j o ur na l ho me pag e: www.elsev ier .com/ locate / landurbplan
esearch Paper
rban land market and land-use changes in post-reform China: case study of Beijing
infeng Dua,∗∗, Jean-Claude Thill b,∗, Richard B. Peiserc, Changchun Fengd
Harvard University, The Real Estate Academic Initiative, MA, USAUniversity of North Carolina at Charlotte, Department of Geography and Earth Sciences, NC, USAHarvard University, Graduate School of Design, MA, USAPeking University, Department of Urban and Economic Geography, Beijing, China
i g h l i g h t s
Land use changed dramatically in post-reform Beijing from 1992 to 2008.Cultivated land more than halved while two types of urban built land doubled.Land price, cultivated land availability and location are significant dynamic factors.Land reforms affected the spatial–temporal dynamics of land use changes.Conformity of land price mechanism with reform goals depends on land use types.
r t i c l e i n f o
rticle history:eceived 5 September 2012eceived in revised form 21 January 2014ccepted 21 January 2014vailable online 18 February 2014
a b s t r a c t
China launched an ambitious urban land-use reform to improve land-use efficiency, raise funds for infras-tructure construction, and facilitate foreign investment in the late 1980s. Few studies have systematicallyevaluated the effects of the market mechanisms on land-use changes in Chinese cities. This paper takesthe city of Beijing as a case study to quantify the magnitude of land-use changes and model it in relationto spatial and market drivers through time and across the metropolitan region. The analysis uses unique
district-level land survey data from 1992, 1996, 2001, 2004 and 2008, as well as land granting records.Analysis reveals that, since the urban land market was established in 1992, the spatial pattern of landuses has changed dramatically in Beijing, land price has exerted a significantly influence on land-useconversion, and the municipal recognition of an open market for land assets has set the spatial–temporalpattern of land uses on a course that sharply differs from past trends in Beijing.
hina
. Introduction
China’s urban land reform put an end to the so-called Threeo’s (no price, no time limit, no transferring) land-use system and
ntroduced the paid-to-use land granting mechanism with respecto land supply for urban expansion. This reform has enabled lando play a critical role in economic development. On the one hand,he reform has been instrumental in the creation of new entities
n the form of village and town enterprises (VTE) to take advan-age of land at near zero price, as well as large labor surpluses,specially in the early reform period (late 1970s and early 1980s)
(Yang & Wen, 2010). Moreover, it generated huge revenue forlocal governments once urban land markets became widely estab-lished in the late 1980s. According to the China Land and ResourcesStatistical Yearbooks, from 1993 to 2011, the total revenue fromland grants amounted to 13.16 trillion yuan in nominal terms. Onthe other hand, land reform has been blamed for triggering seri-ous problems in land use, such as the inefficient use of urbanland, land hoarding and loss of cultivated land (Jiang, Liu, & Li,2007; Liu & Jiang, 2005; Lu, 2007; Zhang, 2000; Zhou, 2011; Zhou,2006).
Ironically, China’s urban land reform, specifically the establish-ment of a land market, was expected to preserve cultivated landand to promote more efficient use of the land by substituting aland pricing system to replace the old free-of-charge land-use sys-
tem (Li, 1999; Zou, 1994). Empirical studies have confirmed thatthis reform has indeed promoted industrial relocation and urbanspatial restructuring (Wu, 1997; Wu & Yeh, 1999; Yeh, 2005). How-ever, critics have pointed out that this reform has also engendered a
hinese version of urban sprawl and land-use inefficiency (Deng &uang, 2004; Zhang, 2000). So far, these criticisms are either unsub-
tantiated or based on anecdotal evidence. Therefore, this paperims to study the regulating role of market mechanisms on theetropolitan land economy. In particular, we study the impact ofarket reforms on the nature and magnitude of changes in urban
and uses in the Chinese context. We do so thanks to a unique dataeries on land-use changes within the Beijing Metropolitan Areahat spans two decades.
Several bodies of literature are relevant to the study of urbaniza-ion and land-use change. Analysis of land-use dynamics conductedn various geographical contexts has underscored that populationrowth, economic development, income growth and infrastructureevelopment go hand-in-hand with urban expansion (Baskent &adiogullari, 2007; Bilsborrow & Ogendo, 1992; Serneels & Lambin,001; Serra, Pons, & Sauri, 2008; William & Turner, 1992). Con-idering the rapid expansion of urbanization and of the Chineseconomy (gross domestic product grew 140.6 times and urbanopulation expanded from 17.92% to 52.57% of China’s populationetween 1978 and 2012, according to the 2013 China Statisticalearbook), a considerable strand of literature on China’s land-se changes de facto argued about the degree to which urbanxpansion has been balanced. Socio-economic factors of land-se changes have all been examined and employed as drivers of
and-use change, including population growth, economic devel-pment, and infrastructure investments (Fang, Zhang, Zhang, Luo,
Li, 2002; Gu, 1999; He, Shi, Chen, & Xu, 2002; Piao & Ma,006). Natural conditions, especially topography, that act as pre-onditions to constrain or facilitate land development have alsottracted researchers’ attention (Du, 2010). Furthermore, urbanlanning policy and major events, such as the 2008 Olympic Gamesnd the 1990 Asian Games in Beijing, have been found to trig-er remarkable land-use change (Kuang, Liu, Shao, & Sun, 2009).hile these studies contribute to explaining urban expansion and
and-use change, they fail to interpret the emergence of urbanprawl and the persistence of land-use inefficiencies in Chineseities.
Land-use conversion modeling also has limited interpretativeower for this issue. It assumes that the present discounted valuef net returns determines the land-use type of a particular landarcel; land owners would decide how to use their land in ordero maximize their total income (Chakir & Le Gallo, 2013; Plantinga,996; Plantinga & Irwin, 2006; Segerson, Plantinga, & Irwin, 2006).his theory, however, is built on two premises. First, it assumeshe existence of a perfectly working market mechanism to prop-rly reflect future land-use returns and also that expected future orlternative land-use returns would be fully capitalized into price.econd, the theory postulates that land owners have the right toecide how to use their land holdings. Yet, today’s China is stillransitioning from one regime to another and neither conditionrevails. The land market is still underdeveloped and land pricing
s still distorted, while owners and users of rural land lack the free-om to use their land for nonagricultural development. When rural
and is proposed for urban development, it must first be expro-riated and go through an approval process. An extensive bodyf literature has now accumulated on the institutional aspects ofand-use change in Chinese cities. In a recent study, the authorsave shown that the land pricing system helped to constrain urbanxpansion when it was initially introduced; however, it promotedrban sprawl because of local governments’ intervention in landarket. More detailed discussion can be found in Du, Peiser, Thill,
nd Feng (2012), Tao, Su, Liu, and Cao (2010), Zhou (2004) and
thers.
