Shrinkage Does Not Follow Population Decline on a Regional Scale: Planning and Reality of Residential Area in Japan

Abstract

This study first examines changes in the amount of residential area on a regional scale, which is said to shrink following population decline, using Japan as an example, where population decline is already underway in most of the country. Second, it examines how the regional governments have set future targets for residential areas and whether these targets are consistent with the current trend. Third, it examines how prefectures, Japan’s regional governments, are responding to this situation through regional land use planning. The methodologies of the paper are a combination of collecting and sorting out the related data, reviewing and extracting information from planning documents, and conducting interviews with persons in charge. In conclusion, residential areas have continued to increase on a regional scale, even as the population has declined. Shrinkage does not follow population decline on a regional scale. It was also found that the amount of residential area changed more closely with the change in households than with the population. National and regional land use plans generally set targets for more shrinkage than the actual trend. The underestimation of the amount of residential area was due to the rapid decline in the number of persons per household. Cases of the three prefectures showed that specific control is left to the individual acts, while the measures are indicated in the wording by the prefectural land use plans.

1. Introduction and Purpose

1.1. Research Background

It has been taken for granted that cities will grow and expand their residential areas, and modern land use planning systems have been established in various countries in response to such urbanization and population growth [1]. In recent years, however, many cities and regions have seen their populations stagnate or decline. The term “shrinking cities” is representative of this tendency, and related research has been accumulated [2,3,4,5]. For example, Mallacha et al. [5] conducted a comparative study of shrinking cities in Japan, the U.S., and Germany and summarized their similarities and differences.

The shrinking phase gives the land use planning system a different role for controlling urban and residential areas. There are various efforts and associated policies at the micro level in municipalities or districts to reduce the size of urban and residential areas. Typical examples include efforts in East Germany to shrink apartment complexes in response to population out-migration due to the worsening employment situation following the end of the Cold War [6] and land banking efforts to stock vacant land in declining areas in the United States [7,8]. In these cases, the demand for residential area decreased as the population declined, and measures were taken to address these phenomena.

However, when we look at this phenomenon on a wider scale, we see a different situation from the relationship between population decline and shrinkage than in small neighborhoods or districts. As this study demonstrates, population decline does not necessarily lead directly to a decrease in residential area. Even is the population has declined, the amount of residential area may increase. Since each dwelling unit mostly corresponds not to an individual person but to a household, the amount of residential land is considered to be proportional to the number of households, not the population.

Despite this situation, studies on shrinking cities to date have mostly used population as an indicator, and few have focused on households to conduct specific research. In addition, most of the specific planning for shrinking cities has been done at the micro level, and few studies have taken a macro and regional perspective. A region or a country where population and households are declining over the long term needs land use planning and a control system to fit the new condition, which was not anticipated when the system was developed in the era of population growth and urbanization. Integrated control at the broader regional level should coordinate and, if possible, optimize individual efforts to address the various measures associated with population and household declines and to maintain the sustainability of urban areas in a whole city or region.

1.2. Purposes of the Study

Based on the discussions above, this study first examines changes in the amount of residential area on a regional basis, which is said to shrink in a phase of population decline, using Japan as an example, where population decline is already underway in most of the regions. Second, it examines how regional plans set future targets for residential areas and the degree to which these targets are consistent with the current trend. Third, it examines how prefectures, Japan’s regional governments, are responding to this situation through regional land use planning.

This study examines the existing fixed idea that a declining population results in declining residential areas based on an empirical analysis. The results of the paper will clarify the necessity of taking changes in households, rather than population, into account in regional land use planning during a period of population decline. The paper also points out the current status and challenges of the national land use planning system, which is responsible for regional land use planning in Japan.

Japan has the world’s most advanced population decline in the country as a whole, where most prefectures are experiencing population decline [9]. Under Japan’s land use planning system, plans are created at three levels: national, prefectural, and municipal. Prefectures that have land use plans in place identify the amount of residential land throughout the prefecture and set targets for this amount. Thus, it is appropriate to study the planning and reality of residential areas in prefectures with declining populations in Japan to clarify the issues of land use planning of so-called shrinking cities and regions.

1.3. Methodology

After performing a comprehensive review of existing studies, discussions, and critiques on regional land use planning system and urban shrinkage of both global and Japanese-specific point of views, research was conducted by collecting and sorting related data, reviewing planning documents, and conducting interviews with persons in charge. Regarding the statistical data on the changes in population, households, and land use, the land use-related data described below were ordered from the national and prefectural governments as necessary for analysis, in addition to statistical data available on the internet, such as the National Census and Basic Resident Registers, which are generally open to public. To compare the targets of plans with the current trends, plan documents of prefectures were collected both from their websites and by e-mail when not available to the public. E-mail interviews with persons in charge were used to inquire about how targets were set and achieved by the specific prefecture.

