Vous êtes sur la page 1sur 11

GeoJournal 49: 257–267, 1999.

© 2000 Kluwer Academic Publishers. Printed in the Netherlands.


257

Diminishing farmland and urban development in China: 1993–1996

Sun Sheng Han & Chun Xing He


School of Building and Real Estate, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260,
e-mail: bemhanss@nus.edu.sg; e-mail: bemp7004@nus.edu.sg

Received 17 August 1999; accepted in revised form 26 February 2000

Key words: China, farmland loss, land policy, regional development, urbanization

Abstract
Farmland conservation is a new perspective in China’s development policy making. In this paper, we seek to explore
the distribution patterns of farmland loss between regions and among city-size groups, and to examine the factors and
institutional background that influence on farmland conservation. A random sample of 179 out of 666 cities in 1996 was
selected. Data were collected from various sources for the period 1993–96. Descriptive statistics, correlation analysis and
nonparametric tests were used to analyze the data. Three-dimensional maps were plotted to present the distribution patterns
of farmland loss, urban population growth and real estate development. The empirical results reveal that regional and
hierarchical variations did exist in farmland loss, so did in urban population growth and real estate development. Farmland
loss was more severe a problem in cities along the coast, especially cities with a nonagricultural population over two
million. Urban population growth and real estate development were two main factors causing farmland loss in cities. Poor
legal ground for land administration, weak law enforcement, and the drive towards fast revenue making were the underlying
problems that prevented effective farmland conservation. In China’s current transition from a planned to a market economy,
farmland conservation can succeed only if government officials were no longer to trade off sustainability of development
with short-term, fast revenue making.

Introduction eas are excluded. On average, every Chinese had 1.17 mu1
arable land in 1996, which was far below the world average
A drastic change in China’s development is the addition of of 3.75 mu per person (Zhang 1998). Even worse, China’s
a new policy perspective towards land resources. Tradition- arable land has diminished at an alarming rate since 1949.
ally, the Chinese leadership took China as dida wubo renkou Between 1949 and 1989, there was an average annual loss of
zhongduo (i.e., a huge territory with variety of resources 10 million mu arable land, or, about 12 mu per 1000 people.
and with a large population), which meant a strong point In 1992, the loss of arable land registered 20 million mu,
of the country. This was true especially under the centrally or, 17 mu per 1000 people (China Real Estate Information,
planned economic system in which the collective strength 1997, p. 52).
of the economy as measured by sizes of land and popu- Land conservation, thus, has been put on the priority
lation meant more resources for manoeuvre by the central list of the government agenda. Laws and regulations were
government, driven by the development goal of equity. To- issued to provide legal ground for land use planning, trans-
day, however, China is pursuing economic efficiency by action, and administration. The date of 25 June was set to
introducing economic reforms, which, together with eco- be a ‘Land Day’ in order to raise public awareness of land
nomic globalization, have placed more weight on individual conservation. High-ranking government officials, including
contribution and initiatives. Not only absolute volume of the Prime Minister, addressed the importance and policy
production but also per capita input and output have be- issues in resolving land problems. The speech delivered
come important measures of economic performance. The by former Prime Minister Li Peng, on the sixth Land Day
incorporation of the population factor, thus, changed the in 1996, which also celebrated the tenth year of the Land
government’s perception on China’s land resources. Now, Administration Act, was an important statement to show
China is perceived as dishao wuxi (i.e., tiny land with scarce the commitment of the central government towards land
resources). Land scarcity is seen as a constraint to national conservation (China Real Estate Information, 1996, p. 4).
development, which calls on better policies to deal with the Studies on the diminishing farmland have accumulated
limited resources (Johnson, 1997). quickly for the last five years. Land scarcity, as measured
Indeed, China occupies about 7% of the world land area by the average land area per person, is a problem widely
but supports almost a quarter of the world population. The recognized by researchers (Xu, 1999; Jiang, 1998; Liu,
vast territory of China becomes tiny when it is divided by 1998). This scarcity is even more severe when the distrib-
the huge population size, more so if the nonarable land ar-
258