Further studies have pointed to the heavy reliance of the localovernment fiscal system on land revenue as a critical source ofistortion, as these government entities are drawn into becoming
lanning 124 (2014) 118–128 119
full-fledged land developers (Cao, Feng, & Tao, 2008; Lichtenberg &Ding, 2009). It has been found that land grant revenues account for60–80% of local governments’ extra-budgetary revenue, which doesnot need to be shared with the central government. In some regions,land revenues are even comparable to the total budgetary income oflocal governments (Jiang et al., 2007; Liu & Jiang, 2005; Zhou, 2007).While astonishing, such data fall short of serving as evidence ofendemic problems in the land management system. Indeed, heavyreliance of local governments on land revenue does not necessar-ily lead to land-use problems, provided that sound governance is inplace. Tao et al. (2010) pointed out that fiscal incentives have causedlocal governments to turn land development into a money-makingbusiness and made them less cautious in expropriating rural land(Tao et al., 2010). The relationship between land pricing and theland hoarding practices of local governments have been clearlydemonstrated (Du & Peiser, 2012). However, the impact of China’sland pricing system on the pace of urban expansion has so far notbeen studied.
According to the urban land rent theory, urban spatial expansionresults mainly from three fundamental forces, that is, populationgrowth, income rise, and decline in transportation costs. The urbanspatial boundary is determined by the location of where urbanland rents equate to agricultural land rents (Brueckner, 1987, 2000;Edward & Kahn, 2003; Wheaton, 1974). Urban land reform, and par-ticularly the establishment of a land market, has brought improvedeconomic efficiency with respect to land resource management inChina. In principle, higher land prices should help to constrain theurban boundary and slow down the pace of urban expansion, whileholding all other factors constant. However, the reality is at variancewith this expectation and ample evidence of land-use inefficiencyand urban sprawl has been reported (Cao, 2004; Yu, 2010; Zhou,2006).
Therefore, this paper posits that the newly established urbanland markets have an influential role on urban land developmentand redevelopment processes and on the distribution of land uses.A more developed market economy for land and higher land priceswould be expected to be more effective in constraining urbanexpansion. This paper examines the specific impact of land pricesand land markets as a new counterbalancing factor for administra-tive intervention on land use by statistically analyzing their effecton land conversion and land development. Moreover, the paneldata analysis enables us to effectively control the impact of otherpotential factors on land use and to focus on the factors of inter-est in this study. We model the extent to which land price impactsthe pace of urban expansion. The analysis uses unique district-levelland survey data assembled for years 1992, 1996, 2001, 2004 and2008, as well as land granting records within the Beijing Metropoli-tan Area since the 1992 urban land reform. These land-use data arebased on the local governments’ periodic land-use surveys usingdetailed land classifications, supplemented by a systematic fieldvalidation of the interpreted aerial image; therefore, these dataprovide the most accurate and systematic record of actual landuse. Beijing is well suited to the purpose of this study becauseit is a typical mega city undergoing the transformation from aplanned economy to a socialist market economy and because ithas relatively good data records that are unavailable in many othercities. Beijing has a total territorial area of 1.641 million hectares.During our study period, more than half of its cultivated landwas lost, while both urban/town land and mining/industrial landdoubled.
The rest of this paper is organized as follows. Section 2 reviewsthe critical phases experienced by land market development inBeijing. Section 3 measures the magnitude of land-use changesbetween 1992 and 2008. Section 4 models the influence of land
price on urban land-use changes. Section 5 discusses our mainfindings, while Section 6 draws the conclusions of the paper.
120 J. Du et al. / Landscape and Urban Planning 124 (2014) 118–128
2B
bE
2
liAlaaoeX(aiPiu1ui
result of this reform. However, land developers and local govern-
Fig. 1. Annual number of granted land parcels in Beijing.
. Major stages of urban land market development ineijing
The post-reform period can be divided into four stages markedy specific patterns of urban land market development in Beijing.ach stage is now discussed in some detail.
.1. Initial experimentation with the land market (1992–1996)
It is not until 1992 that the Beijing Municipality granted the firstand parcel for private development by tender. This date symbol-zes the beginning of urban land-use reform in the capital region.lthough auctions were also introduced for land granting, most
and transactions continued to be conducted via administrativellocation through 1996. Less than 200 parcels were granted annu-lly to the private sector during this period in all eighteen districtsf the Beijing metropolitan region (Fig. 1). These grants were prettyvenly distributed between the urban core (including Dongcheng,icheng, Xuanwu and Chongwen districts), the inner-ring suburbs
including Chaoyang, Haidian, Fengtai and Shijingshan districts),nd the outer suburbs (the remaining ten districts/counties, includ-ng Changping, Shunyi, Tongzhou, Daxing, Fangshan, Mentougou,inggu, Miyun, Huairou and Yanqing) (Fig. 2). In terms of land areanvolved, more land was granted in the outer suburbs than in therban core and inner-ring suburbs (Figs. 3 and 4). Overall, less than0 km2 of land were granted annually during this period, which
nderscores the sluggishness of the newly established land market
n Beijing.
Fig. 2. Spatial distribution of granted land parcels.
Fig. 3. Total land area granted annually in Beijing.
2.2. Rapid growth of the land market (1997–2001)
From 1997 to 2001, the number of land parcels granted annuallyto private developers grew steadily, from 374 parcels in 1997 to763 parcels in 2001 (Fig. 1). This growth was mainly concentratedin the inner-ring suburbs, as shown in Fig. 2, while granted landparcels in the core area and outer suburbs grew at a slow pacecompared to the previous period. Likewise, granted-land area grewdramatically in the inner-ring suburbs, while granted-land area inthe urban core and outer suburbs maintained a stable level (Fig. 4).During this period, the dominant land granting mode was throughdirect negotiation between the government and developers.
2.3. Enhanced land marketization (2002–2004)
At the request of the Ministry of Land and Resources, Beijingproceeded to establish the Land Consolidation and Banking Centeron February 28, 2002 with the stated goal of strengthening landmanagement and of creating a more transparent and fair land mar-ket. Since July 1, 2002, all land granted for commercial, tourism,entertainment, and so-called commodity housing purposes havehad to be transacted in this center and to go through a regulatedand open tender, auction or listing (TAL) process. Direct negotiationwas no longer allowed for these types of land uses. Land marketsand land development underwent significant transformation as a
ments anticipated the new requirements and rushed to close on anumber of land grants via negotiation, especially in the inner-ring
Fig. 4. Spatial distribution of granted land area.
ban Planning 124 (2014) 118–128 121
ssSgie
2
algbotc(dtuc
3
loItvwdcrfitfcainWaa
3
dmbdp1atdn
iMgt
Table 1Definition of different land use types.