2. Existing Studies

2.1. Research Trends in Land Use Planning Systems and Regional Land Use Planning

2.1.1. Studies of Shrinking Cities and Japan’s Positioning

Originally, the mechanism of declining demand for land use through a decrease in population and households was recognized during the period of urbanization and rapid population growth. Klassen, who is well-known for his spatial cycle model, mentioned it long before the term “shrinking city” had been coined [10]. He predicted the shrinking process in terms of a shrinking population (less demand for services, less traffic, more vacant houses, and bankruptcies), fewer households (indicators related not only to housing but also furniture, cars, etc. are affected) and less demand for land (treatment of vacant land and financial problems). However, it seems that it was not until the 2010s that the term “shrinking cities” began to be frequently observed in the studies of several cities around the world. Many studies have been published comparing various shrinking cities in the U.S., Europe, and Asia [2,3,11,12].

Japan has long been noted as the country with the fastest population decline among developed countries due to aging and subsequent natural attrition [13]. According to estimates based on the Population Census (Census of Population), the population of Japan as a whole began to decline in 2008. However, many municipalities in the rural areas of Japan were already in the process of population decline. Japan’s position in studies regarding shrinking cities is, according to Mallacha et al. [5], characterized by a gradual but nationwide movement due to demographic trends.

Grossmann et al. [3] characterized Japan as having low fertility rates, an aging population, a high concentration in large cities, and widening economic disparities, which distinguish it from the shrinking situation in Western countries due to the decline of local industries and major political events such as the end of the Cold War. As reported by Zetter [14], Japan’s population decline was also originally limited to certain cities with declining economies. In the 21st century, however, as Matanle [15] pointed out, even core cities outside rural regions have begun shrinking and declining. Furthermore, Buhnik [16] and Seta [17] reported on the case of shrinking cities in the Osaka metropolitan area, which is recognized as a megacity in the world. Population decline in Japan is now already a nationwide trend.

2.1.2. Urban Shrinkage and Regional Land Use

In terms of land use on a regional scale, many studies on population decline have focused on the relationship between urban centers and suburbs rather than on Japan or its regions as a whole or on specific local areas or prefectural governments. Recently, Kato and Takizawa [18] analyzed the population change caused by the coronavirus pandemic in the shrinking Osaka metropolitan area, which they termed the “urban exodus,” and found that a clear population increase occurred in the municipalities on the fringe of the metropolis from summer to fall 2020.

As Hebbert [19], Sorensen [20], and Millward [21] have long pointed out, land use regulations in Japan have generally been lax, allowing urban sprawl in the suburbs. This has led to the current situation pointed out by Yokohari et al. [22], in which the boundaries between urban and agricultural land is blurred in Japan. In other words, looking only at the interior of an urban area does not provide an accurate picture of the situation of residential areas. In addition to the discussion of “perforated cities” [4,23] observed in Germany, where the boundaries of urban areas are clearly defined, a broader view is needed.

Apart from these studies, quantitative analyses using various types of data, mainly satellite data as in Ohashi et al. [24], have emerged in studies of land use beyond the metropolitan area. However, no one has yet discussed the relationship with the regional land use planning system, specifically identifying residential areas rather than the entire developed areas obtained from satellite and other data.

In addition, there is little existing literature that strongly relates households to the problem of urban shrinkage due to population decline, which will be analyzed later in this paper. Hartt and Hackworth [25] examined what types of household size declines are more acute in shrinking core cities than other locales like suburbs in the U.S. and found that elderly households and households with school-aged children were under-represented in shrinking cities, while households with pre-school-aged children were over-represented. They argued that plans and development strategies for shrinking cities should reflect a range of demographic changes, including outmigration and internal household restructuring. Hijino [26] proposed some guiding principles based on the analysis of policies for attracting young households and closures and mergers of public schools of four cities in Japan. However, the author did not address the land use situation associated with shrinkage and exclusively referred to population and household dynamics related to their own issues.

2.1.3. Regional Planning for the Phase of Population Decline

According to Metternicht [27], who analyzed the role of land use and spatial planning tools, processes, and approaches with numerous examples from around the world, land use planning evolved from a top-down, expert-driven approach in the 1960s and 1970s to a different approach to improve socio-economic opportunities through sustainable management of local land resources.

The transition in land use planning is considered to have led to an overall decline in the importance of regional planning in many countries, as it is difficult to take even local suitability and characteristics into account. In recent years, Harrison et al. [28] stated that “regional planning is dead”, and there have been various discussions about the significance of regional planning [29,30].

However, even in recent years, interest has been expressed and various studies have been conducted, not only in specific areas, but also regarding the objectives and overall system of regional planning [30,31,32]. In light of these and other discussions by Purkarthofer [33], regional planning is becoming strategic, soft, and consensus-based rather than normative, as it was in the past.

