. The rate was Y 5,000 per mu. Only cultivated land was charged this fee. Since 1987, with adoption of the cultivated land conversion tax, the rate was cut by half, to Y 2,500 per mu. /b The vegetable land fund was implemented in 1986. The rate is Y 7,000 per mu. On vegetable land, no land reclamation fee is collected. jc The land use tax was implemented in 1989. The rate in Hangzhou ranges from Y 3 per m1 to Y 0.5 per i2. /d The cultivated land tax was implemented in 1987. The average rate is Y 6,000 per mu, or Y 7-10 per in. Source: Hangzhou Municipal Government. - 90 - Table 3.25: REVENUE FROM LAND IN HONG KONG (HK$ million) Total Property Total Total Land revenues land Crown Property income land government as X of premium rent rates tax revenue revenue total revenue 1974 318.6 34.7 368.9 141.1 863.3 5,305.8 16.3 1975 287.4 40.8 407.9 148.0 884.1 5,973.1 14.8 1976 345.9 68.2 534.4 147.7 1,096.2 6,724.6 16.3 1977 557.3 60.5 618.7 252.7 1,489.2 7,575.8 19.7 1978 1,831.3 89.4 722.9 243.2 2,886.8 10,232.6 28.2 1979 2,007.8 92.5 806.9 238.5 3,145.7 12,557.0 25.5 1980 2,845.2 136.0 889.3 297.8 4,168.3 16,796.1 24.8 1981 10,769.1 158.4 986.1 330.9 12,245.2 30,290.3 40.4 1982 9,676.5 238.5 1,050.9 766.2 11,732.1 34,312.9 34.2 1983 5,048.2 180.2 696.9 779.5 6,704.8 31,097.6 21.6 1984 2,267.1 231.1 1,155.5 601.9 4,255.6 30,399.7 14.0 1985 4,267.2 235.3 1,222.2 850.8 6,575.5 36,342.5 18.1 1986 3,895.0 272.4 1,770.0 701.5 6,638.9 43,695.0 15.2 1987 756.0 342.6 1,188.0 902.7 3,189.3 48,602.0 6.6 1988 461.0 491.3 1,373.0 895.1 3,220.' 60,875.0 5.2 1989 365.0 632.1 1,517.0 869.5 3,383.6 72,658.0 4.7 1990 212.0 767.8 1,663.0 953.1 3,595.9 82,429.0 4.3 Sources: Director of Accounting Services, Annual Departmental Rewort, 1974-90; Commissioner for Inland Revenue, Annual Departmental Report, 1974-90; and Hong Kong Government, Hons Song Yearbook, 1974-91. - Too much time has often been spent on the "upstream" stages of prop- erty taxation--discovery and valuation, important, but also costly-- to the neglect of collection.35/ * Property tax policies--on the rate structure and method of periodic revaluation--have generally led to decisions to impose low nominal tax rates, often in response to political pressures from those assessed, which have proved uniformly heavy around the world. * The costs of administering the tax are not negligible and, if low tax rates are adopted, the program may generate little net revenue, once such costs are deducted. * The imposition of a tax will require not only the completion of a plot-by-plot cadaster but also a computerized system linking the industrial and commercial bureaus, the state land acquisition authorities, the building admiristrations, and the tax authorities to update the cadaster. * Property valuation is a major roadblock and procedures should be simplified, by omitting or charging a flat licensing fee for most small properties. The bulk of revenues are likely to be concen- trated among major land users and CBD users such as hotels, depart- ment stores, and office buildings (and, therefore, valuation efforts should also be focused on this small group). 35/ See footnote (c), Table 3.19 for a cautionary tale involving China. - 91 - * Authorities should avoid giving tax concessions to users who are self-evidently located on land too valuable to justify current use. However, the efficiency benefit of property taxation in encouraging the relocation of such occupiers should not be exaggerated. F. The Legal Underpinnings of Price and Property Rights Reforms 3.98 Properly functioning property markets presuppose a legal framework in place, broadly defined to cover ownership, instruments, institutions, and resolution-dispute mechanisms comparable to those found in market economies. As leasehold systems are adopted more widely, China's authorities will face challenges in each of these areas. 3.99 Currently, who actually represents the state as the titleholder is uncertain. China should assign that role to either a national or local insti- tution. The issue may appear academic; it is not. Particularly for foreign investors, it contributes to a perception of opaqueness and lack of sophisti- cation. As always, those investors have many options across the Pacific Rim; these concerns should not be taken lightly. The report recommends that the local State Land Administration, which is already the key management agent in land-related issues, be clearly designated as the agency to act on behalf of the state with respect to urban land and conversions to urban land. 3.100 Market-driven property transactions, including land use rights transfers between enterprises and institutions, require: (a) easy access to a systematic body of information, including land title documents that summarize all present rights, obligations, and encumbrances (mortgages, etc.); land use master plans and zoning or equivalent site-specific use restrictions; (b) dis- pute-resolution systems by an independent judiciary with a credible mandate to render judgments against Chinese nationals bnd Government itself; and (c) property markets that allow leaseholders to sell their use rights, and mortgagors who hold property gained by default to dispose of these with rela- tively little interference from the public sector. And, quite obviously, lease documents, and the tender documents that will precede them, must follow a format and level of detail consistent with international best practices. 3.101 Chapters IV and V deal with the information requirements of urban land management reform. Anticipating those discussions, one can simply assert that fragmentation of authority in record-keeping (leading to relevant docu- ments being assembled and filed by different agencies, reporting different sets of rights and obligations.); the lack of mechanisms to rapidly access data (which often cannot be retrieved using modern data-processing techniques); and the perception that information is, in general, proprietary in nature and not available on demand; all these hinder the emergence of markets. The Bank study team was often forced to work with tourist city maps that omitted scale measures, and was sometimes denied access to large-scale master plans. If similar information in competing cities of Asia can be obtained as a matter of right, and with a minimum of Governmental interference, then China's cities - 92 - vill have to do a far better job of acknowledging consumer rights in this field. Secrecy and land market development are simply incompatible.36/ 3.102 Anecdotal evidence, gathered from foreign-funded enterprises which has not yet been complemented by any sysrematic study, suggests that property dispute-resolution procedures are unsatisfactory, relying primarily on admin- istrative arbitration methods and then, as a last resort, on a court system that has a reputation for limited independence, when confronted with litigants linked, directly or indirectly, to Government. The presumption exists that arbitration and judiciary procedures, when put to the test, will rule against the plaintiffs, be they foreign or domestic. Again, this judgment is subject to verification possible only after an extensive review of experience to date. Until such a time, one can only reccrd that prevailing anecdotal evidence presents a negative verdict, with arbitrators and judges often viewed as act- ing on behalf of the interests of local authorities. That this will have a chilling effect on urban land management reform hardly needs underlining. 3.103 The accumulation of huge interenterprise unpaid debts, documented earlier, and the almost total absence of successful default judgments result- ing in lenders acquiring real estate collateral, suggests not only that the judicial system is weak but also, upon further investigation, that the dis- posal of collateral property returned to the lender is not without problems. The transfer of property from one "owner" to another requires the fulfillment of onerous reviews and permissions, and renders the option often so problema- tic that overseas mortgage holders simply fail to foreclose on a property that would otherwise be taken and disposed of in a market economy. The Bank's previous report on housing reform stressed the pivotal role of financial intermediaries in the development of a commercial real estate market. This report can only reaffirm those conclusions. If foreign and domestic financial intermediaries feel existing procedures discriminate against them in property- related lending--making default judgments difficult and property disposal sub- ject to heavy public sector interference--then the reform effort will be stalled by what might appear, superficially, to be a mere technicality. 3.104 Legal rights and obligations, properly recorded, open to public inspection, and easily enforceable, may seem peripheral to the overall reform of real estate markets. Nothing could be further from the truth. A building incorporating one fundamental structural flaw may be rendered unusable though all other elements are in order. However difficult it may be to move in a coordinated fashion across all aspects of urban land management reform, an effort must be made to meet this goal as expeditiously as possible. 36/ As noted, this penchant for secrecy extends to otherwise routine docu- ments like municipal budgets. Foreigners also complain that procedures for issuing permits and permissions that accompany investments in China are often unpublished and unavailable for scrutiny. - 93 - IV. REINVENTING THE CHINESE CITY: PLANNING WITH PROPERTY MARKETS IN MIND A. Introduction 4.1 How have urban planners responded to the emergence of market and quasi-market pressures? And what remains to be done in the reform of urban planning? Those are the key issues to be addressed in this chapter.l/ The key messages can be summarized from the start: (a) actual development and its accompanying small-area infrastructure investment requirements are not syn- chronized, due to the rigidities imposed by rarely revised Master Plans; and (b) monitoring of actual small-area development is critical to any systemic reform which accommodates actual development with supportive infrastructure investments. However, at the margin, particularly at the district level of coastal cities, planners are beginning to accommodate developer-driven demands for more flexible norms and regulations that allow new eevelopments and rede- velopment in old core-city areas to achieve rates of return that are roughly comparable with other investment opportunities, and certainly higher than returns from "passive" investments in bonds or long-term savings accounts. In addition, criticism of local planning practices, catalogued below, does not imply that local urban planners are themselves blind to the inadequacies of the resultant "outputs." Planners Ln China operate within a bureaucratic environment that often precludes "first best" solutions because coordinated responses to the issues cited are not easy as long as decision-making author- ity is fragmented between and within layers of government. B. The Orderly City: Local Urban Development Planning in China 4.2 In most countries self-described as "unitary" in character, and with a strong ideological commitment to administer the structure and spatial dis- tribution of economic activity, one would expect local development planning, sensitive to market de. and, to be of secondary importance. Detailed plans, laws, and regulations would be prepared by powerful central ministries, and central budget allocations would help implement the key infrastructure, eco- nomic and social projects that would shape the urban landscape. The outside observers would therefore look first to the central bureaucrats for an under- standing of what is happening and why. In China, with some exceptions, one should direct one's attention to the local level instead. The central author- ities continue to produce laws, regulations, norms and guidelines, many of which influence local planning, and these Beijing-based planners have the power to approve or deny many local practices. They do so within a tradi- tional land use management framework that does not have thoroughgoing market- driven reform at its core. But de facto they exercise such power very selec- 1/ As background, the Bank has produced two reports bearing directly on this topic. One, on urbanization in Zhejiang Province, was based on field work carried out in five cities in 1986, and was issued in 1987. The other, focused on housing, drew upon preparation missions undertaken in 1990, was issued in June 1991; it was based primarily on a study of the provincial municipality of Tianjin (World Bank, China: Zheiianx: Chal- lenges of Rapid Urbanization, Report No. 6612-CHA, August 3, 1987; China: Implementation Options for Housina Reform, op. cit). - 94 - tively. Particularly today, the power that comes with the ability to allocate significant budgetary resources to implement centrally conceived "visions" has been replaced by a decentralized model, characterized by self-reliant cities, enterprises and households.2/ A central policy "vacuum," viewed from the perspective of the requirements for developing a regulated market economy, is detrimental to overall systemic reform, even if local development can be viewed, qu'te properly, as an area that should be largely managed at the local level. 4.3 To understand how cities are built in China, one must first review the role of the local Planning Commission, Construction Commission, Finance Bureau, and Land Administration Bureau. Within the Construction Commission, the City Planning and Real Estate Management Administration Bureaus are also key second-tier institutions. Land allocation follows a set pattern. Invest- ments proposed by licensed enterprises and nonprofit institutions are vetted by the local Planning Commission and, once approved, are passed on to the City Planning Bureau for decisions on a site consistent with the overall or "macro" Master Plan that governs the future spatial development of the city, as well as any detailed control plans that may exist at the local (district, subdis- trict) level. The Bureau works with the enterprise not only in identifying the exact location, but also on all land use and associated facilities' plans. These plans, once finalized, are reviewed by other city and subcity agencies, who are given an opportunity to register objections. Thereafter the site plan is followed by the issuance of various permits and by the acquisition of the site, as necessary, by the local bureau of SLA, in line with land requisition quota guidelines. Very large projects require additional provincial or cen- tral government approvals, complicating the above process (Box 4.1). The 1990 Urban Planning Law codifies this procedure as follows: "A construction project in the urban planned area that needs to apply for urban land shall apply to the urban planning department by presentinr the relevant documents indicating the state approval of the project. The urban planning department shall approve the location and boun- dary of the site to be used by the project, provide condi- tions for designing and planning, and approve and issue the planning license for construction land. The unit or individual undertaking the project may apply to the land administration department at or above county level for use of land only after receiving the planning license for construction land. After the application has been exam- ined and approved by people's government at or above county level, the land administration department will then allocate the land to the unit or individual ut. *rtaking the project." 