Name Land use
Cultivated land Includes farmland and land for vegetable cropsOrchard land Includes land for fruit, tea and other kinds of
perennial plantsForest land Includes the types defined by NLUCS as “forest
land” sensu stricto and “pasture land”Urban and town land Includes the NLUCS types of “urban land”,
“town land”, “transportation land” and “specialland”
Mining and industrial land Is land occupied by mining and manufacturingbusinesses outside settlements
Rural settlement land The NLUCS definition is modified to emphasizeland occupied by construction projects orother non-farming land use types, includingrural settlements sensu stricto, rural roads,animal and livestock farms in rural areas (xuqin si yang), garden and other crop farms andfields (she shi nong ye and shai gu chang)
Water area Includes land for water facilities, fish ponds,fishery aquiculture, and irrigation
J. Du et al. / Landscape and Ur
uburbs, as shown in Figs. 2 and 4. In order to restrict this sort ofpeculation, the Ministry of Land and Resources and the Ministry ofupervision jointly issued a notice stipulating that no land could beranted through negotiation for any reason after August 31, 2004f the development were intended for use in commerce, tourism,ntertainment, or commodity housing.
.4. Land policy as a macro-economic control tool (2005–2008)
After January 1, 2007, land used for industrial developmentslso must be granted through TAL and the prices should not beower than an administratively stipulated minimum standard. Landranting activities became increasingly competitive and mediatedy market mechanisms. The share of land granted to private devel-pers through TAL increased gradually from 13% to 61% from 2005o 2008. Land-use policies were also adopted as macro-economicontrol tools that were as influential as fiscal and monetary policiesZhang, 2007). As a result, overall land granting activities droppedramatically compared to the previous period (Figs. 1 and 3). A spa-ial pattern can also be discerned since the decline was minor in therban core and more marked at greater distances from the urbanenter (Figs. 2 and 4).
. Land-use changes in Beijing in the post-reform era
Land-use changes may encompass different realities, includingand-use conversion among different categories and modificationf land-use conditions within a category (William & Turner, 1992).n this paper, we concentrate on land conversion from one land useo another. Land conversion data are compiled from land-use sur-eys conducted in 1992, 1996, 2001, 2004 and 2008. In accordanceith the 2001 National Land-use Classification Standard (NLUCS),etailed land-use types are collapsed into eight categories, namelyultivated land, orchard land, forest land, urban and town land,ural settlement land, mining and industrial land, water area, andnally unused land. In a broader sense, we also refer to agricul-ural land, built land, and urban built land, which are composed asollows: cultivated land, orchard land, and forest land collectivelyomprise agricultural land; urban built land encompasses urbannd town land as well as mining and industrial land; built landncludes urban built land and rural settlement land. It should beoted that our classification is based on but differs from NLUCS.e pay more attention to the distinction between non-built land
nd built-land, and between urban and rural regions. The meaningnd definition of land-use types are summarized in Table 1.
.1. Change of land-use structure
From 1992 to 2008, the land-use structure of Beijing changedramatically, as shown in Table 2 and in the maps of Fig. 5. Theost striking change is the sharp decline in cultivated land, which
y 2008 had shrunk to half the area it occupied in 1992 in the 18-istrict metropolitan area. Cultivated land lost 2337.7 km2, and itsroportion dropped from 28.4% in 1992 to 14.1% in 2008. More than46 km2 of cultivated land were converted to other land uses annu-lly. Also, all three types of built land expanded dramatically duringhis time frame. Urban/town land and mining/industrial land bothoubled, while the area occupied by rural settlements witnessedearly a 50% increase.
Urban/town land doubled from 3.6% to 7.0% of the total munic-
pal area, and the total addition to its area amounted to 575.3 km2.
ining/industrial land also saw a dramatic increase, with a totalrowth of 556.9 km2, and its share jumped from 3.2% to 6.6%. Itsotal area expansion even exceeded that of urban and town lands
Unused land Includes land left unused, land unsuitable touse, rivers, lakes, and dikes (tian kan)
from 1996 to 2004. For the detailed transition of other land-usetypes, please refer to Table 2.
3.2. Land conversion matrix analysis
The application of Markov chains was introduced to land-useanalysis as a descriptive tool to measure the rate of change betweendiscrete states (land uses) over time by Brown (1970). Bell (1974)confirmed their compact descriptive power for the study of land-use conversion. Markov chains have since been used in a numberof empirical studies (Mullerand & Middleton, 1994; Weng, 2002).
In this section, we calculate the Markov transition matri-ces of land-use conversion for the three periods of 1992–1996,1996–2001, and 2001–2004 based on cross-sectional land-usemaps to reveal the magnitude and patterns of land-use changes inBeijing. For 2008, only land-use statistics are available, but no land-use maps. Therefore, it is not possible to estimate the transitionprobabilities. We include all eight of the land-use types to facilitatethe comparison of conversion magnitudes, although the main focusof this paper is on land conversion between cultivated land andurban built land. For each period, we calculate an in-matrix suchthat element tij is the probability of transition from land-use type i,given that it is in state j at the end of the transition period. Transi-tion probabilities are estimated by unit of land area and rounded tothe nearest 1/10 percent in Tables 3–5. In order to reveal a full pic-ture of land conversion, we also estimate the out-matrices (whereeach entry tij is the probability of transitioning to land-use type j,given a state i at the beginning of the transition period) during thesame periods, and present them in Appendix.
3.2.1. Land-use conversion from 1992 to 1996Based on the conversion matrix between 1992 and 1996
(Table 3), we can see that more than 90% of the land that was culti-vated in 1996 was inherited from land that was already cultivatedin 1992. Surprisingly, over 40% of the orchard land area existingin 1996 was converted from other land-use types since 1992, ofwhich cultivated land alone contributed 23.9%. Table 3 shows thatcultivated land is the most important source of contribution to thegrowth of the three built land types: It contributed 6.2%, 12.7%
and 24.5% toward land area used respectively by urban/town, ruralsettlements and mining/industry in 1996. Conversely, both min-ing/industrial land and rural settlements had only a small amountof area converted to urban/town land.
122 J. Du et al. / Landscape and Urban Planning 124 (2014) 118–128
Table 2Land use structure in the city of Beijing, China: 1992–2008 Unit (%).