Recent case studies of regional planning in the phase of population decline include a comparison of national and regional level planning in Central and Eastern Europe by Nowak et al. [34] and Nowak et al. [35]. While the discussions and plans addressed in these studies include the land use sector, there is little mention of population decline in the studies.

2.1.4. Studies on Regional Planning for Shrinking Cities in Japan

Although the boundary between urban and rural areas is blurred in Japan, research on land use planning in Japan tends to deal with specific cities or metropolitan areas, regardless of central or rural regions.

Tateishi et al. [36] focused on the governance systems of mega-cities that were ahead of the shrinking Tokyo suburbs, noting that their strategies included both growth orientation and shrinkage response, but there were problems with implementing inter-sectoral and inter-municipal coordination in growing metropolitan suburbs. Hoshi [37] described the long-term evolution of land use planning and the operation of the land use planning system in Sapporo, focusing on the actual population distribution and urban area formation at each stage. The author concluded that the basic tenets of urban planning have been consistent for almost half a century.

In addition, the location optimization planning system, which aims to consolidate urban areas and form compact cities, was introduced in 2014, and many Japanese cities have subsequently developed these plans. Many discussions and papers on this location optimization planning system have been published, such as Wang and Homma [38] and Yoshida et al. [39]. However, location optimization plans are currently formulated by single municipal governments, and there is not a case of plural municipalities.

No papers seem to have studied land use planning during a period of population decline for an area larger than a single metropolitan area, such as an entire region or a country.

2.2. Regional Land Use Planning System in Japan

Next, this section provides an overview of Japan’s regional land use system and reviews general critiques of the system.

2.2.1. Basic Characteristics of the National Land Use Planning (NLUP) System

Most of the research on the land use planning system in Japan is concerned with the systems and projects implemented under the City Planning Act within city planning areas. However, when discussing the regional scale, it is necessary to include other areas than the 997 city planning areas [40] throughout the country, which subdivide actual urban and metropolitan areas and cover only 27% of the national land area. The National Land Use Planning (NLUP) Act enacted in 1974 covers all national territories of Japan and stipulates plans for all three tiers of government, i.e., central, prefectural, and municipal governments.

The main characteristics of Japan’s regional land use planning system based on the NLUP Act can be summarized based on the text of the law (Japanese Law Translation) [41]; an explanation by the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) [42]; and discussions by Nakamura [43,44].

First, the NLUP system is a hierarchically constructed system. Figure 1 shows the basic structure of the NLUP system drawn by the author, based on the articles of NLUP act and related acts.

National land use planning corresponds to a three-tier system of government. The NLUP Act stipulates that prefectural plans (NLUP-PPs) shall be based on the national plan (NLUP-NP), and municipal plans (NLUP-MPs) shall be based on the NLUP-PP of their prefecture. Additionally, opinions shall be obtained from the prefectures when formulating the NLUP-NP and from the belonging municipalities when formulating the NLUP-PPs.

It also guarantees a certain degree of democratic procedures. The NLUP-NP requires a cabinet decision. For NLUP-PPs and NLUP-MPs, the members of the councils involved in their enactment are appointed with the approval of each prefectural or municipal assembly.

On the other hand, the NLUP system has inherent flexibility and weak control over the actual planning contents. First, the means to realize the content stipulated in the NLUP Act are indirect. The most important means of realization of the NLUP-NP and NLUP-PPs are the five areas designated in the Land Use Master Plan (LUMP) formulated at the prefectural level: urban areas, agricultural areas, forest areas, natural park areas, and nature conservation areas. Each area is specifically planned and controlled based on individual corresponding acts. Although the LUMP of a prefecture is stipulated to be based on the NLUP-NP and the NLUP-PP of the corresponding prefecture, the influence of these on actual land use control is indirect and left to individual acts.

Among the three tiers of the NLUP system, the NLUP-NP is required to be enacted as an integral part of the National Plan of the National Spatial Strategies based on the National Spatial Planning Act, while the NLUP-PPs and NLUP-MPs are voluntary plans. Since the NLUP Act was enacted, 47 prefectures have formulated their own plans, but after the Fifth NLUP was formulated in 2016, only 24 prefectures, or about half of the 47 prefectures, have formulated NLUP-PPs. In prefectures where NLUP-PPs are not formulated, the LUMPs are formulated based solely on the NLUP-PP. Regarding the municipal plans, 844 or merely 48.5% of the total 1741 municipalities in Japan have formulated municipal land use plans [45].

The contents of the plans by the NLUP system differ among tiers. The NLUP-NP and NLUP-PPs consist of a statement of existing and preferable land use and current and target amounts for each land use category. Although specific land use zoning based on maps and other information is not excluded in the NLUP Act, such content is not included in any plans in practice. On the other hand, some NLUP-MPs draw specific land use zoning maps in some municipalities. In such cases, the regulatory power of the NLUP-MPs is not explicitly secured by the NLUP Act but mostly provided individually by bylaws or other original guidelines of these municipalities.