ution of land is taken into account, as more than 60% of large, 19 large, 67 medium, and 72 small cities, representing
the arable land is in the area north of the Huaihe River, 82%, 52%, 43%, 34%, 18% of the populations, respectively.
where water shortage presents a further constraint on the The period for study focuses on a three-year span, from
productivity of the land (Zhang, 1998). Deforestation and 1993 to 1996, though discussions are extended to the period
desert encroachment are two of the reasons for farmland 1978–1998. The broad coverage of the period for discussion
loss (Chen X, 1996). Construction in both cities and coun- from 1978–1998 is to provide a context for the study in the
tryside is another source encroaching farmland (Lu, 1997; three year period. The year 1992 witnessed Deng Xiaop-
Ru, 1995; Liu, 1995). In many cities, such as Beijing and ing’s Southern Tour, during which he re-confirmed China’s
Dongguan, Guangdong Province, arable land was acquired commitment to the reform and open policy. In 1993, eco-
by developers but was kept vacant for a long time (Chen YQ, nomic activities resumed to a normal level after the cool-off
1996; Ye and Li, 1999). Real estate speculation, which is a induced by the 1989 student demonstrations and the conse-
new phenomenon in China, is a cause for fast reduction of quent crackdown. The year 1996 is selected because this is
farmland. the latest year for which data can be gathered for this study.
Despite the large number of articles published on the Data are collected from various city almanacs, China
subject, there is little systematic study on the diminishing Statistical Yearbooks, China’s Urban Statistical Yearbooks,
farmland on a national scale. What spatial pattern can one China Real Estate Yearbooks, China Land Price, collections
discern in farmland loss between cities along the coast and of land laws and regulations, and journal articles in Chi-
noncoastal regions? Is there any difference among various nese language. Farmland loss is the difference between the
city size groups in farmland loss? How does urban devel- amount of farmland in 1993 and that in 1996. The growth
opment influence on the size of arable land? What is the of nonagricultural population is measured by the difference
institutional background relevant to farmland conservation? between the nonagricultural population in 1996 and that in
1993. Real estate development is represented by the price
level of first grade commercial land parcels. Parametric and
Objectives of study non-parametric statistics, as well as ArcView GIS, are used
in data analysis and presentation.
In this paper, we seek to answer the above questions with
three objectives in mind:
(1) to explore the distribution pattern of farmland loss in Patterns of urbanisation and farmland loss
terms of location of cities and the size distribution of
cities; Urban population growth
(2) to examine the relationship between farmland loss and
urbanization; and Since 1978, Chinese urbanization has proceeded at a phe-
(3) to investigate the institutional background relevant to nomenal rate. The number of cities increased from a mere
farmland loss. 193 cities in 1978 to 668 in 1998, registering a growth of
The location of cities is grouped into coastal and noncoastal 246% in 20 years. The size of nonagricultural population,
regions. Coastal cities are those cities found in the coastal which is often used to represent the size of urban popu-
provinces (viz., Liaoning, Hebei, Beijing, Tianjin, Shan- lation, grew from 80 million to 200 million in the same
dong, Jiangsu, Zhejiang, Fujian, Guangdong, Guangxi, and period, recording a 150% increase. The level of urbaniza-
Hainan). Noncoastal cities are those located in the rest of the tion, measured by the proportion of urban population out of
provinces, which are sometimes further divided into western the total population, grew from 18% to 30% in the period
and central regions.2 The size categorization of cities follows 1978–1998.5
the official five statistical groups.3 Urbanization is discussed The distribution of cities has changed drastically in the
in terms of nonagricultural population growth and real estate last two decades. In 1978, 36% of the cities were in the
development. The institutional background relevant to farm- coastal region, whilst the majority, about 64% of the cities
land loss refers to the legal ground of land use control, and were in the noncoastal provinces. By 1998, 300 cities,
the official and nonofficial channels through which farmland or about 45% out of the 668 cities, were found in the
is converted into urban use. coastal provinces. The increase of cities in the coastal region
reflected the influence from the unbalanced regional devel-
opment policy, which granted promotional measures and
Data and methods generated favorable environment for rapid economic growth
along the coast (Fan, 1995; Hayter and Han, 1998).
This study is based on a random sample of 179 Chinese The summary statistics of the sample cities in non-
cities, out of the 666 national total, representing 27% of the agricultural population growth confirm the uneven distribu-
population in 1996.4 The sample cities spread over 28 of tion of urban growth (Table 2). On a national scale, analysis
the 31 provincial administrative units under the communist of the sample cities reports changes in nonagricultural popu-
government (Figure 1). In the coastal region, the sample lation a range of 370,000 people, with the highest of 310,000
represents 34% of the population, whilst in the noncoastal increase and the lowest of almost 60,000 decrease. Nega-
region, 21% of the population is represented (Table 1). In tive growth in nonagricultural population was attributable to
terms of size groups, there are 9 extra-super large, 12 super the relocation of state enterprises in the resource extraction
259

Figure 1. Distribution of the 179 sample cities.

Table 1. Distribution of sample cities among the regions and city-size groups

City size group Coastal region Noncoastal region All China


Total Sample Total Sample Total Sample

Extra-super large cities 6 6 5 3 11 9


Super large cities 11 8 12 4 23 12
Large cities 21 11 23 8 44 19
Medium cities 90 36 105 31 195 67
Small cities 170 40 223 32 393 72
Sum 298 101 368 78 666 179

Note: In the column Total are the numbers of cities in 1996.

sector, and to the re-definition of city boundaries. Nat- Hierarchical distribution of cities among various city-
ural increase and migration to the cities brought in positive size groups represents another facet of Chinese urbanization.
growth. For years, China applied an urban policy ‘to contain the
Table 2 also presents the distribution of non-agricultural development of large cities, to grow medium-sized cities rea-
population growth among the sample cities in the coastal and sonably, and to encourage the development of small cities,’
the noncoastal regions. The coastal region accommodated with the long-term goal of rural-urban integration (Han and
cities with both the highest and the lowest population in- Wong, 1994). Small cities grew fast especially in terms of
creases. According to the value of coefficient of variance, the number, as more and more large towns and county seats
coastal cities were more diverse than noncoastal cities in were converted to cities. In terms of population, however,
nonagricultural population change. Although the mean value the small cities were a losing group in attracting migrants.
of nonagricultural population increase of coastal cities was Evidence from the sample cities show that extra-super large
higher than that of noncoastal cities, the Mann-Whitney Test cities, with a nonagricultural population of two million and
yielded no significant difference between the coastal and above, had a mean increase of 168,000 people over the three
noncoastal regions.6Indeed, as we can discern from Figure 2, years from 1993 to 1996. The mean values of nonagricul-
cities in the noncoastal provinces but close to the coastal tural population increase decreased over the city-size groups.
region also demonstrated huge increases in nonagricultural Small cities only registered a mean increase of 13,000 people
population. (Table 3).
260
Table 2. Summary statistics of the sample cities in urban population growth, 1993–1996

N Minimum Maximum Range Mean Std. Dev. c.v.

