Geographical Review of Vol. 76, No. 5, 349-366, 2003

The Land Conversion Process due to Landowners' Land-Use Decisions in the Rural-Urban Fringe of City

TSUTSUMI Jun Faculty of Law and Letters, Ehime University, , Ehime 790-8577, Japan

Abstract: This paper concerns the land conversion process in the rural-urban fringe of Maebashi city, a medium sized city in Japan. The author investigated land-use change , land ownership change, and landowners' land-use decisions. Land ownership change was examined from 1980 to 1993 using land assessment data from the Maebashi municipal office . The information about landowners' land-use decisions was obtained by interviewing selected landowners. The landowners in the sample area made land-use decisions based on two types of factors, called initial and decision factors. The initial factors were land re-adjustment , land inheritance, need for a larger income, request to sell their land, and failure of a non-agricultural business. The decision factors were the existence of their successors in agriculture, payment of inheritance tax, intention to keep farming, possession of land with good access to roads and their desire to utilize the land. Based on their decisions, their behavior can be divided into three categories: land utilization, transaction (or sale), and abandonment. Traded land was converted to urban use, such as housing, parking or shops; and was scattered throughout the area . Utilized land was converted by the owners themselves; and was distributed along the main roads leading to central Maebashi, close to the owners' houses. Utilized and traded land were based on different decisions by the landowners. Much of the land in the inner fringe is in demand. Thus, both land being utilized and land being traded tended to be converted to urban use.

Key words: land conversion, inner fringe, land ownership, land-use decision, land inheritance

tween several types of land-use. Introduction In general, there have been a large number of land-use studies, particularly on issues that con As a city grows, urban land demand in cern how rural areas change under high pres creases, because some centrifugal functions, sure of urbanization. The spread of urban land such as factories and residences, move from use extended into rural areas, gradually giving the central area to the suburbs; and some cen rise to land-use conflict between rural and ur tripetal functions, such as commercial and ban land. In general, although the proportion of office activities, move into the built-up area. part-time farmers expanded steadily, most of Therefore in the central area of the city, due to the farmers had a tendency to spare their farm high demand of urban land, we often find many land as much as possible. Conversely, some high-rise buildings, and an increasing number farmers reduced their proportion of land dedi of parking lots in transitional land-use (i.e., to be cated to farming, and instead turned to renting converted to urban use in the immediate fu their properties for apartment houses, parking ture). In addition, we can also find general lots, and so on (Fukuhara 1971; Nagashima tendencies of land-use change from rural to 1972; Harada 1976). Moreover, in a body of urban and of land-ownership change from literature on the conversion of prime land to farmer to non-farmer in the rural/urban fringe urban land through an Application Roll for Ur area. Thus, the built-up area of the city ex ban Development (nouchi-tenyo), regional char pands outward as a result of competition be acteristics of land-use change were also ana-

349 102 TSUTSUMI J. lyzed (Takahashi 1977). decisions for land-use and for farming manage During the 1960s and early 70s, rapid indus ment). trialization occurred in Japan, causing conflict In the 1980s, there were strong arguments for between rural and industrial land-use. In this incorporating behavioral approaches into geo period, factory location in agglomerated areas graphical studies (Bryant et al. 1982; Golledge formed a striking contrast with that in the and Stimson 1987). Especially land-ownership small local cities; capital-intensive divisions re was emphasized as important because it was a mained in highly industrialized areas, while la suitable index for the degree of urban influence bor-intensive divisions dispersed into rural ar (Pond and Yeates 1993, 1994a, 1994b). Other eas. Small local cities were characterized by an sample studies following the behavioral ap influx of factories from large companies based proach include Tanaka (1982), who investi in big cities. When a factory moved to a rural gated the urban growth process from the view area, much farmland was replaced not only by a point of how the farmers transferred their land factory itself but also by company houses, wel ownership, using a case study of Mitaka city in fare facilities, and shops attached to the factory the western suburbs of . Nakamura (Takahashi 1968). Recently land-use issues (1995) focused on the farmers' recognition have come to be investigated in terms of legal processes to choose crops which were most suit regulations such as the impact of the Urban able for particular natural conditions. Pyle Planning Act.1 These studies provide good ex (1986, 1989) analyzed the decision-making amples that identify the main factors of land process of landowners and revealed that their use change, but they do not explain the process decisions were closely related to shape patterns of land-use change. of land-use. Ganderton (1994) also analyzed the Undoubtedly, there are many factors that relationship between spatial pattern of land-use cause land-use change. It would be more appro and land-use decisions of landowners, and the priate to consider urbanization as the context of characteristics of decision-making processes of multiple causes. Kikuchi2 analyzed interactions landowners. Tsutsumi (1999) investigated the between several factors affecting urban land land conversion process in the CBD of Maebashi use change by using sample areas situated in city from the viewpoint of changing roles of different zones in the Auckland metropolitan decision agents in an urban land market. These region in New Zealand. In a series of studies, he studies were notable examples of the behav regarded development stages and the different ioral approach. strength of potential urbanization forces in the From these general observations, this paper same metropolitan region as important. In one attempts to analyze the land conversion sample area situated close to the built-up area process at a micro level by investigating the of Auckland where urban pressure was high, decision-making processes of landowners in the definite conversion to urban land-use was iden rural/urban fringe area of Maebashi city. tified due to increased interactions between It is a medium-sized city in Japan (Figure 1) such factors. In another sample area situated in with a population of approximately 280,000 in the city's countryside, most of the agricultural 2000. The trends in population decentraliza land-uses were sustained, although the area tion and associated rural land-use changes have was under urban pressure. Kikuchi's works been similar to those of other cities of almost were noteworthy because he tried to synthesize the same size in population. That means Mae a number of factors such as urbanization, bashi is a typical medium-sized city with one change in farming management, innovations in core built-up area. farming, land-ownership change, regulations, Many of the residential estates in Maebashi and increase of urban newcomers in the origi were developed by land re-adjustment pro nal rural communities (konjuka)3, which were grams. Land re-adjustment programs are un usually treated separately in previous studies. dertaken either by public housing associations However, the studies were insufficient in that (e.g., prefectural or city housing associations), they did not address behavioral factors (e.g., or government institutions. One of the most