Year Cultivated land Orchard land Forest land Urban and town land Rural settlement land Mining/industrial land Water area Unused land Total
.2.2. Land-use conversion from 1996 to 2001The comparison of Table 4 to Table 3 reveals that all land-uses
xperienced much higher rates of conversion from 1996 to 2001han during the previous period. For instance, cultivated land over-hadowed all other land-uses as the main source of new built land.ll three types of built land encroached on a higher proportion ofultivated land than during the previous period (10.7%, 18.1% and7.8% for urban/town land, rural settlement, and mining/industrial
and, respectively).
.2.3. Land-use conversion from 2001 to 2004As shown by the conversion matrix for the 2001–2004 period
Table 5), land conversion dramatically slowed down in comparisono the previous period, even after accounting for the shorter lengthf the third period. By 2004, land for urban/town, rural settlement,nd mining/industry had encroached only 6.2%, 7.4% and 13.1% ofheir respective areas on land-used for cultivation in 2001.
.3. Spatial heterogeneity of land-use changes
Changes in land use were characterized by significant hetero-eneity across the Beijing metropolitan area over the study period.e calculate here the rate of land-use change from 1992 to 2008
or each of the six main land-use types (i) as follows:
hange ratei(%) = area2008,i − area1992,i
area1992,i× 100
The calculated rates are mapped in Fig. 6. Because urban/townand is the predominant type in the four inner districts of the cityf Beijing, and given that very little land-use change was experi-nced in these districts, we focus on changes in the other 14 districtsn this section. The 1992–2008 dynamics of change in each of theix land-use types displays a marked gradient from the core of theetropolitan area toward the outlying districts. Departures from
hese trends are largely the result of idiosyncratic conditions andircumstances in specific districts.
Cultivated land area declined in all districts, but it did so mostramatically in the inner suburban districts around built-up areas,ecause these areas were more accessible from the urban core, andherefore most easily encroached on by urban expansion. Although
able 5and use conversion in-matrix: 2001–2004.
2001 land uses (%) 2004 land uses
Cultivated land Orchard land Forest land Urban/town
Cultivated land 95.2 15.8 4.1 6.2
Orchard land 1.0 76.7 0.6 1.3
Forest land 1.5 5.4 93.6 2.4
Urban/town land 0.1 0.1 0.1 81.9
Rural settlement 0.6 0.5 0.3 3.4
Mining/industrial land 0.4 0.4 0.3 4.0
Water area 0.4 0.1 0.1 0.6
Unused land 0.7 1.0 0.8 0.3
4.4 43.1 2.4 0.61.5 3.4 66.9 0.81.6 4.2 3.3 58.2
it is located far from the urban core, the Mentougou District expe-rienced a dramatic percentage decrease in cultivated land. It isa mountainous region with only 90.5 km2 under cultivation in1992; by 2008, this number had declined to 18.2 km2. Orchardland decreased significantly in the four western suburban districts(Changping, Haidian, Shijingshan, and Fengtai) nearest to the urbancore but increased significantly in the outer suburbs, as orchardsmoved outward under pressure from urban growth. Contrary tothe perception that forests in the inner suburbs would be put undertremendous stress by urban expansion and suffer sharp losses, themap reveals in fact a dramatic increase in the districts near theurban core. These districts have experienced unprecedented eco-nomic growth and improvements in the standard of living since1992. Given these new conditions, local government has respondedto people’s aspirations for more open space by promoting for-est planting campaigns. The establishment of the Olympic Parkin Chaoyang District and the green belt policy to manage urbanexpansion also contributed to this growth of forest land.
Spatial patterns for the three types of built-land are sharplycontrasted. The area of rural settlements decreased in Shijing-shan District and increased slowly in the inner suburban districts,including Haidian and Fengtai and Chaoyang. Conversely, ruralsettlements increased dramatically in two inner-ring suburban dis-tricts (Changping and Daxing), and two outer suburban districts(Yanqing and Huairou). The possible reasons for these contrast-ing trends are different for inner and outer suburban districts. InHuairou and Yanqing, the rather moderate settlement of new ruralpopulation took on the form of an explosive growth because thesedistricts were rather sparsely settled by small rural communitiesuntil 1992. Improvements in the living conditions of farmers maybe the main reason that contributed to this growth. Conversely,rural economic development and the emergence of various typesof informal development zones at the onset of the economic reformmay have played a significant role in rural settlement expansion inChangping and Daxing.
Mining/industrial land area increased in all districts. This
increase was least pronounced in the four inner suburban districtswhich were already quite industrialized in 1992, and in the south-west area. It increased fastest in the outlying districts of Huairouand Daxing. Finally, urban/town land area increased in all fourteen
land Rural settlement Mining/industrial land Water area Unused land
124 J. Du et al. / Landscape and Urban Planning 124 (2014) 118–128
ges fr
diptd
4u
vboFcooishifltoiault
Fig. 6. Rates of land use chan
istricts. This growth was slowest in Shijingshan and Haidian, andn the western mountainous region, but grew fast in the easternlains, such as Tongzhou, Shunyi and Huairou. Topographic condi-ions constitute the fundamental prerequisite for urban economicevelopment and spatial expansion of the urban frontier.
. Dynamics of urban expansion since the establishment ofrban land markets
Given the massive encroachment of urban built land on culti-ated land, a critical question is to determine the influence exertedy the newly established paid-to-use land market mechanismsn the pace of urban expansion in the Beijing metropolitan area.or this purpose, we estimate multivariate models of annualizedhanges in land development over the 1992–2008 period. The focusf these models is specifically on each of the two componentsf urban land development, namely urban/town land and min-ng/industrial land, respectively. In contrast to the past land-useystem based on administrative allocation, the land pricing systemas caused the market-driven land price to become a major factor
n regulating land use and land development. Given the imper-ect nature of pricing mechanisms, we expect that the effect of theand market on development is also mediated in some complemen-ary ways, namely through demand for development and supplyf developable land. Demand for development, which translatesnto development pressure, is anticipated to be location dependent,
nd more precisely inversely related to distance to the developedrban core, as posited by the bid-rent theory. As a finite resource,
and availability at the start of the development process regulateshe magnitude of development that can take place. The success or
om 1992 to 2008 in Beijing.
failure of the new land pricing mechanism to affect land develop-ment can be used as a measure of the successfulness of the landreform. If the new pricing mechanism has a significant impact onthe levels of urban expansion, as predicted by urban land markettheory, then the reform has achieved some success.
