Typology of Small- to Medium-Sized Korean Local Cities with Population Decline from the Perspective of Urban Compactness

Abstract

This study examines urban structure typologies for small- to medium-sized cities in South Korea facing population decline, with a focus on urban compactness as a sustainable strategy. Population reduction and aging trends have become prominent issues in South Korea, especially impacting smaller cities, where decreased population density affects urban service functionality and infrastructure maintenance. This research applies and adapts Japan’s urban structure evaluation framework, specifically designed for the Japanese compact city model, to analyze the spatial conditions of 15 small- and medium-sized cities in Gyeongsangbuk-do province, South Korea. Using various indicators such as population density, accessibility to daily services, public transport, and local economic activity, this study conducts a typological classification based on principal component analysis and clustering methods. The findings suggest distinct urban structure patterns within these cities, offering strategic insights for urban policy aimed at enhancing urban compactness and sustainability. The implications highlight the need for tailored policies that address the spatial reorganization of services and infrastructure to maintain urban viability amidst demographic shifts.

1. Introduction

Globally, the phenomenon of population decline is deepening in numerous cities. Countries in Asia prominently experiencing this issue include Japan, South Korea, and China. Japan is one of the first and most severely affected nations in Asia by population decline. The country’s population started to decrease in the early 2010s, primarily due to low birth rates and a rapidly aging population [1]. South Korea is also facing a population decline issue, with its fertility rate recently dropping to one of the lowest in the world. This decline, starting in the early 2020s, has led to a rapid transition to an aging society, raising sustainability concerns regarding the labor market, economic growth, and social security systems [2].

The challenges of population decline and aging are resulting in a reduction in the working-age population and tax revenues, leading to a deterioration in urban service functions. In the medium to long term, these issues are leading to urban problems such as deterioration and decreased safety in city districts, particularly in small- and medium-sized cities. Consequently, sociodemographic policies aimed at increasing the population are being pursued, along with efforts to reorganize the spatial structure of cities that have expanded during periods of growth. In Asia, Japan, confronting the population issue head-on, has implemented the 2014 Location Optimization Plan [3]. This urban planning framework aims to guide the long-term reduction of sprawled city areas and represents Japan’s version of the compact city policy to respond to population decline.

In South Korea, the “Guidelines for the Establishment of Urban and County Master Plans”, revised in 2018, mandate that the impact of population decline on cities be assessed and incorporated into the formulation of master plans, which have traditionally been based on the assumption of population growth [4]. Furthermore, the 5th Comprehensive National Territorial Plan, announced in 2019, includes strategies for reorganizing national territorial space in a compact manner in preparation for an era of population decline [5]. However, the development of specific methodologies or spatial models has not yet been realized.

This research focuses on South Korea and aligns with the perspective that the consolidation of urban structures is essential for the sustainability of small- and medium-sized cities in the context of ongoing population decline. Furthermore, urban compactness contributes to enhancing urban resilience by enabling cities to adapt to social, economic, and environmental challenges. This resilience ensures that cities can better respond to demographic shifts and resource constraints while maintaining long-term sustainability. Before proposing specific spatial structure solutions, it is crucial to understand the current spatial situations of Korean cities following urban expansion. Therefore, this study aims to evaluate the urban structure of small- and medium-sized cities in South Korea experiencing population decline and to derive implications for the compressive reorganization of expanded urban structures.

As a methodology for evaluating urban structures, the urban structure evaluation system used in Japan will be utilized. [6] The methodology employs the urban structure evaluation system used in Japan, adapting it to the Korean context by modifying and applying specific indicators for evaluation.

The study targets small- and medium-sized cities in Gyeongsangbuk-do, South Korea, which have been identified as facing significant population decline risks based on analyses of local extinction risk due to population decrease in various related studies within Korea [7,8].

Specifically, the spatial scope of the research includes fifteen cities in Gyeongsangbuk-do with a population of less than 100,000, all experiencing population decline. These cities are Yeongju-si, Yeongcheon-si, Sangju-si, Mungyeong-si, Gunwi-gun, Uiseong-gun, Cheongsong-gun, Yeongyang-gun, Yeongdeok-gun, Cheongdo-gun, Goryeong-gun, Seongju-gun, Yecheon-gun, Bonghwa-gun, and Uljin-gun. Ulleung-gun, a distant island region, was excluded from the study due to its unique circumstances (Figure 1).