350 The Land Conversion Process 103

Figure 1. Study area in 1995. notable features of the housing policy in Mae which accounts for 40.5% of the ADFU. When bashi is that most land re-adjustment was un partially developed works are completed, the dertaken by the Maebashi Municipal Office it total developed area will increase to 2,532 ha self. In general, land re-adjustment works con (54.7% of the ADFU), placing greater pressure stitute 40% of total housing, which has tended on rural areas. The areas currently under de to be easily converted to urban use. velopment are situated in the rural-urban A proportion of land developed under re fringe areas in the southern, south-eastern, adjustment works to the total `Area Designated northern, and western suburbs. for Urbanization' (shigaika-kuiki) (simply, First of all, the author examined the outline ADFU) is almost 50% in Maebashi city, which of the urbanization process for the city by using ranks as one of the highest in Japan.4 The first Digital Land Information to select a sample work on land re-adjustment in Maebashi was area for micro-level analysis. In order to obtain restoration from damage caused during World land-ownership data, land assessment rolls War II. Most of the built-up area seen today is were used. The Property Tax Division of Mae a result of this work. After the General Urban bashi City Hall houses the rolls that are cross Plan of Maebashi was decided in 1957, one of referenced with the data at Maebashi's Legal the important issues in the city's development, Affairs Bureau (homukyoku). By using the rolls, several land re-adjustment works were taken it was possible to compile an array of informa up in the suburbs surrounding the already de tion about land-ownership including the name veloped built-up area. Most of the early land and address of landowner, the lot size, in some re-adjustment projects were relatively small in cases, the date of sale or inheritance, and so scale (about 20 ha). After the late 1970s, the forth. The information of landowners in the scale of works increased due to the high eco urban land market could be obtained by land nomic growth: principal roads were widened trade and land inheritance data from the files. with rapid development of industrial and resi The data of 1980 and 1993 were examined for dential estates. As of the end of 1997, the the sample area. The land-use data for the area ADFU covered 4,627ha. The area completely was mainly obtained by the author's field sur developed by land re-adjustment was 1,871ha, vey conducted in August, 1993, while some

351 104 TSUTSUMI J. parts were supplemented by aerial photos, topo tions (Table 1) (Kikuchi and Tsutsumi 1998). graphic maps, and residential atlases. To ob The vertical axis of the matrix represents land tain information on the decision-making use in 1976 and horizontal axis land-use in processes, the author interviewed the selected 1989. The number of grid points indicating agents (e.g., landowners, land buyer, and inter land-use change to `vegetable fields' were 656 mediate actors) during the period September, (34.5%); 'buildings' were 561 (29.5%); mulberry 1993 to August, 1994 for the sample area. The fields were 227 (11.9%); traffic-use were 79 selection of landowners in the sample area was (0.4%); and 'other uses'5 were 222 (11.7%). based on land-trade and land-inheritance data When using Digital Land Information, the fol from the rolls. Thirty landowners were selected lowing two matters should be kept in mind: (1) as interviewees whose causes, conditions and truncation error, and (2) category error (Bito behavior in regard to land-use decisions was 1992). Truncation error occurs because only closely examined. Their land accounted for the primary land-use category is recorded for 43.6% of the total land-use change, 62.8% of the each grid point. Secondary and tertiary catego total inherited land, and 45.0% of the total pur ries are not shown in cases of multiple land-use chased land in the sample area. The GIS pack for a single section of land. For example, the age was used to assist in representing, process major uses for a section of land maybe vegeta ing and visualizing (mapping) spatial data. ble, mulberry and paddy field in 1976, but the land-use code only indicates 'vegetable.' After Urbanization in Maebashi City development, the vegetable fields are converted to building use, so that the major uses of this In order to view the tendency of land-use land become mulberry, building (from vegeta change in Maebashi city, the mesh method was ble field), and paddy field in 1989. However, the taken for the years of 1976 and 1989 with land-use code identifies 'mulberry' only. Land Digital Land Information data obtained from use code information records a change in land the National Land Agency of Japan. The data use from 'vegetable' to 'mulberry.' Actually, were compiled from 100m intervals, giving a according to the information from the Agricul total of 18,700 grid points for Maebashi city in tural Affairs Division in Maebashi City Hall, the each year. There were 1,904 sections that un proportion of mulberry fields to the total rural derwent land-use change between those years, land has been in decline during the same period. which was 10.2% of the total number of sec Therefore, 'vegetable to mulberry' changes

