The Trend of Conversion of Agricultural Land in Japan since 1980

北海道教育大学大雪山自然教育研究施設研究報告 第38号
平成 16 年 3 月
Reports of the Taisetsuzan Institute of Science No. 38
March
2004
The Trend of Conversion of Agricultural Land in Japan since 1980
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
Hokkaido University of Education, Asahikawa Campus Asahikawa, 070-8621, Japan
日本における 1980 年以降の農地転用の動向
氷見山幸夫・新谷奈美・高橋賢市
北海道教育大学旭川校地理学教室
要 旨
各都道府県の「土地利用動向調査」報告書に掲載されている農地転用に関するデータをデジタル
ファイル化し，1980 年∼ 2001 年の農地転用動向を分析した．その結果，1986 年∼ 1991 年のいわ
ゆるバブル経済期およびその前後の農地転用の動向に，次のような明瞭な特徴を見出した．
１）総農地転用はバブル期前は漸減し，バブル期に急増，バブル崩壊後は速いスピードで減少した．
いずれの変化もきれいな直線状をなす．
２）１）の直線状増減が，農地の他用途への転換が全農地に占める割合が非常に小さいことによる
ということを，数学的に証明した．
３）回帰分析の結果，総農地転用届出面積は，1991 年∼ 2001 年の間の直線的減少が続くと仮定す
ると，2012 年にはゼロになることが示された．これは，2002 年∼ 2011 年の 10 年間に 73,073
ha の農地が他用途に転用された後，農地の転用による減少が，少なくとも届出を要するもの
についてはなくなることを意味する．なお，バブル経済期前の 1980 年∼ 1986 年の転用届出
面積の減少動向を将来に延長した場合は，転用面積がゼロになるのは 2015 年との結果が得ら
れ，先に得た 2012 年という結果と非常に近いことが明らかとなった．
４）農地の他用途への転換を，転換件数と転換面積の両方から転換目的別に見た結果，バブル経済
期に見られた農地転用の増大のかなりの部分が，地方自治体等の公的機関が関わったものであ
ることが示された．
以上のように，
「土地利用動向調査」報告書に掲載されている農地転用に関するデータは，これ
までの農地転用の実相を詳細に明らかにしてくれるだけでなく，将来予測にも非常に有用であるこ
とが示された．
Keywords: conversion of agricultural land ( 農地転用 )，Survey of Land Use Trend ( 土地利用動向
調査 )，Agricultural Land Act ( 農地法 )
1. Introduction
Each of the 47 prefectures of Japan has been conducting an annual survey called the
since 1980. The Survey report, which consists of two booklets with
− 43 −
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
statistical tables and two maps corresponding to them, is an invaluable information source for
the study of land use trend (Himiyama, 1989). However, as the Survey report is primarily aimed
at the use in the prefecture it is made in, it has not been fully used for the national-level study.
This paper intends to show how the Survey report can be used nationally and can yield useful
results. The data of conversion of agricultural land to other uses, which are in a statistical table
in the Survey report, have been digitized and processed in order to analyze the trend of the
conversion of agricultural land. The time range of the study is 1980-2001, i.e. all the existing
reports have been used. There are 10 variables related with conversion of agricultural land, as
shown in Table 1 (a). The printed statistical tables are scanned and digitized with a software
called eTypist, and are put into EXCEL files, as shown in Table 1 (b). Arc View GIS has been
used for mapping.
2. Trend of All Reported Conversions of Agricultural Land
2.1 Agricultural Land Act
With its rather limited area of flat land for its large population, Japan has long been one of
the top food importing countries in the world for decades. It is hence natural for the government
to protect good agricultural land by controlling its conversion to other uses. The owners or
buyers of agricultural land have to report to the local government when they are to convert the
land to other uses, and outside urbanization promotion areas they even need to get permission
to do so from the local government according to the Agricultural Land Act. The Act came into
effect in 1952.
Agricultural land in Japan has been declining since 1955, when it reached 6 million ha, and
the cultivated land in 2003 was 4,736 thousand, a 26 thousand ha (0.5%) decrease in a year. This
decrease is attributed to farm abandonment, urban development and afforestation, as well as
natural disasters.
