Housing, Taxation, and Urban Development in China
John E. Anderson
College of Business Administration
University of Nebraska-Lincoln, USA
This paper was prepared for presentation at the International Symposium on Urban Development and Land Economics, Peking University-Lincoln Institute Center for Urban Development and Land Policy, July 9-10, 2010
China’s fast growing urban areas are a fascinating study in urban development and the complex interactions between housing markets and local government taxation. This paper explores several important issues in the intersection of urban development, housing, and taxation with the aim of highlighting some important policy options currently considered in China.
Economic liberalization reforms have not only accelerated general economic growth in the country as a whole, they have also created a secondary market for private sector long term land leases and created transferable land use rights that has let loose a dynamic process of rapid urban development. Within this contemporary China context, the focus of this paper is on analyzing several critical aspects of land development, housing, and taxation. The following section provides an overview of recent land and housing market developments in China. This material sets the context for analytic modeling of several key aspects of land and housing market dynamics. A dynamic model of land development is employed to investigate the impacts of Chinese policy related to rural-to-urban land transition, development fees, and property taxation. The final section provides a summary and conclusions.
2. Land and Housing Market Developments
Background and Context
As the People’s Republic was established in 1949 its stated commitments included the eventual elimination of private ownership of property and the means of production, as well as the establishment of a centrally planned economy monopolized by the state sector. These two commitments were formally included in the Constitution of the People’s Republic of China adopted in 1954 (Articles 8, 10, and 15). With these two important initial commitments, the allocation of land resources was clearly to be affected in subsequent years. Land is both a
property asset and an important factor input or means of production. Lin and Ho (2005) describe the history of land reforms in the Mao era as follows. The new Chinese state abolished the old land system by which rural land had been owned by landlords who had allocated the land to individual farm households. Zhang and Li (2006) report that under the old system during the 1930s-40s, between 50 and 60 percent of the arable land was owned by absentee landlords who rented the land to tenant peasants and farm laborers. The rent paid for the land in that era could amount to 40 to 50 percent of the value of the output produced. These changes were invoked by the Land Reform Law of the People’s Republic of China in 1950. The state also confiscated all suburban land owned by landlords and expropriated the suburban land that had been owned by urban industrial and commercial entrepreneurs. Suburban land acquired by the state in this way was then allocated to peasants for agricultural use. Ownership by the state was accompanied by a limited allocation of suburban land use rights to peasants. The peasants were then subjected to an agricultural tax on their output. The state owned land was not subject to sale or lease. Table 1 provides an overview of these and other key milestones in China’s land allocation system over the period 1949-2009. Each of these key milestones plays a part in the developing story of the vibrant contemporary Chinese land market.
In 1950 land reforms in the rural areas permitted limited private ownership of land alongside state ownership of suburban land. In 1956 regulations on agricultural cooperatives were first promulgated that permitted peasants to retain ownership of their private land, but required them to turn over land use rights on those lands to the cooperatives. Later that same year, the regulations took private ownership of land from the peasants and ended the era of private land ownership. A rural land system centered on collective ownership was formalized legally in 1962 and was in effect until reforms in 1978 when the household responsibility system
was implemented. In that reform use rights to agricultural land were contracted by the collectives to individual farm households, although the ownership of the land remained collective. Even after the dismantling of the rural commune system in 1983, land ownership was collective.
In the cities the socialist state initially formed temporary partnerships with private industrialists, thus not completely eliminating the private sector. But, eventually by 1958 most urban land had become state owned. The small amount of remaining privately owned urban land was eventually taken by the state during the Cultural Revolution. In the Communist command and control economy that was established, land is a means of production owned by the state and allocated to urban uses by state entities. Urban land was allocated at a zero price (unpriced) with no time limit on its use. The state could acquire rural land and convert it to urban use if that was deemed to be in the public interest. The local government would act on behalf of the state, expropriate the rural collectively owned land, and allocate it to the state unit that had need of the additional land. While the state unit paid the rural collective a fee in compensation for the land taking, the fee was undoubtedly not reflective of the true value of the land transferred. The state paid for the resettlement of the displaced rural peasants.
This land allocation regime resulted in a serious misallocation of land resources, as would be expected. At a zero price for land, users who were given rights to land use had no incentive to economize on land. Prices could not explicitly recognize location advantages. Transfer of land use was not subject to prices reflecting recognition of value from competing uses. Rent seeking behavior could run rampant. Lin and Ho (2005) provide a comprehensive review of the Chinese land allocation system over the past fifty years, contending that China’s staggering and sometimes uncontrolled process of land development raises both important theoretical and
practical questions about the dynamics of land use change and the ability of the socialist state to manage land resources.
As Lichtenberg and Ding (2008b) describe the current system, all land in China is under public ownership, but increasingly market forces are resulting in private control. This results in a two-track system of allocation with some land use rights allocated via market mechanisms and other land use rights allocated via non-market means. The transformative institutional change making market allocation possible was the introduction of long-term leases for transferable land use rights. Urban land, in particular, is owned by the state and its use is administered by local government officials. Those officials lease out land use rights to private entities under long-term leases. An up-front conveyance fee is charged for the land use rights transferred with the lease mechanism. Historically, those fees were determined by negotiation between the government official(s) and the private entity wishing to lease the land. Increasingly, however, market forces are being observed as the leases are being allocated via competitive auctions or tenders.