Table 1. Number of land-use changes measured in 10,000m2 lots (100m•~100m square interval) for

1976 to 1989

Source: Digital Land Information of Japan.

352 The Land Conversion Process 105

were excluded from the data because they were ing to data constraints, the conversion from not considered as indicative of the changes oc rural to urban use6 was analyzed. Little change curring during this period. Category errors was observed in the central area, where by 1980 arise because the varied possible urban uses of almost all land was urbanized. A high rate of land were partitioned into only three catego change was identified in two zones; the inner ries. For example, both low-rise housing and fringe zone (two to four kilometers from the high-rise buildings were placed in the common CBD); and the isolated residential and industrial category of 'building uses.' Thus, land-use parks zones (six to seven kilometers from the change from low-rise housing to high-rise build CBD). ings would not appear as any change since both Land-transaction is a good predictor of im types of land-use were recorded under the same pending land-use change. The number of land category. Keeping in mind these problems, the -transactions in Maebashi city is illustrated in following two general patterns of land-use Figure 3. The number of transactions in the change were identified: (1) general agricultural inner fringe during 1980-84 was greater than use (e.g., paddy, mulberry) to vegetable fields, other zones in Maebashi city. In other words, and (2) from various uses to urban use. the land market in the inner fringe zone was The author examined the spatial characteris more active than in the CBD or rural areas. tics for urbanization of Maebashi city from After 1985, the number of transactions in CBD 1976 to 1989 using the same data set. Figure 2 exceeded that of the inner fringe. This dra illustrates the spatial pattern of land-use matic change was driven by the Bubble economy changes aggregated by the third-order grid on creating a strong demand in urban land-use (e.g. the data set (i.e. one interval equals 1 km2). Ow office space), which was concentrated in a very

Figure 2. Conversion from rural to urban land-use in Maebashi city for 1976 to 1989 .

353 106 TSUTSUMI J.

level of 50 transactions per square kilometer. The number of land-transactions for rural land increased temporally in 1988 and 1991 due to the development of some industrial estates. The sample area of Rokku is illustrated in Figure 4. Rokku is situated about three kilome ters south-southeast of the CBD. This area was one of the most active in terms of land-use change, indicated by a count of the number of grid points (>15) identifying change in the Rokku area (see Figure 2). Such activity in the inner fringe makes Rokku a suitable area for investigating the relationship between land -transaction and land-use change. The bound ary designating land controlled for urbaniza Figure 3. Number of land-transactions in Mae tion effectively divides Rokku into two areas. bashi city by year for 1980 to 1992. The area designated for urbanization contained many urban-oriented facilities such as a gar bage disposal factory, large-scale warehouses, and gas tanks that were in existence before small area. After the Bubble economy, the num 1980, which is the starting point of this article. ber of transactions in the CBD fluctuated. Con This area is easily accessible because of two versely, the number of the transactions in the major roads leading to the CBD of Maebashi. inner fringe was generally stable, around the One part of this area was selected as the sample

Figure 4. Sample area of Rokku in the rural/urban fringe in 1993.