Nation-wide 179 −5.99 31.40 37.39 4.35 5.14 1.18


Coastal 101 −5.99 31.40 37.39 4.94 5.93 1.20
Noncoastal 78 −1.25 21.56 22.81 3.59 3.79 1.06

Notes: Urban population growth is measured by the difference between the size of nonagri-
cultural population in 1996 and that in 1993.
c.v. stands for coefficient of variance.

Figure 2. Regional distribution of non-agricultural population growth in the sample cities (in 10,000 people)

Further examination of Table 3 reveals that the lower Figure 3 shows a three dimensional map of the first grade
level the city-size group, the more diverse of population commercial land prices of the sample cities. Apparently, the
change. Small cities demonstrated a 1.38 coefficient of vari- coastal region accommodated all the high commercial land
ance, medium sized cities 0.83, whilst the larger groups were prices. Shanghai, Beijing, Tianjin and Wenzhou were among
below 0.55. The nonuniform growth of nonagricultural pop- the highest. Evidently, cities formed groups as we discern in
ulation was confirmed by the Kruskal–Wallis Test, which Figure 3, in areas of the Beijing-Tianjin area, the Yangtze
yielded a highly significant χ 2 value.7 River Delta, and the Pearl River Delta.
Summary statistics of commercial land prices by region
Real estate development show that the coastal cities rendered a mean value of com-
mercial land price almost three times as high as that rendered
Rapid increase of the number of cities and urban popula- by noncoastal cities (Table 4). However, the values of coeffi-
tion have led to the expansion of urban areas. Between 1985 cient of variance reveal more variations among the coastal
and 1998, the built-up areas of cities expanded from about cities than that among the noncoastal cities. The coeffi-
9000 km2 to 21,000 km2 . The expansion of urban built-up cient of variance was 1.37 for coastal cities, while for the
areas was particularly fuelled by real estate development. In noncoastal cities, it was only 0.76.
the 1990s, most of the Chinese cities introduced the concept Table 5 shows commercial land price variations among
of land-user right and land-use fees.8 Housing reform also the five city-size categories. By the mean value, it is clear
spurred the commercial housing market. An understanding that larger cities had higher commercial land prices. The
of the pattern of real estate development among Chinese extra-large group was especially outstanding, as its mean
cities is an indispensable part to understand farmland loss. value was two times higher than that of the super large group,
Studies on China’s regional property markets reveal wide and was almost six times higher than that in small cities.
variations between the coastal and the noncoastal regions However, commercial land prices were more varied among
(Han, 1998). Using the total volume of property transaction the extra-large cities than that in other city-size groups.
as a yardstick, Han (1998) found that the coastal provinces Commercial land price in Shanghai showed the highest, with
developed a much more active property market than the a value of 35,192 RMB yuan per m2 (about $4240 USD per
noncoastal provinces did. An active coastal property market m2 ).
was largely attributable to the regional development policy,
which allowed some selected coastal cities and provinces to
introduce market forces ahead of other provinces.
261

Table 3. Summary statistics of sample cities in population growth, by city-size group,


1993–1996

City size group N Minimum Maximum Range Mean Std. Dev. c.v.

Extra-super large 9 6.28 31.40 25.12 16.85 8.93 0.53


Super large 12 2.64 14.68 12.04 9.70 4.01 0.41
Large 19 1.40 15.56 14.16 6.07 3.95 0.65
Medium 67 0.40 22.75 22.35 4.46 3.69 0.83
Small 72 −5.99 10.12 16.11 1.34 1.85 1.38

Notes: Urban population growth is measured by the difference between the size of nonagricul-
tural population in 1996 and that in 1993.
c.v. stands for coefficient of variance.

Figure 3. Regional distribution of commercial land prices of the sample cities (RMB yuan/sq.m).

Table 4. Summary statistics of commercial land prices, by region, 1996

N Minimum Maximum Range Mean Std. Dev. c.v.

Nation-wide 176 126 35,192 35,066 2,055 3,033 1.48


Coastal 98 500 35,192 34,692 2,812 3,839 1.37
Noncoastal 78 126 4,000 3,874 1,105 842 0.76

Notes: Land price is in RMB yuan/sq. m.. $1 US dollar was about 8.28 RMB yuan as in
October 1999.
c.v. stands for coefficient of variance.

Table 5. Summary statistics of commercial land prices, by city-size group, 1996

City size group N Minimum Maximum Range Mean Std. Dev. c.v.