2.2.2. Existing Critiques of the NLUP System

The above basic characteristics of the NLUP system in Japan have been introduced by, for example, Hebbert [19], who regarded it as a loose land use control framework. Ishida [46] critiqued the NLUP-NP but did not evaluate NLUP-PPs or NLUP-MPs.

In terms of a hierarchy of planning systems, the NLUP system can be considered as centralized in comparison to systems of many other developed countries described in ESPON [47]. As for the public participation in the enactment of the plans, characteristics of procedural planning instruments defined by Dühr [48] or of participatory systems discussed by Metternicht [27] are considered to be lacking in the NLUP system, since the public involvement is limited.

Nakamura [43], however, positively evaluated the feedback mechanism of each of the plans based on the different scales described above. Nakamura [44] also evaluated the flexibility of the system, as many cross-sectional issues of land use which cannot be specifically addressed by the individual departments in charge are described in the NLUP-NP and NLUP-PPs. He suggested, for example, that the NLUP system can be used for the consolidation of residential areas by coordinating respective controlling measures created by individual acts for urban areas and agricultural areas [44]. However, as mentioned above, statements to identify the specific areas for controlling land use are not observed in the NLUP-NP and NLUP-PPs.

Applying this to discussions by Berisha et al. [31] and Nowak et al. [35], it can be said that these plans have a strategic element but not a regulatory element that is responsible for specific control of land use. Contrasting this with the Italian case reported in Todella et al. [49], where individual regulations are integrated at the end in Italy, only individual controls are applied in Japan according to the five areas recognized by LUMPs in respective prefectures.

Several international studies in recent years on land use planning systems have evaluated them along conceptual axes such as conformative or performative [31] or whether they are substantive or procedural [48]. It can be said that the NLUP system in Japan is substantive or breakdown-based, judging from the characteristics of the planning system from the national government to the municipalities, and is performative or outcome-oriented in the sense that it addresses specific issues described in the wording of the plan. However, the system alone is regarded as incomplete as a land use planning system because it leaves specific control to individual acts, as described above.

Jain and Arai [50] argued that there is pressure to review the planning system created during Japan’s population growth phase, including the NLUP system, in order to restructure Japanese society in an era of aging and a declining population. However, the NLUP system continues today as it has since it was established in 1974, with no major changes.

3. Transitions in Population, Households, and Residential Areas on a Regional Scale

3.1. Data on Residential Areas

Regarding data for residential areas, the Land Use Status Survey (LUSS) [51] is used as the basis for the current status and target values in the NLUP system. These data were collected in this paper by reaching out to related sections of prefectural governments. The LUSS is based on various existing statistics to determine the area of land use. For residential areas, data from property tax records were used to determine the amount of residential area owned by the private sector, and areas used for public housing and portions of residential use in village districts were acquired from some other surveys.

National Land Survey Data [52], which are widely used for land use studies in Japan, can only determine whether the land is used for construction or not, while the amount of residential area is unknown. Data on land use subdivided by detailed kind of use, including residential area, are published based on the Urban Planning Basic Survey as stipulated in the Urban Planning Act but are limited to the urban planning areas. Some prefectures and municipalities conduct more detailed surveys by combining data and statistics on their own, but whether or not they do so, and the content of their surveys, varies between prefectures and municipalities.

3.2. Overview of Land Use in Japan

Looking at Japan’s total land area (Figure 2) using the LUSS, forests cover 250,000 square kilometers or 66% of the country’s land area, and farmland covers 44,000 square kilometers or 12%. The next largest area is developed land, defined as “land necessary for the site of a building and for the maintenance or utility of the building”, including residential and industrial area, accounting for 5.2% or 20,000 square kilometers. Other types of land are roads, water surfaces, wilderness, and others.

Among developed land, residential areas, the subject of this study, account for 12,000 square kilometers, or 61% of the total developed land. The rest is industrial areas, commercial areas, public land, and others (Figure 3).

Long-term trends across the country related to residential areas are shown in Figure 3, including the whole developed land, residential areas, and roads. The figure also includes the trends in population and households as indicators of urban shrinkage. The data for land area were extracted from the LUSS using 1985 as the base year, when the survey’s current categories of land use were first used. For population and households, the five-year National Censuses were used, and the interpolated corrected population calculated related to the censuses is also shown.

The population of the whole country has continued to decline from 1.06 in 2008 to 1.04 in 2020, when the 1985 figure was 1.00. The number of households, however, has continued to grow since 1985, reaching 1.46 in 2020. Developed land, residential areas, and roads all continue to grow at a similar pace, though not as fast as households, with residential land at 1.30 in 2020.