354 The Land Conversion Process 107

area in which a micro-level examination of either land re-adjustment, or abandonment.8 land-use analysis was performed. Sites for This type of land-use change appeared in the large-scale facilities were excluded because south-west part, with relatively bad access to they did not undergo land-use change during main roads. After the remarkable increase in the study period. The area controlled for urban urban land-use during 1980-93, urban land-use ization was also excluded due to insufficient accounted for more than half of the total land transactions. area (Figure 5(b)). Studies of land-use change on urban areas are Process of Land-Use Change in the numerous, but few examine ownership change, Rural-Urban Fringe which is a fundamental issue in the conversion process. Changes in land-ownership almost al In 1980, land-use in the southern part of the ways precede alterations in land-use. Rural Rokku area was predominately rural, and in landowners play a pivotal role in the process particular, paddy fields. Shops, rented offices, because they decide how their property will be factories and high-rise apartment houses ap used, or when it will be sold for use by someone peared along the main road leading to central else. Buyers and sellers of rural property are Maebashi. The eastern part of the Rokku area the actors in the land market who set the course was mainly residential (Figure 5(a)). By 1993, for changes in the urban structure (Bryant et al. the most striking change that occurred was the 1982). decrease in rural land-use. At the same time, In 1980, most major landowners were those the number of roads increased with progress on who lived in the sample area. They accounted the land re-adjustment program (Table 2).7 for 81.1% of the total sample area. Area-orient Along newly built roads there was a rapid in ed usage had already dispersed from the city crease in area-oriented usage such as shops, center to the Rokku area by 1980. For example, apartment houses, and parking lots (with a ca some head offices located in the Maebashi city pacity greater than 10 cars). Another notable center opened branch offices or warehouses in change was the increase in wasteland and va the sample area especially along some of the cant land on which there are no buildings. main roads. Moreover, people in other areas Wasteland (and vacant land) was the result of such as Maebashi city (excluding the Rokku

Table 2. Development of land re-adjustment in Rokku-machi, Maebashi city

Sources: Data from the Urban Planning Division of Maebashi City Hall.

355 108 TSUTSUMI J.

Figure 5. Land-use in Rokku area, Maebashi city for 1980 and 1993.

area), (excluding Maebashi lived in the Rokku area dropped to 77.8%. On city), and Tokyo came to own residential lots the other hand, land owned by people in Mae which were relatively smaller parcels. bashi city increased. Most of the land In 1993, the rate of land owned by those who - ownership change was observed in the rela- 356 The Land Conversion Process 109

tively broader land that extended over several urban land demand (Tsutsumi 1996). Such a parcels and more than 1,000 square meters in relationship was analyzed in the sample area of area. Two reasons are given for this type of Rokku. land-ownership change: l) The Rokku area pro The major land buyers were individuals and vides easy access to the city center and corporations in the sample area, which were sufficient space for those who want to open followed by individuals and corporations in branch offices or warehouses. 2) Since under Maebashi city and Maebashi Municipal Office taking the land re-adjustment, some amount of itself (Figure 7). There were some examples of land was needed by Maebashi Municipal Office corporations domiciled in the Kanto region, es (who conducted the program) for public facili pecially in the Tokyo metropolis and Yoko ties after the program was completed. hama city that purchased land in the sample Figure 6 shows the number of land-transac area. As for land-uses undergoing land tions in the Rokku area for the years 1982 to -transactions, two tendencies are identified. One 1991. The graph shows two peaks in the num is a tendency for urban land-uses such as resi ber of land-transactions, which occurred in the dences and shops, and the other is for waste years 1984 and 1991, respectively. The first land. Some rural lots with easy access to roads peak was due to a planned land re-adjustment were replaced by shops, restaurants, and program that was announced at the end of pachinko parlors. Even lots without easy acces 1984. Since change in land-use was not permit sibility were changed to residential use through ted until land had been reallocated (a period of real-estate firms. In addition, other notable about six months after the announcement), transactions in the sample area included irregu transactions intended for the first half of 1985 larly shaped lots smaller than 20-50 square must have been completed before the end of meters, purchased by office workers who al 1984. The announcement caused a 'run-in' ready lived in the sample area. The sale of effect for transactions in that year. The second small lots was a countermeasure by a land peak was due to a greater than average number owner to the reduction of land due to the land of land inheritances during that year. There is re-adjustment program. The reduction rate was a close relationship between land-transactions around 20-30% in the sample area. When a and land-use change in an area ranging from landowner has several parcels of land in the the city center to the inner fringe under strong same project area, they can be easily integrated to form a new lot. The number of office workers who have purchased small parcels of land to taled more than 100. The characteristics of the relationships be tween land-use and land-ownership in the Rokku area are summarized as follows. In 1980, the rate of rural land-uses in the area was more than 50%. On the other hand, the rate of urban land-uses which was distributed mainly along the newly built roads remained at around 40% (Figure 8). Most of the major landowners were farmers who owned more than 50% of the total land in the Rokku area. By 1993, the most striking change that occurred was the decrease in rural land-use. At the same time, the number of roads increased with progress on the land Figure 6. Number of land-transactions in the re-adjustment program. Along newly built Rokku area of Maebashi city by year roads there was a rapid increase in area for 1982 to 1991. oriented usage such as shops, apartment houses, and parking lots. After urban land-uses

357 110 TSUTSUMI J.