2/ The sample of cities used to document this study is biased in favor of cities with provincial status or with powers to report directly to the central government. The roles of provinces may be understated. (For greater details, see Annex 4.) - 95 - Box 4.1s LAND REQUISITION REGULATIONS IN CHINA Where a project requiree less than 3 mu (2,000 m2) of cultivated land or less than 10 mu (6,700 m2) of uncultivated laud, local governments at the local (county) level can make the appropriate authorizations. Projects requiring up to 1,000 mu of cultivated land or 2,000 mu of uncultivated land require provincial government authority, except for local authorities who have special autonomy, including many of the large metropolitan coastal cities (Guangzhou included). If land requisition for a project requires more than 1,000 mu of cultivated land or 2,000 mu of uncultivated land, then the State Council (Cabinet) itself must grant the necessary approvals. Local authorities and representatives of the SLA in Beilina suggest that the practice of subdivid-in a urolect into ohases. each small enough to avoid awprovals from higher-level authorities, is rare for very small Proiects. Projects reauirina provincial or State Council anRroval, however, are freauentlv divided into dis- crete seaments with the precise intent of avoiding scrutiny. The degree to which such prac- tices are followed are unfathomable to foreign analysts, and apparantly poorly monitored by the SLA. Physical quota controls on land use requests, though intuitively attractive as a transitional device during periods when enterprises can avoid "hard budget constraints," are no substitutes for land prices that have a direct bearing on enterprise after-tax profits. Quotas are but one of many devices cited in this report as "changes" that are labeled "reforms" without advancing the ultimate objective of introducing a regulated market econ- omy. The quota system is not only a second-best "market" proxy tool, it is misdirected because it assumes Implicitly that the loss of cultivated land is driven by urban (city and town) construction requirements. In fact, a careful review of recent evidence suaRests that the loss of cultivated land has very little to do with urbanization reauirements (Table 1). This is consistent with the well-established fact that urban development requires very lit- tle land, in terms of cultivated land conversion. Meat-cleaver tactics, though labeled as reform, are both inadequate and inappropriate. Too much administrative control is being devoted to what is a minor problem. Once again, prices, not bureaucratic quotas, is the simple and only reasonable method to deal with the competing claims of agriculture and high "value-added" expansion of urban areas. Table 1: STRUCTURE OF LOST CULTIVATED LAND IN CHINA: 1988 AND 1989 1988 1989 '000 ha Percent '000 ha Percent Total Cultivated Land Loss 676.3 100.00 417.3 100.00 State Construction 71.2 10.52 51.2 12.27 City 1.36 6.0 1.4 Designated towns 9.0 1.32 5.3 1.27 Iadependent factories 14.9 2.20 10.9 2.62 Railway 3.5 0.52 3.7 0.89 Hi hwa 16.3 2.41 8.4 2.02 Bydroelectrlc projects 13.0 1.93 11.7 2.80 Other 5.2 0.77 5.1 1.23 Collective Construction 29.0 4.29 22.7 5.44 Roads 7.6 1.12 4.9 1.17 Irrigation 9.1 1.35 13.2 3.17 TVEs 7.7 1.14 2.9 0.69 Farmer housing 22.0 3.25 15.3 3.68 Other 4.6 0.68 1.7 0.41 Agricultural Restructurins 394.8 58.37 231.1 55.38 For fruit trees 119.4 17.65 69.2 16.57 For forest 155.3 22.97 93.4 22.39 For grassland 99.5 14.71 60.0 14.37 For flishponds 20.6 3.04 8.5 2.04 Ratural Disaster Related 159.4 23.56 97.0 23.24 Source: State Land Administration Bureau. - 96 - 4.4 In theory, almost all decisions regarding urban land development, including the uses, location, density, timing, and size of the real estate development are controlled by city agencies responsible for the close monitor- ing of the "city planning area," which is that part of the city proper subject to strict urban growth management.3/ This state of affairs needs to be modified to deal with an emerging regulated market economy. The current sys- tem is not effectively organized to deal with an alternative demand-driven model. whereby individual developers or investors would prepare investment Droiosals. seek out apirolriate sites for Rh'sical develol,ment. and t-euest Dermits to be processed by agencies interested Primarilv in ensuring that certain performance indicators are met. Such performance indicators could, for example, deal with health, safety, and environmental concerns; or they could involve assessments as to the ability of the existing infrastructure network to accommodate the proposed investments, requiring remedial steps. As noted belcy, in Section C, this formal structure of planning controls has begun to change at the margin, particularly among southern coastal cities and in the Special Economic Development Zones, which are given broad autonomy and encouraged to experiment with market-driven planning tools. This was not clearly apparent to international observers until 1991/92. 4.5 For example, according to the Bank's earlier Zhejiang and Housing reports, even the emergence of the REDCs, building commercial estates with properties for sale to a number of buyers, did not appear to fundamentally affect the physical planning system. In market economies, such multioccupant subdivisions originate with developers, who draw on various independent agents, including architects, brokers, contractors, and bankers, to satisfy perceived demand. Given assumptions about market trends, developers decide what to build, how much to build and when, where to locate the development, and how to price the resulting product. They assume a financial risk and are thus rewarded or penalized, depending on the astuteness of their judgment. 4.6 The REDCs, instead, appeared to be builder-centers, with land, proj- ect financing and subcontractors as the only elements outside a self-contained system. Land continued to be allocated in a noncompetitive manner, and pro- ducers were encouraged by public policy (and rewarded by an absence of compet- ing alternative delivery systems) to focus on gross output measures, with little regard for variety in unit size or type of units marketed. A cat of norms and regulations left little latitude, in any case, encouraging the con- struction of similarly configured walk-up apartments containing four to seven floors; with the associated production of obligatory, comprehensive "support facilities and infrastructure"; utilizing land rather indifferently, so that 3/ Between theory and practice, good connections and practical political considerations play an undocumented role. - 97 - developments seemed to differ little in intensity regardless of the implicit value of the land.A/ 4.7 With residential greenfield developments largely in mind, those earlier Bank reports concentrated on the housing delivery system's ability to produce enormous amounts of floor space, while issu'-.g a series of warninges (a) In complete disregard for underlying land values, gross population densities (i.e., the ratio between population and the total built-up area) across "Socialist" suburbs were converging over time, due to the fact that land use standards in housing estates tended to be uniform from suburb to suburb, independent of the city or city loca- tion involved. (b) Master Plans, governing the location of various types of land uses and the location of trunk infrastructure and citywide social ameni- ties in China, were rarely revised, static products that ignored alternative-growth scenarios, neglected the cost implications of the plans, and, over time, promoted implementation of major investments within a time frame and in dimensions that ignored actual demand. In addition, Master Plans were criticized for encouraging an unnec- essary separation of land uses and promoting self-standing satellite zones costly to service with connective infrastructure; and discour- aging the more compact, continuous expansion of cities. (c) Both the Master Plans, in advocating "city beautiful" models, and associated regulations involving compensation arrangements to exist- ing occupants, conspired to promote a policy of dedensification of inner cities, with an accompanying neglect of the package of infra- structure investments needed for redevelopment, and a failure to recycle downtown property through a change to more profitable uses. That same strategy, in neglecting the value of land (and governed by less burdensome compensation rules for the displaced), was held rasponsible for the excessive demolition of structures to improve cross-city transportation facilities, even when these were otherwise justified, and for the neglect of historic neighborhoods. 4/ Even so, by contrast with their Soviet counterparts, the Chinese compa- nies have had the advantage of being able to acquire building materials through the market, and the good fortune to adopt "appropriate" building technologies that have avoided the use of large, prefabricated panels requiring factory construction and installation using cranes. They are also largely driven by the type of constrained profit maximization model described in Chapter III. - 98 - (d) At the project level, in greenfield sites, the very low gross resi- dential floor to land area ratios,5/ averaging loes than 1:1, were criticized as wasteful, and the relative absence of high-density redevelopment projects was noted, again with disapproval. 4.8 Yet some countertrends, viewed favorably in the cited Bank reports, were identified, but labeled exceptional in scale, both within any one city and in terma of all cities involved. By 1990, efforts at making redevelopment projects more profitable emerged in places like Beijing, with the promotion of mixed-use development (including office and commercial space) and the creation of new financial incentives to deter former residents from automatically returning to a redevelopment area at little or no cost. Similarly, note was taken of the changing land allocation process in Guangzhou, where real estate development companies were entering into a more compotitive system of site selection, and where land allocation was no longer viewed as a free good (req- uisition of land from farmers excepted), but one for which ehe city should receive compensation in cash or in kind. It is now clear that such changes, at the margin, are more significant than anticipated by the earlier cited Bank reports. C. Chinese Planning Practices in Response to Emergina Land Markets: A Reassessment 4.9 The Bank's current reassessment, based on a review of continuing and accelerating experiments across many cities and benefiting from greater expo- sure to the vanguard cities of the South, reconfirms the validity of many key earlier conclusions, while modifying others. The reliance on "macro" Master Plans has remained unchanzed. even thouRh most were prepared in the late 1970s or early 1980s. predatina the emergence of the reforms catalogued above, and based on outdated assumptions concerning nopulation and emplovment growth. at the municipal and small-area level. New or amended Master Plans (Shenzheng Shanghai) prove only that the Design Institutes in charge, reporting to the local Urban Planning Bureaus, continue to be dominated by architect-planners with little access to "feedback" based on systematic monitoring of relevant small-area demographic and economic indicators, and that these agencies still show little apparent concern about the economic costs or consequences of the actions outlined in the Plan. These same static planning exercises, still concerned with an Immutable final product, are being replicated as new plans are developed. Mere updating of plans using inappropriate assumptions and tools does not represent systemic reform. 4.10 A distillation of international best practices (Annex 9) suggest that China's urban planners charged with developing macrospatial, metropolitan plans should develop Structure Plans, which: 5/ The gross residential floor to land area ratio is the ratio between the total floor area, excluding community facilities, and the total land area of the estate, including green space and that occupied by all mandated community facilities, whose existence, location and layout reflects norms, not land values. In contrast, gross building floor to land area ratios incorporate all building stock in the numerator, and automatically yield higher floor to land area ratios. - 99 - (a) begin with an examination of the global configuration of the exist- ,1A city, including detailed (small area) information on present land use, land prices, and population and employment densities. This exercise, which should be updated at frequent intervals (at least every two to three years) can benefit from tools not available to the original group of Chinese master planners, including satel- lite imagery and the computer-assisted digitization of local maps upon which one can superimpose both the present and emerging trends in location behavior, and the existing and projected network of trunk infrastructure; (b) develop alternative scenarios of expected expansion and redevelop- ment, with the associated basic infrastructure networks and other facilities needed to support related urban configuration of the city, clearly spelled out, and backed up in each case by enough data to allow the cost consequences of each scenario to be estimated; (c) subject each scenario to widespread review both by all the affected public authorities, and by the ultimate "beneficiaries" or consumers of urban land. In no way should this process be viewed as an exclu- sionary or elitist exercise driven by the preferences of the few; (d) judge scenarios by multigle criteria, including the infrastructure costs of producing additional developable land in each case and the degree to which the proposed macrospatial plan conforme with evi- dence of consumer preferences regarding desirable densities, loca- tions, and land use patterns. These preferences should be disre- garded only if the consequences of endorsing them, in any one instance, create serious problems which cannot be mitigated by com- plementary public policies and investments. The Structure Plan. then, would not be "owned" by the planners, but emerge. instead.-as a "living" document subiect to extensive public review. 4.11 The Structure Plan, with its vision of the future spatial configura- tion of the city, would become a reference framework used to coordinate the demand-driven, phased introduction of infrastructure investments, and to develop and apply complementary regulatoti instruments (pLrformance indicators in developed areas, subdivision regulations affecting the urban development of "raw" land, environmental controls, and historic preservation of selected neighborhoods). Unlike Master Plans, these Structure Plans would be prepared relatively quickly, and periodically reviewed and updated to accommodate unex- pected changes in user preferences that are not self-evidently damaging in their consequences, with the ultimate goal--to establish an internationally competitive city--borne in mind (Box 4.2). 