4. Planning and Reality of Regional Land Use in the Phase of Population Decline

4.1. Enactment of NLUP-PPs and Changes in Population and Households

Next, I compare plans and targets of residential areas under the NLUP-PPs of some prefectures with declining populations and actual conditions.

As noted above, it is not mandatory for a prefecture to enact an NLUP-PP. Originally, all prefectures had formulated this plan, but since 2015, when the Fifth NLUP-NP was formulated, only 25 prefectures, or about half of the 47 prefectures in Japan, have enacted it. Figure 4 shows the relationship between the formulation of these plans and changes in population and households based on the information and data the author collected. Population has decreased in 41 prefectures, but only two prefectures have seen a decrease in households. There seems to be no relationship between whether or not an NLUP-PP is enacted and changes in population or households.

4.2. Targets of NLUP-PPs Related to Residential Areas and Their Achievement

4.2.1. Population and Household Trends and Targets of NLUP-PPs

Here, residential areas targets in the NLUP-PPs are compared to the actual changes as measured by the LUSS. In addition, how residential lands have changed relative to the population and households is examined.

In this section, population and household figures based on the Basic Resident Register, which is updated annually, are used for prefectural population and households rather than the census conducted every five years. The population by prefecture is larger for both population and households in municipalities with declining populations because the Basic Resident Register do not adjust the number of residents who have moved but don’t change their registration, while the census directly asks citizens about their place of residence and household status at the time of the census. However, in 2010, only about 1% of all municipalities had an error margin of more than 10% [53]. Looking mainly at the transition here, this paper does not regard this as a serious problem.

First, the target and actual transition of residential areas in the whole nation is shown in Figure 5. The Fifth NLUP-NP was enacted in 2015, and it stipulated that the residential area, which was 11,600 square kilometers in 2012, the base year, should be the same in 2024, the target year. The NLUP-NP describes that the development of residential areas will be curtailed while also “creating a favorable living environment” for residential areas.

The actual trend for residential land has increased from the base year of 2012 to 1.03 in 2020, the latest year for which information is available. During this period, the population has continued to decline, while the number of households has continued to rise. The population based on the Basic Resident Register was based only on Japanese residents until 2012 and began to include foreign residents in 2013, so this paper shows the Japanese population based on the base year of 2012 and the total population, including foreign residents, based on the base year of 2013. The same trend is seen in both cases.

The Fifth NLUP-NP assumed that the number of households would decrease during the planning period quoted above, but in fact, it has continued to increase. While the actual amount of residential area shows a trend similar to the increase in households, the target amount appears to be lower, dragged down by the decline in population.

4.2.2. Cases of Prefectures Setting Decreasing Targets for Residential Areas

Next,

Table 1. Base year amounts and targets for residential areas in the Fifth NLUP-NP and subsequent NLUP-PPs.

Prefectures	Condition of Base Year				Actual Amount of Residential Area of Base Year, Year of Enactment of Plans and Planned Amount of Residential Area of Target Year (See Note)																						Residential Area of Target Year
																											Amount (Base Year = 100.0)	Planning Period (Years)	Annual Rate of Increase/Decrease (%)
					Unit: Hundred ha (for National Plan (5th) Ten Thousand ha)
	Population	Annual Change	Households	Annual Change	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024	2025	2026	2027	2028	2029	2030	2031	2032	2033
National (5th)	128,226,483	−0.4%	56,412,140	0.7%	116			*									116										100.0	18	0.0%
Nagano	2,145,962	−0.4%	825,012	0.7%	327				*									338									103.4	13	1.1%
Iwate	1,309,009	−0.7%	509,702	0.7%		236			*									237									100.4	12	0.9%
Niigata	2,364,632	−0.6%	859,516	0.6%	310					*								312									100.6	13	1.0%
Kyoto	2,542,740	−0.2%	1,132,893	0.7%	161					*								161									100.0	13	0.0%
Gifu	2,058,652	−0.5%	765,098	1.0%		247				*								247									100.0	12	0.0%
Aomori	1,368,246	−1.1%	579,497	0.4%		200				*									185								92.5	13	−1.2%
Osaka	8,673,898	−0.1%	3,990,017	0.7%		344				*										353							102.7	14	1.1%
Hokkaido	5,463,045	0.0%	2,727,383	0.7%			607			*								607									100.0	11	0.0%
Hyogo	5,655,361	−0.1%	2,460,392	0.5%			378			*								379									100.4	11	0.9%
Yamanashi	861,615	−0.3%	350,033	0.5%			118			*									124								104.8	12	1.1%
Shiga	1,421,779	0.2%	548,478	0.9%			153			*										157							102.6	13	1.1%
Shizuoka	3,803,481	−0.2%	1,518,772	0.6%			369				*								369								100.0	12	0.0%
Chiba	6,254,106	0.1%	2,735,874	1.2%				527			*							551									104.6	10	1.2%
Akita	1,056,579	−1.3%	425,771	0.2%				182			*									182							100.0	12	0.0%
Miyazai	1,135,652	−0.6%	517,715	0.3%				175			*									178							101.5	12	1.0%
Okinawa	1,454,023	0.4%	610,129	1.6%				105			*									108							103.1	12	1.1%
Saga	842,457	−0.6%	325,221	0.7%					108			*									115						106.5	12	1.2%
Oita	1,183,961	−0.6%	530,704	0.6%					157			*									155						98.7	12	−1.0%
Tokushima	770,057	−0.8%	332,780	0.5%					100				*								98						98.4	12	−1.0%
Aichi	7,532,231	0.3%	3,214,669	1.4%						577			*										594				102.9	13	1.1%
Kumamoto	1,810,343	−0.4%	767,976	0.8%					243					*									251				103.3	14	1.1%
Miyagi	2,319,438	−0.2%	980,808	0.9%						289				*										292			101.0	14	1.0%
Yamagata	1,106,984	−1.0%	413,685	0.4%							175			*								177					101.1	11	1.0%
Saitama	7,390,054	0.2%	3,353,979	1.4%									541			*										551	101.8	13	1.0%
Ehime	1,369,131	−0.9%	655,255	0.2%									157				*						163				103.8	10	1.1%