Japan is a small country for its large population, and low flat land is a battle zone between
different types of land uses. The Agricultural Land Act is intended to protect agricultural land
from such conflict, and conversion of agricultural land to other uses is put under control by it.
This paper first views the gross national trend of conversion of agricultural land, then
examines conversions for different purposes, at national, district and municipal scale levels.
2.2 Trend of all reported conversions of agricultural land
Figure 1 shows the area and number of all conversions of agricultural land reported to
the local authorities in Japan according to the Agricultural Land Act during 1980-2001. At the
starting year of 1980, the number was about 310,000, while the area was about 2,000 ha. Both the
number and the area were in steady decline till 1986, but then they started to increase sharply.
The increases continued till 1991, reflecting what was called the bubble economy in Japan.
The number and the area reached their peak values of 300,000 and 29,000 ha, respectively, in
1991, then started to decline. The increase rate of the area during 1986-1991 was larger than
that of the number, i.e. the mean area of conversion for each report was larger than in other
− 44 −
The Trend of Conversion of Agricultural Land in Japan since 1980
Table1 Sample statistical table of
(a) In original printed form
(b) In Excel form
Figure 1 Trend all reported conversions of agricultural land in Japan
periods. After the collapse of the bubble economy in late December 1991, the area of converted
agricultural land entered a linear declining phase, while the number marked a small peak in
1996, although it was also in an almost linear declining phase.
Figure 2 and Figure 3 respectively show the number and the area of reported conversions
of agricultural land to other uses in each prefecture during 1980-2001. Both maps show high
values in the prefectures in the Kanto and Chubu District, particularly in Saitama, Chiba, Gunma
and Ibaraki in the Tokyo Metropolitan Area, and Aichi Prefecture, which is the centre of the
Chukyo Metropolitan Area. Another large metropolitan area, the Kei-Han-Shin Metropolitan
− 45 −
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
Area, does not show large values, because their agricultural areas were already small by 1980.
Hokkaido, Tohoku and Kyushu District show high values of the areas for their numbers of
conversions, because they occupy peripheral areas and the unit of a conversion of agricultural
land is generally large. It is particularly so in Hokkaido, where the average area of a conversion
is 0.377 ha. as against the national average of 0.075 ha.
2.3 Regression analyses of the trend
If there is a persistent motivation for change of the population of a group, whether it is
economic, social or human, and if it surpasses the constraint to change, and if the change of
the population is proportional to the population itself, then the change of population can be
expressed as follows:
N = N0ekt ................................................. (1)
where N is the population size at time t, No is the initial population size, i.e. at t = 0, and k is the
change coefficient. This formula is known by ecologists or demographers. Himiyama (2001) and
Himiyama et. al. (2002) showed that this formula could be applied to the case of the increase of
the number of reported land transactions for golf courses in Japan during the 1980s, and proved
that the increase followed a beautiful exponential curve. Figure 4 shows three regression lines
of the area of reported conversions of agricultural land for 1980-1986, 1986-1991, and 1991-2001.
As can be easily understood, the three regression lines all have very high R2 values, indicating
extremely good lineality of the changes.
In order to understand the regression lines in Figure 4, it is useful to note that an annual
Figure 2 Number of reported conversions of
agricultural land by prefecture (1980-2001 total)
Figure 3 Area of reported conversions of
agricultural land by prefecture (1980-2001 total)
− 46 −
The Trend of Conversion of Agricultural Land in Japan since 1980
Figure 4 Regression lines of the area of reported conversions of agricultural land for 1980-1986,
1986-1991, 1991-2001
change of 1,000 ‒ 2,000 ha is considerably small in comparison to the total area of agricultural
land in Japan, which is 6,400,000 ha. In another words, k in (1) is very small. According to the
Maclaurin s theorem of progression,
ex = 1 + x/1! + x2/2! + x3/3! + ................ (2)
Hence if k is very small, N is expressed as
N = N0ekt = N0 (1 + kt) ............................... (3)
This formula explains the linearity of the trend of the area of converted agricultural land
in 1980-1986, 1986-1991 and 1991-2001. The very high R2 values for each of the three periods
indicate the relevance of formula (3). The regression analyses showed that the area of reported
conversions of agricultural land decreased linearly during 1991-2001, and if it continues, the area
of reported conversions of agricultural land would reach zero, after a conversion of 73,073 ha
during 2001-2011. The regression line of 1980-1986, i.e. the period preceding the bubble economy,
also indicated that the area of reported conversion of agricultural land would diminish by 2015,
which is surprisingly close to 2012.
