Taxing Publicly Owned Land in China

A Paradox?
Yu-Hung Hong, January 1, 2005

After spending more than a decade on restructuring central-provincial fiscal relations, the Chinese government is advancing its efforts to reform local public finance. In 2003 the central government issued a directive to ameliorate the real property tax system in China. To fulfill this mandate, tax authorities are reviewing international property taxation experiences, sending officials overseas to study pertinent models and inviting foreign experts to China for consultation. Yet comparable cases from which the government can draw relevant lessons for tailor-making a Chinese property tax system are few. The danger is that when public officials are under pressure to move the reform forward, they may be tempted to adopt concepts that do not match the country’s conditions.

One recent proposal that may develop into such a scenario is to establish an ad valorem property tax system in which leasehold land would be taxed as if it were freehold. This article explains what the Chinese government’s current proposal entails, why it may not be consistent with existing land tenure arrangements and, more tentatively, how the establishment of a land rent system could mediate potential contradictions of taxing land that is not private property.

China’s Property Tax Reform Proposal

The Chinese property tax system currently has as many as nine property taxes, depending on the definitions (see Hong 2003; 2004). The central government has proposed to consolidate three of these taxes into a single levy to simplify the existing tax structure. One of them is the Township and Urban Land Use Tax (LUT), which all land users (except foreign entities, government and nonprofit agencies, and agricultural industries) are required to pay. To collect this tax, local governments divide their jurisdictions into different taxing zones according to population size or land use. Land in different zones is taxed at an array of tax rates preset by the central government, ranging from 0.2 to 10 yuan per square meter (1 yuan = US$0.122). Some Chinese officials have admitted that the tax rates for the LUT have been set too low; hence its collections have little impact on local revenue. The government plans to eradicate this tax.

The other two taxes, the Building (or House) Tax and Urban Real Estate Tax (URET), will also be subject to reform. While the Building Tax is imposed on income-generating properties held by Chinese nationals, the URET is levied on all real estate owned by foreign entities and overseas Chinese. Both are ad valorem taxes whose bases can be the discount original purchasing cost, assessed capital value or gross annual rental value of the property.

When the assessed capital value (or the purchasing cost for the Building Tax) is used as the basis for tax assessment, the tax rate is 1.2 percent for the Building Tax and 1.5 percent for the URET. If an estimated rental value is used instead, the tax rates for the Building Tax and URET will be 12 and 15 percent, respectively. In some locales, like Beijing, if actual rental value is available because individual property owners rent their dwellings to another party at the market rate, the Building Tax rate will be 4 percent of gross rental income of the property. In view of this discrepancy in taxing local- and foreign-owned real estate, the government would replace these two levies with a single property tax as part of the upcoming reform.

The proposed new property tax would be imposed on both land and buildings at a uniform rate. The tax base would encompass all properties, domestic and foreign, located in rural as well as urban areas. As some public officials argue, a standardized property tax could have at least three advantages. First, the new property tax system may ease tax administration. Instead of administering the collection of the LUT, Building Tax and URET separately, local tax bureaus will be able to concentrate their effort on just one tax.

Second, the new property tax would be a value-based tax, which allows the government to capture future land value increments if property reappraisal can be done regularly. Third, one key purpose for creating the new property tax is to convert selected real estate development charges into a unified tax. Many scholars argue that some local governments might have abused the current system of user charges, thereby making payments for public services unduly cumbersome.

Collecting these charges through the new property tax may lower the transaction costs of doing business. As well-intentioned as the proposal may sound, policy designers might have underestimated the importance of one fundamental matter: the integration of the new property tax system with the current land tenure arrangements.

Property Taxation and Public Leaseholds

As specified in the Chinese Constitution, urban land is owned by the state and rural land is owned by collectives. Local governments, empowered by the state, can assign land use rights to users through a set of leasing arrangements. Lease terms are 40 years for commercial land, 50 years for industrial land and 70 years for residential land. If a local government wants to lease an urban land site to a private entity, it must be assigned through a bidding process. The winning bidder must pay the total set of leasing fees (including a “conveyance fee,” expropriation costs if land is acquired from the collective, and various land allocation charges) in a lump sum and immediately to obtain the land use rights.

The payment of the market-determined conveyance fee allows the lessee to transfer or rent the land use rights to another party and to use them as collateral. In the past, land rights were allocated mainly to private entities through negotiation, but this method failed to collect proper fees due to personal connections or corruption and it was suspended by the central government in 2002.

Users of land assigned administratively to public agencies or state-owned enterprises are not required to pay the conveyance fee, but must compensate the state for any allocation costs. The assignment of the land rights has no term limit. According to the law, if a state-owned enterprise wants to transfer its land rights to a private entity for commercial purposes, it must pay the conveyance fee to the state before doing so. For the transfer of rural land into urban uses or to nonmembers of the collective, the state will first expropriate the land from the collective with compensation and then lease the use rights to interested users for the payment of the conveyance fee and other leasing charges.