Ownership of Land and the Transfer of Development Rights
Despite state ownership of land, the institutional innovation permitting the transfer of land use rights, including development rights, the real estate sector has flourished in recent years. Deng (2005) chronicles how the transfer of land use rights evolved in recent years in China. Local officials in Shenzhen were the first to initiate land leasing in 1987 despite the fact that the nation’s constitution explicitly banned land transactions at the time. In 1988 the PRC constitution was amended and in 1990 a ground lease system was formally approved by the central government. Public land leasing began nation-wide in 1992. At first this innovation was only applied to foreign and private firms. State owned enterprises (SOEs) and collectively
owned enterprises (COEs) continued to operate on land that had been allocated to them by the central government in the past.
Liu (1998) provides a descriptive background in the development of the housing sector of the Chinese economy over the period of the 1980s and 1990s. Wang (2010) describes the process by which a city takes agricultural land for urban developed purposes and discusses the various forms of threat that have a bearing on land use rights. Furthermore, Chan (2003) discusses the problems of compensation related to land acquisition in China.
Land development in China occurs in both transparent and disguised ways. Since land leasing is such an important revenue source for local governments, they have a strong incentive to engage in land development, whether legal or illegal. Lin and Ho (2005) report that over the period 1995-2002, “…nearly a million cases of illegal land occupation and transaction were uncovered involving a land area of 189,000 hectares, an equivalent of 42 percent of the land acquired legally through land conveyance.” (p.411)
The Peoples’ National Congress approved a constitutional amendment in 1988 that legalized the transfer of development rights, touching off a new era in urban development. The land market began to develop at this point as leasehold property rights were recognized. This process has not been without difficulties, however. Zhu (2004) describes the fundamental institutional change that has taken place during the transformation from a centrally controlled system to a market-oriented system in China over the past twenty-five years. Part of the difficulty in transition to a more market-oriented land market, however is the fact that under socialist system principles land use rights tended to carry with them effective ownership right for existing tenants of the land. While land leasing as an institutional innovation provided a means for development of green fields, they were not fully effective in driving redevelopment of
underutilized urban land. Zhu developed a case study using Shanghai data to illustrate how rapid urban redevelopment was facilitated by clarification of land development rights. Nevertheless, he emphasizes that the “informal and insecure institution of the land development right” causes hasty and non-optimal land redevelopment. His policy prescription is to clarify property rights in order to have more orderly development and redevelopment processes. Zhang (2006) further discusses the problem of China’s asymmetry in property rights and the difficulties posed for economic growth.
With the introduction of transferable development rights, the next important step in China was the innovation of long-term land leases. With the advent of transferable development rights and 40, 50, and 70 year leases (depending on land use) the land market has flourished.
Peterson (2006) states that “China used land leasing as part of a strategy to jump-start urban infrastructure investment after decades of neglect.” Indeed, leasing has been a major source of infrastructure finance for Chinese municipalities in recent years. Anderson (2009a) examines infrastructure finance in China, with an emphasis on the use of development fees to finance infrastructure projects. Peterson (2006) observes that cities which were pioneers in land leasing have more recently shifted their strategies to rely more heavily on producing income streams generated by the infrastructure that has been created. In this way, they can recover the capital costs of developing the infrastructure. Bird (2004) provides sound advice for financing urban infrastructure in China. Zhu (2005) provides evidence from Shenzhen that over the time period 1988-99 very little of the land allocated by leasing for development was subject to market forces. Of the 36.4 square kilometers of land leased during that time period of fast economic development nearly 98% of the land was allocated by negotiation and only 2% was allocated by
tender or auction. Zhu concludes that “The old system of free allocation of land survives in the guise of land leasing by negotiation.” Xie et al (2002) provide evidence on price differentials between land sold in private transactions and land sold via open tender in Shenzhen over the period 1990-92. They found that open tender prices were from four to six times greater than private transaction prices, confirming Zhu’s generalization. Rao et al (forthcoming) provides evidence from prefecture-level cities regarding land leases via auction and negotiation.
Anderson (2009b) develops a model of the optimal leasing trajectory over time and demonstrates that lease prices should rise exponentially over time, at a rate equal to the discount rate. This model mimics the familiar Hotelling Rule for optimal extraction of a non-renewable resource. Under this rule the undiscounted price of land is rising at a rate equal to the interest rate.
Real Estate Market Development
Deng et al (2009) examine the factors driving Chinese housing prices. They use a sample of data from each of 30 provinces or cities during the period 2000-2005. Fundamentally, they find that house price variations are caused by changes in household disposable income, new-build supply, housing units sold, unemployment, and land prices. The most important of these factors is the land price, with an especially strong impact after the adoption of the new system of granting land use rights in 2002.