Note: (left) Actual Amount of Residential Area of Base Year, (*) Year of Enactment, (right) Planned Amount of Residential Area of Target Year.

shows residential areas in the base year and their target amounts in the NLUP-PPs of 25 prefectures after the Fifth NLUP-NP was formulated. In the base year, the population decreased in 19 prefectures, but not a single prefecture experienced a decrease in households. In the target year of the NLUP-PPs, 17 prefectures expect an increase in residential area compared to the base year, five prefectures expect a flat increase, and three prefectures expect a decrease. The status of population increases or decrease in the base year is not necessarily linked to the increase or decrease in residential area targets from the base year to the target year set at that time.

Since this study focuses on residential area shrinkage, the paper first examines in detail the eight prefectures where residential area targets are expected to decrease or remain flat from the base year.

Figure 6 compares residential area targets and actual amounts, as well as population and household transitions, for each of the eight prefectures using the same method shown in Figure 5.

Since the base year for Kyoto Prefecture is 2012, its population and households are shown both for Japanese households from 2012 and for total households from 2013, as in Figure 5.

As can be seen from these graphs, first of all, the population is declining in all prefectures, while households are increasing in all except Akita Prefecture. The transition of residential areas tends to be much more closely related to households than to population.

However, in some cases (Oita and Gifu), residential area and households grew almost identically, while residential area grew more rapidly than households in some other cases (Aomori and Tokushima). However, households grew more rapidly in the other cases (Kyoto, Akita, and Shizuoka). It can be inferred that in addition to the growth of households, other factors are influencing the growth of residential areas.

In contrast, except for Akita Prefecture, the residential area targets in the plans fall between the trend of population increase/decrease and actual residential area values. In many prefectures, as is the case with the national trend, the actual amount of residential area tends to be close to the increase in households, while the targets appear to be lower, dragged down by the decline in population.

4.3. Planning of Residential Areas in Prefectures with Shrinking Targets

Next, this paper examines and discusses in detail the reasons why actual residential area amounts do not follow population decline in the three prefectures where residential land target amounts have decreased relative to the amount in the base year.

First, as noted above, the continued increase in actual residential area amounts despite the continued decline in population, and the pace of this increase outpacing even the increase in households, is thought to be directly caused by the rapid decline in the number of household members. Figure 7 shows the number of persons per household for the nation and for the three prefectures calculated from the National Census and the Basic Resident Register. The census figures are for “general households,” excluding households residing in institutions and dormitories, and are available only every five years. On the other hand, the Basic Resident Register does not take into account people who move without changing the registration as described above, and numbers including foreign households are available only after 2013. Regardless of these differences, the number of persons per household declined rapidly in both data.

Figure 7 also shows the number of persons per household, which are calculated by the population and households in the target year indicated in each plan.

These show that the decline in the number of persons per household is more rapid than assumed by the national and prefectural governments. For example, the projection in the NLUP-NP is 2.32 persons per household in the target year of 2025, but the actual figures had already decreased to 2.21 in 2020 according to the census and 2.08 in 2023 according to the Basic Resident Register, showing that the downward trend is already below the projection for the target year. The same trend is observed for the three prefectures shown here.

Table 2. Processes to set target amount of residential area in the three prefectures with declining residential area targets.