4.12 Structure Plans should form the basis for infrastructure development and financing plans, drawing on a variety of cost-recovery mechanisms that allow the overall scheme to be implemented in an efficient manner. Clearly, as noted earlier, this Structure Plan approach presupgoses that what matters is process. not static outcomes, and that sensitivity to monitoring indicators and user preferences is critical to olanning. It also presupposes that these are local exercises guided by very broad central guidelines and not otherwise subject to centrai approval, including submission to the State Council (Cabi- - 100 - Box 4 2: SHANGHAIs AN ALTERNATIVE APPROACH TO THE MASTER PLAN? An excellent case study of an alternative approach was produced in 1989 using Shanghai as a model, by a c,mbined team from the Shanghai City Planning and Design Institute and French consultants. The conceptual approach of the report consisted of a comparison betweea the objectives of the Master Plan and the urbualzation trends of the city, followed by the formulation of alternative spatial development scenarios which took into account the inconsistencies noted between objectives and trends. These scenarios differed in that one forcibly reorle ted trends, another guided trend development, while yet another simply accommodated emerging trends. The costs and benefits of each was then examined, with prior- ity placed on the best way to encourage potential investment promoting a more competitive city. The trends, countering Master Plan assumptions, and often identified using satel- lite Imagery Included: (a) the fact that the agglomeration already greatly surpassed the limits set by the Plan, particularly in the West, lad by extension of the city and the de:- sification of the perl-urban areas and supported by extensive interregional transport investments, as well as various economic development zones; (b) the fallure of a city center dedensification strategy, due to relocation financing problems, the relative attractiveness exercised by the core in terms of amenities and employment, and the growing Impact of previ- ously neglected "temporary residents"; (c) the failure of the east shore of the Huangpu River to develop as desired, given the delay In Installing cross-river Investments, the presence of existitg storage and industrial uses incompatible with the amenities required to turn Pudong into a high-quality tertiary center! and the degree of commuting imposed on newly located Pudong residents. Design Institute managers, interviewed during the current sector study, claimed that a new plan, using newer tools and analyses, would be prepared by 1993, and be more accommodating of actual growth trends. A review of work done through August 1992 suggests that traditional practices prevailed and that the 1989 exercise was never internalized. net). It is unlikely that central bodies have any comparative advantage in vettinz urban plannir.2 exercises, Rarticularlv if the emerging model of a reg- ulated market economv is to have any meaning. Requiring a "stamp of approval" from the State Council virtually guarantees that cities will continue to pro- duce static plans subject to minimal change over time. 4.13 This suggested reform approach contrasts sharply with China's exist- ing macrourban spatial planning, which still responds to central guidelines, partly because most such plans are at least a decade old and are legally bind- ing documents approved by the State Council (Cabinet). The existing system neglects present land use and densities, and has no mechanism to monitor them at the municipal level and, where possible, adapt to them, by modifying poli- cies, regulations and infrastructure development phasing. Consumer prefer- ences are conspicuously ignored, even though a regulated market economy nas been endorsed as the overall key objective of reform. Instead, the Master Plan resembles the architectural design plans for an office building or an apartment complex, proprietary in nature, "cast in concrete," and nothing less than a "product" meant to be built as is. 4.14 China's existing urban plans, which carry the unusual disadvantage of being implemented in a rapid and single-minded manner, have imbedded within them several specific abstract and outdated ideas which previous Bank reports did not fully explore and subject to critical scrutiny. Six items are partic- ularly striking: - 101 - (a) the building of trunk infrastructure without regard to "rate of return" calculations, particularly in terms of timing, scale and design, given self-evident demand; and in terms of foregone options, involving alternative subsector infrastructure investments during a five- to ten-year iiorizon; (b) the failure to tackle the challenge (and value-added opportunities) of redevelopment and historic preservation in a systematic manner, covering a whole pre-1949 city core; (c) the related failure to assess the negative consequences of a central area strategy undergirded by assumptions that dedensification is an undisputable "good"; (d) the adoption of administrative control mechanisms as a means to preserve valuable, urbanizable land for the inherently lower-value cultivation of food supplies intended to guarantee each city a means of "self-sufficiency" in a very narrow geographic sense; 'e) the creation of discrete satellite towns and economic development districts, far from the CBD, as greenfield islands intended to off- set pressure on the existing built-up areas, but with total disre- gard of the costs of the local and connective infrastructure invest- ments involved; and (f) the related adoption of a segregated land use strategy that, in its extveme form, designates such towns and districts as single- or limited-purpose activity centers, devoted to electronics here, transport equipment there, or (Pudong, Shanghai) the creation of a central business district built de novo. Concomitantly, there is a tendency to funnel foreign-funded industrial activities to particu- lar zon6., as if they could not otherwise be accommodated by the existing u.rben fabric. Worse still, where sector team members were able to interview local officials, there was little evidence that comprehensive feasibility studies, with potential market demand in mind, have ever been conducted to validate these decisions to pro- mote special-purpose, segregated development islands. In areas like Pudong. Shanghai. and the various sDecial economic zones established in Guangzhou or HanRzhou. the investors are expected to be primarilv or significantlv offshore in nature. They are free to ignore the Rresumptions of local 2lanners, who still think in terms of assign- inm sites to enterDrises Datientlv waitinR in a queue. Given the magnitude of infrastructure investments involved, and the danger that these may lrain resources away from other locations with high- priority infrastructure investment requirements, this is a very serious flaw that demonstrates all too graphically how the planners that shape each city have failed to grasp the "new rules of the game" that accompany a demand-driven economy. 4.15 In this report, only five examples will be used to illustrate the harmful effects of abstract Master Plan efforts devoid of economic content that were documented during the course of the sector study: (a) Pudong, Shanghai; (b) Fuzhou; (c) Shenzhen; (d) Hangzhou; and (e) Guangzhou. - 102 - 4.16 Pudong, on the east side of Shanghai's Huangpu River, appears to be a physical planner's dream: a deceptive "blank slate." Centered across from the famed Burd, the heart of the west-side CBD, and paralleling the heavily urban'zed old city, both north and south; it seems to be an obvious choice for introducing the land use changes required to create a restructured Shanghai, ready to face the international competition as a Pacific Rim center, yet free of the messy requirements imposed by redevelopment constraints. 4.17 Pudong, as currently gnvisioned, will expand from a built-up area of 38 km2 in 1990 t* more than 100 km2 in the year 2000. The urban population is expected to double from 600,000 to 1.2 million. Currently a center of more than 2,000 industrial sites (400,000 workers), several dormitory suburbs, and many rural communities, Pudong is to be administratively transformed into spatially and use-segregated "islands," whose (1990-2000) infrastructure costs (including cross-river connections) will total Y 60 billion or $10 billion. In addition, significant if undocumented costs will be incurred for relocation of existing users, replacement of market gardens and food production, and cleaning up hitherto unpublicized polluted sites. There will be five self- contained districts, each separated by a greenbelt 2-3 km in widtht * Liuiiazui, located across from the Shanghai Bund, is to be the "gol- den triangle" of Pudong, and, covering 28 km2, will become a center for commercial services; * Wai Gaoaiao, encompassing 75 km2, and located 20 km from the center of Shanghai, on the Yangtze River, wv!l1 be developed as a state-of- the-art port, capable of accommodating 10,000-ton class ships. Within the area, a 10-km2 free-trade zone will be created to accom- modate export processing, storage, and entrep8t trade; * Oinaningshi-Jinzaiao and Zhouiiadu-Liuli (21 km2 and 34 kma, respec- tively) will be set aside for high-technology, nonpolluting enter- prises; * Beicai-Zhanaiianz will be the fifth and final area, covering 17 km, and designed by the physical planners as a science, research and educational center. 4.18 It has proved difficult for the sector study team to determine how the size, locations, and uses of the users could be so confidently determined by pla.ners lacking information about real estate development as a market- driven economic activity. The Pudong scheme is extremely ambitious, costly, and driven by physical planning parameters, uninformed by marketing feasibil- ity studies. Only the plans regarding the planned new CBD, at Liujiazui, have been reviewed carefully by the Bank, and the remarks that follow do not neces- sarily reflect a more generalized skepticism concerning other component parts of the Pudong scheme. The Liujiazui plan is driven by a "city beautiful" model that attempts to address the perceived lack of green space on the west side of the river (Puxi) by prescribing extravagant uses of "CBD" land: fully 30 percent of the district will be devoted to open space, including the very valuable riverfront area facing the Bund. Floor to area ratios are planned with limited regard to relative land prices: the choicest sites, facing the vest bank, and having greatest access to the existing center, will be set - 103 - aside for low-rise developments; densities will increase toward the periphery, while implicit land values would dictate the exact opposite.6/ 4.19 More generally, key transport nodes, created by the Nanpu and Ningpu bridges, the extension of the subway into Pudong, the reconstruction of Yanggao Road (Pudong's spine),7/ and the completion of the citywide inner- ring road, seem to have limited bearing on land use plans, with the potential real e .ate development impact of key "nodes" severely underestimated. And finally, all statements to the contrary, it is not clear that Pudong's devel- opment strategy will necessarily contribute in an optimal manner to the self- evident objectives of modernizing Puxi. Particularly with major cross-river investments in place, Pudong would appear to be ideally suited to relocate low-value Puxi land users to the east bank. If Shanghai's objective is to be at the cutting edge of a foreign-financed technological transformation of industry, backed by the requisite commercial and personal services, then existing sites in Puxi (including the development zones of Hongqiao, Caohejing, and Minghang) and other radevelopable plots on the west side seem prima facie potentially more efficient ways to achieve that objective. These alternatives do not seem to have been seriously studied. Perhaps the most important lesson, overall, is that past practices of ignoring trade-offs, and 1/ This "vision" is graphically described in a recent official document, Shanghai ProDertv 1991: "The center ... together with parts of the Nanjing Road-Bund form the Shanghai CBD.... (T]he riverside area will be mainly arranged with open grounds as well as low-density buildings with public open space, amusement parks, recreation centers, exhibition hall, etc. ..., while the central part will be composed of high-density and high-standard modern comprehensive developments for finance, trade, office, and hotel services. In order to leave a clear field of vision from the skyscrapers in the center of Liujiazhui Finance and Trade Zone to the Water Front, the building heights will be designed to fall gradu- ally from the central part of the riverside area." (Source: Office of the Shanghai Municipal Committee for Land Use Reform and Shanghai Munici- pal Statistical Bureau in conjunction with Jones, Long Wooten, Hong Kong, Shanghai Property 1991, Hong Kong, 1992.) 7/ The reconstruction of Yanggao Road provides another example of planning where costs and benefits appear to have been ignored. Now barely more than a local road carrying little traffic, local planners initially pro- posed a 35 m wide highway plan, with land reserved for a 50 m wide road when demand justified this. The plan has sincQ been revised by central authorities and a 50 m wide road will be built in one stage. The step was taken without the benefit of economic analysis, according to Shanghai Construction Commission sources. - 104 - market research, in making infrastructure investments or assessing the feasi- bility of the plans for land development, apparently continue to prevail.8/ 4.20 Fuzhou's new Master Plan, still under preparation and not made available to the sector study team, covers 30 years (1990-2020). It envisions the creation of a 3 km2 park along the most important commercial street (May First Road), where two proposed light rail lines will intersect. The city's growth will promote dedensification of the center city, prevent incremental suburbanization, and focus growth in five self-sufficient satellite cities, amply separated from one another and the existing city by greenbelts, acting as buffer zones. Light rail systems will be built to connect each satellite with the city center. There appears to have been little analysis of alterna- tives undertaken, to compare benefits and costs, measured in terms of devel- opable land made available per unit of investment required (on- and off-site). The scheme was developed with the assistance of Singapore consultants, and provides an example of the fact that foreign expertise, in and of itself, may not be appropriate; in fact, it may "validate" concepts that will harm the economic competitiveness of the city in the future. 4.21 Shenzhen, often viewed as in the vanguard of real estate reform, produced a second comprehensive Master Plan, approved by the State Council in 1989. Given a variety of geographic constraints, and the saturation of the existing CBD, the new Plan calls for a new urban downtown, in Fujian district, to the west of the existing core, covering 44.5 km2. If intended to extend the CBD, whose functions are primarily tertiary in nature, the plan is quite inadequate. The new district is zoned primarily for residential uses, though considerable space will be devoted to governmental institutions. This area will be built using relatively low floor to land area ratios, and dominated by green space and recreational activity centers. As a result, per ;apita land consumption (124 in) will resemble levels found in towns and townships rather than large city centers. Ironically, this plan was developed at the same time the city was planning to carry out expensive land reclamation projects along its coastline. 4.22 Hangzhou Municipality, as part of work financed under the prepara- tion of the Zhejiang Multicities Project, engaged foreign consultants to eval- uate its 1983 Master Plan, given trends since then.9/ The Master Plan was developed around at least two premises: (a) that the interregional transport investments (roads, railroads) would be spatially dispersed, and thus have limited district-level impacts; and (b) that urban expansion could proceed in several directions from the city center, promoting, for example, the develop- ment of "virgin" agricultural land in the southeast quadrant of the city (300 ha), while accommodating the infrastructure requirements of already 8/ Even in 1992, the Shanghai Government proposed World Bank financing of a cross-river link between Pudong and Chongning Island, involving 20 km of tunnels (4 tunnel lanes) and 18 km of surface roads (8 lanes wide). The total cost of the project (1992 values) will total Y 8 billion for the tunnels aline. No land use marketing feasibility study or physical plan accompanied the proposed scheme. 