Extra-super large 9 1,014 35,192 34,278 8,706 10,817 1.24


Super large 12 975 5,250 4,274 2,800 1,330 0.48
Large 19 330 5,500 5,170 2,330 1,399 0.60
Medium 67 126 8,226 8,100 1,761 1,312 0.75
Small 69 140 4,035 3,895 1,269 869 0.68

Notes: Land price is in RMB yuan/sq. m.. $1 US dollar was about 8.28 RMB yuan as in October
1999.
c.v. stands for coefficient of variance.
262

Farmland loss there were wild variations in farmland loss in medium cities,
as represented by the high coefficient of variance.
Farmland loss in the period 1993–1996 was very much
influenced by the drastic changes in the establishment of
economic development zones in 1992, when farmland loss Influential factors
reached a phenomenal peak.9 A total of 20 million mu of
arable land were converted to nonfarm use. The reason be- Correlation analysis
hind this large-scale farmland loss was the overheated trend
towards setting up economic development zones. According On a national scale, the correlation coefficients between
to statistics from the Land Administration Bureau, there was farmland loss and real estate development, and between
a total of 117 economic development zones in 1991. After farmland loss and urban population growth were both sig-
the Southern Tour of Deng Xiaoping in the Spring of 1992, nificant at a high confidence level.12 These results sug-
however, the total number of economic development zones gest that farmland loss was closely related with real estate
increased to 2700 by the end of 1992. Tough measures to development and the growth of nonagricultural population.
eliminate most of the economic development zones were The same was true among coastal cities. Both non-
introduced in 1993, leading to a drastic reduction of the agricultural population growth and real estate development
development zones to about 500 nation-wide. Farmland loss was significantly correlated with farmland loss. For the non-
was subsequently controlled. In 1994, about 6 million mu coastal cities, however, no significant correlation can be
of arable land was converted to urban use, whilst in 1997, detected between either farmland loss and real estate de-
farmland loss was contained to about 2 million mu. velopment, or farmland loss and urban population growth.
Figure 4 shows farmland loss in the sample cities across The lack of close relationship between farmland loss and
the nation. Apparently, coastal cities, especially the larger urban development in noncoastal areas was attributable to
ones such as Beijing and Shanghai, lost more farmland. In the conversion of farmland to other agricultural use, such as
Beijing, where farmland loss was a severe problem, a total of housing for the villagers, or developing fruit farms.13 The
200 km2 arable land was encroached by urban development. latter did occur in Heilongjiang province in 1996 (Yearbook
The same amount of arable land was encroached in Heyuan, of the Land of China, 1997).
a small city with 140,000 people in Guangdong Province. In In order to analyze the hierarchical characteristics of
Zhuhai, about 150 km2 of arable land was lost in the period farmland loss, the sample cities are conveniently grouped
of 1993–1996, whilst in Shanghai, 80 km2 arable land was into big and small groups, which include the extra-super
converted to urban use (Figure 4).10 large, super large and large cities in the former, and medium
There were a few cities that showed increase in farm- and small cities in the latter. Farmland loss was signifi-
land. These cities included Jiamusi, Heilongjiang Province, cantly correlated to real estate development in both big and
Xuzhou, Jiangsu Province, Chongqing, Akesu, Xinjiang Au- small groups. However, urban population growth showed no
tonomous Region, among others. There are two reasons for significant correlation with farmland loss in the same two
urban farmland increases. First, farmland increase in such groups.
cities as Jiamusi and Xuzhou was the result of new develop-
ment of uncultivated land, which exceeded farmland loss by The impact of urban population growth
urban encroachment (Statistical Yearbook of Xuzhou, 1998;
Economic Statistical Yearbook of Jiamusi, 1997). Second, Urban population growth influenced farmland loss because
farmland increase in Chongqing was caused by changes in larger population size led to the expansion of the built-up
the administrative boundary. In 1996, i.e., one year before areas in providing more housing and employment opportu-
Chongqing was designated the fourth municipality directly nities (Li and Zhang, 1998). In Beijing, urban construction
under the Central Government in 1997, five urban districts, was responsible for only 37% of farmland loss in the period
one county level city and 16 counties were added to its ad- 1993–1995. The remaining 63% of farmland loss was at-
ministrative area. Chongqing’s land territory increased from tributable to conversion to other agricultural uses, including
23,113.95 km2 in 1995 to 82,403.00 km2 in 1996.11 fruit farms, and also to natural losses caused by sandstorm
Table 6 shows the summary statistics of farmland loss by and flooding (Chen X, 1996; Wang and Zhang, 1996).
region. The mean value of farmland loss in cities along the Housing construction was one of the main activities that
coast was a positive 1.11, whilst the mean value for non- occupy farmland on the fringe areas of cities. Again, in
coastal cities was negative 1. These mean values indicated Beijing, housing construction has quickly expanded to the
that farmland loss was a problem of coastal cities, less so to fringe areas beyond the fifth ring road for the last five or
noncoastal cities. Indeed, the overheated real estate develop- ten years. In addition, migrants settled in the fringe areas,
ment was observed in coastal cities, especially the southern further pushed up the conversion rate of farmland towards
coastal cities, rather than in noncoastal areas. non-farm use. Due to the high price level in the city, which
The mean values of farmland loss, as tallied by city-size can go as high as 8000 RMB yuan per m2 for the building
groups, showed that the extra-super large cities were the area of an apartment, the municipal government is seriously
cities where large amount of farmland disappeared (Table 7). looking at the far suburbs for housing the low and middle in-
In the groups of medium, large and super large cities, there come residents. This may lead to more construction projects
were average increases in the size of farmland. However, in the far suburbs.14
263

Figure 4. Regional distribution of farmland loss of the sample cities (in 1,000 ha).