In order to improve market competition for land and make land use rights more transparent, the government adopted a new method of granting land rights on July 1, 2002. Deng et al (2009) report that on that date China’s Ministry of Land and Resources implemented a new set of regulations: Regulations Regarding the Granting of State-Owned Land Use Rights through (1) Invitations to Tender, (2) Auctions, or (3) Listings. Since that date, China’s Ministry
of Land and Resources has implemented additional regulations such as Decree 39 which expands the scope of mandatory land grants via tender, auction, or listings, to industrial land. The new system has been fully implemented as of August 1, 2004. In the interim from 2002-2004 both the old system and the new system were simultaneously in use. Most recently concern has focused on the potential for a housing asset price bubble in China. According to Bloomberg News (2010) banks in China have been directed to conduct stress tests that include scenarios in which house prices drop by 50 to 60%. Dyer (2010) outlines the case for a Chinese housing asset bubble in dramatic terms.
As Chinese cities have rapidly grown in an era of automobile transportation, rather than bicycle transportation, the cities have been developed in highly decentralized patterns. For example, Fu et al (1999) find that urban redevelopment in Shanghai was highly decentralized in the early 1990, at least in part due to the need for districts to pay resettlement costs for displaced residents. This contributed to a spatial mismatch between the supply of redevelopment sites and the market demand for commercial space. This and other concerns brought on by rapid development have lead to urban growth controls of various forms.
Taxes and Other Methods of Controlling Urban Growth
During the period of time when the government owned all real estate, there were no explicit prices for real estate in transactions. Property taxes were implicit, not explicit--the government did not tax itself with explicit property taxes. Starting with the housing privatization trend of the 1980s, however, the government began to tax the housing rental revenue stream. Gu and Trefzger (2000) report that by the 1990s property owners in the rental housing market were paying from 37% to 60% of their rental income in taxes and fees. In 1994, as part of tax
assignment reforms, the VAT was applied to the rental value of land. Many observers indicate that the number of taxes and fees, and the aggregate level of taxation that results, is too high. That criticism, however, must be examined in the context of the public services provided.
Gu and Trefzger (2000) indicate that taxes and fees on rental housing in Beijing, for example, amount to about 31% of rental income for individual owners and about 45% for rental companies. These total amounts include a 12% property tax, a 5.5% operating tax, a 33% land usage tax, and other fees. The income tax is applied to a gross revenue base, with no deductions for depreciation or maintenance expenses. Cruz (2008) chronicles the various transactions costs related to real estate across Asia, including China. He also provides information on the security of property rights and the transparency of real estate markets.
Given that the political-administrative system was still highly centralized, enabling direct control in many ways, the main financial control employed centered on the payment of taxes and fees for development. Western methods of urban growth control such as growth boundaries and sophisticated zoning systems were not particularly applicable. Establishment of tradable development rights has been implemented in limited areas of China, primarily to preserve agricultural land in critical areas.
Farmland Conversion and the Dynamic Balance Policy
A major policy concern in China has been the preservation of farmland in order to provide food security for the country. Lichtenberg and Ding (2008a) describe the two fundamental laws that govern Chinese farmland preservation policy: (1) the Basic Farmland Protection Regulation passed in 1994, and (2) the New Land Administration Law passed in 1999. The first of these forbade the removal of prime-designated farmland from cultivation, while the second required
each province to designate 80% of its cultivated farmland as prime and tightened regulations on conversion of farmland to developed uses. Tan and Beckman (2010) outline the three quotas stipulated by land use plans at all five administrative levels in China (central, provincial, prefecture, municipal and township). The three quotas are: (1) a quota stipulating the maximum amount of land permitted for construction within the planning period, (2) a quota stipulating the minimum amount of farmland that is permitted during the planning period, and (3) a quota stipulating the maximum amount of converted farmland for construction during the planning period. The first quota provides a restriction on the maximum amount of construction land in every administrative district during the planning period, which is allocated by an upper-level authority. The second quota determines a minimum amount of farmland in each jurisdiction during the planning period, which is similarly allocated by an upper-level authority. Finally, the third quota gives authority to each upper-level government entity to determine the annual quota of farmland conversion for urban development at each sub-level, taking into account local economic conditions, national industrial policies, and other factors.
A recent innovation in the implementation of these quotas is significant in that it permits some flexibility in meeting the quotas. In fast growing cities the need to convert land to developed use, while complying with the dynamic balance requirement, has led to experimentation with an innovation called the linkage between urban land taking and rural land giving (LUTRG). This innovation was tested in pilot cities in 2006 and the central government has recommended its adoption nationwide in 2009. Under this innovation, a project links the legal urban development that takes farmland and converts it to developed use, with the reclamation of the same amount of farmland obtained by adjusting parts of the rural construction land. For example, farmers’ residential land may be converted to cropland, or existing farmland
may be cultivated more efficiently or intensively. The result is that the total amount of farmland is not reduced and the total productivity of the farmland is maintained.