9/ Groupe Huit, Zheiiana Multi-Cities Proiect: Final Report, November 199'. - 105 - urbanized or urbanizing areas, particularly the northeast quadrant. In fact, the interregional infrastructure investments that actually took place during the decade created a strategic node of highways, railways, canals, and air- ports in the "northeastern extension," a development unforeseen by planners. Yet the Master Plan dictated a fixed development plan for that northeastern corridor, provided it with the trunk infrastructure linked to those assump- tions and then, ignoring the benefits of ongoing monitoring of actual economic activity location, did little more. Recent initiatives to build spatially segregated economic development zones, to the south, and far from the city center, graphically illustrate a failure on the part of municipal planners to set demand-driven priorities with budget constraints in mind. 4.23 The consultants pointed out specific deficiencies in the planning process, zeroing in on the problems of the "Northern Extension" (Box 4.3). Local planners maintained that trunk infrastructure investments had accommo- dated (ir. the main) developments that actually took place in different subur- ban extensions of the city. They agreed, however, that district-level control plans, with accompanying secondary and tertiary investments, were inadequate and had underestimated the cumulative impact of plot-by-plot locational choices that took place during the decade. Blame was placed on individual locational decisions that took place immediately after the Cultural Revolution (1966-76) and chus, by implication, unlikely to have continued since then. The foreign consultants, aided by modern monitoring tool- such as those afforded by satellite imagery, suggested a different unfolding set of events. The end result, however, is one and the same. In summary, major infrastruc- ture investments are being scattered across the municipal landscape, while major quadrants of the city develop in a disorderly fashion, with planners failing to redirect public investments to rectify problems created by past (and present) development trends subject to inadequate monitoring. Box 4.3, HANCZHOU: THE URBAN PLANNING PROBLEMS OF THE NORTHERN EXTENSION Overall, substantial outlays have been granted for facilities and organization of the road system within the city, over the last 10 years, and are programmed for the next five years; however, outside a few main axes, little effort has been expended en or Is pro- grammed for the outlying areas, In particular to the north of the city, where a huge zone of unanticipated and poorly monitored housing and industry has been allowed to emerge (i.e., premises built at random according to the land requisition possibilities available; facto- ries lining the roadsides, leaving huge gaps of cultivated areas and producing noncontrolla- ble discharges of liquids and gases). The ad-hoc way In which industrial companies have set up in the northern zone is resulting in envirormental problems and excessive use of land, In difficulties in services being provided by the road system and other infrastructure networks, and in operating infra- structure service costs which are higher because of the low densities involved. A redevel- opment plan for the Industrial fabric is required: it will have to rationalize land occu- pancy, while absorbing new demand, including that from units being displaced from the old city. Source: Hangzhou City Planning Bureau; mission assessments. 4.24 An assessment can also be made of infrastructure planning in Guangzhou, whose reputation for being in the vanguard of reform is often con- fused (and enhanced) by the achievements of its Pearl River Delta neighbors. - 106 - Figures 4.1 and 4.2 illustrate the existing land uses and the requirements imposed by the Master Plan in terms of trunk infrastructure requirements that partly shape those land use patterns. A quantitative analysis of the Plan reveals that 72 percent of the new areas to be developed under the Plan are located at between 10 and 42 km from the city center, while most of the eco- nomic, physical, and demographic growth is actually taking place within a 10 km radius of the city center. The Plan does not block small area develop- ment, which is facilitated by the district plans developed by lower-level authorities and described below, but it directs long-term investment in trunk infrastructure away from the areas with the greatest, self-evident potential for market-driven growth.lO/ Finally, Box 4.4 on Tianjin's decision to build its Third Ring Road ahead of demand, while crowding out necessary infra- structure investments in core areas, merely reinforces the contention that the problem of misdirected Master Plan exercises is not confined to a few cities. Box 4.4: TIANJIN'S THIRD RING ROAD: WHEN PLANNERS AND MARKET FORCES COLLIDE Tianjius urban population is largely contained within 9 or 10 km from the CBD, with approximately 80 percent living within 7 km of the city center. Most of this area has been historically poorly serviced with lnfrastructurel funding for secondary and tertiary sewer connections, for example, is still a problem to this day. Nevertheless, driven by the Master Plan, prepared before 1980, the local government decided to build a 100 km long Third Ring Road, 11 to 12 km from the city center, in an area largely reserved for greenfield uses. The expressway has three lanes in each direction, and, for a distance of approximately 25 km, is paralleled by a provincial expressway, with two lanes in each direc- tion, linking Beijing and the Tianjin port area of Taegu. World Bank funds were used in the latter case. The Ring Road was costly to build, is expensive to maintain, and is grossly underutilized, though exact figures on the latter were not provided to the sector team. 4.25 In contrast to this rather bleak reassessment of macrosoatial olan- nina and associated trunk infrastructure investment locational oriorities. the sector study team found significant if selective modifications at the district and subdistrict level. favoring market forces and the more efficient use of land. 4.26 At the district level, when actual real estate projects are being designed or vetted, "realism" sets in, and the local planners find it neces- sary to take economic feasibility and "local realities" into account. The result is ad-hoc but generally sensible planning. See Figures 4.3, 4.4, and 4.5. The key problem lies in the fact that the two exercises--Master Plan and district-level plan----are sequential and that no mechanism exists to allow for "feedback" that recasts the Master Plan's priorities, including the nature, location, and phasing of infrastructure investments. If local plan- ners could follow Hong Kong and Singapore's example in developing more flexi- IQ/ Notwithstanding the fact that most of the newly setup zactories are located in peripheral areas, some 55.7 percent of industrial establish- ments, 58.1 percent of the workforce, and 53.1 percent of the total industrial output were, in 1984, found in the inner core (S. Li, "A Com- parative Study of the Urban Land Use Patterns in Guangzhou and Hong Kong," Hong Kong: Department of Economics, University of Hong Kong, Discussion Paper No. 79, December 1986, processed). - 107 - ble macro Structure Plans that allow for such feedback, the problem whereby infrastructure investments and actual spatial development diverge would be largely solved. 4.27 The undermining of the macrospatial "command economy" model at the localized level can best be understood by examining metropolitan trends in a spatially segmented manner. Land use systemic reform has to be seen as a phenomenon proceeding at a different pace, with different results, in four spatially distinct areas: the "old" (1949/50) city center; the suburban "Socialist" districts developed thereafter; the satellite towns, largely out- side the city proper, assumed to bear the burden of accommodating much of the displacement required by the planned "dedensification" of the old city; and the peri-urban areas, variously defined to include the zones of the city proper that are not closely supervised by city planners (who focus on a more spatially compact "city planning area" and the even smaller "built-up area") and the suburban counties operating under the control of county authorities, who act largely in their own interest, except for the few satellite towns d^veloped in their territory by municipal authorities. The aggregate outcome of this spatially segregated battle of plan versus market will not become clear for some time. 4.28 Typically, within the "old city," in the municipalities studied by the sector study team, and the near suburbs, redevelopment has emerged as a mechanism that is, on a site-by-site basis, raising floor to land area ratios from very low levels (0.3-0.6) to much higher ones (2.5-10.0), while trans- forming land uses. Unfortunately. the scope of such redevelopment is still limited. Table 4.1 provides a typical example of land use transformation at the city block level in areas in or near the old city. Notably, land devoted to industrial uses declines, as does land (but not necessarily building space) dedicated to residential uses. Other uses show dramatic increases. This varies city by city (Tables 4.2-4.4). Population densities, however, have not changed very much, because redevelopment has led to much higher consumption of floor space per capita. Using Shanghai as an example, on a district-by-dis- trict basis, a decade of redevelopment has led to either insignificant alter- ations in densities or to decreases that rarely exceed 9-10 percent of those recorded earlier. Redevelopment is accompanied by selective upgrading of local infrastructure; street widening projects are now often the catalyst for this intensification of land use, though other infrastructure upgrading typi- cally lags behind and hinders redevelopment. As an example, Shanghai Munici- pality has requested Bank assistance in redeveloping nine parcels in the city, covering only 34 hectares, and characterized as having very great "value- added" potential. Of the nine, however, only one (covering 3.5 hectares) is rated as currently well-equipped for redevelopment. 4.29 Redevelopment, broadly defined, requires a change in land uses. China's "old city" areas have typically devoted too little land to the ser- vices sector, too much space to factories and warehouses, and, arguably, excessive land to accommodate residential uses. Change has taken place in various ways. Physical redevelopment is one mechanism, but, in the absence of redevelopment strategies endorsed at the municipal level, the scope of land use change induced by this tool is still limited. Alternative_y, land use is altered as residential structures are converted to commercial uses, with or without official permission. And many factories and institutional land users, - 108 - Table 4.1: TYPICAL LAND USE CONVERSION IN CITY CENTERS Before redevelopment After redeveloDment Land Area Street 8.0 18.0 Residential 55.0 30.0 Office 20.0 25.0 Commercial 12.0 27.0 Industrial 5.0 0.0 Total 100.0 100.0 Z FAR 2 FAR Floor Area Residential 49.7 0.6 24.3 2.5 Office 24.1 0.8 36.4 4.5 Commercial 21.7 1.2 39.3 4.5 Industrial 4.5 0.6 0.0 0.0 Total 100.0 0.7 100.0 3.1 m2 2 Increase Floor Area per Hectare Residential 3,300 7,500 127.3 Office 1,600 11,250 603.1 Commercial 1,440 12,150 743.8 Industrial 300 0 -100.0 Total 6.640 30.900 365.4 Note: It must be noted that while the residential area generally decreases, the residential floor space increases due to a much higher floor area ratio. The population density increases only slightly due to increases in floor consumption per household. _ 109 - Table 4.23 SHANGHAI CITY: 23 REDEVELOPMENT PROJECTS Land Approved Of whichs No. of area building Public Gross District blocks (ha) area (mi) use (m2) Housing PAR Yangpu 10 24.03 563,452 50,111 513,341 2.34 Yangpu 8 8.89 216,870 13,797 203,073 2.44 Hongkou 65 32.40 523,624 98,689 424,935 1.62 Hongkou 14 25.04 322,464 3,100 319,364 1.29 Zabei 16 18.04 624,265 220,361 403,904 3.46 Zabei 1S 21.73 528,040 51,827 476,213 2.43 Putuo 26 30.14 746,886 25,758 721,128 2.48 Putuo 15 12.18 330.837 45,364 285,473 2.72 Jin'an 9 7.05 211,935 30,602 181,333 3.01 Jin'an 6 4.82 159,200 13,800 145,400 3.30 Jinlan 9 10.27 364,800 46,900 317,900 3.55 Changning 12 14.51 269,861 52,465 217,396 1.86 Changning 10 48.06 349,380 32,620 316,760 0.73 Xuhuei 26 66.40 1,767,800 213,500 1,554,300 2.66 Xuhuei - _ _ _ _ _ Luwan 12 8.56 448,182 13,420 434,762 5.24 Luwan 4 3.56 206,462 - 206,462 5.80 Nanshi 4 23.89 811,848 72,304 739,544 3.40 Nanshi 9 - _ _ _ _ Nanshi 8 - - - - - Huangpu 8 4.96 303,420 184,579 118,841 6.12 Huangpu 6 7.51 1;a,610 325,446 165,164 6.53 Huangpu 1 5.51 1J7,596 8,224 129,372 2.50 Total 293 377.55 9.377.532 1.502.867 7.874.665 2.48 Source: Shang Rongling and Zhong Yongjun (eds.), 1991, One Apartment for Each FamilX--The Objective of Housing for the Year 2000, pp. 131-132. - 110 - Table 4.3: REDEVELOPMENT PROJECTS IN FUZHOU Land Building Name of projects (ha) (10,000 m2) PAR Use type 1 Donghu 1.70 5.00 2.94 Housing 2 Jianhai 2.14 18.60 8.69 Mixed 3 Taohuashan 0.63 1.15 1.81 Housing 4 Qixinjing 5.00 11.00 2.20 Housing 5 Zhujing 4.00 9.56 2.39 Housing 6 Mingdong 0.37 3.00 8.11 Mixed 7 Yuanhong 1.40 11.70 8.36 Mixed 8 International 0.39 3.60 9.26 Mixed 9 Xianshi 0.80 2.56 3.20 Retail 10 Twin Towers 0.59 2.08 3.52 Office 11 Hua Lian 0.22 2.20 10.00 Mixed Total 17.24 70.45 4.09 Source: Fuzhou Land Administration Bureau. Table 4.4: HOUSING REDEVELOPMENT IN HANGZHOU Land Building Year of Name of projects (ha) (10,000 m2) FAR construction 1 Chenghuan 14.10 15.55 1.10 1981 2 Jiangchen 23.00 23.30 1.01 n.a. 3 Dongyuan 37.40 41.92 1.12 1988 4 Changqing 11.70 14.37 1.23 n.a. 5 Huazhang 13.75 20.79 1.51 n.a. Total 99.95 115.93 1.16 Source: Selected Works of Hangzhou Housing Plan, 1990. - ill - with land to spare, have established joint-venture arrangements with partners eager to accommodate market demand. Occasionally the original occupant has been forced (by mergers or other mechanisms) to simply relocate to suburban locations or to remain in situ but switch to other economic activities. A note of caution is necessary on this point. The original occupant'always reports to, and is "owned" or supervised by a local industrial bureau; that bureau may insist that the property be reused for activities over which it can continue to claim "ownership" or supervisory rights. This intervening agent can thus block the emergence of appropriate reuse, from the perspective of citywide comparative economic advantage. Evidence that this may be happening is suggested by items such as are shown in Table 4.5, particularly within the old-city districts. Local authorities confirm this hypothesis is correct and constitutes a major barrier to developing a citywide industrial relocation strategy. Tabl% 4.5: SHANGHAI: CHANGING INDUSTRIAL FLOOR SPACE (1987-90) Industrial Building stock building Land area 1987 1990 1987 1990 Change (km2) -- (10,000 MI) -- - (10,000 m2) - (Z) City proper 748.71 152.30 172.56 44.90 48.22 7.4 Huangpu La 20.46 12.34 13.39 1.63 1.76 7.7 Nanshi La 27.92 12.28 14.23 3.20 3.18 -0.6 Luwan La 8.05 7.64 8.29 2.03 2.14 5.6 Xuhuei 46.64 16.99 18.81 4.32 4.61 6.8 Changning 28.82 11.69 13.88 3.07 3.20 4.3 Jingan Ia 7.62 8.27 9.08 1.98 2.04 3.1 Putuo 29.88 14.50 16.73 4.13 4.50 8.9 Zabei 27.95 13.32 14.04 4.02 4.20 4.4 Hongkou 23.48 14.47 16.19 3.58 3.78 5.6 Yangpu 59.63 24.72 27.72 9.62 10.46 8.7 Minghang 43.08 6.17 7.51 2.93 3.38 15.6 Baoshan 425.18 9.90 12.69 4.39 4.97 13.1 la These four districts lie entirely within the old "1949" city. Source: Shanghai Statistical Yearbook 1989, p. 433. 