Table 6. Summary statistics of farmland loss, by region, 1993-1996 (in 1000 ha)

N Minimum Maximum Range Mean Std. Dev. c.v.

Nation-wide 179 −33 20 53 0.19 5.06 26.63


Coastal 101 −29 20 49 1.11 4.81 4.33
Noncoastal 78 −33 5 38 −1 5.17 5.17

Notes:
Farmland loss is measured by the difference between the size of farmland in 1996 and that in
1993.
c.v stands for coefficient of variance.

Population pressure called for rapid economic devel- proportion of land that were transacted under free market
opment. City governments were keen to attract foreign prices, an active land market presented only in the first tier
investors by setting up numerous development parks in the property market, i.e., the transfer of land from government
urban fringe.15 In Shanghai, development parks were estab- to users with user-rights attached.
lished by not only the municipal government, but also by As land could be acquired at a price far below free market
district governments and county governments. Even town- price, through negotiation with the government, it became
ships had their own development parks. On a national scale, a shortcut towards fast revenue making for those who had
development parks occupied 15,000 km2 in 1992 (Tian, good connections with government officials. Thus, illegal
1994). A great portion of these development parks was set approval of farmland conversion and illegal occupation of
up by local governments without the required approvals. In farmland was common among provinces and cities all over
1993, 78% of 2804 economic development parks, which the country (Yang, 1993; Tian, 1994; Lu, 1997). Accord-
were under investigation, were set up illegally. These de- ing to the China State Land Administration Bureau (1997),
velopment parks occupied 11.43 million mu (or 7620 km2 ) the total registered cases of land use without approval in
of land, which were virtually farmland (Yang, 1996). 1996 were 240,111, in which 296.23 km2 of land, includ-
ing 118.49 km2 of farmland, was involved. 154,850 of these
The impact of real estate development cases, which involved 186.40 km2 including 89.02 km2 of
farmland, were unauthorized occupation. 8,587 cases in-
China’s real estate market is still in its infant stage. Most cluded unauthorized approval, i.e., the approving authority
of the transactions of land were being made under arranged did not have the power and should not have made such an
prices rather than free market prices.16 Further to the small approval. Illegal land approval and occupation were found
264
Table 7. Summary statistics of farmland loss, by city-size group, 1993–1996 (in 1000 ha)

City size group N Minimum Maximum Range Mean Std. Dev. c.v.

Extra-super large 9 0. 20 20 4.44 6.27 1.41


Super large 12 −29 3 32 −2.42 8.81 3.64
Large 19 −33 3 36 −1.58 7.73 4.89
Medium 67 −18 15 33 −0.03 4.23 35.25
Small 72 −13 20 33 0.76 3.38 4.45

Notes: Farmland loss is measured by the difference between the size of farmland in 1996 and that
in 1993.
c.v stands for coefficient of variance.

Figure 5. The channels of transfer for land collectively-owned from farmland to construction sites.

in all provinces with a minimum of 3 cases in Tibet and a (or, 1160 km2 ) of converted land parcels left idle after being
maximum of 42,962 cases in Shaanxi Province. acquired (Real Estate Periodical, 1998 (31), p. 19).
Another problem in real estate development that had a di-
rect bearing on farmland loss was ineffective use of the land
converted. A survey of 55 cities in the mid-1990s revealed Institutional background relevant to farmland
that on average, industrial land areas accounted for 27.5% conservation
of the built-up areas. This is much higher than the plan-
ning standard of 15% industrial land use currently applied in The land law systems
many cities around the world (Real Estate Periodical, 1996
(26), p. 13). Even land was tagged as industrial development Laws and regulations for transfer of collectively owned land,
zones or residential development districts, a large propor- including farmland were vague. According to ‘The Land Ad-
tion of the parcels were left vacant for a long time period. ministration Act in PRC,’ which was first endorsed in 1986
Idle construction land accounted for 20–30% of the total ur- and subsequently revised in 1988, urban land was owned
ban construction land in 1996 (Real Estate Periodical, 1996 by the state. Collectives were owners of land in suburbs of
(23), p. 15). According to Mr. Zhou Yongkang, Minister cities and in the countryside, unless for those parcels spec-
of Land and Natural Resources, 40% of urban construction ified otherwise. User-rights of state-owned and collectively
land was utilized ineffectively. There were 1.74 million mu owned land could be transferred according to the Act. User-
265