Wang et al (2009) provides an insightful view of the growth and development of the development rights market in Zhejiang in relation to farmland preservation. In particular, they focus on the clash between development and preservation interests and the ways in which that conflict is manifest in bargaining between central and local governments. The central government has a broader interest in food security and farmland preservation while the local government has a more narrow focus on economic development and tax revenue generation.
3. Analysis of Selected Land Development Issues
In this section we consider specific land development policy issues and use a dynamic model to analyze each policy. Specific policy issues considered include: improper pricing of land use rights, optimal development, agricultural land conversion, threats to development, development fees, and ad valorem taxes.
Development Policy Context
In an attempt to cool the overheating housing market in China, the government has implemented a system of up-front fees levied on developers of new housing units. In this section we consider the effect of the fee system and examine what would be involved in the transition from that system to a property tax system.
We can begin with a simple model of the development process in which we assume that the local government owns the land and acts as the developer. This is consistent with Lichtenberg and Ding (2008b) in their characterization of Chinese cities as land entrepreneurs. While the primary allocation mechanism for land is controlled by the government entities that
operate in a highly bureaucratic centralized structure according to Lichtenberg and Ding, urban land is under the control of municipal officials. Economic forces in rapidly urbanizing areas place local officials in a position of being responsive to the demands of the market, and they apparently respond in quite entrepreneurial ways as urban developers. Peterson (2006) contends that local officials try to acquire as much land as possible and then either sell it at market rates, use it as collateral for loans they obtain to build infrastructure projects, or provide the land at sub-market rates to strategic development investors. In any case, the city officials as economic developers must decide both when to develop the land and how best to develop it.
Analysis requires a model, so we begin by describing a basic land development model. Suppose that S is the structural density and T is the time of development chosen. The value of the property is the sum of the pre-development income stream w plus the post-development rent stream R (S,t ) , net of development cost C (S ):
, , . (1)
The post development rent stream R (S,t ) represents the upper envelope of the bid-rent curve in the city. We assume that this function is strictly increasing and concave in S , and strictly increasing in t . That is, 0, 0, and 0. Furthermore, we assume that the cost function is strictly increasing in S and weakly convex. That is, 0, and 0. Finally, the discount rate in this model is r .
Cities as Developers--Improperly Pricing Land Use Rights
Zhu (2005), Peterson (2006), and other authors contend that the market rents reflecting highest and best use are often not used in determining prices, however. If that is the case, and the
property is valued in a way that discounts the development rent stream derived from the granting of land use rights, we can modify the value expression as:
, 1 , . (1')
In this expression, the rent stream is discounted by the factor 0,1 . The larger , the more heavily the rent stream is discounted from its market rate. The discounted non-market value can also be written as,
, , , , 2
and illustrates the rent-seeking opportunity for the city to provide to potential land users. A rent seeker wishing to have the advantage of a favorable lease price granted by a government official would presumably be willing to pay up to the amount of the second term on the right-hand-side of equation (2) for that favor. This is the economic motivation for the development of illegal and black market activity described and documented in Lin and Ho (2005). They describe three contexts in which black markets develop (see Figure 3 on p. 421). First, there is a black market in the rural collectively owned land sector allocating agricultural land for construction purposes used for town-village enterprises or for housing. Second, there is a black market in the urban state owned land sector allocating land for commercial uses or for state units. Third, there is a black market between the rural and urban sectors affecting the allocation of land.
In China, local governments actively play the role of developers. As such, they
aggressively seek to maximize profits according to Tao (2010). They do this both in acquiring new land for development as well as in the redevelopment of urban villages. Furthermore, the empirical evidence in Tao et al (forthcoming) indicates that local governments deliberately subsidize land development by foreign investors as they underprice the land. They may be
willing to do this in exchange for future VAT tax revenue as they earn 25% of the tax generated
by the business. In fact, Tao et al (forthcoming) finds that land leases via negotiation, rather than via auction, lead to greater VAT tax revenue subsequently. Whether this strategy is effective will require analysis, perhaps along the lines used in Anderson and Wassmer (2000).
Optimal Land Development
The usual developer’s problem is to select S and T in such a way as to maximize the value of the property, V (S,T ). The two first-order necessary conditions for this optimization problem are the structural density condition,
, , 3
and the timing condition,
, . 4
Equation (3) requires that the discounted present value of the added rent due to greater structural density equal the marginal cost of the added density. Equation (4) requires that the developer wait to develop until the time when the annualized cost of development, including the
opportunity cost of the undeveloped land, equals the land rent that is derived from development. The left-hand-side of this expression is the marginal benefit of waiting an additional period to develop, which is the capital cost saved plus the pre-development income earned, while the right-hand-side is the marginal cost of waiting, which is the foregone developed rent.