Figure 7. Distribution of traded land in the Rokku area of Maebashi city during 1980 to 1993.

increased remarkably during 1980-93, urban land owned by farmers decreased remarkably land-uses accounted for more than half of the to the level of 33.2% (Table 3). On the contrary total land area. Conversely, the rate of rural the land owned by office workers increased to land-uses dropped to about 30%. In 1993, the 38.3%. In addition, the land belonging to shop-

358 The Land Conversion Process 111

relationship between land-ownership and land - use change, it has been established that land -transaction is one of the most important factors in land-use change. However, land transaction alone does not completely explain why rural land-use was reduced so remarkably during 1980-1993. There is another cause of land-use change: landowners' utilization of their own properties. Landowners often provide some residential units on their land with easy access to main roads. Thus, land-use change is primar ily brought about in two ways: transaction and utilization. In this section, the author focused on land owners' land-use decisions as the reasons for those decisions differed between these two cate Figure 8. Percentage of each type of land use gories. Thirty landowners were selected as in in the Rokku area of Maebashi city, -terviewees whose causes, conditions and behav 1980 and 1993. ior of land-use decisions were closely examined. Of the thirty owners, 26 were farmers and the rest were non-farmers. Their land accounted for 43.6% of the total land-use change, 62.8% of the total inherited land, and 45.0% of the total Table 3. Percentage of each type of landowners in the sample area of Rokku for 1980 purchased land in the sample area. The rate of and 1993 their own land to the total land of Rokku was 42.0% in 1980, which dropped to 34.1% in 1993. They all have a close relationship with relatively larger land transfer, such as transac tion and inheritance of more than 1,000 square meters. In the interview landowners were asked about their reasons for disposing of land - holdings. Items of question were land-use on their own land in 1980 and 1993, the reason for land-use change and for land transfer (land - ownership change). The 30 landowners made their land-use deci sions on their own land based on several causes Sources: Data from Property Tax Division of Maebashi and took on different behavior. There were City Hall. three types of behaviors: utilizing, trading (or selling), and abandoning land. They all sold keepers, normal offices, and Maebashi Munici some pieces of their land between 1980 and pal Office also increased. In general, the origi 1993. For 20 owners their behavior was trad nal land owned by farmers was replaced by ing only; a combination of trading and utilizing other agents newly entering the area, such as was taken by four owners; three owners took a office workers, shopkeepers, and factory own combination of trading and abandoning; three ers. owners took a combination of utilizing, trading, and abandoning. (Figure 9). The landowners Land-use Decisions of Landowners can be divided into four groups based on their behavior: (1) trading, (2) trading and utilizing, In the approach used so far in discussing the (3) trading and abandoning, and (4) trading,

359 112 TSUTSUMI J.

Figure 9. Land-transfer time for landowners and their land-use in the Rokku area of Maebashi city for 1980 and 1993. utilizing and abandoning. The 'trading' group more effectively. of 20 owners can be further subdivided into five ' Trading 2' includes seven owners and is categories based on the causes of their deci characterized by owners who sold some part of sions. their land to pay inheritance tax. These seven ' Trading 1' includes six owners and is charac owners consisted of one full-time farmer, five terized by owners selling some part of their part-time farmers, and one non-farmer whose land because the impending land re-adjustment farmland was rented to another farmer. The program was expected to adversely affect their full-time farmer had been a non-farmer before farming management.9 They managed rice he retired from his regular job. After retire farming, in combination with wheat growing, ment he turned to full-time farming. The char sericulture and truck farming. Since their pri acteristics of this type of owner were that most ority was farming, they decided to sell their of them combine vegetable farming with rice land to purchase land in the rural farmland area growing as it is easy to keep with a non-farm (several kilometers away) that could be farmed regular job. After land inheritance, they sold

360 The Land Conversion Process 113

Figure 10. Land-use decision making processes of landowners in the Rokku area of Maebashi city dur ing 1980 to 1993.

'Trading 5' con only the minimum amount of land needed to sists of two non-farm land pay inheritance tax, and tended to keep their owners who have sold out their own land due to former farming management. the failure of their non-farm businesses. Before ' Trading 3' includes three owners and is char they started the businesses, they had been acterized by owners who sold some part of their farmers with more than 2,000 square meters of land due to requests to sell in accordance with farmland. rapid increase in urban land demand. Two of ' Trading and utilizing' includes four owners. the three owners have kept their former farm The decision-making processes of this type of ing management for the remainder of their landowner are similar to those of the 'Trading land. The other owner, who was a non-farmer 1' type, but differed as to whether they utilize (owner number 16 in Figure 9), sold all of his some part of their land or not. They also sold land in the Rokku area (about 3,000 square their land to purchase farther rural land where meter) to Maebashi Municipal Office which was they could farm more effectively . Moreover, the conductor of the land re-adjustment pro they have constructed some buildings , e.g., gram. apartment houses, shops and parking lots in the ' Trading 4' includes two owners and is char Rokku area, so that they could supplement acterized by owners who sold some part of their their income by renting such properties . That land due to need for a larger income . Landown is, owners in this type adapt the way of land-use ers in this type are full-time farmers growing to suit the occasion. only rice. Since their farming managements ' Trading and abandoning' includes three were simple and were not efficient enough, the owners. Their reasons for selling land were income was not sufficient for them. inheritance, just the same as those of the 'Trad-