rights of collectively owned land can be used as shares for areas can only be converted to nonagricultural construction
investment in collective enterprises. With the approval from land after state requisition. The state acquires farmland from
the authorities at county level or above, enterprises owned collectives based on a standard requisition fee-schedule for
by townships or by village councils could convert farmland compensation. The state then releases the land to developers
into nonfarm use. or end-users at a price that can be accepted by both parties.
Based on the above, ‘The Temporary Regulations for This is the only legal way to convert farmland owned by
User-Rights Transfer of Urban State-Owned Land in PRC’ collectives to nonfarm use.
(1990) and ‘Urban Property Administration Act in PRC’ In practice, however, there were several more or less
(1994) outlined the rules for land use and transaction in ur- illegal channels through which farmland owned by collec-
ban area. The Urban Property Administration Act made it tives was converted to nonagricultural construction land in
clear that ‘collectively owned land in urban planning areas suburban and rural areas (Figure 5). First, some towns and
should be acquired by the state before being released for villages used their farmland as shares to set up joint venture
development.’ However, there are no detailed regulations for enterprises. Their partners could be state-owned enterprises
user-rights transfer of collectively owned land in rural areas. or foreign investors. In return the towns and villages received
Without a clearly defined legal bound, collectively dividends from the joint venture enterprises. Second, since
owned land parcels entered the land market for real estate rural enterprises and farmland were both collectively owned
development directly. Planning and control over farmland by towns or villages, some farmland parcels were directly
conversion into urban land were out of the question, as it converted for the construction of rural industries. The town
was vague whether the state had to acquire the parcels before or village was better off by so doing, as rural industries
they were released to the market. Thus, collectively owned yielded higher return to the collective than the farmland
land parcels entered the land market through a variety of could do. Third, in some towns and villages, the people in
channels, being investment shares in collective enterprises, charge simply ignored the laws and regulations. They sold or
production sites of rural enterprises, and parcels for sale or leased farmland to real estate developers for a quick return.
for lease. In fact, collectives in towns and villages were big In such transactions, the town or village obtained cash from
suppliers of land parcels only second to the state. It was very the sale or lease.
difficult to differentiate legal from illegal transactions on the Profit maximization was the root cause for illegal conver-
basis of the current laws and regulations. As a result, state sion of farmland owned by collectives in suburban and rural
control over farmland conversion was severely undermined. area. As shown in Figure 5, collective economic units can
Direct provision of collectively owned land to the real estate receive only compensation premium from the state, when
market caused loss of government revenue from land and their farmland is acquired by the state. By eliminating state
distorted land allocation of market mechanism. involvement, collectives can get revenue in the forms of
dividends, profits or sales and rents, which are significantly
Land use control in rural areas greater than the compensation premium from the state. The
state, however, lost not only revenue but also control over
Illegal occupation of farmland for nonfarm use was intensi- farmland conversion by this exclusion.
fied by weak control and the lack of supervision on land use
in suburban and rural areas. Land administration authorities
did not have the means to stop illegal land use, reflecting Summary and conclusions
the weak legal system and poor law enforcement practice
in China today. In some cases land administration author- Land scarcity has become a new policy perspective to Chi-
ities violated land laws intentionally, driven by the desire nese leaders in making up policies for national development.
to reap profit from land sales for the local community and In the 1990s, attention was especially given to the conserva-
in many cases for the officials involved. They set up by- tion of farmland, as rapid economic growth led fast devel-
laws to encourage illegal transfer of collectively owned land opment of cities, which encroached considerable amount of
to construction land (Qian, 1996). The weak control over farmland in almost all provinces across the country.
farmland conversion was largely a reflection of the poor legal The purpose of this paper is to examine the spatial
system and law enforcement measures nation-wide. China dimension of urbanization and farmland loss, and to ex-
has been transforming herself from ren zhi (government by plore the factors and institutional background that influence
man) to fa zhi (government by law) for the last 20 years. A on farmland conservation. By using random sampling and
nation-wide legal framework is only evolving recently. Fur- statistical techniques, findings of this paper confirm that
ther, economic and political liberalization have motivated farmland loss has been a problem nation-wide. However,
local officials to seek rapid economic growth of their ar- coastal cities registered more cases of farmland loss than
eas/regions, at the expense of national interest and interest noncoastal cities. The three-dimensional map shows that
of other areas/regions. mega cities along the coast, as well as the cities surround-
ing those mega cities, were the cities where farmland loss
Channels of farmland loss was severe. Further to the regional variations, we detect
that cities with a nonagricultural population of two million
In accordance with the current land laws and regulations in and over were definitely in heavy demand of farmland for
China, farmland owned by collectives within the planning urban expansion. Cities below the two million mark were
266

very diversified in farmland encroachment. Indeed, quite a standards. Criteria in evaluation for granting city status include the number
few cities even gained more farmland through cultivation of of nonagricultural population, the amount of GDP per annum, and the per-
centage of population working in the agricultural sector. For more details,
nonarable land, not mentioning those having increased their see footnote number 2 in Han and Yan (1999).
farmland stock through boundary adjustment. 5 Based on data from China Statistical Summary (1999, p. 31). The calcula-
Regional concentration of great farmland loss along the tions of Chinese urbanization level are always confusing. For a discussion
coast was closely related to urban population growth and real on the various population categories in China, see Ma and Cui (1987), Chan
and Xu (1985).
estate development. Not only the three-dimensional maps 6 The Mann–Whitney U of 3497.5 was computed, with a z value of −1.28.
show similar patterns among farmland loss, urban popula- 7 The χ 2 value was 98.42, which was significant at 0.01 level of signifi-
tion growth and commercial land prices between regions, the cance.
8 With the land user right and fee schemes, land is no longer used free of
correlation coefficients also demonstrate close ties. Large
charge. Rather, tenants have to pay for their use of a piece of land.
population size of cities demanded expansion of the built- 9 Economic development zones refer to jinji kaifa qu. The essence of these
up areas by housing construction and industrial growth. The economic development zones was to set up enclaves where preferential
overheated real estate market in the early 1990s led to a large policies and tax breaks could be offered to investors, both domestic and
amount of conversion of farmland into nonagricultural use. foreign, in the hope of promoting local development.
10 For the purpose of this paper, absolute values of farmland loss in 1000 ha,
Yet in most of the cities, utilization rates of the converted rather than per capita values, were used in discussions.
farmland were low. 11 After adjustment for the change of administrative area, total agricultural