Figure 1 illustrates the case where the marginal benefit of waiting to develop is rising over time and the corresponding density demanded curve is rising. Turnbull (2005) has
developed this graphical method of illustrating the dynamic aspect of urban development and has shown that when the marginal benefit curve is rising the density demanded curve must also be rising. The figure shows the simultaneous determination of the optimal time to develop T 1 and
the optimal structural density S1 in a growing city where the current land use has a lower structural density than the anticipated future land use ( 0). In this case, the marginal benefit of waiting to develop equals the marginal cost of waiting at time T1, indicating the optimal time to develop the land. At that time, the optimal structural density, seen in the lower panel of the figure, is S1. Figure 2 illustrates the case of a declining city where the marginal benefit is falling over time and the corresponding density demanded curve is declining and the current land use has a higher structural density than the anticipated future land use ( 0). Projects that are developed later in this case are developed at a lower density. This basic model and the graphical illustrations in Figures 1 and 2 provide a context for examining several policy issues concerning Chinese development issues.
Agricultural Land Conversion
A major policy concern in China has been the retention of agricultural land in the face of growing urbanization pressure. As we think about the conversion of agricultural land for developed purposes, equation (4) indicates that the conversion process will be slowed if the marginal benefit of waiting to develop is increased. That can happen if either (1) the pre-development income stream is increased via higher commodity prices for the agricultural products produced on the land, or (2) the capital cost saved by waiting is increased, which requires either a higher discount (interest) rate or a higher capital cost of development.
Hence, as potential policy responses it would appear that the most promising indirect method of slowing the conversion of agricultural land to develop uses is to supplement the predevelopment income stream in some way. Removal of government restrictions artificially
keeping commodity prices low will relieve development pressure and slow the development process. Another policy option is to encourage farmers to produce higher value specialty crops.
Effects of Threats to Development
Restrictions to land use rights carry with them economic impacts. In a private ownership economy, a government restriction on land use is equivalent to a discriminatory tax or an exaction. Any threat by a city government to prohibit development, impose a moratorium on building, or limit the density of construction effectively creates increased risk and increases the cost of holding undeveloped land in the city. As a result, there is a greater incentive for those owning or holding the land use rights to develop the property more quickly. Furthermore, Turnbull (2005) has demonstrated that threats to development have an impact on both the timing and the structural density of development. Any threat to development has the effect of raising the marginal cost of waiting to develop. The threat of a restriction by the government on the allowed use of the land creates a regulatory risk that has the effect of increasing the marginal holding cost of undeveloped land. That can be illustrated as an upward shift of the marginal cost curve in Figures 1 and 2.
Figures 3 and 4 illustrate the impacts of threats to development for both growing and declining cities. In Figure 3 a development threat causes the marginal cost of waiting to develop to rise, which has the impact of speeding the optimal development time from T1 to T2 and thereby causing the optimal structural density to be lowered from S1 to S2. In Figure 4 the development threat raises the marginal cost of waiting, causing development timing to be accelerated and structural density to increase.
One particular threat to development possibilities is the threat from farmer resistance. Wang (2010) describes how resistance by farmers to taking of land for urban developed purposes can add to the cost of development and also impose threats to the very possibility of taking the land for development. While land taking may be mandatory via eminent domain, it requires compensation for the farmers who had been using the land. That compensation adds to the cost of development, but the political resistance from the farmers may also add to the cost and reduce the probability of being able to take the land for developed use.
Suppose that at some point in time the government restricts the range of potential developments that may be permitted. Turnbull (2005) suggests that in such a situation we can consider the development process to be subject to a stochastic survival model. Let p(t) be the probability density function describing the likelihood that regulations have not been implemented by time t. Suppose further that the probability of undeveloped land avoids regulatory restrictions falls over time so that 0. In this case, the present value of a parcel of undeveloped land can be written as,
In this expression, the first term w/r is the value of agricultural rents in perpetuity, which puts a floor on the value of the land. Even if development is never permitted, the land has this minimum value. Beyond the agricultural land use value, development adds potential value given by the second and third terms on the right-hand-side of equation (5). These terms are the same as those in equation (1) with the exception that they are weighted by the probability of development restrictions not having been implemented by time T.
The impact of a stochastic development restriction is to leave the structural density first order condition unchanged, but to alter the timing condition. The modified timing condition is given as,
, , . (6)
Comparing this condition to equation (4) we see that the new second term on the right-hand-side increases the marginal cost of waiting to develop (since 0). The effect is therefore to speed development. Waiting to develop involves a rising probability that restrictions will limit development; therefore the developer will act sooner. Thus, we see that the uncertainty
introduced into the problem by the government’s potential restrictions will cause development to accelerate. Figures 3 and 4 illustrate the effects of a stochastic development restriction for both growing and declining cities.