4.30 In the post-1949 "Socialist" suburban belt of development, compari- sons of early and recent project-level development show evidence of improved land use, but the trends are far less dramatic. Floor to land area ratios that were once uniformly below 1:1 are now often above that, but the absolute improvement in land use efficiency is typically low (Table 4.6-4.8). The driving force behind any change is probably the rising cost of land requisi- tion, and the increased proportion of suburban land development carried out by profit-motivated real estate development companies. - 112 - Table 4.6: NEW HOUSING PROJECTS IN SHANGHAI (1980-90) Land Building Of which: K of 2 of area floor areat Nonresi- high- nonresi- No. Project District (ha) (10,000 m ) Housing dential rise dential PAR 1 Shiguang Yangpu 45.33 49.79 43.18 6.61 23.06 13.28 1.10 2 Changbai Yangpu 34.20 51.44 45.46 5.98 - 11.62 1.50 3 Gongnong Yangpu 65.87 68.75 59.68 9.07 7.00 13.19 1.04 4 Mingxing Yangpu 26.22 27.59 23.86 3.73 45.48 13.52 1.05 5 Quyang Hongkou 78.37 107.32 87.58 19.74 38.00 18.40 1.37 6 Fengzbeng Hongkou 17.80 15.59 12.86 2.73 34.90 17.53 0.88 7 Yuaguang Hongkou 23.33 24.44 20.84 3.60 29.50 14.75 1.05 8 Pengpu 456 Zabei 56.31 51.10 46.00 5.10 - 9.98 0.91 9 Guanguong Putuo 28.41 31.53 27.16 4.37 22.78 13.87 1.11 10 Hutai Putuo 74.20 71.52 59.11 12.41 20.76 17.35 0.96 11 Changfeng Putuo 45.50 47.80 43.25 4.65 - 9.73 1.05 12 Xianxia Changuing 74.40 82.85 73.62 9.23 40.63 11.14 1.11 13 Tianling Xuhuei 97.10 116.30 105.60 10.70 51.50 9 20 1.20 14 Kangjian Xuhuei 94.25 74.30 66.10 8.20 - 11.04 0.79 15 Dezhou Nanshi 120.00 110.58 99.32 11.26 18.00 10.18 0.92 16 Shanguan Nanshi 119.40 125.03 111.23 15.84 - 12.67 1.05 17 Shauggang Nanshi 72.00 82.00 74.00 8.00 5.00 9.76 1.14 18 Xuaye Nanshi 28.44 28.50 26.66 2.84 26.00 9.96 1.00 19 Weifang Huangpu 58.70 75.90 66.10 9.80 - 12.91 1.29 20 Zhuyuan Huangpu 37.35 44.85 40.03 4.82 - 10.75 1.20 Source: See Table 4.2. Table 4.7: NEW HOUSING DEVELOPMENT IN HANGZHOU Land Building Year of Name of projects (ha) (10,000 m) FAR construction 1 Hemu 15.82 15.96 1.01 1980 2 Caohui 74.24 61.15 0.82 1979 3 Caohui-5 10.02 10.80 1.08 1982 4 Yujia-Lin 21.49 26.81 1.25 n.a. 5 Cuiyuan-1 10.98 15.22 1.39 1985 6 Cuiyuan-2 8.71 11.14 1.28 1986 7 Cuiyuan-3 11.59 12.16 1.05 1987 8 Cuiyuan-4 11.65 13.82 1.19 1988 9 Caihe-2 21.23 23.02 1.08 1984 10 Desheng 19.54 20.57 1.05 n.a. 11 Pujia 18.00 20.23 1.12 n.a. 12 Shibang-3 4.50 3.68 0.82 n.a. 13 Gudang 33.30 36.40 1.09 1988 14 Hangyang 18.90 21.93 1.16 n.a. Total 279.97 292.87 1.05 Source: Selected Works of Hangzhou Housing Plan, 1990. - 113 - Table 4.8: NEW HOUSING DEVELOPMENT IN FUZHOU Land Building Year of Name of projects (ha) (10,000 mI) FAR construction I Wangdian 53.90 61.50 1.14 1982 2 Shanchajie 16.41 22.28 1.36 1982 3 Puxia 7.86 11.63 1.48 1980 4 Yangqiao 8.81 11.82 1.34 1985 5 Yangxia 25.29 19.96 0.79 1979 6 Shanghai 14.30 19.14 1.34 n.a. 7 Liming 9.93 13.26 1.34 1986 8 Fenghuang 7.30 8.54 1.17 n.a. 9 Wufeng 2.87 4.23 1.47 n.a. 10 Pingdong 2.98 4.96 1.66 1981 11 Pingxi 11.47 13.10 1.14 1983 12 Station 7.00 21.00 3.00 1979 13 Xiaolu 2.49 4.20 1.69 1985 Total 170.61 215.62 1.26 Source: Fuzhou Land Administration Bureau. 4.31 Furthermore, and unlike evidence gathered in earlier Bank assess- ments, it is now clear that newer housing and commercial projects, Particu- larlv in southern provinces, typically devote less land to "public facilities" that are donated to the local neighborhood authorities, but--at the same time --provide greater attention to landscaping and other factors meant to respond to consumer demand. Clearly these regionally segmented developments are mar- ket-driven. In areas, particularly in southern China, where individuals are buying houses, and relatively high incomes generate high demand for retail space, the "new look" project is more likely to emerge, simply because market forces are forcing developers and planners to change their practices more rapidly than elsewhere in China. For example, free-standing retail space located within the inner confines of a project, and implicitly disconnected from the larger community of potential customers, is giving way to ground- level shops that are incorporated directly into residential structures, par- ticularly those adjacent to streets with heavy pedestrian traffic or otherwise considered comparatively accessible. Since these projects are closely super- vised and regulated developments, such changes can be assumed to be "vali- dated" by urban planners and their supervisory authorities (Box 4.5). 4.32 Less apparent is the degree to which approvals may have been obtained for the spontaneous conversion of ground-floor dwellings built in earlier periods for residential use and now devoted to commercial uses. Some of this "retrofitting" can hardly be characterized as unregulated, however. Many older estates were built with deep setbacks from the street. Often one finds that this open space is now occupied by recently built and highly visi- ble low-rise commercial facilities that block easy pedestrian access to the residential structures behind them. - 114 - Box 4.5t GUANGZHOU: INCREASING DENSITY IN NEW DEVELOPMENT Like many other Chinese cities, land use density of new urban development in pre- vious decades was not very high, particularly among industrial and institutional projects. According to a comprehensive urban land survey in Guangzhou in 1982, the general floor area ratio within the city built-up area was 0.33, with an average building height of 1.89 sto- ries . Although there are no comparable data for more recent years, available information on individual projects tndicates that much higher FARe are the norm. For example, among 50 new urban housing projects in Guangzhou complated in the nast 10 years, the average floor area ratio is more than 2.0 with 2.53 for four central u;ban districts, and 1.91 for four suburban districts. Compared with many other cities In China, land value Is clearly under- stood by both developers and planners. Here, we do not see large setbacks along major streets. We do not see significant amounts of unused land wlthin housing districts. In order to save or intensively use land, many new building types have been created, and vari- ous innovative design and planning efforts have been made. Most apartment buildings along main streets are designed as mixed-used structures with leased retal.l space on the ground floor, leased offices on the second and third, and a landscaped terrace and housing on the fourth and above floors. The area between buildings Is either used for additional retail space, or for well-maiutained playground spaces. This contrasts with many old-type projects which leave vacant land ln between structures. Urban land use patterns consist of both different types of buildings and the way they are being arranged In space. In the past 30 years, new urban areas were developed in the form of identical apartment buildings placed in very similar arrangements across the country. Such land use patterns are now undergoing a series of changes in Guangzhou. One of these changes allows farmers to build preferred housing types within planned urban devel- opment areas. Farmers' houses, such as those In Tianhe and Shiqiao, reflect their preferences: 2-4 stories, with a small yard on the ground floor, and several balconies on upper levels. Each family li given a standard plot. They choose to use as much as they can, which results in very high floor area ratios--from 1.5 to 2.5. They provide an alter- native housing type for urban dwellers. In some new housing projects, like Five Ram Vil- lage, a townhouse type of building was added to attract additional consumers. These new building types, and distinctive treatments along the main streets, have begun to create a different streetecape within the new projects. Another new development In Guangzhou is that, unlike the past, farmers' interests are being carefully considered, and farmers' villages are being preserved during the process of urban expansion. In China, even though the Constitution states that rural land is col- lectively owned, rural property rights have not been well-protected. Whenever there is state land expropriation, only minimal compensation has been required. Both the economic interests and physical structures are sacrificed. Such practice has met increasing resis- tance from farmers. Instead of imposing government terms on farmers, in recent years Guangzhou has been experlmenting with a "land readjustment" process, which has been widely practiced in Japan, Korea and Taiwan (China). In the Tianhe Railway Station area, besides the normal minimal land compensation for crops and structures, a parcel of prime land along the main street will be reconveyed to farmers for commercial development. And their village will also be maintained in the new area plan. This approach will reduce unnecessary delays during the land requisition process, and also protect farmers' interests. It also creates a more diverse urban environment. Many designers and planners would agree that most farmers' houses are more interesting, architecturally, than many modern apartment buildings. Unfor- tunately, most urban planners pay no attention to integrating such structures into new development. For example, in Minghang, Shanghai, the sector team was told that well-built village houses would lie sacrificed for future urban expansion. Source: Zhang Guixia, 1990, wReview of Urban Housing Development in Guangzhou," in Xu Shaoji (ed.), Develooment of Guanazhou Real Estatu Sector. 4.33 The satellite towns and economic development districts found in China's cities are, by contrast, creatures of the command system promoting a supply-driven, norm-restricted pattern of real estate development, unrespon- - 115 - sive to market forces.ll/ In fact, during the 1980s many satellite areas were simply relabeled as economic development zones, requiring internal and connective infrastructure to attract joint-venture investors. The Master Plan of each city, developed on the basis of other criteria, somehow seemed to dovetail, at the local level, with spatial strategies to att::act investors. Some might suggest that this spatial coincidence was driven by the lack of financial resources to fully implement the proposed satellite towns. Among the cities visited, the following examples of expedient conversion from satel- lite town to development magnet can be documented, though many more exist: Guangzhou: Huangpu Economic Zone (25 km from city center) Hangzhou: Xiasha Economic Development Zone (19.5 km from city cen- ter) Fuzhou: Mawei Economic Development Zone (20 km from city center) Chengdu: Longquanyi Industrial District (19 km from city center) Shanghai: Minghang (30 km from city center) and Wujing (20 km from city center) These are few in number, within any one metropolis, but they are typically disconnected from the built-up area, sited 20 to 40 kilometers away from the city center, and make relatively extravagant use of land, whether measured on a per-capita or a per-employee basis. They have been built at enormous cost, represented by such indicators as developable land per yuan of local and con- nective infrastructure invested, and& have proved to be insignificant appen- dages to the urban network to which they belong. In both absolute or relative terms, few people work there, and most of these workers have to commute from their "old city" or suburban estates. Many forcibly relocated factory mana- gers complain that their freight-related costs have grown dramaticallyg in addition, if meant to be self-contained communities, and if intended to con- serve land, the satellite cities have failed. Unfortunately, but most reveal- ing of local monitoring practices, up-to-date information on "planned" versus "actual" outcomes is hard to find (Table 4.9), at the municipal level.12/ Again, such islands of development emerged as visions of planners, without the benefit of the sobering impact of comprehensive feasibility studies that incorporated marketability into their development equation. Perhaps state- 11/ As noted, urban planners should not be held fully responsible for this, particularly when heavy industry projects authorized by the Planning Com- mission force planners to find suburban locations that protect residen- tial land users from the environmental effects of massive heavy industry investments. 12/ It is interesting to note that the most recent assessments (1992) made by specialists outside the Bank have encountered similar problems in docu- menting the same conclusions, have to cite data published in 1985 or 1986. They conclude that "in 1983 the total population of the satellite towns reached 580,000 or only 8.7 percent of Shanghai's total urban popu- lation. About half of the 430,000 industrial workers in these towns still have their household registration ... in Shanghai [city properl, and a substantial number of workers commute to work from the central city." K. Fung, Z. Yan, and Y. Ning, "Shanghai: China's World City," op. cit. - 116 - owned enterprises could be directed to locate there, but, as locational flexi- bility becomes the i=orm, and foreign-funded ventures fill the void, this neglect of potential customer interest could prove to be an expensive experi- ment. Table 4.9: SHANGHAI: BASIC SITUATION IN MINGHANG AND WUJINGs 1980-90 Total Land Number Industrial Satellite town Built-up area per of Land Land area town residents area person workers area per worker (10,000) (ha) (mWp) (10,000) (ha) (m2/p) Minghang 1980 6.02 770 127.91 5.83 495 84.90 1990 12.52 1,818 145.21 n.a. 1,124 n.a. Wujing 1980 1.18 300 254.24 2.34 239 102.14 1990 2.88 381 132.34 n.a. n.a. n.a. City proper 1980 584.00 i4,100 24.14 176.97 3,579 20.22 1990 706.01 21,600 30.59 n.a. n.a. n.a. Sources Shanghai Institute of Urban Planning and Design, 1986, "Impacts of Satellite Town Development on Expansion and Redevelopment of Shanghai," Urban Redevelopment Plan, p. 154; Minghang District Plan- ning Office, 1991. 4.34 Even in Guangzhou, where the co3ts of such a strategy have become obvious, the local Design Institute is busy planning more such satellite townsl Yeung, Deng, and Chen 13/ note that the Huangpu Economic and Tech- nical Development Zone, 25 km from the Guangzhou city center, lacks a sound economic development rationale (from a marketing perspective) and has required large infrastructure investments, ignoring the relative benefits of investing ir, the city core and its near suburbs. As they rightly stress, "... as much as one third of the investment has been consumed in infrastructure ... (while] the overall objective of using the zone to revitalize technology in Guargzhou might be jeopardized by the ... (fact] that many of the enterprises set up so far were not large consumers of land,14/ public services, or environment- ally polluting and could be located, in fact, in more central parts of the city." These authors then report (uncriticilly) on Guangzhou's plan to develop satellite towns within a framework stressing the need to locate these 13/ Y. Yeung, Y. Deng, and H. Chen, "Guangzhout The Southern Metropolis in Transformation," in Y. Yeung and X. Hu, eds., China's Coastal Cities, op. cit. 141 Bank estimates suggest most leases fall below 1 ha in size. - 117 - 20 to 80 km from the city centers, populated with 200,000 to 500,000 resi- dents, too distant to ever be absorbed by the emerging metropolis. The lesson here is all too simple: Master Plan practices adapt reluctautly to the requirements of a market economy. 4.35 In spite of the trends over the last decade, most cities' Master Plans do not anticipate "exurban sprawl," and planners are yielding only slowly to such market economy signals. On the contrary, economic growth is expected to be confined to selective "old city" redevelopment, carefully regu- lated suburban development, and isolated exurban towns and zones. Satellite imagery confirms that a vast amount of pe.l-urban development has emerged in the last decade, often equal in scale to the officially requisitioned land meant to follow, however imperfectly, the Master Plan guidelines governing the growth of the officially sanctioned city. If China has a developing country equivalent of an "informal" land and real estate market ("black market" trans- actions aside), then one finds it at the margin, emerging, for example, in the suburban counties and even within the strips of land belonging to the city proper but outside the planning area clos2ly supervised by municipal planners. It bears repeating that, from the vantage point of the year 2000, the de facto urbanized area of China's cities will far exceed anything found in the color- coded planning maps. Add to that the fact that, in selected areas of China (the Pearl River Delta of Guangdong Province; and the conurbation radiating from Shanghai into Jiangsu and Zhejiang Provinces, along particular transport corridors), cross-jurisdictional megapolitan areas are emerging. This unfore- seen pattern of urban development dramatically illustrates the challenge faced by the conventionally trained Chinese planners focused on relatively limited portions of overall urban development, defined in purely spatial terms.15/ 4.36 Peri-urban development is poorly understood, since few local or out- side researchers have invested much time in studying the phenomenon. Never- theless, even if one confines attention to any one metropolitan area, and ignores the development and implications of an interconnected set of cities creating an unprecedented "urban field," the fact remains that peri-urban development is the fastest growing part of China's urban economy.16/ 4.37 It is ironic that, after all the arguments about the benefits of agglomeration economies have been documented, and given that, across coun- tries, cities have always been labeled as the engines of modern economic growth, in China the market economy operates with fewest constraints away from the supervision of the city planners and their related agents of "control." Here, in the far suburbs, an unfettered market economy is emerging and a 15/ This report did not have the mandatti or resources to explore the implica- tions for urban planning and regulatory practices created by the emer- genue of cross-county or even cross-provincial conurbations. 