The weak legal system of land laws and regulations, land in urban and rural areas still decreased by 39.53 km2 in 1995 and
70.50 km2 in 1996 (Statistic Yearbook of Chongqing, 1997).
and poor law enforcement explain the root cause of farm- 12 The table below presents the correlation coefficients of farmland loss with
land loss in China. Procedures for conversion of farmland real estate development and urban population growth, grouped by region
into nonagricultural land-use are defined vaguely in nonur- and city-size categories. Only the positive values of farmland loss (i.e.,
ban planning areas. Even for the areas falling in the urban GD3_6) are included in computation in order to limit the distortions that
farmland increase may bring in. Farmland increase was caused by either
planning boundary, collectives may ignore the procedures of changes of administrative boundaries, or conversion of nonarable land into
land requisition, that is, being acquired by the state before arable land, or by a combination of both. Correlation between farmland
releasing the land to the market. No severe penalties are loss and real estate development (COMLP) and between farmland loss and
attached with violation of the land requisition procedures. urban population growth (FN6_3)
Thus, it was quite common that land parcels were transferred Data groups COMLP FN6_3
through channels other than those defined by the law, such
Nation-wide 0.410∗∗ 0.304∗∗
as investment shares, sites for rural industries, and sale or
(150) (153)
lease.
Coastal cities 0.389∗∗ 0.317∗∗
Farmland loss in China is by and large a problem typical
(89) (92)
of the development of a transitional economy. In the current
Noncoastal cities 0.113 0.226
transition, China’s legal framework is evolving but remains (61) (61)
in its infant stage. Regulative details are lacking in the exist- Big cities (i.e., extra- 0.357∗ 0.294
ing laws and regulations. Even some of the regulative details super large, super (40) (40)
present, mechanisms for law enforcement are absent. The large and large cities)
weak legal ground of land use control is further undermined Small cities (i.e., 0.236∗ 0.019
by the desire of officials to achieve rapid economic returns. medium and small (116) (119)
It is possible to resolve the farmland loss problem only if a cities)
sound legal basis is made available, and corrupt practices are
Notes:
minimized. People in charge at various government authori- Numbers within parentheses are the number of cases.
ties must make their minds clear that farmland conservation ∗ Correlation is significant at the 0.05 level (2-tailed).
∗∗ Correlation is significant at the 0.01 level (2-tialed).
is critical to China’s sustainable development, which cannot 13 The accounting methods used in China Statistical Yearbooks exclude
be traded off by any high-speed economic leap forward. pure fruit or tea farms (i.e., farms specialized in fruit or tea production
only without involving other agricultural plants such as grain or cotton),
or forestry land, from farmland.
14 As in 1996, 70% of the construction projects were in the suburban areas
Acknowledgements
in Beijing.
15 This is another name of economic development zone, or kaifa qu. In many
The authors wish to thank the two anonymous referees for cases, development parks could be smaller than economic development
their constructive comments on an early draft of this paper. zones in size of the area.
16 The arranged price of land is usually much lower than free market price.
The two parties involved, i.e., the party such as local government agency or
state enterprise representing landowner and the buyer, negotiate for an ac-
Notes ceptable price, which is not determined by the market conditions of demand
and supply.
1 1 mu equals 666.67 m2 .
2 For details of regionalization in China, see Han and Wong (1994).
3 These five groups are extra-super large (2 million and over), super large References
(1–1.99 million), large (0.5–0.99 million), medium (0.2–0.49 million), and
small (<0.2 million) cities according to the size of nonagricultural popula- Chan K.W. and Xu X.Q., 1985: Urban population growth and urbanization
tion. in China since 1949: restructuring a baseline. China Quarterly 104: 583–
4 Cities refer to officially designated cities by government classification
613.
267