Effects of Improper Pricing of Land Leases
Consider the case of a Chinese city that is selling land leases to developers at less than their full market value. If the city is discounting the developed rent stream as in equation (1'), by pricing the land lease at a level below the highest and best use of the land, the development incentives will be altered. In the optimal development model, the first order conditions are modified to be:
1 , ,
1 , . 8 Clearly, discounting the post-development rent stream has implications for both the optimal
structural density and the optimal timing of development. The effect on structural density is seen in equation (7) as discounting the bid-rent stream reduces the left-hand-side of the equation
which requires a lower structural density to reduce the marginal cost of development on the right-hand-side of the equation. Since the development cost function is assumed to be convex this requires a lower structural density S . The effect on development timing is seen in equation
(8) and Figure 1where discounting the bid-rent stream reduces the right-hand-side of the
equation (the marginal cost of waiting to develop) and a corresponding reduction in the marginal benefit of waiting on the left-hand-side of the equation requires a smaller development cost, which necessitates a lower density of development and will also result in faster development (smaller T ) in a city in which the density demanded is rising over time. For a declining city with density demanded falling over time development will be sped up, but structural density will be higher as illustrated in Figure 2. Consequently, we can see that the practice of under-pricing land rents has serious consequences for the development process in Chinese cities.
Effects of Development Fees
Now we wish to consider the effects of levying a development fee to slow the development process, as has been the recent practice in China. Suppose that the city imposes a development fee on the developer, charging up-front for the land lease and other fees. If we add this fee F to the cost of development in equation (1) we have
, , . (9)
The first effect to note is that the value of the property is reduced by the fee. We should expect to find, therefore, that market prices of homes would be reduced as a result of the up-front lump-sum fee. Beyond the impact on property value, we need to consider the effects on the optimality conditions for timing and structural density, equations (3) and (4). There is no impact on the structural density equation (3), but there is an impact on the development timing equation (4).
Since the fee F affects the marginal benefit of waiting to develop in equation (4), shifting MB upward in Figures 5 and 6, the optimal time to develop is delayed. Therefore, housing
development is slowed, as Chinese cities intend. The delay in development has a further effect, however, as it impacts structural density. The direction of the impact depends upon whether the city is growing with a dd curve that rises over time, as in Figure 5, or is a declining city with a dd curve that falls over time as in Figure 6. In the case of a growing city, the structural density increases due to the slower development process, whereas in a declining city the structural density is reduced. Consequently, a lump-sum development fee will reduce the value of the
property, slow development, and alter structural density in a way that depends on the slope of the dd curve.
An alternative form of development fee is an increasing function of the structural density, which we can denote as a variable fee F (S ), where 0. The rationale for this
specification is that the development fee is likely to be greater the more intensely developed a parcel of land. The taller the high-rise apartment building built on the land, the greater the land lease payment required and the greater are the other fees charged by the city.
Incorporating such a fee in the model of property value gives the modified value expression,
. (10) Maximizing this value expression yields the first order conditions,
, . (12)
The fee affects the structural density equation (11) by increasing the marginal cost of density. As a result, the marginal benefit of structural density must also increase.
Figures 7 and 8 illustrate the effects of a variable development fee. Since the variable fee has the effect of shifting the marginal cost curve downward and the marginal benefit curve upward, there is an unambiguous delay in the development process. Hence, the variable fee has the intended effect of slowing development. Note also that the variable fee has a stronger slowing effect than a lump-sum fee since it shifts both curves rather than just the marginal benefit curve as in Figures 5 and 6. The slower development then has an impact on the density of development, raising the density in a growing city as illustrated in Figure 7 and reducing density in a declining city as illustrated in Figure 8.
Another effect must also be acknowledged. The variable development fee is a function of structural density; hence it acts as a tax on structures as noted by Anderson (2009). Like any tax, a tax on structural density results in less structural density. Consequently, introduction of the fee then requires a lower level of structural density, other things being equal, causing the dd curve to shift downward in Figures 7 and 8. That has an indirect effect on the timing of development as well. Equation (12) has two offsetting effects. First, the marginal benefit of waiting to develop is increased by the fee (on the left-hand-side). Second, with lower optimal density of development the marginal benefit may be reduced. Assuming that the development fee increases the marginal benefit more than the lower structural density reduces it; the net effect will be an upward shift in the marginal benefit curve and delayed development timing. Furthermore, the reduced structural density also reduces the marginal cost. The total effect on timing and structural density is therefore ambiguous.
In a rapidly growing Chinese city the dd curve is rising (as in Figure 1) the development fee unambiguously delays the time to develop. For a given marginal benefit curve, the fee shifts the marginal cost curve downward and thereby increases T. But the effect on structural density is ambiguous. In a declining city the dd curve is falling (as in Figure 2) and a development fee has the unambiguous effect of delaying development and reducing structural density.
Conversion of Up-Front Lump Sum Development Fees into Annual ad valorem Taxes
Next, we wish to consider what may happen as Chinese cities replace the up-front lump-sum development fee with a continual (annual) fee. To do so we begin with the results of Anderson (2010) who has shown that a development fee F(S) is equivalent to a property tax applied to the structure value only (not land value). If we take equation (7) above and set it equal to the comparable first order condition for a property tax on structural value only, we can derive the condition,
This expression indicates that the fee applied to the flow of rent income should, on the margin, be a fraction of the marginal cost of added structural density. For example, if the tax rate applied to the structure is .02 and the discount rate is .05, then the marginal fee should be 40 percent (2/5) of the marginal cost of structural density.