Figure 5 and Figure 6 show the change of the number and the area of reported conversions
of agricultural land during 1986-1991 in each district, respectively. The regression lines for all
districts, particularly those of the area, show very high R2 values, suggesting that the bubble
economy had a uniform effect on people s attitude towards agricultural land throughout the
country all through 1986-1991. The increase rate of the Kanto District is relatively low in the
case of area, and even minus in the case of number, reflecting the fact that the agricultural land
suitable for conversion is more limited than in other districts.
− 47 −
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
3. Trend of Reported Conversions of Agricultural Land for Different Purposes
The report of the
shows the number and the area of converted
agricultural lands by their conversion purposes, such as residential use, industrial use,
afforestation, and other uses, including schools, parks, sport facilities, roads, railways and other
public uses. This chapter examines the trend of conversions by their purpose.
Figure 7 shows the trend of the number of reported conversions of agricultural land in the
whole country by purpose. The conversions for residential use occupy the highest position, then
other uses, industrial uses and afforestation follow. One interesting aspect of the graph is the
response of each case to the bubble economy. Other uses show by far the largest response to
the bubble economy, i.e. marked increase during 1987-1992 and sharp decline after it, indicating
that the bubble economy was driven by public bodies. Industrial use showed more moderate
response, afforestation showed little response, and residential use showed no response as far as
the number of conversions is concerned. In every case, decline since the late 1990s is evident,
and even the conversions for residential uses marked a record low of 90,000.
Figure 8 shows the trend of the area of reported conversions of agricultural land in the whole
country by purpose. It shows a similar picture to Figure 7, but there are some noticeable differences
between them. As in the case of the number of conversions, the conversions for residential and
other uses keep higher positions than industrial uses and afforestation, and in every case both
the number and the area kept declining after the bubble economy, and reached record lows in
2001. In the case of industrial use, the area obviously responded to the bubble economy, with its
marked peak of over 6,000 ha in 1991 and the sharp decline that followed. This indicates that
there were some investors strongly motivated by the bubble economy and converted larger-thanusual agricultural lands to industrial uses. A similar trend is observed in the case of residential use,
suggesting a similar reason. As for afforestation, there was a small peak of 3,000 ha in 1987, i.e. in the
middle of the bubble economy, but then started to decline, and stays below 1,000 in recent years. 4. Distribution of Reported Conversions of Agricultural Land for Different Purposes
Figure 9 shows the total area of reported conversions of agricultural land for residential use
during 1980-2001. The pattern resembles that of Figure 3, i.e. the case of the area of all reported
conversions, except Hokkaido, which is less prominent here. As expected, Kanagawa (9,100 ha),
Chiba (8,800 ha) and Saitama (8,600ha), which are all adjacent to Tokyo, and Aichi (8,900 ha)
in the centre of Chukyo Metropolitan Area show particularly large areas of conversions for
residential uses. Hokkaido may appear to have large area of conversion, but it is only due to the
large administrative territory it occupies.
Figure 10 shows the total area of reported conversions of agricultural land for industrial
use during 1980-2001. The Chubu District and the Kanto District in central Japan show higher
values, while western Japan show relatively low values. Niigata Prefecture (5,200 ha) in the
Chubu District and Ibaraki Prefecture (3,500 ha) in the Kanto District show particularly high
values. The high value of Hokkaido is simply due to the large area of its territory.