16/ The Ministry of Agriculture and the State Council Development Research Center estimate that, by the year 2000, half of all industrial production will be acr"unted for by township and village industrial enterprises (TVIEs) (Chl..a Daily, June 2, 1992). Most of this is assumed to be peri- urban in nature. -118- vibrant real estate market along with it.17/ The flexibility offered by this peri-urban economy has become self-evident not only to the local communi- ties, but to "outside" investors as well. Casual empiricism, buttressed by satellite imagery of the cities studied in this report, suggests that ringing every city proper is a new economy whose development is supported by outsid- ers--seeking subcontracts or investing directly in land for factories, hous- ing, warehouses, and solid waste disposal sites. A massive "leapfrogging" effort by otherwise constrained "urban" investors is complementing and often underwriting relatively unregulated, locally financed urbanization. Some municipalities have simply tried to accommodate these trends (Box 4.6), while providing elements of land use management regulation in rural areas. In most cases, county and township administrators have given little attention to pern- urban land use management. Box 4.6: GUANGZHOU: SUBURBAN COUNTY URBAN LAND DEVELOPMENT POLICIES Compared with other cities, such as Shanghai, the suburban counties in Guangzhou enjoy more autenomy. Guangzhou's city government has allowed counties and districts to set up their own export Industrial zones to attract foreign and joint-venture investors. About 57 hectares of land have already been allocated for building industrial parke or residential quarters. Among 19 suburban towns, as many as 5 hectares of land were allocated for such purposes in every town. Even In villages, about 0.8-1.0 hectare of land area per village was devoted to attracting foreign investors. The county and district real estate companies have been encouraged to develop commercial real estate Including residential, retail, and Industrial buildings to attract consumers from the central city. According to city offi- cials, many commodity housing units built in Panyu county, for example, were sold to resi- dents from the central city, who obviously were attracted by the relatively low prices. Such development has gradually integrated suburban counties, once separated by huge institutional barriers and income gaps, Into a real metropolitan region. Much-improved transportation, such as several new bridges crossing the Pearl River, facilitates such development. In fact, most of this infrastructure was financed and built by the county gov- ernments. They certainly understood the Importance of these transportation projects. This contrasts with some highway projects In Shanghai (connecting the city center with satellite towns) that followed the Master Plan, and whose utilization Is very low. 4.38 On an experimental basis, whose scope is not possible to establish at this time, county land use planning (including industrial zoning) and cen- 17/ Vogel summarizes prevailing foreign perceptions in one paragraph: "'Com- pared to the counties of the Pearl River Delta,' observed one of Guangzhou's leading economic officials, 'Guangzhou is like a tired old man.' After reforms, the counties could establish new factories on for- mer marshes or farmland without dislocating very many people or facili- ties. Guangzhoul'sJ ... large population was hard to move and its heavy industry plants ... expensive to replace.... Compared to the small, usu- ally cohesive, group of county leaders who could act quickly, Guangzhou was slowed down by complex issues and four layers of bureaucracy above its enterprises.... [Given prevailing intergovernmental tax-sharing arrangements) ... counties or small cities that began with hardly any industry and grew rapidly had virtually a tax holiday for five year..... Guangzhou had no such advantage." (E. Vogel, One Stei Ahead in China: Guangdong Under Reform, Cambridge, Massachusetts: Harvard University Press, 1989.) - 119 - tralized servicing of wastewater discharge is being introduced. During this decade, township planning regulations will have to become routine. Technical assistance, based on work funded or otherwise supported by multilateral and bilateral assistance, may be able to play an important role in this case. The balkanization of responsibilities, so often identified in sector studies as obstructing progress in reaching particular objectives, is here and in land management, generally, of great importance. Land development issues in and around cities require that institutions allow multidisciplinary teams to work together. The problems at hand cannot be resolved by independent initiatives led by those whose core responsibilitiea are labeled as "agriculture," "urban," "transport," "industry," "environment," or "finance." - 120 - V. INFORMATION AS INFRASTRUCTURE: ITS ROLE IN DEVELOPING AN INTERNATIONALLY COMPETITIVE CITY A. Introduction 5.1 There is little "glamoutr" attached to promoting the systematic gath- ering, recording, accessing, and use of information. However, no fundamental reform of urban land management can succeed if this area is neglected. The systems involved can be divided into two component parts: legal and planning- related. 5.2 Note has been taken of the fact that legally binding documents, such as Master Plans, land use and building regulations, and the various laws, licensing requirements, and regulations governing property development are not always easily accessible in China. This section focuses on the systematic recording and easy retrieval of site-specific legal data that provides very basic information on the boundaries of each site, the "ownership" rights asso- ciated with each plot, the transparency of rules governing allowable uses, the "encumbrances" associated with that property (including mortgages), and the procedures and costs of transferring control from one party to another. It bears restating that what is traded on the market is not land per se but a bundle of legallv defined rights and obligations which Provide control over Particular parcels of Property, including stratified interests in buildings, land-use rights, mineral rights, and easements, each of which yields or con- strains income-generation possibilities (Annex 10). And it is worth emphasiz- ing that resolution-dispute mechanisms associated with this bundle of recorded rights and obligations are as important an element as the systematic recording of information. 5.3 The introduction of land cadasters in China offers an opportunity to provide not only an accurate land use register, and a necessary instrument for land use rights markets, but also a means by which to strengthen significantly the efficient management and financing of urban infrastructure programs. Modern digital technology, when applied to a cadastral database (or initially to a subdistrict database) offers a powerful means of storing sharing, and using land information to help planners to forecast the impact of planning and investment decisions, and the "timely adjustments" that are inevitably required by unpredictable market-led development (Annex 11). B. Land Administration and Registration 5.4 Prior to 1949, a fairly complete land title cadaster was maintained in China, which recorded the current land description, title and encumbrances, as well as the historical records of ownership, transfer and subdivision or consolidation. In many cities today, the old cadasters exist intact in archives. Following 1949, they fell into disuse, given the abolition of pri- vate property rights and the effects of the Cultural Revolution. The need for a cadaster was seen as unnecessary since titles were now vested in only one owner, the state, and inalienable rights were allocated to users administra- tively (without charge and in perpetuity). Moreover, the trained cadastral surveyors were redeployed to other tasks such as engineering surveying. Reg- istration recommenced, with limited resources, around 1982. - 121 - 5.5 The Land Administration Law and the Constitution, duly amended in 1988, renewed the need for an operational cadaster, which became recognized, once again, as an important instrument of management. Local governments nationwide received a mandate to establish up-to-date cadasters over a five- year period. Following the enabling legislation, relevant central, provincial and municipal implementing regulations were also promulgated. 5.6 As noted earlier, in order to strengthen land administration and management, the 1986 Law established SLA. The staff of SLA are drawn from several existing agencies previously concerned with land administration in various ways, including the Ministries of Agriculture (MoA) and Construction (MoC) and the National Bureau of Surveying and Mapping (NBSM) and their respective line agencies at lower tiers of government. The new and tradi- tional roles of these institutions with respect to land use regulation and information have not, however, been fully reconciled to date, delaying the issuance of implementing regulations and leading to some administrative confu- sion at central and, more particularly, at local levels. Particular anomalies lie in th,e separate registration required for land use rights and building rights, respectively, and in the terms of lease contracts which span both land use and building control conditions. Different local governments have chosen to try to resolve this issue with different administrative combinations and procedures,l/ but operations are clearly hampered by the plethora of maps, registers (and the separate staffs) being separately maintained. 5.7 In terms of land information, SLA is empowered to compile an inven- tory of different classes of land use in China (including urban land) and to maintain a register of user rights. In 1988, a national five-year program of land use classification, and land use rights (and building rights) registra- tion began. Eight primary classes and 46 subclasses of rural land use, and 10 primary classes and 24 subclasses of urban land use are recognized by SLA's I/ Land use rights are the purview of SLA, while building rights are those of MoC. Both types of rights embody controls that would normally be part of "zoning" and subdivision control regulations. The organizational arrangements differ in the degree to which the respective administrative units concerned have been integrated. The units combined generally have different line reporting responsibilities (to SLA, MoC, MoA, NBSM, etc.), which naturally hampers unified management. In Shenzhen, the most advanced organization has been formed, with the creation of a completely integrated land administration system, combining land administration, urban planning, and surveying and mapping units under a unified struc- ture. In Shanghai, a collaborative arrangement exists among the agencies concerned, under the leadership of a committee chaired by a Vice-Mayor. In Guangzhou, urban land and rural land are now administered jointly, and urban land and building administration personnel have been combined into one unit. - 122 - system.2/ Further class subdivisions, i.e., a third, and an even more detailed, fourth tier, can be recognized by local land bureaus. Another agency of government, MoC--through either its Urban Planning or Real Estate Management Departments--also exercises a close interest in urban land use classification through regulation ("zoning" and subdivision control) and building controls, and maintains detailed plans and records.3/ 5.8 At the local level, it is clear that the land administration bureau must register land use rights and issue title certificates. However, the responsibility for compilation of the cadaster has not been clearly defined. Consequently, a conflict developed between NBSM, hitherto the surveyor of fundamental land information including boundaries, and SLA, as to the respon- sibility for surveying and mapping the boundaries of land parcels. This con- flict has hindered the efficient allocation of resources to the topographical and cadastral mapping program now ongoing countrywide. Cadaster Administration 5.9 Land Use Rights. The transition from free allocation of land use rights in perpetuity to the system of time-bound leaseholds for payment must pass through several phases before it will apply to all urban land. This will involve initial registrations and then reregistration as the rights are amended or transferred from time to time. Different local governments are following different timetables. In Guangzhou Municipality, for example, it has been decided that all new user rights should, beginning on July 1, 1992, be applicable for a fixed period. In general, those rights purchased by joint-venture, foreign, or domestic investors will be transferrable. In Shanghai, the decision to limit the period of all user rights has not been taken yet, but transferrable leases are already a feature of all foreign investments requiring land, as well as domestic commercial real estate ven- tures. User rights to most new land offered for development in the extensive Pudong Special Economic Zone (200 ki2) will be in the form of assignable 2/ SLA's rural land classification system conforms to the "Technical Regula- tions for Land Use Investigation" issued in 1984 by the National Agricul- tural Classification Commission. The urban classification system is defined by a separate regulation issued by SLA in 1989 entitled "Survey Regulations for Urban Land Cadaster," which applies to all urban, town- ship and independent industrial and mining development areas. 