This result indicates that converting from an up-front lump-sum fee to a continual fee implicitly involves selecting the tax rate to apply to structures. Policy makers must make the decision on the structural tax rate in order to set the fee. Alternatively, as policymakers set a fee, they are implicitly fixing a tax rate on structures.
Of course, the most efficient way to tax property is to tax the land and not the structures. Anderson (2010) rehearses the case for land taxation, demonstrating that if the objective is to minimize the excess burden of taxation (i.e. the welfare loss over and above the revenue generated) the tax should be applied to land, not improvements. Short of moving to that extreme, land should be taxed at a higher rate than structures in a graded, or two-rate tax system.
4. Summary and Conclusions
This paper provides an overview of land and housing market developments in modern China since 1949. Key milestones in China’s land allocation system are documented along with a description of the amazing transformation that has taken place. In particular, with the advent of transferable development rights and long-term land leases real estate markets have emerged and supported phenomenal urban development. Table 1 summarizes the key milestones included in the paper’s narrative.
In addition, the paper employs a standard dynamic model of urban land development and uses that model to analyze several current policy issues in China. Those issues include: improper pricing of land use rights, optimal development, agricultural land conversion, threats to development, development fees, and ad valorem taxes. In each policy case, the model is used to gain insight regarding both the timing of urban development and the optimal capital intensity of development. Model insights are used to inform the policy debates.
Figure 1: Optimal Timing and Structural Density in a Growing City
T 1 T 1
Figure 2: Optimal Timing and Structural Density in a Declining City
T 1 S
Figure 3: Effect of a Development Threat in a Growing City
T 1 S
S 1T 2 S 2
Figure 4: Effect of a Development Threat in a Declining City
T 1 S
S 1T 2 S 2
Figure 5: Effect of a Lump-sum Development Fee in a Growing City (dd curve is rising)
T 1 T 2
T 1 T 2
Figure 6: Effect of a Lump-Sum Development Fee in a Declining City (dd curve is falling)
T 1 T 2 S
S S 2
Figure 7: Effect of a Variable Development Fee in a Growing City
T 1 T 2 T 1 T 2 S
S S T 3 S T 3
Figure 8: Effect of a Variable Development Fee in a Declining City
T 1 T 2 S
S S 2
T 3 S 3
Table 1: Key Milestones in China’s Land Allocation System
1949 PRC established with stated goals of (1) eliminating private property and the means of production, and (2) implementing a centrally planned economy
monopolized by the state sector
1950 Land Reform Law of 1950 adopted, abolishing old system of land allocation 1954 Constitution confirms commitments of 1949
1956 Regulations adopted on cooperatives, allowing limited private ownership but requiring land use rights to be controlled by cooperatives; Later in the same
year, private ownership of land ended
1958 Most urban land was state owned by this date
1962 Rural land system centered on collective ownership formalized
1960-63 Famine after fast transformation to the commune system
1978 Reforms introduced the “household responsibility system” in which land use rights were contracted by the cooperatives to individual farm households, with
ownership retained by the cooperatives
1983 Rural commune system dismantled
1987 Shenzhen initiates land leasing system
1988 PRC constitution amended to legalize the transfer of development rights 1990 Ground lease system formally approved
1992 Public land leasing system instituted nationwide
1994 Basic farmland protection regulation adopted
VAT applied to rental value of land
Tax assignment reforms implemented
1996 Dynamic balance of cultivated land policy implemented to result in no net loss of farmland
1998 Land Management Law of 1998 reaffirms dynamic balance policy and tightens rules on farmland conversion
New restriction on residential plots (one per household) implemented
1999 New land administration law adopted
2002-04 New method of granting land rights implemented via: (1) invitations to tender,
(2) auctions, and (3) listings
2009 Central government recommends implementation of LUTRG—linkage
between urban land taking and rural land giving policy, which facilitates land
Anderson, John E. 2010. “The Path to Property Taxation in China.” Chapter 9 in Local Public Finance and Property Taxation in China, edited by Joyce Man and Yu-Hung Hong, Lincoln Institute of Land Policy, Cambridge, MA.
Anderson, John E. 2009b. “The Chinese Municipality Land Leasing Problem.” Conference paper presented at the All China Economics International Conference, City University of Hong Kong, December, 2009.
Anderson, John E. 2009a. “Financing Urban Development in China.” The Chinese Economy, Volume 42, Number 2, pp. 48-62.
Anderson, John E. and Robert W. Wassmer. 2000. Bidding for Business: The Efficacy of Local Economic Development Incentives in a Metropolitan Area, W.E. Upjohn Institute for Employment Research, Kalamazoo, Michigan, USA.
Bird, Richard M. 2004. “Getting it Right: Financing Urban Development in China.” ITP Paper 0413, International Tax Program, Institute for International Business, Joseph L. Rotman School
of Management, University of Toronto.
Bloomberg News, 2010. “China Said to Tell Banks to Test for 60% Home-Price Drop.” August 4, 2010. http://noir.bloomberg.com/apps/news?pid=20670001&sid=amL1TmzgvMCw
Chan, Nelson. 2003. “Land Acquisition Compensation in China—Problems and Answers.” International Real Estate Review, Volume 6, Number 1, pp. 136-52.