Chen X., 1996: Who treasures land resources (Guotu Zhiyuan Shui Lai Qian B., 1996: Problems in the utilization of collective-owned land as non-
Zhenxi). China Real Estate 1996 (9): 11–15. agricultural construction land (Liyong Jiti Tudi Jinxing Feinong Jianshe
Chen Y.Q., 1996: Studies on the problems and solutions in utilization of de Luogan Wenti 1–3). Beijing Real Estate 1996 (7): 20–22; (8): 18–20;
land in urban and rural integrated area in Beijing (Beijing Chengxiang (9): 19–20.
Jiaochuodai Tudi Liyong Wenti yu Duice Yanjiu). Economic Geography Real Estate Periodical 1996 (26): 13; 1996 (23): 15; 1998 (31): 19
16 (4): 46–50. Ru Y., 1995: Discussion on the appropriately strengthening administration
China Real Estate Information 1997 (1): 52; 1996 (8): 4. and planning of urban land by governments (Lun Zhengfu Dui Chengshi
China State Land Administration Bureau, 1995: Yearbook of the Land of Tudi Guanli de Shidu Qianghua yu Tudi Guihua). China Real Estate
China 1994–1995. Zhongguo Dadi Chubanshe, Beijing. 1995 (7): 20–22.
China State Land Administration Bureau, 1997: Yearbook of the Land of State Statistical Bureau, P.R. China, 1994, 1997: China Statistical Yearbook
China 1997. Zhongguo Dadi Chubanshe, Beijing. 1994, 1997. China Statistical Publishing House, Beijing.
Chongqing Municipal Statistical Bureau, 1997: Statistical Yearbook of State Statistical Bureau, P.R. China, 1994, 1997: Urban Statistical Year-
Chongqing 1997. China Statistical Publishing House, Beijing. book of China 1994, 1997. China Statistical Publishing House, Beijing.
Fan C.C., 1995: Of belts and ladders: state policy and uneven regional State Statistical Bureau, P.R. China, 1999: China Statistical Summary 1999.
development in Post Mao China. Annals of the Association of American China Statistical Publishing House, Beijing.
Geographers 85 (3): 421–449. Tian F., 1994: The overview of studies on the problems in real estate indus-
Han S.S., 1998: Real estate development in China: A regional perspective. try in China currently (Guanyu Dangqian Wuoguo Fangdichan Ye Wenti
Journal of Real Estate Literature 6: 121–133. de Yanjiu Zhongshu). Urban Affairs 1994 (2): 51–54.
Han S.S. and Wong S.T., 1994: The influence of Chinese reform and Pre- Wang F.X. and Zhang F.M., 1996: Potential Land Resources and Devel-
reform policies on urban growth in the 1980s. Urban Geography 15: opment (Tudi Houbei Ziyuan Kaifa Qianli Jiqi Changshi). Beijing Real
537–564. Estate 1996 (6): 22–23.
Han S.S. and Yan Z.X., 1999: China’s coastal cities: Development, planning Wang X.J. and Wang G.X., 1995: China Land Price. Zhongguo Wujia
and challenges. Habitat International 23 (2): 217–229. Chubanshe, Beijing
Hayter R. and Han S.S., 1998: Reflections on China’s open policy towards Xu B.M., 1999: Analysis on the effectiveness of land market operation
foreign direct investment. Regional Studies 32 (1): 1–16. in China (Dui Wuoguo Tudi Shichang Yunzuo Youxiaoxin de Fenxi).
Jiamusi Municipal Statistical Bureau, 1997: Economic Statistical Yearbook China Real Estate 1999 (1): 7–10.
of Jiamusi 1997. Jiamusi Municipal Statistical Bureau, Jiamusi. Xuzhou Municipal Statistical Bureau, 1998: Statistical Yearbook of Xuzhou,
Jiang R.G., 1998: Total land supply is out of control and its solutions (Tudi 1998. China Statistical Publishing House, Beijing.
Gongying Zongliang Shikong ji Duice). Real Estate World 1998 (6): Yang J.R., 1993: Problems and solutions in real estate development
13–14. (Fangdichan Kaifa Zhong de Wenti yu Duice). Urban Development 1993
Johnson T., 1997: Clear Water, Blue Skies: China’s Environment in the New (8): 21–25.
Century. World Bank, Washington D.C. Yang Y.X., 1996: Handling the three great relationships appropriately to
Li R.H. and Zhuang Z.M., 1998: To strengthen Beijing land administration protect cultivated land resource (Zhengque Chuli Sanda Guanxi, Qieshi
and to enforce laws and regulations (Qianghua Shoudu Tudi Guanli, Baohu Gengdi Ziyuan). Beijing Real Estate 1996 (10): 23–24.
Jiada Zhifa Lidu). Beijing Real Estate 1998 (2): 20–22. Ye J.A. and Li X., 1999: Study on the economic development, urban ex-
Liu W.X., 1995: Cultivated land conservation and urban development in pansion and cultivated land loss in Zhujiang Delta the case of Dongguan
Beijing (Beijingshi de Gengdi Baohu yu Chengshi Kaifa). Beijing Real (Zhujiang Sanjiaozhou Jingji Fazhan, Chengshi Kuozhang yu Nongtian
Estate 1995 (11): 27–29. Liushi Yanjiu Yi Dongguan Weili). Economic Geography 19 (1): 67–71.
Liu Y.Y., 1998: Analysis of the effects of urban land allocation mechanism Yearbook of China Real Estate Market, 1996, 1997: Zhongguo Jihua
on urban construction in China (Shixi Wuoguo Chengshi Tudi Peizhi Chubanshe, Beijing.
Jizhi Dui Chengshi Jianshe de Yingxiang). China Real Estate 1998 (3): Zhang L.L., 1998: The national and municipal conditions, and the future of
10–14. land – thinking before the coming of land day (Tudi Guoqing, Shiqing
Lu Y.H., 1997: Current problems in non-agricultural construction land yu Weilai – Tudiri Jianglin Zhiji de Sikao). Beijing Real Estate 1998 (6):
and the ideas to strengthen administration (Feinong Jianshe Yongdi 7–9.
Xianchun Wenti ji Jiaqiang Guanli de Yijian). China Real Estate
Information 1997 (1): 47.
Ma L.J.C. and Cui G.H., 1987: Administrative changes and urban popula-
tion in China. Annals of the Association of the American Geographers
77: 373–395.

Vous aimerez peut-être aussi