2. Literature Review

2.1. Overview of Urban Structure Evaluation Sytem in Japan

The urban structure evaluation system adopted in this study, introduced in Japan in 2014 alongside the Location Optimization Plan mentioned earlier, is a technique for analyzing the current situation and urban structure of a city before establishing a Location Optimization Plan aimed at practicing urban consolidation. Local governments in Japan conduct an urban structure evaluation based on the “Handbook on Urban Structure Evaluation” published by the Ministry of Land, Infrastructure, Transport, and Tourism [9]. This preliminary evaluation informs the development of a Location Optimization Plan tailored to each local government, based on the evaluation results. The evaluation of urban structures involves an objective, quantitative assessment of the city’s density from various perspectives, incorporating macroscopic considerations such as population distribution, aging trends, regional economy, and finances. This process clarifies the challenges of promoting urban consolidation by overlaying the distribution of urban functions and public transportation networks with infrastructure facilities.

Urban structure evaluation first sets the fields of evaluation from the perspective of deriving measures to ensure the city’s sustainability, considering future urban challenges like loss of daily living functions due to population decline and urban decay, as well as the direction of policies like the Location Optimization Plan. Subsequently, the evaluation criteria are established by considering the direction and policies each evaluation field should aim for. Once the indicator values are calculated, the current national average and the average for cities of similar sizes are presented for comparison with the city’s indicator values, thereby evaluating the city’s density. Additionally, indicators that allow for future projections are used to compare the projected future of cities without plans and the desired future outlined in the plans, identifying urban structures that need improvement. Beyond current status evaluation, it also involves setting targets for desired urban structures, using these targets in the planning process, and conducting regular evaluations to compare with these targets, thus monitoring the progress of the plan. This comprehensive approach is utilized throughout all stages of developing and implementing the Location Optimization Plan.

The Ministry of Land, Infrastructure, Transport, and Tourism (MLIT) first identifies problems arising within cities due to population decline, including the following: (1) deterioration of daily living functions, (2) reduction in outings and worsening health, (3) decreased stability of city areas, (4) decline of the local economy, (5) deterioration of local finances, and (6) increased energy consumption and carbon emissions. It then sets out urban structure elements that can respond to these issues as fields of evaluation and indicators for assessment.

Each city conducting an urban structure evaluation first determines the directionality it should aim for and the evaluation fields that can address potential problems for its realization. Criteria are then established regarding from what perspectives each field should be evaluated, leading to the setting of specific indicators. The MLIT provides examples of evaluation areas and indicators related to the six urban problems listed, with the main contents outlined as follows [9] (pp. 9–10) (

Table 1. Examples of Urban Structure Evaluation Indicators.

Evaluation Field	Evaluation Criteria		Example of Evaluation Indicators
Convenience of Life	By properly guiding residence and urban functions, realize a city where daily necessities such as medical care, welfare, and commerce, as well as public transportation services, are satisfied within walking distance.
	Direction of Response	Appropriate guidance for residence	Population density in residential guidance areas
		Proper distribution of urban functions Improvement in public transportation service level	Population density within walking distance to life service facilities Modal share of public transportation
Health and Welfare	By creating a pedestrian-friendly environment, realize a city where citizens can live healthily.
	Direction of Response	Increase walking and promote health	Population density of metabolic syndrome and its prediabetic population Modal share of walking, cycling
		Improvement in urban living convenience Creation of a pedestrian-friendly environment	Ratio of medical facilities within walking distance for the elderly Percentage of areas without parks within walking distance for the elderly
Safety and Security	Realize a city with low risk of disasters and accidents.
	Direction of Response	Encouraging residence in safe areas Improvement in pedestrian environment safety Securing urban safety	Population density in natural disaster risk districts Average distance to the nearest evacuation site Number of traffic accident fatalities per capita
Local Economy	Realize a city with a vibrant service industry and a sound real estate market.
	Direction of Response	Activation of the service industry	Average sales per person in tertiary industry
		Formation of a sound real estate market	Average residential land price
Administrative Operations	Realize a city where administrative services are operated efficiently, and local government finances are sound.
	Direction of Response	Efficiency of urban management Stable tax revenue acquisition	Administrative expense ratio per citizen Tax revenue per citizen
Energy and Low Carbon	Realize a district with high energy efficiency and low energy consumption and carbon dioxide emissions.
	Direction of Response	Energy saving in the transportation sector	CO2 emissions per citizen from automobiles Modal share of public transportation
		Energy saving in the domestic sector	CO2 emissions per citizen from the household sector

).