Given the panel nature of the data structure we rely upon, itis customary to first test whether pooled ordinary least squares(OLS) estimation is appropriate. Tests indicate that there is asignificant individual district effect. Therefore, a correlated ran-dom effects (CRE) modeling approach (Wooldridge, 2012) isemployed:
Ytt = ˇXtt + ut + εtt (1)
ut = a + �X̄t + rt (2)
where Yit is the dependent variable, i.e. the relative annu-alized rate of change in urban/town land area (utrelrate) andmining/industrial land area (mirelrate); Xit denotes the vector ofindependent variables, including the interaction variables withtime dummy variables; X̄t denotes the time average of the time-varying explanatory variables; i is the index for each district; andt is the time period; ui and ri are the coefficients specific to eachdistrict to capture unobserved effects; �it is the error term. The CREmodel unifies fixed effects (FE) and random effects (RE) approaches;it assumes that ui is correlated with Xit whenever � /= 0 and ri isuncorrelated with Xit. Therefore, it provides a formal way to account
for FE and RE effects by testing the significance of parameter� .
We use panel data from thirteen of the eighteen districtsthat form the Beijing metropolitan area calculated for the four
J. Du et al. / Landscape and Urban Planning 124 (2014) 118–128 125
Table 6Variable summaries.
Variable Definition Max. Min. Mean Std. err.
Utrelrate Relative annualized rate of change in urban and town land area;Utrelrate = [(change in urban and town land area during aperiod)/cultivated land area in first year of this period]/time length of thisperiod × 100
12.62 −0.03 2.10 3.02
Mirelrate Relative annualized rate of change in mining and industrial land area;Mirelrate = [(change in mining and industrial land during aperiod)/cultivated land area in first year of this period]/time length of thisperiod × 100
6.57 −1.89 1.20 1.33
Lnprice Log of average land price in end year of each period 9.18 0.00 5.62 2.22Lnculand Log of cultivated land area in first year of each period 6.53 2.75 5.27 0.92Disring Distance of each district’s urban and town land centroid to Beijing’s ring
roads; the ring road is the 3rd, 4th, 5th and 6th ring road respectively forthe first, second, third and fourth period; if the centroid is inside the ringroad, the value is defined negative; if the centroid is outside the ring road,the value is positive
77.99 −9.83 30.81 25.68
T1 1992–1996 period, taken as the base periodT2 1 if during 1996–2001 period, 0 otherwiseT3 1 if during 2001–2004 period, 0 otherwiseT4 1 if during 2004–2008 period, 0 otherwise
pBvXdlmtP
alooiceomeTaewtdtacb
othto(mpBrt
moderately significant; its time-average variable (lnculandbar) isalso marginally significant. Therefore, the new land pricing sys-tem has significantly constrained the pace of urban/town landexpansion during the last period, 2004–2008, while it failed to
eriods: 1992–1996, 1996–2001, 2001–2004, and 2004–2008.ecause our interest is in new urban development rather than rede-elopment, the four urban core districts of Dongcheng, Xicheng,uanwu, and Chongwen, which were already overwhelminglyeveloped in 1992, are excluded from the analysis. Along the same
ine, Shijingshan district is excluded as most new urban develop-ent was done through redevelopment, particularly as a result of
he administrative relocation of a large industrial complex to Hebeirovince. Thus, the panel dataset contains 52 observations.
The formal definition of all variables used in the models as wells their summary statistics is listed in Table 6. Since urban/townand and mining/industrial land expand primarily at the expensef cultivated land, the dependent variables of interest use the areaf cultivated land at the start of each period as the basis (denom-nator) against which change in developed area is assessed, whileultivated land area is employed as a proxy for developable landndowment. The mean price of transacted properties at the endf each period is employed as a direct measure of the develop-ent of the land market. Finally, distance to the ring roads is
mployed as a proxy for distance to the developed urban core.o control heteroskedasticity, the land price and cultivated landrea variables are log-transformed. Aside from overall longitudinalffects, we introduce time dummy variables and their interactionith explanatory variables, taking the first period as basis, to fur-
her test the evolution of the effects of the explanatory variablesuring the process of land reform. For instance, T2lnprice denoteshe interaction of period T2 with the land price variable (lnprice),nd so on. Finally, the time-average of the explanatory variablesarry the “bar” suffix, such as lnpricebar, lnculandbar and disring-ar.
The possibility of spatial effects is tested with a Moran’s I statisticn model residuals; these tests fail to detect any spatial autocorrela-ion. A modified Wald test indicates there is significant group-wiseeteroskedasticity, while Pesaran’s test, Friedman’s test and Frees’est all indicate that there is no cross-sectional correlation. Therdinary least squares (OLS) estimator, generalized least squaresGLS) estimator, and panels corrected standard errors (PCSE) esti-
ator for heteroskedastic panels are employed. The GLS estimatorresents the most significant p-value while OLS presents the least.
ecause the GLS estimator tends to introduce overconfidence forejecting the null hypothesis while the PCSE estimator generateshe same coefficients as OLS but with more accurate standard errors
(Beck & Katz, 1995), only PCSE models are presented. The estima-tion results are given in Table 7.
Results in Table 7 show that the two models of land develop-ment have a good fit, given the limited number of predictors. Inthe urban/town land model, the time period dummy variables ofthe third and fourth periods, their interactions with cultivated landendowment, with the distance to the ring roads, and with landprice at the fourth period are all significant. In the mining/industrialland model, the third and fourth period of time dummy vari-ables are also significant. With respect to the interaction variables,only that of the cultivated land at the fourth period is signifi-cant. The cultivated land variable in the first period (lnculand) is
xhibit a significant effect on the growth rate of mining/industrialand.
. Discussion
Since the initiation of urban land reform in 1992, the land mar-et has experienced several stages of development in Beijing. Thearket mechanism has gradually dominated land distribution. We
nd that land use has been profoundly altered since the introduc-ion of land reform. The city lost more than half of its cultivated landrom 1992 to 2008, while urban/town land, and mining/industrialand both doubled. The magnitude of cultivated land loss dur-ng 1996–2008 in Beijing is even more than the total agriculturaland loss due to urbanization in several European countries as
whole, such as Italy and the Netherlands, between 1990 and000 (European Environment Agency, 2010). Typically, the mostertile land is cultivated for grain production. Therefore, duringhe process of land-use conversion, the most productive culti-ated land experienced the most pronounced loss. It should beoted that the problem is endemic to all rapidly growing suburbanetropolises. The suburbs of Orange County and San Bernardino
ounty next to Los Angeles, for example, have witnessed a dramaticoss of prime farmland and orange grove to suburban develop-
ent over the last thirty years (Land Use Services Division anddvance Planning Division, 2007; The California Department ofonservation, 2011).