Owing to a long bureaucratic process and high transaction fees, many users have transferred their land rights to other parties without going through the proper procedure and registration. As such informal exchanges have gained in popularity, the official land leasing record is no longer reliable. Hence, any future attempt to identify the actual landholders, delineate their land rights, and estimate the leasehold value for tax purposes would no doubt be a difficult task.

The design of the new property tax system must take these unique land tenure arrangements into consideration. Aside from the extensive informality involved in land transaction and possession—a topic that is beyond the scope of this article—the most basic question is: How can the government convince lessees to pay property tax on lands that they do not own?

Certainly not all property tax systems are based on the premise that property owners should be taxpayers; occupiers are sometimes liable for tax payment. In some countries, such as Australia, the Netherlands and United Kingdom, taxes paid by occupiers are referred to as rates, a council tax or a user tax to avoid any confusion. Despite the different names, the calculation of these levies is still based on either the capital or rental value of the property, which is the same approach as for the property tax.

More fundamentally, since the supply of land is fixed, the landowner (the state government in the case of China) would bear the ultimate tax burden even if land users paid the property tax directly to the government. This is because the new tax would dampen the demand for land use rights and in turn reduce the fees that local governments could receive from leasing public land.

Because the Chinese government is both the landowner and property tax collector, lessees who leased land in the past and paid the entire leasehold value without anticipating the additional property tax burden would wonder why they should pay more land tax to the government. Thus it is essential to have a rationale for taxing leasehold land, so as to convince lessees to comply with their property tax obligation.

One way to analyze the matter is to treat property rights as a bundle of rights, which includes the right to own, use, develop, transfer, bequest and benefit from land. This bundle also comprises the right to exclude others from enjoying these privileges.

Viewing the Chinese land tenure arrangements through this lens, the government holds the ownership of land and leases other attributes of the bundle of land rights to private entities. So long as the privileges and obligations of holding the leased land rights are fully delineated and recognized, both legally and by the society, there is no reason why leasehold rights cannot be regarded as private property of the lessees for a specific period of time as stipulated in the lease.

In 1988 the Chinese National People’s Congress amended the Constitution to acknowledge the transferability of the right to use land. Further amendments are needed to explicitly recognize leaseholds as private property and empower the state to establish special legislation for the enforcement and protection of leasehold rights. In this way, the implicit contradiction in imposing property tax on leased public land would be clarified and resolved.

One technical issue remains, however: valuation of leasehold rights for tax purposes. Since the new property tax will be value-based, assessors will face the challenges of estimating the leasehold value of land independently, based on market data that normally reflect a combined value of land and all improvements. Most property valuation methods presume that land is freehold, and that developed real estate markets are present. Neither of these assumptions can be applied to China. Although there are practices that separate land and building values for tax purposes, the divisions are generally based on crude assumptions. How can assessors modify the existing (or invent new) valuation techniques to accommodate these special Chinese conditions?

More important, leasehold value is highly sensitive to the lease term and conditions, both of which can vary significantly from one case to another. At this moment, time-tested mass appraisal techniques for assessing large numbers of leasehold sites do not exist. Do these issues imply that property assessment for tax purposes under the Chinese leasehold system requires a case-by-case approach? If so, do local governments have the capability to carry out such detailed property appraisals for the collection of the new property tax? The Chinese government must find ways to deal with these practical matters if it decides to tax leasehold rights as private property.

It is also extremely important to educate would-be taxpayers and public officials about the distinctions between freehold and leasehold systems. Lessees must recognize that they possess only the leased land rights that are not designed to last in perpetuity. If the rights and obligations of both the state and lessees are not clearly delineated, taxing leasehold rights as if they were freehold could complicate the implementation of future land and tax policy. For example, in Canberra, Australia, and Israel, lessees are requested to pay the entire leasehold value up front, and thereafter they pay an annual property tax (or rates in Australia) for leasing public land. Lease terms in both cases are long and renewable—99 years in Canberra and 49 years in Israel with four automatically renewable terms totaling 196 years.

This method of collecting leasehold charges and taxes is tantamount to the payment system for land in countries where land is freehold. Due to this similarity, lessees have developed the perception that land is privately owned (Hong and Bourassa 2003). This view, albeit legally a fiction, has engendered the expectation that any government’s attempt to exercise its rights as the landowner to retake land for public uses or to demand additional payments from lessees for enlarging or extending land use rights would constitute an infringement on private property.

This expectation has added conflict to government efforts to redistribute land and land value between private landholders and the state on behalf of the public. As Neutze (2003) argued, had the Canberra government provided enough public education about its leasehold system, it would have spared the Australian capital from many intractable disputes over land ownership.

The Chinese government has no immediate plan to give fee simple deeds to private landholders. Thus, if local governments continue to collect all leasehold charges up front and then levy the new property tax on both land and buildings, they may be at risk of creating the same mistaken expectations, that is, that land is privately owned. This may put the government and lessees at odds with each other when there is a later need to reallocate land from private to public uses. Designing a real property tax that will not add more complications to the already unsettling land tenure system is a critical task that policy makers should not overlook.