The actual evaluation of urban structures is conducted based on urban planning surveys and other foundational research held by each local government, which means that the specific fields of evaluation and indicators can vary between local governments. Once the fields of evaluation and indicators are established, the density of the urban structure is assessed by comparing the current values of each indicator for the city with the average values of a comparative group of cities of similar size. This comparison analysis identifies aspects of the urban structure that require improvement. Additionally, the evaluation of urban structures can be used to set targets for the desired urban structure and for establishing objectives. Regular comparisons with these targets allow for monitoring the progress of plans and can be utilized in various processes.

The results of each field of evaluation and indicator are displayed using radar charts. The central red line represents the average values for cities of the same size as the city being evaluated, while each blue dot represents the result of an individual indicator for the city. This visualization allows for a relative understanding of the current status level of the city being evaluated by comparing it with the average values of other cities of similar size (Figure 2 and Figure 3).

2.2. Example of Urban Structure Evaluation Sytem in Japan

Ichihara City, located in the northwestern part of Chiba Prefecture along Tokyo Bay, is a representative industrial city of the prefecture. The population of Ichihara City decreased from 280,416 in 2016 to 268,285 in 2021. Both the population of children aged 0–14 and the working-age population aged 15–64 have been declining, while the elderly population aged 65 and over has been increasing significantly. The aging rate, which was only 7.2% in 1985, surged to 20.9% by 2010, demonstrating a rapid acceleration of population aging.

In response, Ichihara City established a Location Optimization Plan in 2018. The plan set evaluation fields such as living convenience, health and welfare, safety, regional economy, administrative operations, and energy and low carbon. Using relevant indicators, the city conducted an urban structure evaluation. This evaluation included a comparative analysis with national averages and other local urban areas of approximately 300,000 residents to assess the current state of the city (Figure 4 and

Table 2. Summary of Ichihara City’s Urban Structure Evaluation Results (Source: [10] (p. 16)).

Evaluation Fields	Evaluation Criteria
Living Convenience	The percentage of the population within walking distance of essential service facilities is lower compared to the national average and cities with a population of 300,000. However, the population density within walking distance is higher than in other cities. The percentage of the population within the service areas of public transportation is higher compared to the national average and cities with a population of 300,000, indicating a well-developed public transportation network. The modal share of public transportation is lower than that of the three major metropolitan areas but higher than that of cities of similar size.
Health Welfare	The modal share of walking and cycling is low, indicating a high dependence on private vehicles. The percentage of the population within walking distance of childcare and elderly welfare facilities is lower compared to other cities, suggesting relatively inadequate welfare support for young children and the elderly.
Safety	The number of traffic accident fatalities is higher compared to the national average and cities with a population of 300,000, and the average distance to emergency evacuation points is longer than in other cities, indicating a need for enhanced safety measures. The vacancy rate is lower than that of the three major metropolitan areas and cities with a population of 300,000, and is similar to the national average.
Regional Economy	The sales revenue per employee in the tertiary industry is decreasing compared to cities with a population of 300,000, but the efficiency of stores across the city is increasing. The average residential land price remains lower than the national average, the three major metropolitan areas, and cities with a population of 300,000.
Administrative Operations	The fiscal strength index is high, and the per capita expenditure is lower compared to the national average, the three major metropolitan areas, and cities with a population of 300,000, indicating efficient urban management in terms of finances. The per capita tax revenue is stably secured compared to the national average and cities with a population of 300,000.
Energy Low Carbon	The per capita CO2 emissions from automobiles are higher than the national average, the three major metropolitan areas, and cities with a population of 300,000, indicating the need for a shift to public transportation to reduce environmental burdens. In the household sector, per capita CO2 emissions are increasing compared to the national average.

).

Based on the results of the above urban structure evaluation, Ichihara City has organized the tasks required for the establishment of a Location Optimization Plan for the reorganization of the city’s structure from various perspectives, including population trends, land use, living convenience, transportation accessibility, disaster safety, and fiscal soundness, as follows (

Table 3. Challenges for Urban Structure Improvement (Source: [10] (pp. 19–20)).