Land conversion analysis for Beijing shows that urban expansionccurred primarily at the expense of cultivated land. Conversion ofultivated land to mining/industrial use consistently outstrippedonversion to rural settlements and to urban/town land, the lat-er accounting for the least. The magnitude of cultivated land lossas most massive between 1996 and 2001, as a result of the con-
ergence of recent market liberalization and of the fast populationrowth of the metropolitan area. The dramatic loss of cultivatedand alarmed public officials causing them to pay more attentiono land preservation. The Land Administrative Law was revisedn 1999 in order to guarantee the nation’s food sufficiency (U.S.mbassy, 1997; Zhang, 2000). This reform had four components:1) it withdrew the local governments’ authority to approve agricul-ural land requisitions; (2) it decreased the standard land area thatrovincial governments could approve; (3) it improved the com-ensation criteria in cases of agricultural land requisition, and (4)
t introduced an approval process for the conversion of agriculturaland for non-agricultural purposes. Throughout the study period,he pace of conversion from other land uses to built-land use expe-ienced a steady decline. The primary source of newly built land wasand under cultivation, and particularly so for mining/industrialand.
Correlated random effects panel data regression modelingnabled us to isolate the effects of socio-economic factors anddministrative power on land use specific to each district and tooncentrate exclusively on the impact of the land market, the cul-ivated land endowment, and its spatial pattern. The regressionesults point to sharp differences between urban/town develop-ent on the one hand and mining/industrial development on the
ther. Mean rates of urban/town land development acceleratedrom 1992–1996 to 2001–2004, before tapering off in later years.side from the influence of the fast pace of economic developmentnd population growth, several innovations in land managementay have significantly contributed to this trend. The land banking
ystem established in 2002 and the promotion of TAL land granting
odes hastened urban expansion in the third period. Other regula-
ory measures that followed the rash of clandestine and speculativerants in 2004 triggered a relative softening of granting activitiesnd land conversions.
lanning 124 (2014) 118–128
Conversely to urban/town land, mean rates for min-ing/industrial developments consistently dropped throughthe study period (although this trend is not always statisticallysignificant). This trend is tied to large expanses of mining/industrialland being part of various development zones formed by expro-priation of land tracks at the beginning of the study period.Land in these development zones was quickly developed at first,especially for infrastructure construction. Subsequent industrialland development occurred at a slower pace.
At the start of the study period, the annual pace of expansion ofurban/town development tended to be positively linked to averageland price, although this relationship was statistically insignificant.From 1996 to 2004, this relationship reversed (although insignifi-cantly) as a direct effect of land reform measures. The land marketreforms attenuated the pace of urban/town development. Thistrend deepened even more and became statistically significant after2004: land prices set by the market effectively controlled the paceof urban/town expansion in Beijing.
Development of mining/industrial lands proceeded quite differ-ently from urban/town lands. Indeed, while mining/industrial landexpansion was originally more pronounced where land prices werelower, the reverse pattern emerged after 1996 and was reinforcedfurther later in the decade, although the price effect failed to besignificant during any time period. Until the 2002 reform was spear-headed by the central government, local governments used lowindustrial land prices as a competitive tool to attract investment(Tao et al., 2010). With socio-economic development and land con-straints becoming more significant, new industrial developmentstended to cluster in the districts that had been most successful inattracting new businesses, which was also reflected in higher landprices.
Beijing’s multiple ring roads were built outward from the urbancore in conjunction with the expansion of the urbanized area. Thesesuccessive rings mark the position of the urbanization front asit was pushed further into the outlying parts of the metropoli-tan area, concomitantly with the growth of the city from 1992to 2008. Distance to this front is a meaningful predictor of thepace of urban/town land development. The directional effect of itsimpact on urban expansion is also captured by assigning a negativevalue for land located within the ring roads and close to the urbancore. The development gradient was negative for each time period(although not significant during 1992–2001), and increasingly soacross the four time periods being considered. Clearly, land reformeffectively constrained new urban/town development to the dis-tricts adjacent to the built-up portion of the metropolitan area.With regard to mining/industrial land, the pace of developmentalso had an inverse relationship to distance from the ring roads,although in a statistically non-significant way. Land reform hasenabled industrial and urban developments to proceed across themetropolitan area according to their own business logic, industrialprojects seeking more peripheral locations than urban projects.This uncoupling of industrial land development from urban landdevelopment is one of the striking outcomes of land market reform.It marks a radical departure from past trends, and points towardincreasing land-use segregation, following in this respect modelsthat prevail in other parts of the world, most notably in NorthAmerica.
Land availability has been a constraint on Beijing’s growthtrajectory for some time. We account for this factor by mea-suring cultivated land area in each district in the initial year ofeach period. After controlling for land price and distance to theurban core, urban/town land expanded faster (although not sta-
tistically significant) in the districts with more cultivated land atthe start of the study period. By 1999, policies of cultivated landpreservation were enforced; urban development became largelyan in-fill phenomenon and large agricultural districts had become
ban P
tctdhmitaciit2
6
Raugtilclrt
bslhvettTC
A
A
J. Du et al. / Landscape and Ur
he least likely to change as far as urban/town development wasoncerned (and this new element of differentiation among dis-ricts was highly significant henceforth). For mining/industrialevelopments, once again a contrasting evolution unfolded. At first,igher land prices resulting from the establishment of the landarket helped to shield cultivated land from encroachment by min-
ng/industrial developments. In the following periods, however,he prevailing “race to the bottom” land pricing strategy aimedt attracting industrial investments placed the districts with moreultivated land in a stronger competitive position to accommodatendustrial developments. Therefore, these districts experienced anncreasing pace of mining/industrial land development and thisendency became more significant during the period from 2004 to008.
. Conclusions
Urban land reform constitutes a significant part of China’seform and Opening Policy. The newly established land market actss a counterweight to the administrative power in regulating landse. As urban land reform matures, its increasing impact on urbanrowth is expected. This paper aimed to analyze the influence ofhe newly established land market on the pace of urban expansionn post-reform China, taking Beijing as a case study. In particu-ar, we have quantified the spatial–temporal properties of land-useonversion across the metropolitan area over the two decades fol-owing reform enactment and have interpreted this dynamic inelation to spatial processes and market mechanisms enabled byhe reform.
This study indicates that after the reform, cultivated land haseen more than halved in Beijing and has become the primaryource for three types of built-land expansion. For a country withimited per capita land resources, this is a severe problem thatas long been a source of critical public concern. The land con-ersion exhibits significant spatial heterogeneity. Urban/town landxpansion pace had a negative gradient with the distance from
he ring roads as well as distance from the urban core. This nega-ive gradient increased in magnitude with each successive reform.hus, land management instruments can be said to be effective.onversely, this tendency is not pronounced for mining/industrial
ppendix 1 Land use conversion out-matrix: 1992–1996.