− 48 −
The Trend of Conversion of Agricultural Land in Japan since 1980
Figre 5 Change of the number of reported
conversions of agricultural land in each
district during 1986-1991
Figure 6 Increase of the area of reported
conversions of agricultural land in each
district during 1986-1991
Figure 7 Trend of the number of reported
conversions of agricultural land in Japan
by purpose
Figure 8 Trend of the area of reported
conversions of agricultural land in Japan by
purpose
− 49 −
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
Figure 11 shows the total area of reported conversions of agricultural land for afforestation
during 1980-2001. In contrast to the cases of residential and industrial uses, Kyushu shows
its high concentration, with Kagoshima (3,000 ha), Kumamoto (2,600 ha) and Oita (2,300 ha)
Prefectures in Kyushu occupy the highest three positions. This is due to the extensive tree
planting in marginal agricultural areas covered by infertile volcanic ashes or having poor
geomorphological conditions.
Figure 12 shows the total area of reported conversions of agricultural land for other uses
during 1980-2001. Here, other uses include schools, parks, roads, railways and other public uses.
Hokkaido (154,000 ha) shows the highest value, not only because of its large territory, but also
because of its heavy economic dependence on public development schemes. Chiba (9,900 ha) and
Saitama (9,000 ha) in the Kanto District and Aichi Prefecture (8,800 ha) follow. 5. Conclusions and Acknowledgements
The conversion of agricultural land in Japan since 1980 has been studied in detail by using
the Survey of Land Use Trend. It has demonstrated its valuable use in the study of land-use
changes in Japan. In the study, all the data of conversion of agricultural land in the statistical
tables of 47 prefectures from 1980 to 2001 have been digitized and analyzed. The analyses have
shown the characteristic trends of the area and the number of conversions of agricultural land
during 1980-2001 at national, district, prefectural and municipal levels. The most outstanding
feature of the trends has been found to be the effect of what is called bubble economy' of Japan
during 1985-2001. The main findings are as follows:
1) Total yearly conversions of agricultural land decreased gradually during the period preceding
the bubble economy, increased sharply during the bubble economy, then declined more rapidly
than before. In all three periods, the changes were linear.
2) The linear changes found in 1) have been proved mathematically that they are the result
of the fact that the conversions of agricultural land occupy only small portions of the total
agricultural land in Japan.
3) Through the analyses of the number and the area of conversions of agricultural land by their
purpose, much of the excessive conversions of agricultural land during the bubble economy has
been brought about by local authorities and other public bodies.
Bubble economy' may take place anywhere, and in fact the economic situation of some
developing countries, including China, shows a similar symptom to it. The fact found in this
study is very likely to be applicable to China, hence helps people to understand what can happen
in the near future.
It is hoped that the Survey will be continued with improvements as it goes along, but
excluding the destructive modifications that governmental data sometimes undergo.
This study has been sponsored by the JSPS Science Research Fund (Kagaku Kenkyu Hi, No.
13851003). The authors are grateful to the prefectural officials in charge of the Survey of Land
Use Trend for their kind instructions and help. Without their devotion to the Survey and kind
guidance, this study could never have been accomplished.
− 50 −
The Trend of Conversion of Agricultural Land in Japan since 1980
Figure 9 Area of reported conversions
of agricultural land for residental use
(1980-2001 total)
Figure 10 Area of reported conversions
of agricultural land for industrial use
(1980-2001 total)
Figure 11 Area of reported conversions
of agricultural land for afforestation
(1980-2001 total)
Figure 12 Area of reported conversions
of agricultural land for other uses
(1980-2001 total)
− 51 −
Yukio HIMIYAMA, Nami SHINTANI, Kenichi TAKAHASHI
References
Each Prefecture of Japan (each year since 1980)
Himiyama, Y. (1989) The recent trend of land transactions in Japan. in
R. D. Hill ed.,
Hong Kong University Press.
Himiyama, Y. (2001) Some thoughts on human aspects of LUCC in Japan and China. in
, Shi Peijun ed., Beijing Normal
University, 17-25.
Himiyama, Y., J. Kumagai, Y. Kakuchi, H. Shirakawa (2002) The Trend of Recreational Developments in
Japan since 1980. Reports of the Taisetsuzan Institute of Science, No.36, 73-82.
− 52 −