Categories		Aomori Pref. NLUP-PP (2017)	Tokushima Pref. NLUP-PP (2020)	Oita Pref. NLUP-PP (2019)
Population	Actual	(a) National Census/Prefectural Survey	(not used)	(not used)
	Estimates	(b) Projection of Prefectural Population on the Vision (prefectural strategy)	(not used)	(not used)
Households	Actual	(c) National Census	(c) National Census *	(c) National Census *
	Estimates	(d) Estimates of IPSS ** for the whole nation	(d) Estimates of IPSS **	(d) Estimates of IPSS **
Other variables		(e) Number of vacant houses (Housing and Land Survey) (f) Average area per residential lot (Housing and Land Survey)	(e) Total number of houses (Housing and Land Survey) (f) Owner-occupied house rate, rebuilding rate, building ratio by building type (Annual Report of Building Statistics) (g) Data on building permits and development permits (owned by prefectural and municipal governments)	(e) Total number of houses (Housing and Land Survey)
Process to Set Target Amount of Residential Area		(1) Number of households in the prefecture is calculated by regression analysis using (a), (b), (c) and (d). (2) number of vacant houses in target year is esimated by actual trend of (e). and added to number of households calculated by (1). (3) Average area per residential lot in target year is projected by actual trend of (f). (4) (3) is multiplied by (2) and residential area in target year in the prefecture is calculated.	(1) Number of new housing units is estimated from (c), (d) and (e), and gross Increase in residential area by new housing units are calculated taking (g) into consideration. (2) Gross decrease in residential area is estimated based on (c), (d) and (f). (3) Net increase/decrease is calculated based on (1) and 2).	(1) Growth rate of households in target year is calculated by (c) and (d). (2) Average houses per household in target year is calculated based on trend obtained by (c), (d) and (e).

Note: * Regarding the actual number of households, the National Census was used in all prefectures, although Tokushima and Oita prefectures answered the results acquired by the research of IPSS. ** IPSS: National Institute of Population and Social Security Research.

shows the results of e-mail surveys asking the departments in charge of the three prefectures in July 2024 about the specific methods used for calculating residential area for the target years in the three prefectures mentioned above.

First, it was found that residential areas for the target years were calculated in each of these three prefectures using different methods. The simplest process for setting a target amount of residential area was found in Oita Prefecture, where only changes in the number of households and the ratio of households to housing units were considered. Tokushima Prefecture used various data to set the target of residential area due to new development, the conversion of residential land to other uses, and the increase in the number of households. Aomori Prefecture estimated the number of future households in Process 1 and calculated the residential area by taking into account the area of vacant houses that were not occupied but remained residential land.

Although each of the three prefectures used various methods to calculate the residential land area for the target year, the results showed a similar underestimation, as seen in Figure 6. The reason for this can be attributed to the underestimation of the number of households in the target year, regardless of the decrease in the number of persons per household, as shown in Figure 7.

Next, this paper discusses how the three prefectures setting declining residential area targets in their NLUP-PPs attempted to achieve said targets.

As described above with Figure 1, the NLUP Act stipulates that NLUP-PPs shall be based on the NLUP-NP. The statements for residential areas are set forth in the NLUP-PPs of Tokushima and Oita Prefectures in exactly the same wording as the national NLUP-NP, as quoted above.

In Aomori Prefecture’s NLUP-PP, the phrase “the number of households is expected to decrease during the planning period” is deleted in accordance with its own projections, and statements are added regarding prefectural characteristics such as high snowfall and snow cover, and the quality of housing stock with enhanced security features against crimes. However, no special mention of the amount of residential area is found, as in the other two prefectures.

E-mail surveys for the three prefectures in July 2024 also asked about the means of achieving the target amount of residential area in detail. The results are shown in

Table 3. Approaches to try to realize targets in the three prefectures with declining residential area targets.

Means	Aomori Pref. NLUP-PP (2017)	Tokushima Pref. NLUP-PP (2020)	Oita Pref. NLUP-PP (2019)
Other uses and these breakdown after conversion of residential areas	- To “Water Surface”, “Road” and “Others” - “Others” category includes conversion to open space as snow dumping sites, disaster prevention sites, etc., taking into account natural conditions such as winter snowfall. - No specific breakdown to certain categories was made.	- To other “Developed Land” (174 ha) and “Others” (692 ha) is assumed.	None in particular
Relationship with the five areas designated in the Land Use Master Plan(LUMP)	None in particular related to the target of residential area	None in particular related to the target of residential area	None in particular
Evaluation regarding categories or areas shown above	Not specifically evaluated	Not specifically evaluated	Not specifically evaluated
Other means/methods of advancing residential area reduction	- The conversion of land from other uses to residential areas will be curbed through the effective and efficient use of underutilized land, vacant houses and existing housing stock. - Urban functions and residences will be concentrated in urban centers and living centers, etc., in light of the declining population and households.	- Priority will be given to land use upgrading, effective use of underutilized and vacant land and existing housing stock. - Conversion from natural land use, etc. will be restrained, while securing the necessary sites.	None in particular
Measures to address the rising actual amount of residential area	None in particular	Promote effective use of underutilized land and vacant houses and effective use of existing housing stock	None in particular

.