1992 land uses (%) 1996 land uses
Cultivated land Orchard land Forest land Urban/town
Cultivated land 80.3 5.6 3.7 0.8
Orchard land 7.0 76.9 9.1 0.5
Forest land 1.7 1.9 90.4 0.1
Urban/town land 1.9 0.4 1.2 92.7
Rural settlement 8.4 1.6 2.2 1.6
Mining/industrial land 6.5 1.2 1.8 3.0
Water area 6.2 1.4 2.5 0.8
Unused land 2.4 1.9 25.4 0.1
ppendix 2 Land use conversion out-matrix: 1996–2001.
1996 land uses (%) 2001 land uses
Cultivated land Orchard land Forest land Urban/town
Cultivated land 66.9 7.1 8.6 2.0
Orchard land 13.1 54.1 17.6 1.4
Forest land 1.9 2.2 85.3 0.6
Urban/town land 2.3 0.6 5.2 76.7
Rural settlement 9.1 2.5 3.9 4.7
Mining/industrial land 9.9 1.9 6.4 13.0
Water area 10.7 3.1 5.2 2.4
Unused land 4.3 2.7 33.3 0.6
lanning 124 (2014) 118–128 127
land development, thus leading to ever greater land-use seg-regation between industrial and urban developments. The landpricing mechanism enacted in Beijing in1992 had different effectson the pace of expansion of urban/town development and min-ing/industrial land development, respectively. Land price becamean effective regulator in restraining urban/town expansion. Theland price mechanism for mining/industrial land developmentis more complex because of the combined effect of preservingcultivated land and attracting industrial investment. The enforce-ment of TAL granting modes for certain types of profit-orienteddevelopment in 2002 shifted new urban development away fromlargely agricultural districts. Scarcity of developable land and fiercecompetition to attract industrial investment forced mining andindustrial development to turn to these districts more frequently.
Research reported in this paper is based on the city of Beijing.However, the pace of economic development and the modalitiesof urban land reform show significant regional variations acrossChina. It can therefore be expected that the interplay between land-use change, economic development, land market mechanisms,and urban land regulations would also exhibit meaningful hetero-geneities. Comparative studies across multiple metropolitan areaswould contribute to the portrayal of a more comprehensive contex-tualized theory of land markets in China. Soil quality is an importantfactor influencing land productivity. Illuminating the evolution ofsoil quality and quantifying how it affects land-use change asidefrom the magnitude of cultivated land loss would be helpful formeasuring how land-use changes affect China’s food productioncapacity in the process of urbanization. Research into the dynamicsof urban expansion will inevitably encounter endogeneity prob-lems. We have attempted to control for endogeneity by regressingthe pace of urban expansion on land price. A structural equa-tion and/or instrumental variables modeling approach with moredata on demand and supply of land may help to provide morecomprehensive findings. Moreover, since local governments playa significant role in China’s land-use change, study of these insti-tutional factors will further improve our understanding of China’surban expansion and land use conversion.
Appendices. Conversion out-matrix of land use changes inBeijing during 1992–1996, 1996–2001 and 2001–2004
land Rural settlement Mining/industrial land Water area Unused land
askent, E. Z., & Kadiogullari, A. I. (2007). Spatial and temporal dynamics of landuse pattern in Turkey: A case study in Inegol. Landscape and Urban Planning, 81,316–327.
eck, N., & Katz, J. N. (1995). What to do (and not to do) with time-series cross-sectiondata. The American Political Science Review, 89(3), 634–647.
ell, E. J. (1974). Markov analysis of land use change – An application of stochas-tic processes to remotely sensed data. Socio-Economic Planning Sciences, 8(6),311–316.
ilsborrow, R. E., & Ogendo, H. W. O. O. (1992). Population-driven changes in landuse in developing countries. Ambio, 21, 37–45.
rown, L. A. (1970). On the use of Markov chains in movement research. EconomicGeography, 46(Suppl.), 393–403.
rueckner, J. K. (1987). The structure of urban equilibria: A unified treatment of theMuth–Mills model. In E. S. Mills (Ed.), Handbook of regional and urban economics.Amsterdam: North Holland.
rueckner, J. K. (2000). Urban sprawl: Diagnosis and remedies. International RegionalScience Review, 23, 160–171.
ao, D. (2004). China cancels 4,800 development zones. Beijing: China Daily.ao, G., Feng, C., & Tao, R. (2008). Local land finance in China’s urban expansion:
Challenges and solutions. China & World Economy, 16(2), 19–30.hakir, R., & Le Gallo, J. (2013). Predicting land use allocation in France: A spatial
panel data analysis. Ecological Economics, 92(0), 114–125.eng, F. F., & Huang, Y. (2004). Uneven land reform and urban sprawl in China: The
case of Beijing. Progress in Planning, 61, 211–236.u, J. (2010). (Study on urban expansion of Beijing city) Beijing Cheng Shi Kuo Zhan Yan
Jiu. College of Urban and Environmental Sciences, Peking University.u, J., & Peiser, R. (2012). Land supply, pricing and local governments’ land hoarding in
China. The Real Estate Academic Initiative, Harvard University. Working paper.u, J., Peiser, R., Thill, J.-C., & Feng, C. (2012). Land pricing, urban expansion and urban
land use efficiency in post land reform Beijing. The Real Estate Academic Initiative,Harvard University. Working paper.
dward, G. L., & Kahn, M. E. (2003). Sprawl and urban growth. In J. V. Henderson,& J. Thisse (Eds.), Handbook of urban economics (pp. 2481–2527). Amsterdam:Elsevier.
uropean Environment Agency. (2010). In A. Jones, M. Erhard, C. Iversen, A.Lükewille, & J.-E. Petersen (Eds.), The European environment: State and outlook2010 (soil) Luxembourg, (p. 14).
ang, X., Zhang, W., Zhang, L., Luo, H., & Li, Z. (2002). Jin Bai Nian Lai Beijing Cheng ShiKong Jian Kuo Zhan Yu Cheng Xiang Guo Du Dai Yan Bian [The urban expansionand the evolution of urban fringe in Beijing in the 20th century]. City PlanningReview, 26(4), 56–60.
u, C. (1999). Beijing Tu Di Li Yong Fu Bei Bian Hua Ji Zhi Yan Jiu [Study on phenomenaand mechanism of land use/cover change in Beijing]. Journal of Natural Resources,14(4), 307–312.
e, C., Shi, P., Chen, J., & Xu, X. (2002). Beijing Di Qu Cheng Shi Hua Guo ChengYu Ji Zhi Yan Jiu [Process and mechanism of urbanization in Beijing area]. ActaGeographica Sinica, 57(3), 363–371.