Cruz, Prince Christian R. 2008. “Transactions Costs and Housing Affordability in Asia.” International Real Estate Review, Volume 11, Number 1, pp. 128-50.
Deng, Changrong, Yongkai Ma, and Yao-Min Chiang. 2009. “The Dynamic Behavior of Chinese House Prices.” International Real Estate Review, Volume 12, Number 2, pp. 121-134.
Deng, F. Frederic. 2005. “Public Land Leasing and the Changing Roles of Local Government in Urban China.” The Annals of Regional Science, Volume 39, pp. 353-73.
Dyer, Geoff. 2010. “China Property Risk is Worse than in US.” Financial Times, May 31, 2010.
Fu, Yuming, Tsur Somerville, Mengdi Gu, and Tongchen Huang. 1999. “Land Use Rights, Government Land Supply, and the Pattern of Redevelopment in Shanghai.” International Real Estate Review, Volume 2, Number 1, pp. 49-78.
Gu, Anthony Yanxiang, and Joseph W. Trefzger. 2000. “Make It Simple and Light: Some Thoughts on Real Estate Related Taxation in China.” International Real Estate Review, Volume 3, Number 1, pp. 142-61.
Lichtenberg, Erik and Chengri Ding. 2008b. “Local Officials as Land Developers: Urban Spatial Expansion in China.” Unpublished manuscript, Department of Agricultural and Resource Economics, The University of Maryland, College Park.
Lichtenberg, Erik and Chengri Ding. 2008a. “Assessing Farmland Protection Policy in China.” Land Use Policy, Volume 25, pp. 59-68.
Lin, George C.S. and Samuel P.S. Ho. 2005. “The State, Land System, and Land Development Processes in Contemporary China.” Annals of the Association of American Geographers, Volume 95, pp. 411-36.
Liu, Hongyu. 1998. “Government Intervention and Performance of the Housing Sector in Urban china.” Journal of the Asian Real Estate Society, Volume 1, Number 1, pp. 127-49.
Peterson, George, E. 2006. “Land Leasing and Land Sale As An Infrastructure-Financing Option. World Bank Policy Research Paper 4043.
Tan, Rong, and Volker Beckmann. 2010. “Diversity of Practical Quota Systems for Farmland Preservation: A Multicounty Comparison and Analysis.” Environment and Planning C: Government and Policy, Volume 28, pp. 211-24.
Tao, Ran. 2010. “Urban Village Redevelopment in China: Experiences, Challenges, and Solutions.” Unpublished conference paper presented at the International Symposium on Urban Development and Land Policy, Peking University, July 10, 2010.
Tao, Ran, Fubing Su, Mingzing Liu, and Guangzhou Cao. Forthcoming. “Land Leasing and Local Public Finance in china’s Regional Development: Evidence from Prefecture-Level Cities. Urban Studies, forthcoming.
Turnbull, Geoffrey K. 2005. “The Investment Incentive Effects of Land Use Regulations.” Journal of Real Estate Finance and Economics, Volume 31, pp. 357-95.
Turnbull, Geoffrey K. 2004. “Urban Growth Controls: Transitional Dynamics of Development Fees and Growth Boundaries.” Journal of Urban Economics, Volume 55, pp. 215-37. Turnbull, Geoffrey K. 2002. “Land Development under the Threat of Taking.” Southern Economic Journal, Volume 69, Number 2 (October) pp. 290-308.
Wang, Hui. 2010. “The Effect of Negotiations on Compensation in Land Expropriation—Insights from National Survey Data.” Unpublished conference paper presented at the International Symposium on Urban Development and Land Policy, Peking University, July 10, 2010.
Wang, Hui, Ran Tao, Lanlan Wang, and Fubing Su. 2009. “Farmland Preservation and Land Development Rights Trading in Zhejiang, China.” Peking University-Lincoln Institute Center for Urban Development and Land Policy working paper number w024, 2009.10.
Xie, Qingshu, A.R. Ghanbari Parsa and Barry Redding. 2002. “The Emergence of the Urban Land Market in China: Evolution, Structure, Constraints and Perspectives.” Urban Studies, Volume 39, pp. 1375-98.
Zhang, Xiaobo. 2006. “Asymmetric Property Rights in China’s Economic Growth.” DSGD Discussion Paper Number 28, International Food Policy Research Institute, Washington, DC.
Zhang, Hongye and Xiaofeng Li. 2006. “Impacts of China’s Rural Land Policy and Administration on Rural Economy and Grain Production.” Review of Policy Research, Volume 23, pp. 607-624.
Zhu, Jieming. 2005. “A Transitional Institution for the Emerging Land Market in Urban China.” Urban Studies, Volume 42, pp. 1369-90.
Zhu, Jieming. 2004. “From Land Use Right to Land Development Right: Institutional Change in China’s Urban Development.” Urban Studies, Volume 41, pp. 1249-67.