According to this, Oita Prefecture did not assume any specific measures in connection with its NLUP-PP, while Aomori and Tokushima Prefectures made some assumptions about the means of achieving a decrease in residential area and its subsequent use. However, in terms of specific assumptions with quantities, Tokushima Prefecture only provided a brief breakdown into broad categories, and no post-evaluation has been implemented.

Thus, the NLUP-PPs of the three prefectures provided general direction for the assumed decrease in residential area in response to population decline but left the specific means of achieving this entirely to individual acts.

5. Discussion

This paper has firstly shown that in the Japanese case, residential areas have continued to increase on a regional scale even as the population has declined (see Figure 5 and Figure 6) for both the whole country and several prefectures.

It was also found that the amount of residential area changed more closely with the change in households than with the population. However, the relationship between the amount of residential area and the number of households was not necessarily uniform. In the eight prefectures studied in this study, the growth of residential area was greater than that of households in many prefectures, while in others, the growth of households was greater. In all prefectures, however, the change was closer to that of households than to that of population. As long as the number of households continues to grow, residential land will also continue to increase, even as the population declines. Thus, shrinkage is not necessarily linked to population decline.

Regardless of this situation, it was found that NLUP-PPs of shrinking prefectures generally set targets for more shrinkage than the actual situation. The underestimation of the amount of residential area was found to be due to an underestimation of the number of households because they did not consider the rapid decline in the number of persons per household, as shown in Figure 7.

The case studies for the three prefectures that assume shrinking residential areas individually set targets for the amount of residential area in different ways, as shown in Table 2. However, the national projection of IPSS was commonly used for the number of households, resulting in a common underestimation. Aomori Prefecture removed the reduction in the number of households from its NLUP-PP, which was indicated by the NLUP-NP, but this was not reflected in the actual target amount. A more accurate projection of the amount of residential area would have been possible if the decrease in the number of persons per household and the resulting increase in the number of households had been considered.

Regarding specific measures of control in residential areas by NLUPs, existing studies have critiqued the NLUP system as a loose and indirect control mechanism, although the individual acts should be based on NLUPs. The case studies of the three prefectures showed that, as can be seen from Table 3, while the policy is indicated in the wording by the NLUP-PPs, the specific control is left to the individual acts for each of the five areas. If the number of households begins to decline, the whole planning system, including NLUPs and individual acts, will need to be improved in order to control for the amount of residential area on a broad scale.

To summarize the discussion above, the originality of this paper, especially when compared to existing studies, is that it shows the drastic decrease in the number of persons per household affected the amount of residential areas, which was not taken into account in regional plans. However, the limitation of this paper is that it is only a piecemeal survey of the analysis of the main factors contributing to the change in the amount of residential area and the planning methodology used to guide the targets.

6. Conclusions

The main conclusion of this study is that shrinkage does not follow population decline on a regional scale. In fact, as Klassen et al. [10] stated, population decline is followed by household decline, but the time lag is considerably longer than recognized. This misunderstanding led to an underestimation of the amount of residential land in the Japanese prefectures that served as case studies for this study.

This misunderstanding needs to be corrected before dealing with actual urban shrinkage in the future. However, Japan’s current land use planning system is divided into the NLUP systems at a regional scale and systems based on individual and segmented acts that directly control land use separately. Linking these systems is needed to deal with the shrinkage.

Although this study served as a benchmark showing the relationship between the shrinkage of residential area and population and households, and the efficacy of regional planning during a phase of population decline, there are several issues to be addressed for future research. First, regarding setting targets for the amount of residential area, there are factors other than households and population that have influence, like socio-economic and cultural characteristics of regions and their residents, changing urbanization trends, and related policies affecting land use. More precise forecasting and estimation methods should be developed by including these factors as variables. The survey methodology for the LUSS also needs to be more closely examined. In relation to the phenomenon of shrinkage, attention should also be paid to the case of Akita Prefecture, where residential areas have begun to shrink on a regional scale.

As for research issues in the planning system, it is necessary to conduct research on individual acts, which was not conducted this time, especially on the existence and content of target values and control methods. If these studies can be conducted, a proposal for an integrated planning system that reconstructs the relationship between individual acts and the NLUP Act may be considered.

The Japanese national government formulated a new Sixth NLUP-NP in July 2023. It is expected that new NLUP-PPs will be also formulated based on this new NLUP-NP, and it would be of great significance to investigate what contents will be included in the plans for areas where the population is further declining.