iang, S., Liu, S., & Li, Q. (2007). Tu Di Zhi Du Gai Ge Yu Guo Min Jing Ji Cheng Zhang[Land reform and economy development]. Management World, 9, 1–9.
uang, W., Liu, J., Shao, Q., & Sun, C. (2009). Nian Yi Lai Beijing Zhu Cheng Qu Tu Di LiYong Kong Jian Kuo Zhang Te Zheng Yu Ji Zhi Fen Xi [Spatio-temporal patternsand driving forces of urban expansion in Beijing central city since 1932]. Journalof Geo-Information Science, 11(4), 428–435.
and Use Services Division and Advance Planning Division. (2007). Final environmen-tal impact report and appendices for county of San Bernardino 2007 general planprogram. http://www.sbcounty.gov/Uploads/lus/GeneralPlan/FinalEIR2007.pdf
i, L. (1999). Urban land reform in China. New York: St. Martin’s Press, Inc.ichtenberg, E., & Ding, C. (2009). Local officials as land developers: Urban spatial
expansion in China. Journal of Urban Economics, 66(1), 57–64.iu, S., & Jiang, S. (2005). Tu Di Rong Zi Yu Cai Zheng He Jin Rong Feng Xian [Financial
risk of land financing by local governments – Case study of a developed area inEast China]. China Land Science, 19(5), 3–9.
u, D. (2007). Wo Guo De Cheng Zhen Hua Jin Cheng Yu Kong Jian Kuo Zhang[Urbanization process and spatial sprawl in China]. Urban Planning Forum, 4,47–52.
4.4 85.2 0.3 2.20.6 1.3 60.1 34.60.4 0.4 0.1 93.7
Mullerand, M. R., & Middleton, J. (1994). A Markov model of land use change dynam-ics in the Niagara Region, Ontario, Canada. Landscape Ecology, 9(2), 151–157.
Piao, Y., & Ma, K. (2006). Beijing Cheng Shi Jian Cheng Qu Kuo Zhang De Jing JiQu Dong [Economic driving force of urban built-up area expansion in Beijing].Natural Resource Economics of China, 7, 34–37.
Plantinga, A. J. (1996). The effect of agricultural policies on land use and environ-mental quality. American Journal of Agricultural Economics, 78(4), 1082–1091.
Plantinga, A. J., & Irwin, E. G. (2006). Overview of empirical methods. In K. P. Bell,K. J. Boyle, & J. Rubin (Eds.), Economics of rural land-use change. Burlington, USA:Ashgate.
Segerson, K., Plantinga, A. J., & Irwin, E. G. (2006). Theoretical background. In K. P.Bell, K. J. Boyle, & J. Rubin (Eds.), Economics of rural land-use change. Burlington,USA: Ashgate.
Serneels, S., & Lambin, E. F. (2001). Proximate causes of land use change in Narok Dis-trict, Kenya: A spatial statistical model. Agriculture, Ecosystems and Environment,85, 65–81.
Serra, P., Pons, X., & Sauri, D. (2008). Land cover and land use change in a Mediter-ranean landscape: A spatial analysis of driving forces integrating biophysicaland human factors. Applied Geography, 28, 189–209.
Tao, R., Su, F., Liu, M., & Cao, G. (2010). Land leasing and local public financein China’s regional development: Evidence from prefecture-level cities. UrbanStudies, 47(10), 2217–2236.
The California Department of Conservation. (2011). Historic land use conversion1984–2010. http://redirect.conservation.ca.gov/dlrp/fmmp/product page.asp
U.S. Embassy. (1997). China’s farmland loss rings alarm-satellite photographs reveal aserious problem.
Weng, Q. (2002). Land use change analysis in the Zhujiang Delta of China usingsatellite remote sensing, GIS and stochastic modeling. Journal of EnvironmentalManagement, 64, 273–284.
Wheaton, W. C. (1974). A comparative static analysis of urban spatial structure.Journal of Economic Theory, 9, 223–237.
William, M. B., & Turner, B. L. (1992). Human population growth and global land-use/cover change. Annual Review of Ecology and Systematics, 23, 39–61.
Wooldridge, J. M. (2012). Introductory econometrics: A modern approach. Mason, OH,USA: South-Western, Cengage Learning.
Wu, F. (1997). Urban restructuring in China’s emerging market economy: Towardsa framework for analysis. International Journal of Urban and Regional Research,21(4), 640–663.
Wu, F., & Yeh, A. G.-O. (1999). Urban spatial structure in a transitional economy:The case of Guangzhou, China. Journal of American Planning Association, 65(4),377–394.
Yang, S., & Wen, T. (2010). Jing Ji Bo Dong, Cai Shui Ti Zhi Bian Qian Yu Tu Di Zi YuanZi Ben Hua [The economic fluctuations the change in taxation institution, andthe capitalization of land resources]. Management World, 4, 32–41.
Yeh, A. G. (2005). Dual land market and internal spatial structure of Chinese cities. InL. J. C. Ma, & F. Wu (Eds.), Restructuring the Chinese city: Changing society, economyand space (pp. 52–70). Routledge: Taylor & Francis Group.
Yu, M. (2010). (Idleness of land become the most severe problem in land use) Tu Di XianZhi Cheng Dang Qian Zui Da Wen Ti. Beijing: People’s Daily Online.
Zhang, T. (2000). Land market forces and government’s role in sprawl: The case ofChina. Cities, 17(2), 123–135.
Zhang, W. (2007). (Exploratory research on land use policy’s role in macroeconomiccontrol) Tu Di Zheng Ce Can Yu Hong Guan Tiao Kong Tan Suo. Beijing: GeologyPress.
Zhou, F. (2007). Sheng Cai You Dao: Tu Di Kai Fa He Zhuan Rang Zhong De ZhengFu He Nong Min [The role of government and farmers in land development andtransfer]. Sociological Studies, 22(1), 49–82.
Zhou, Q. (2004). Nong Di Chan Quan Yu Zheng Di Zhi Du [Property rights and landrequisition system: A critical choice for China’s urbanization]. China EconomicQuarterly, 4(1), 193–210.
Zhou, Q. (2011). (Urbanization, land policy and macroeconomic control) Cheng Shi Hua,Tu Di Zhi Du Yu Hong Guan Tiao Kong. The Economic Observer.
Zhou, Y. (2006). Tu Di Shi Kong Shui Zhi Guo [Who responsible for the loss controlof land use?]. City Planning Review, 30(11), 65–66.
Zou, J. (1994). Shen Hua Gai Ge, Jia Qiang Guan Li, Cu Jin Guo Min Jing Ji Chi Xu KuaiSu Jian Kang Fa Zhan [Intensify reform strengthen management and establish ahealthy and speedy economic growth]. China Land, 11, 4–9.