361 114 TSUTSUMI J. ing 2' type. In addition, they abandoned some The initial factors were land re-adjustment, part of their land as well as selling. The differ land inheritance, need for a larger income, re ence between these two types was their attitude quest to sell their land, and failure of a non to farming management. In the 'Trading and agricultural business. Decision factors were the abandoning' type, only one or two persons con existence of their successors in agriculture, pay tinued farming in each family, while other ment of inheritance tax, intention to keep farm members have non-farm regular work. After ing, possession of land with easy access to the death of a farmer, usually a head of the roads, and their own desire to utilize the land. family, they cannot keep their former farming Based on their decisions, their behavior can be management. In this case almost all their land divided into three categories: trading (or sell for cultivation was abandoned and turned to ing), utilizing, and abandoning land (Figure 10). wasteland. Some farmers in the sample area made land Finally, 'Trading, utilizing and abandoning' use decisions based on their intentions to keep consists of three farmers. The decision-making farming. They sold part of their land to pur processes of this group of owners was the same chase new land several kilometers away in ru as those of the 'Trading and abandoning' type, ral farmland areas that could be farmed more but there was a slight difference in the combi effectively. Moreover, if the intention to keep nation of their behavior: They also chose utili farming was high, even when the head of the zation. When the head of the family died, the family died, the surviving members generally decision to rent apartment houses or parking sold only a part of their land to pay inheritance lots was taken by the surviving family mem tax while managing the remaining land for bers because they had no other source of in farming as before. On the other hand, some come-either they had no other job, or could combinations of behavior can be seen in the not farm the land. group where the intention to farm was low. The landowners in the sample area made When the head of the family died, they tended land-use decisions based on two types of fac to utilize, abandon, or sell their land. For exam tors: the initial factors and decision factors. ple, if their desire was to utilize their land, some

Figure 11. Distribution of converted land in the Rokku area of Maebashi city between 1980 and 1993.

362 The Land Conversion Process 115

portion with easy access to roads tended to be which micro-level examination of land-use converted to urban use, such as apartment analysis was performed. The Rokku area is a houses, shops and parking lots. If utilization newly developed residential area adjacent to was not desired, portions of land that were the existing built-up area of Maebashi . In order highly valued tended to be sold, while lowly to obtain land-ownership data, land assessment valued portions tended to be abandoned. In rolls housed in the Property Tax Division of case of non-farm landowners, some principal Maebashi City Hall were used. The selection of reasons for selling were at the request of a landowners in the urban land market was based second party; and the failure of their non-farm on land-trade and land-inheritance data from businesses. the rolls. The data of 1980 and 1993 were The distribution of land converted by land examined for the sample areas . The land-use owners for the three types of behavior is shown data for the area was mainly obtained by the in Figure 11. Most of the traded land was author's field survey, while some parts were converted to urban use such as housing, park supplemented by aerial photos, topographic ing or shops through small real-estate firms. maps, and residential atlases. To gather infor Utilized land, which has been converted from mation on the decision-making processes, the rural to urban use by its owners, was distrib author interviewed the selected landowners. uted along the main roads leading to central Thirty landowners were selected as interview Maebashi and close to the owners' houses. ees whose causes, conditions and behavior in Land abandonment occurred as a result of an regard to land-use decisions was closely exam ongoing agricultural labor shortage (mainly ined. Twenty-six of thirty landowners were caused by inheritance); and as a countermea farmers and the rest were non-farmers. During sure against the reduction of land due to land the interview, landowners were asked about re-adjustment. Most of the landowners, espe their reasons for disposing of landholdings. cially office workers who owned residential Based on the analysis of the land-use deci property in the sample area, purchased small sions of landowners in the sample area, the lots and left them idle. Lots that have been characteristics of the land conversion process reduced in size due to land re-adjustment can of the city is schematically showed in Figure then be combined as all idle lots, so that owners' 12. In general, the built-up area of the city loss of land due to land re-adjustment is mini gradually expands outwards in accordance mized. Abandoned land became scattered with urban growth. In this section the author among other land-uses resulting in urban emphasizes the impacts of land re-adjustment, sprawl. Utilized land and traded land were urban land demand, and several types of land based on different decisions by the landowners. - use decisions as major factors for land-use Much of the land in the inner fringe was in change. demand. Thus, both land being utilized and During Ti, the extent of the built-up area of land being traded tend to be converted to urban the city was limited to the thin strip enclosing uses. the CBD. Almost all of the part of the sample area of Rokku was still in a rural area at this Conclusion stage. Thus, urban land demand was not so high in the sample area. Due to a low level of The objective of this paper was to analyze the urban land demand, the behavior of landowners land conversion process at a micro level by was limited to only two categories: selling and investigating the decision-making processes of abandoning land. Some landowners in the sam landowners in the rural/urban fringe by under ple area sold their land along major roads of taking a case study in Maebashi city. relatively high urban land demand. Others just The city is a medium-sized city with a popula abandoned land with no access to roads and of tion of approximately 280,000 with only one relatively low demand. Toward T2, in accor core developed area. The Rokku area in the - dance with increasing urban land demand, land inner fringe was selected as a sample area in trade from farmer to non-farmer gradually in-