Land tenure reform is a long, controversial process, however, and the Chinese government would be ill-advised to delay the implementation of the new property tax system until land reform is completed. What the government needs is a transition system in which property tax reform can proceed as planned without interfering with its endeavors to restructure land ownership. Establishing a land rent system seems to be an option.

Land Rent System

Under a land rent system, leasehold charges would be paid in the form of an annual land rent, not a one-time leasing fee. Local land bureaus could continue to assign land use rights by public auction, but the bidding would be to determine the amount of annual land rent. Similarly for lands that were assigned to state agencies administratively, users would pay their conveyance fee for transferring land rights to other private parties in annual installments, which would be equivalent to the yearly rental payments. The land rent system has pros and cons (see Hong 2004 for a detailed discussion); four important advantages are discussed here.

First, collecting a land rent is the most straightforward way to characterize the landowner-tenant relationships between the state and lessees. More important, requesting lessees to make their rental payments annually would serve as a constant reminder of their leasehold relationships with the state.

Second, if leasehold charges were paid in annual installments, local officials would no longer be able to generate a large amount of cash instantly to cover short-term fiscal shortfalls. This in turn may lower their incentive to lease land rapidly—a major malady of the current land leasing system.

Third, research using the input-output (I/O) technique and the 1997 I/O Table of China found that collecting land rent could facilitate the transition to the new property tax system (Hong 2004). Had the central government required all land users to pay an annual land rent in 1997, rental income would have added 29.8 billion yuan (US$3.6 billion) to the government treasury, representing a 2.9 percent increase in total tax revenue (see Table 1). This revenue increase would represent a net gain over estimated tax revenue losses under the proposed property tax reform.

The land rent system, however, may generate a cash flow problem for local governments. When leasing fees are deferred and paid by lessees in annual installments, fewer funds would be immediately available for local governments to cover public expenditures. To resolve this problem, local jurisdictions may borrow money from the central government or other financial intermediaries, using perhaps the future land rent collections as collateral. Loans would then be repaid in annual installments by funds gathered from yearly rental payments made by lessees.

Had the government decided to keep the total tax revenue approximately the same, it could have set the new property tax rate at 4 percent, which is the same as the Building Tax rate for personal dwellings rented at market prices, and then discounted the land rent by as much as 47 percent (see Table 1). With a reasonable tax rate and a substantial reduction on rental payment, taxpayers would be less resistant to the reform.

Table 1 also shows several possible combinations of rent level and property tax rate to produce a revenue-neutral shift. If the government were to increase the new property tax rate to deepen the tax reform, it could lower the rent level to avoid antagonizing taxpayers. This approach would provide local governments with an array of options to adopt the new property tax system in stages and at a pace that suits their economies.

Fourth, the proposed land rent system could keep future tenure choices open. If the sociopolitical sentiment of the country favors public leaseholds, local governments could continue to levy the land rent and property tax at the ratio that matches local needs. Subsequent adjustments to the rent-tax ratio could also be made when new circumstances arise.

If central authorities, in response to popular demand, were to grant fee simple deeds to all lessees, it could order local governments to phase out the collection of land rent and raise the new property tax rate accordingly. As shown in Table 1, directing the reform toward either path would not create adverse effects on local government budgets.

This analysis shows that choices available to the Chinese government are not limited to privatizing land ownership and relying solely on real property taxation to recoup land value. Undeniably, the Chinese government may eventually choose to do just that because it is indeed an option, but there are many other possibilities as well. Why, then, should the government make such a decision now, when there may be other viable alternatives that can keep all options open? Recognizing that there are many choices could unleash the creative powers of policy makers and scholars to imagine a unique Chinese system to capture land value.


Director General of State Statistics Bureau. 1999. Input-output table of China, 1997. Beijing: China Statistical Press.

Hong, Yu-Hung. 2003. The last straw: reforming local property tax in the People’s Republic of China. Working paper. Cambridge, MA: Lincoln Institute of Land Policy.

_____. 2004. Assessing property tax reform in China. Report for the David C. Lincoln Fellowship Program. Cambridge, MA: Lincoln Institute of Land Policy.

_____ and Steven C. Bourassa. 2003. Why public leasehold? Issues and concepts. In Leasing public land: Policy debates and international experiences, Steven C. Bourassa and Yu-Hung Hong, eds., Cambridge, MA: Lincoln Institute of Land Policy.

Neutze, Max. 2003. Leasing of publicly owned land in Canberra, Australia. In Leasing public land: Policy debates and international experiences, Steven C. Bourassa and Yu-Hung Hong, eds. Cambridge, MA: Lincoln Institute of Land Policy.

Yu-Hung Hong is a fellow of the Lincoln Institute of Land Policy. This article reports on selected preliminary results of his research funded by the David C. Lincoln Fellowship in Land Value Taxation.