3/ MoC, following from the provisions of the Land Administration Law as amended, has decreed that its line representatives at the local level should register building rights and issue corresponding title certifi- cates. The conditions applying to the building rights often overlap with land use rights. Possibly adding to the confusion, an Urban Planning Act was passed in late 1989 in order to define and empower the urban planning function and to systematize the process of urban planning. The consis- tency between the two laws and the administrative implications have not yet been explored and resolved. - 123 - leases or "market" rents reflecting a realistic rate of return on implicit land values.4/ 5.10 Proof of title ownership is fundamental as a form of collateral security.5/ The cadastra.l registration system established does not appear to provide absolute clear proof of title, since it may be challenged in a Chinese court of law. Nevertheless, the registration system established is intended to e.Jtablish reasonable proof of title ownership, and a meticulous process of examination and field investigation is followed to ensure the issue of a correct certificate (Box 5.1). Before recourse to legal adjudication, any title dispute is first adjudicated by SLA. Box 5.1: TITLE ISSUANCE PROCEDURE IN GUANGZHOU The process employed by Guangzhou's Real Estate Management Bureau (Land Management Division) is typical in many respects. The Land Management Division is organized into three unics: a registration division (103 staff), a land survey division (105 staff), and an archives division (40 staff), and is responsible for registration of all parcels in the urban development area (approximately 600,000 cases). In other municipalities (e.g., Shanghai and Chengdu), this function has been devolved to District Offices (and County Land Administration Offices) in order to spread the workload. All legitimate land users are required to register their interests. The registration division publicly invites, accepts and records applications for land use rights certificates and carries out the examination of the application and supporting documentation, including field surveys to determine rightful ownership. Upon Initial clearance, the parcel concerned is then identified in the field and the plot corners marked and sides surveyed in detail to produce a certified plot diagram indicating the plot location, dimensions, area and existing improvements (buildings, roads, etc.). With the confirmation of the physical aspects of the parcel, the intention to issue a certificate is publicly advertised. If no objections are received, a certificate is pre- pared, and registered. The original is filed in the Archives and a conformed copy Is given to the applicant. (When district- or county-level offices are involved, a further copy ls retained at the local level). Condominium or strata titles are registered through this same procedure. The Archives are well organized and generally well equipped for long-term stor- age and maintenance (lncluding document restoration), and retrieval of the document files (the premises visited were generally housed in new buildings equipped with microfilming capabilities as well as climate-controlled air conditioning and fire protection). Proces- sing of applications reportedly takes between 4 to 8 weeks, depending on the completeness and accuracy of the application. Fees related to the size and value of the parcel are charged for the registration and issue of the certificate (approximately Y 15/plot for households and Y 125/plot for enterprises). Fee income is not sufficient to cover adminis- trative costs in the case of Guangzhou. Proposals are being made by SLA to increase inves- tigation and registration survey fees substantially. 5.11 Building Rights. In two of three cities where the issue was care- fully studied, progress of land use and building rights registration, which 4/ For industrial land users, particularly for Shanghai-owned enterprises, this reform is likely to be relaxed. 5/ In Shanghai, a China Investment Bank loan derived from the proceeds of one of the Bank-supported industrial development credits could not be given to a proposed industrial borrower because of the lack of a valid land use certificate. - 124 - begin in 1988, was reported as well advanced.6/ Progress was greatest in the newer built-up areas of the city and the county areas (including reregis- tration, in some cases of redevelopment), but much less elsewhere. Registra- tions in such areas are being processed on an ad-hoc basis. Notable omissions were also found with respect to the land and buildings occupied by public agencies (administrations, schools, hospitals, etc.). Such gaps reduce the value of the information base and hamper analysis. 5.12 Delays in cadaster preparation (caused in part by irifficient sur- veying techniques) are quite serious in some instances and the resources allo- cated for the preparation of the cadaster may now be inadequate to complote the work in a timely fashion. Building registrations are, on the whole, gen- erally more advanced than those for land use rights because of an earlier start promoted by MoC; however, each building is registered separately and there is no mechanism to determine the aggregate building righto, involving multiple properties, held by one property "owner." One issue which appears not to be a major barrier to cadastral work is emergence of "multiple claim- ants"; instances do obviously exist, but they cover fewer than 5 p' -ent of all cases. In one example, in Shanghai, the issuance of a land use certifi- cate was delayed because the original unterprise occupying the land had been restructured and it was unclear which of the resulting new entities had a claim to the assets of the original structure. Appeals procedures exist to cover such instances, beginning within the local land bureaus and moving on to the Provincial land bureaus; ultimately, cases could be turned over to the civil court system. 5.13 Information Access. The hard-copy cadaster informat4.on stored in the Archives is nominally open to public inspection, but in practice, only "authorized" persons (owners, attorneys acting for owners or purchasers, bona- fide potential developers, etc.) are allowed access. Nevertheless, the Guangzhou archives reported dealing with some 40,000 inquiries during 1990. A digital database has been initiated to improve retrieval and facilitate inquiries, but is far from complete. All new registrations are recorded in the database. Similar systems are being developed in other municipalities. In some cases, the cadastral mapping information is also being digitized for use in information systems for broad municipal planning and administration purposes. Too often, however, the statistical information is kept at the ward or district level. In the absence of automated aggregation of this informa- tion, municipal authorities and potential investors have no was to access the information easily. C. Geographic Information Systems 5.14 The cadastral maps and their parent topographic base maps are a source of fundamental physical information. Other important urban information concerning thematic topics such as land use, land values, demography, owner- ship, social conditions, etc., or also physical data such as underground 6/ The building registration process is also a meticulous discovery process which includes an examination of the structure to ascertain conformity with "zoning" and building controls. Illegal buildings must be rectified prior to the issuance of the building rights certificate. - 125 - infrastructure, etc. is essential to efficient planning. Such information, when overlaid onto the fundamental database, provides a comprehensive GIS, invaluable to urban planners and administrators. Unfortunately, under current arrangements in China, such information is typically stored on separate maps or In tables or other forms of hard copy in inconsistent formats that are difficult to reproduco or manipulate and, therefore, oftea not consulted fre- quently. However, recent advances in planning technology have opened up the opportunity to store urban information in digital form suitable for convenient analysis and manipulation. The results could be readily presented and repro- duced, thereby facilitating the use of current information in support of decision-making. Digital GIS and related applications (such as "zoning" anal- ysis, public services demand modeling, infrastructure network analysis, etc.) are actively under development in China.7/ The desire to improve land man- agement and urban management could benefit greatly (particularly in the face of the currently fragmented responsibilities for urban land use planning and regulation, as well as primary infrastructure planning) from the use of uni- fied databases and the digital technology now available to conveniently acquire, manage, analyze and apply or present the information. This would obviously entail administrative reforms as well. 5.15 The advent of computer technologies and digitizing of data has made the acquisition and storage of information much simpler, and greatly facili- tates its retrieval and manipulation. In addition, computer programs are able to present the maps and tables, perform the analyses, and generate the carto- graph automatically. All this is now possible with a relatively inexpensive, desktop personal computer (PC) system. Thus, the urban planner or administra- tor can have available a "live" database with which to work, as well as the tools to make projections and display results quickly. Hence, a more dynamic planning process is possible that can more accurately drive planned objectives and match expenditure proposals to reality. The information stored can include the location of people and their workplaces, the physical characteris- tics of the urban area, the level of services available and the location of physical infrastructure networks. This information can be used, for example, to forecast where people will live and work in the future, the problems they will face and the amount of services they will require. The information (and projections) can be updated easily and frequently, and used interactively with computerized models of supply and demand for transport and public utilities W/ Many national research institutes, university laboratories, provinces and municipalities are engaged in GIS development for municipal operations. Much of the research, although supported by external donors, is uncoordi- nated and may be duplicated. The Bank is supporting GIS development through the Key Studies Project (Report No. 2210-CHA) in three key state laboratories in the Wuhan Technical University for Surveying and Mapping, Tongji University in Shanghai, and Nanjing University. The project pro- vides for collaboration and research exchange mechanisms which could be usefully applied in the case of GIS development. In addition, applica- tions for GIS are under development in Bank-supported projects in the cities of Tianjin, Changzhou, Luoyang and Shashi, assisted by the National Laboratory of Resources and Environment Information Systems of the Chinese Academy of Science's Institute of Geography. For greater detail on implementing GIS, see Annex 12. - 126 - (such as water supply, power, sewerage and draivage) to optimize infrastrac- ture networks and investment programs. It can be used to help select and appraise investment projects and can also be used to help manage the related O&M requirements of such infrastructure (Annex 12). 5.16 The advantages of an automated GIS have not escaped Chinese planners and engineers. Many GIS facilities have been established and an information engineering industry is beginning to emerge to support the potential demand. Many teething problems are apparent, as might be expected. For example, the technology is currently chiefly in the hands of information engineers, and means rather than ends dominate development. Moreover, the present urban planning system is rooted in the traditional static planning methodologies and regulations. Such traditions are not readily adaptable to the emerging dyna- mic planning methodologies. Exposure to dynamic planning methodologies is badly needed, as outlined in Chapter IV. 5.17 Not the least of the problems is the highly fragmented nature of the municipal administration and the data sources. The potential economies of scale and advantages of integrated planning are not yet been grasped because each agency embraces its particular area of responsibility and attempts, in the process, to develop its own independent GIS. Much duplication of software and experimentation is evident, and few usable results have so far been given into the hands of policymakers. Issues and Constraints in DeveloDing GIS 5.18 The development of GIS systems is a widespread activity in many municipalities and research institutes throughout China. Many are facing similar obstacles as well as local constraints. The major issues are: (a) Technology. The GIS must draw upon many different data sources and support many different users. Each of these sources and users will likely have an existing information system in place or under devel- opment. Many subsystems have similar data needs and a common layer (or layers) of information such as topography, cadaster, population, etc. is called for. Unified, digital surveying and mapping technol- ogy, as well as systematic cooperation, should be used to facilitate topographic and cadastral survey work and the creation of a digi- tized, coordinated database. The sDecialized information needs to be stored in a compatible form and referenced in a manner that can be understood by others. The data structure, its classification and coding should be standardized. Hardware and software interfaces are needed to allow exchanges of information. Database programs should. be relational and data acquisition and entry simplified. Software should be powerful, versatile, user friendly and inexpensive. Hard- ware should be as inexpensive and reliable as possible. An inter- agency central clearinghouse for technology development and evalua- tion would be a useful measure to avoid duplicate or redundant research. (b) MuniciDal Information Systems Policies. Systematic policies on the management of (and access to) information are generally lacking, particularly in the area of data sharing and exchange, as well as - 127 _ updating and archiving. Local practices and security concerns affect this issue. Linkages among various commissions and bureaus information systems (Planning, Finance, Construction) are desirable in order to have the economic, financial and physical data necessary for integrated planning purposes. Guidelines on "best" practices in data sharing and exchange are needed. As mentioned above, data structure, classification and coding should be standardized to facilitate exchanges. (c) ADDlications. Most effort to date has been applied to the informa- tion engineering aspects of GIS. Little effort has been spent on identifying applications for the data or integrating the use of the database into routine work such as macrospatial planning, sectoral planning, annual planning and budgeting, project analysis, O&M plan- ning and management, or program and project evaluation. User requirements must be surveyed, and processes and procedures spelled out and their data needs and reports identified. D. Strategic Information Systems 5.19 Chapter IV and section C of this chapter both acknowledge that com- prehensive systems of information collection and processing, detailed enough to provide substantial inputs into the planning of infrastructure project., must be preceded by less-detailed, land-use monitoring systems, or Strategic Information Systems (SIS). 5.20 The objective of such systems would be to identify subdistrict-level development trends, aggregating these in different forms and comparing them with expectations recorded in individual Master Plans. This, then, is not a parcel-by-parcel based system, and can thus be assembled quickly and inexpen- sively. 5.21 The number of subdistricts in any given city ranges from 60 to 300. In Shanghai's case, the 130 subdistricts were digitized into computer- manipulated maps in three days. For each subdistrict, the SIS would concen- trate on a select set of data including: (a) populat on trends; (b) employ- ment trends, by category; (c) land use trends, by c- igory; (d) land use con- sumption trends, per capita or per employee; (e) 1-. ,rice trends, by land use category; and (f) building floor-space trends, uy category and on a per- capita basis. Where necessary, such as in a path-breaking study on the link between infrastructure deficiencies and public heqlth indicators, undertaken as part of the Tianjin Urban Development Project's preparation, other indi- cators, mapping death rates by category for each subdistrict, car, be incorpo- rated.8/ 5.22 The tools and end-results, which identify small-area trends in urban development, are a necessary part of a planning reform that allows Strategic Planning to replace static Master Plans. The key message drawn from Chapter 8/ A. Bertaud and M. Young, "Geographical Pattern of Environmental Health in Tianjin, China," Washington, DC, Infrastructure and Urban Development Department, World Bank, February 1991. - 128 - IV is that producing immutable products is not what matters. With SIS, made operational even before GIS is in place, this transition can begin. 