363 116 TSUTSUMI J.

Figure 12. Schematic model of the land conversion process in the rural/urban fringe of Maebashi city.

creased especially along the major roads and petition was observed between the types of resulted in the built-up area of the city extend land-use in this area, due to centrifugal land-use ing into the sample area. from the CBD and centripetal use from the During T2 and T3, the sample area was in a countryside. At this time, road networks were typical rural/urban fringe environment. Com improved and urban land demand in the area

364 The Land Conversion Process 117

became generally high. As a result, there were areas and have been documented by authors many land-use options open to landowners in such as Sawa (1990). the area and much of the land in the sample 4. According to certain statistics of urban plan ning in 1990, which covered 80 major cities in area had a tendency of being converted to ur Japan with over 2,000m2 ADFU, the highest ban use. proportion of land developed under re During T3, a serious problem of urban sprawl - adjustment works to the total ADFU was had occurred in the area. At this stage the local achieved in city (74.2%). Behind Fukui city, Kasugai city (northern suburb of government, for example, City Office, tended to city) (66.9%), city (62.4%) and other take some infrastructural investment program cities followed. Maebashi city was the sixth such as land re-adjustment to solve the prob highest in Japan. lem. Along the newly built roads urban land 5. 'Other uses' includes some huge artificially lev demand became much higher and thus much of eled lots with very little vegetation such as the land tended to become urban. This type of schools, factories, residential estates in prepara tion, and sports stadiums. behavior taken by the local government is iden 6. In the data set, the varied possible urban uses tified as important because it facilitates devel of land were partitioned into only three catego opment for the entire area, including land with ries: 'building uses,''traffic facilities,' and 'other out good access to roads. uses.' In this study, the author regards these Finally during T4, in the near future, the three categories together as 'urban use.' 7. The sample area of Rokku (about 45.8ha) are built-up area of the city will further expand. At all included in the projected area for the Rokku this stage, even abandoned land at the earlier land re-adjustment program (133.8ha). stages of development will tend to be converted 8. Morimoto (1991, 1993) are good examples of to urban use due to improved urban infrastruc studies analyzing the increasing process of ture. Thus, almost all land in the sample area land abandonment. 9. According to the author's interview reports, will be merged into a completely urban area. farmers who sold some part of their land could purchase new land of more than three times the Acknowledgments area in rural lots several kilometers away, where they could continue effective farming The author is grateful to Dr. Steven Phillips at the management. National Institute of Advanced Industrial Science and Technology (Japan) for his useful comments during References the preparation of the English manuscript. In addi tion, appreciation is given to Mr. Tadashi Fukuda and other officers in Maebashi City Hall, who helped me in Bito A. 1992. Preparation and use of Digital Land many ways by providing support, materials and in Information of Japan. In Tochiriyou henka to sono formation; and the many landowners in the Rokku mondai (Land-use changes and its problem), ed. area, who kindly explained their long-term decision Himiyama Y. and Okamoto J., 41-57. Tokyo: Tai - making processes on their own properties. meido. (J) Bryant, C. R., Russwurm, A. G., and McLellan, A. G. Editor's Note: This article was first published in 1982. The city's countryside, land and its manage - ment in the rural-urban fringe. London: Longmans. Japanese in the Geographical Review of Japan 68A, 721-740, 1995. Doi H. 1984. Land-use change in the rural-urban fringe: A case study of Matsuyama City. Human Geography 36: 1-21. (JE) Notes Fukuhara M. 1971. Urbanization of Urayasu under the rapid national economic growth. Geographical 1. Some notable examples of land-use studies Review of Japan 44: 751-764. (JE) from the viewpoint of land-use regulations in Ganderton, P. 1994. Modelling the land conversion clude Doi (1984), Katayanagi (1994), Kitamura process: A realist perspective. Environment and et al. (1980), and Nagashima (1976). Planning A 26: 803-819. 2. For example, Kikuchi and Moran (1990), Golledge, R. G., and Stimson, R. J. 1987. Analytical Kikuchi (1994), and Kikuchi (1995). behavioral geography. New York: Croom helm. 3. Lewis and Maund (1976) proposed a conceptual Harada T. 1976. Transfer of land ownership and framework to analyze particular types of rural change in land use in the suburban farming set communities invaded by urban newcomers. In tlement-Case study of Ichikawa Municipal area Japan, such communities are known as konjuka adjacent to Tokyo-. Geographical Review of