5.23 Figures 5.1 through 5.9, using Shdnghai as a case study, provide evidence of various developmentst and are drawn upon to document certain assertions in Chapter IV, many of which were identified by the earlier study undertaken in 1989 by the local Design Institute, with considerable foreign technical assistance. Among these one can highllght the following: (a) Shanghai is expanding more rapidly that anticipated by the Master Plan, and now extends well beyond the planned area meant to contain urban growth (peri-urban included) (Figures 5.1 and 5.2). (b) Shanghai has benefited from the relatively little damage done to it by traditional Soviet planning practices, for reasons discussed earlier; the population density gradient is not unlike that found in a market economy, especiallr within a 4-5 km radius from the city center, and thus, very different from outcomes reported for Moscow, St. Petersburg and Warsaw (Figures 5.3-5.6); (c) While local planning priorities appear to favor the development of Pudong (where the infrastructure cost of opening up serviced land for urban uses will be enormous), it is the Puxi districts which currently contain almost all population and housing in Shanghai (Figures 5.7 and 5.8); (d) As Shanghai restructures its land uses, the disproportionate land area devoted to industrial uses in central locations, inherited from past location practices, and subject to potential conversion to uses more cor,sistent with implicit land values, is illustrated in Figure 5.9. E. Planning Without Facts: The Case of the Temgorarv PoDulation of Urban China 9/ 5.24 Every traDh and table in this report is in one sense incomplete and inaccurate: the resident population of cities and towns without permanent residency rights is missing from the data, because Chinese planners fail to draw systematically upon the temporary registration records kept by the local offices of the Public Security Bureau. Yet, as already noted, on any given day perhaps 66 million such individuals (defined as staying more than three days) fill the cities and towns, whose "permanent" population totals 301 mil- lion. Even if the in- and outflow restricts the average length of stay of such residents, with only 40 percent staying over one year, the fact remains that the stock of such residents present on any one day keeps growing, and yet local planners have no mechanism to track this group except through periodic surveys. These surveys do not yield small area data useful enough for plan- ning public services and infrastructure investments. 9/ This section draws on Michael Rutkowski, "China's Bioating Population and Labor Market Reforms," World Bank, December 1991, processed. - 129 - 5.25 The issue here involves both social and economic considerations. This population must be housed, and provided with services at home and at work, since approximately 70 percent are in the workforce. In Shenzhen, they represent 80 percent of the total workforce. Omitting them from consideration in information gathering, given that they will become a permanent feature of a regulated market economy, is pure folly. 5.26 The maxim "last but not least" applies to information infrastructure in any effort to restructure urban land management in China's cities. Sys- temic reform involves an interconnected set of changes. Neglect any one of these and the reform will fail to achieve the objective of creating interna- tionally competitive cities. And this goal is the only one that makes sense for China as it looks forward to the year 2000. - 130 - VI. MOVING FORWARD 6.1 In the transition to a post-command economy, China has no peers, either in urban or rural land management reform.l/ As only one example, almost all of CB'ina's millions of farmers have effective ownership rights over their farms; in the Russian Federation, only 144,000 private farmers existed in mid-1992. Practices introduced into urban China over the last few years, involving land and building rights registration, land leasing, the development of profit-driven real estate development companies, are rare or nonexistent elsewhere in the post-Socialist world. Municipal governments, with identifi- able budgets and local autonomy over real estate and infrastructure develop- ment, are still a "novelty item" in other post-command economies. The Chinese have moved gradually, but over several years, to set the legal, institutional and price framework needed for real estate reform. There has been no dramatic "big bang" reform, no effort to create a market economy in 500 days; instead, a very deliberate process has been at work, and the report provides evidence that China has accomplished a great deal, while generally maintaining the macroeconomic and political stabilit% that is an undeniable prerequisite for urban real estate reform. 6.2 Within the last year the pace of land leasing activity has quickened in the approximately 30 cities where the reform process is most advanced. Thirty-five km2 were leased from 1988 through 1991; 70 km2 were leased durin& the first six months of 1992. However, in total this is equivalent to only 1 percent of the built-up area of China's cities. In individual cities like Shanghai, most leases involve for-profit redevelopment of previously occupied sites. And there is still a genaral reluctance to allow competitive bids for land to become alL important tool, promoting the emergence of market (rather than cost-recovery) prices. The best that can be said is that land leasing no longer faces the political onus of being a reversible experiment. Authorities at the highest level have endorsed urban real estate market reforms as appro- priate for China. 6.3 China's need to go through yet another readjustment in relative prices, now involving urban land, will be vastly eased by the experiences gathered over the last decade and by the fact that its inflation rates are low, its economy is growing rapidly, and the :iousehold propensity to save is unmatched in the world. China's links to its Overseas Chinese and compatriots in Hong Kong, Singapore, Taiwan Province, Malaysia, Thailand, the Philippines, and Indonesia, along with its growing links with investors from Japan, Korea, and Western countries, will also immeasurably ease the transition to market. More encouraging still, the Chinese at the central level are now taking the initiative in dealing with real estate reform in a comprehensive fashion and, in the last few months, to develop systemic reform plans that have required 1/ The most det&iled review of real estate reform in post-command economies is found in H. Matras and B. Renaud, "Housing Reforms in Transition Econ- omiess A Survey Report on Country Reforms," World Bank, January 15, 1992, processed. - 131 - little prompting from outside advisors. The Bank's previous conclusions and recommendations on urban reform, focused on Zhejiang Province and Housing, were frankly unanticipated and viewed as controversial and audaciousl they were well ahead of contemplated local policy initiatives. Local research on these subjects was meager, at best. Policy analysis on urban land management reforms, however, is now well under way in China, and foreign expertise will clearly play a supplementary role in these efforts, adding useful "nuts and bolts" advice, and working at the margin. China is taking over the task of major policy reform analysis, as the following examples illustrate: (a) CASS, Urban Land Management in China (June 1991), op. cit.; (b) the Research Team on the Intensification of Land Resource Manage- ment, "Towards a Land Management System with Chinese Characteris- ticss A Study Report on the Intensification of Land Use Manage- ment," Beijing, State Land Administration, October 1991, processed; and (c) a comprehensive policy statement prepared by MoC, on the need to integrate land management into overall urban economic development. 6.4 In one report or another, most themes found in this report are now echoed: (a) CASS vigorously promotes a transition to a land market economy, the regularization of "black market" transactions, and the need to revamp the tax and fee treatment of REDCs, which are clearly recog- nized as the key players in the reform effort; (b) the SLA October 1991 report advocates the expansion of land market transactions. It recognizes that urban redevelopment is critical, and that land uses in urban areas need readjusting. It stresses the interconnected nature of urban land management reform, and calls for a Ministry of Land and Housing Administration, able to deal with the myriad issues involved that now cut across competing bureaucratic lines. It recognizes that the peri-urban economy is becoming part of the urban economy and announces various measures to introduce land use fees for urban uses of rural land (TVIEs, rural housing), with experiments already reported in 35 percent of China's counties. It advocates heavy reliance on monitoring of data and its use in making policy decisions. And it calls for greater involvement of financial institutions, including the use of land-based mortgages. Finally, it emphasizes the need to keep environmental issues "front and center" in all debates. (c) the MoC Policy Statement calls for specific action in the following fields: * urban planning should be restructured to make guidelines far more responsive to resource preservatior. in urban areas, allow- ing floor-to-land area ratios to rise, on average, and be more "developer-friendly"; - 132 - * while advocating government monopoly over land requisition, land should then enter the market economy through such mecha- nisms as leasing, and be subject to only limited regulation thereafter; * regulations should be set in place that facilitate the transfer of urban real estate according to market principles, and prop- erty rights reforms should receive particular attention; * REDCs should be encouraged to expand the scope of their activi- ties, and their profitability should not be endangered by "clawbacks" that could act as serious disincentives to their expansion; * at the local level, the planning, regulation, legal, fiscal, and institutional arrangements should be "unified" to treat buildings and the land they occupy as one; * urban and peri-urban planning should be "unified" and jurisdic- tional barriers inimical to that end should be eliminated; 3 the financial sector and the household, as saver, should be encouraged to participate more fully in promoting the develop- ment of the real e3tate sector; and * what governmental revenues accrue from the reform should be kept at the local level, to help promote infrastructure devel- opment. 6.5 What then can the Bank report add in a concluding statement? The key to success for reform is to recognize the need to balance the further unleashing of market reforms with the development of the necessary regulatory, legal, and fiscal framework. The key challenges faced by Chinese authorities include facilitating the redevelopment of densely populated core-city areas, while anticipating the incorporation of peri-urban areas into the year 2000 metropolis, and allowing for a more market-driven densification of the "Socialist" suburban belt. 6.6 This move to market calls for a holistic approach, with a role for both central and local authorities. The center's role begins with continuing to ensure that the reform takes place within a stable macroeconomic environ- ment, with relatively low rates of inflation maintained even as relative price adjustments take place. Then, an overall policy environment is needed that: (a) promotes the emergence of competitive real estate markets; (b) provides for the legal framework that gives investors the confidence needed to make the buying and selling of land use rights to occur at a faster pace; (c) provides for further clarification of the fiscal regime relating to real estate, while insisting on the establishment of regulatory mechanisms to oversee the tran- sition to markets; and (d) sets firm environmental policy guidelines. 6.7 The Center will have to follow up in specific instances, among which one can list as priorities: - 133 - * clarification of state titleholder responsibilitiest * guidelines governing leasehold terms and renewal conditionas * introduction of a unified real estate registration and cadastral policy; * creation of firm rules governing land use rights, vis-&-vis govern- ment authorities, including the guidelines governing expropriation and compensation; * further insulation of the court system from political pressures, to ensure that perceptions of unfairness in real estate dispute resolu- tion are lessened; * development of unified guidelines involving real estate transaction taxes; and the rights of localities to set property taxes, utility rates, and other infrastructure fees, in the context of a thorouah overhaul of municipal finance and related intergovernimental arranae- ments; - strengthening (as necessary) of environmental protection legislation and regulatory mechanisms to see that factories comply with existing and future environmental regulations. At the local level, initiatives are also needed, if only because no central policy, legal, fiscal or regulatory framework can work without follow-up, detailed enactment by provincial, municipal, and county authorities. 6.8 Among the key roles to be played locally, the following can be sin- gled out: * vigorously promote the expansion of urban land rented or leased on .competitive terms, with reasonably unfettered rights over uses and transfers of title; * unify and complete the real estate registration and cadastral pro- grams under way; • implement and publicize rules governing land expropriation, compen- sation, and appeals processes; * monitor the performance of the local court system in dealing with real estate transaction disputes; within the limits set by central policy, review municipal finance arrangements, particularly policies regarding taxes and fees involv- ing real estate transactions, to encourage transactions to take place (particularly core-area development), while ensuring that the government receives a "fair share" based on transparent rules; vigorously promote the monitoring and "information feedback" through the use of coordinated SIS and GIS systems, that can make it possi- - 134 - ble for Structure Plans to be developed and updated, and to redirect infrastructure investments as critical development and core-area redevelopment bottlenecks emerge; * perform periodic regulatory audits with respect to land use develop- ment controls, to facilitate real estate transactions and to allow critical priorities like core-area redevelopment to proceed in a mora "developer-friendly" manner, consistent with transparent "per- formance" criteria involving health, safety, and environment consid- erations; * strengthen the capacity of local environmental protection authori- ties, and provide them with more staff and resources, in dealing with factory pollution among existing and new enterprises, including those located in township areas. 6.9 The report recommends the adoption of a comprehensive reform strat- egy, knowing that such an approach will not be easy to accomplish and that certain "strands" of reform will proceed ahead of others, for a variety of reasons, including political feasibility. The ideal model presented in the report is, in practice, probably unachievable in the short term. Not all meaeures will move forward in the simultaneous fashion that would bring urban land m&.nagement into the market economy as expeditiously as possible. Never- theless, while conceding the bottlenecks created by "practical realities," it bears restating that change is not necessarily systemic reform, and partial reform will not provide the potential benefits that thoroughgoing urban land management can contribute to the modernization and sustainable rapid growth of China's urban economy. 6.10 Finally, looking back with pride, and with satisfaction in comparing Chinese achievements with those of other post-command economies, local author- ities must nevertheless watch out for the real competitors. Hong Kong, Korea, Singapore, Taiwan Province, Malaysia, Thailand, Indonesia, the Philippines, India and Viet Nam will be after the same investors anC export markets as the Chinese seek. To a greater or lesser degree, as pointed out in this report (but requiring much more "fleshing out" through study tours and expert advice), many of these countries are ahead of China in resolving the relevant issues involved. Monitoring their institutions and practices, and adapting them selectivelv is the best advice the Bank can provide, detailed figures and tables aside. That selectivity will be guided by the need to keep the final target in mindt creating a regulated market economy that is internationally competitive.