365 118 TSUTSUMI J.

Japan 49: 616-631. (JE) Review of Japan 49: 314-326. (JE) Katayanagi T. 1994. Tokyo-to Adachi-ku ni okeru Nakamura Y. 1995. Cognitive land conditions and tochiriyou no henka to youtochiiki no taiou agricultural land use in the Chichibu Mountains. (Land-use in Adachi ward, Tokyo and its reac Geographical Review of Japan 68A: 229-248. (JE) tion). Quarterly Journal of Geography 46: 107-125. Pond, B., and Yeates, M. 1993. Rural/urban land con (J) version I: Estimating the direct and indirect im Kikuchi T. 1994. Land use change and its character pacts. Urban Geography 14: 323-347. istics in the Auckland city's countryside, New Pond, B., and Yeates, M. 1994a. Rural/urban land Zealand-A case study of the Clevdon area, conversion II: Identifying land in transition to Manukau city-. Journal of Geography 103: 377 urban use. Urban Geography 15: 25-44. - 397. (JE) Pond, B., and Yeates, M. 1994b. Rural/urban land Kikuchi T. 1995. Sustainable changes of agricultural conversion III: A technical note on leading indica land use in New Zealand. Geographical Reports of tors of urban land development. Urban Geography Tokyo Metropolitan University 30: 147-156. 15: 207-222. Kikuchi T., and Moran, W. 1990. Land use change Pyle, L. A. 1986. Sellers and nonsellers of land for and its characteristics in the Auckland metropoli new rural housing. Professional Geographer 38: tan region, New Zealand-A case study of East 343-349. Tamaki area, Manukau city-. Geographical Re Pyle, L. A. 1989. Persistent landownership at the view of Japan 63A: 741-765. (JE) rural-urban fringe. Urban Geography 10: 157-171. Kikuchi T., and Tsutsumi J. 1998. Sustainability and Sawa M. 1990. The social structure of 'rurban vil changeability of agricultural land use in the pen lages' in metropolitan area, Japan. Geo urban environment: A case study in a sericulture graphical Review of Japan 63A: 653-675. (JE) region in Maebashi city, central Japan. Quarterly Takahashi N. 1968. Urbanization due to industriali Journal of Geography 50: 1-16. (JE) zation in the region of Mishima-, Shizu Kitamura Y., Ueno K., and Ishida N. 1980. The infl oka prefecture. Geographical Review of Japan 41: uence of the City Planning Act to the urbaniza 1-18. (JF) tion area in Hadano city, . Takahashi N. 1977. Urbanization in Shimizu city, Geographical Review of Japan 53: 463-474. (JE) prefecture. Journal of Geography 86: Lewis, G. J., and Maund, D. J. 1976. The urbanization 241-264. (JF) of the countryside: A framework for analysis. Tanaka K. 1982. Changes in landownership and land Geografiska Annaler B 58: 17-27. use of former farms in Nakano-ku and Musashi Morimoto T. 1991. The increase of fallow abandoned no-shi, Tokyo. Geographical Review of Japan 55: cultivated land with the development of inten 453-471. (JE) sive agriculture in Hasaki-machi, Ibaraki prefec Tsutsumi J. 1996. Land transaction and following ture. Geographical Review of Japan 64A: 613-636. land use change in the central area of (JE) city, Japan. Annals of the Japan Association of Morimoto T. 1993. Idle and abandoned farmland in Economic Geographers 42: 118-131. (JE) Kashiwai-cho 4-chome, Ichikawa city, pre Tsutsumi J. 1999. Land Conversion due to Decision fecture. Geographical Review of Japan 66A: 515 Agents in an Urban Land Market-A Case Study -539. (JE) of Maebashi City, Japan. Geographical Review of Nagashima H. 1972. Non-agricultural land-use by Japan 72B: 23-47. farmers in Soka city. Human Geography 24: 38-58. (JE) (J): written in Japanese Nagashima H. 1976. Urban development and its in (JE): written in Japanese with Englis abstract fluence on rural areas around cities. Geographical (JF): written in Japanese with French abstract

366