Category	Task Summary
Population Trends	There are low-density population areas within urbanized zones, requiring measures to encourage residential concentration in these areas. Urban functions are concentrated around train stations, and population density is high, necessitating the induction of higher-order functions to strengthen these areas as city centers.
Land Use	Urbanization is progressing around Yawatajuku Station and Goi Station, and, in the southern region, residential land development is shifting natural land use to building sites. While urbanized zones have been designated in each area, it is necessary to guide residential concentration within these zones, considering future population decline, aging, and living convenience.
Living Convenience	The percentage of the population within walking distance is lower compared to the national average and cities with a population of 300,000. Efforts are needed to improve the insufficient functions of areas not included in the service areas of daily life services through network enhancement. The area around the city hall, which hosts various functions including public facilities, requires maintaining urban functions and actively encouraging residential concentration as a convenient residential area.
Transportation Accessibility	Train stations serve as transportation hubs for railways and buses, providing high transportation accessibility. It is necessary to induce functions required by the local area to these key living hubs. An increase in the elderly population is expected in residential areas within urbanized zones. Strengthening and promoting the use of bus networks connecting key hubs is essential to reduce reliance on private vehicles while maintaining transportation accessibility.
Disaster Prevention	In riverfront areas with concentrated populations, safety measures are required to ensure residents can continue living safely in zones designated as flood-prone areas. Some areas around urbanized zones are designated as special caution zones for sediment-related disasters, necessitating restrictions on residential expansion in these areas.
Fiscal Soundness	Amidst population decline, it is necessary to maintain the population and buildings within urbanized zones to ensure efficient tax revenue generation. With aging public and infrastructure facilities, continuous replacement projects are anticipated. From a long-term perspective, it is essential to concentrate facilities and maintain them efficiently to suppress expenditure.

).

Based on the urban status and challenges identified through the urban structure evaluation as mentioned above, Ichihara City set a goal to create a sustainable city that promotes vitality and interaction, while ensuring a safe and secure living environment for its citizens. The city proposed a future vision that emphasizes residential attraction, urban function inducement, and the establishment of a transportation network through public transit.

3. Establishing an Urban Structure Evaluation Framework for Analyzing South Korea’s Small- and Medium-Sized Cities

3.1. Overview of Establishing the Urban Structure Evaluation Framework

Through the above process, we have broadly reviewed the Japanese urban structure evaluation system. To apply this evaluation system to cities in South Korea, it is first necessary to set the evaluation fields and indicators. This study, while broadly following the categories suggested by Japan’s Ministry of Land, Infrastructure, Transport, and Tourism, has also referred to evaluation cases of actual Japanese cities similar in size to the small- and medium-sized cities of Gyeongsangbuk-do, which are the focus of this analysis. Furthermore, the study has considered the availability of data for each indicator to select evaluation fields and indicators suitable for the domestic context. As a result, the six evaluation fields proposed by the Ministry of Land, Infrastructure, Transport, and Tourism were applied as is, and the evaluation indicators were streamlined to 29. For the first evaluation field, ① “Convenience of Daily Living”, the indicators include elements that assess accessibility to life service facilities. Specific life service facilities targeted include hospitals, welfare, commerce, and public transportation (train stations/bus stops), with the population density within walking distance of each facility and the ratio of the total population to the overall city population selected as indicators. Additionally, the ratio of the population within the overlapped walking areas of the aforementioned facilities to the total population, the average daily distance traveled by private car, and the modal share of public transportation were selected as indicators to measure accessibility.

For the second evaluation field, ② “Health and Welfare”, indicators were selected related to accessibility to health and welfare facilities, and pedestrian friendliness. The indicators target senior citizen centers, hospitals, kindergartens, neighborhood parks, and public sports facilities, with the proportion of the population within each facility’s service area relative to the total city population being chosen as an indicator. For pedestrian-friendliness, the walking practice rate among the elderly population aged 65 and above was selected as an indicator.

For the third evaluation category of ③ “Safety”, indicators related to disaster and traffic accidents were selected. In terms of disasters, the ratio of the population within natural disaster risk areas designated by each local government to the total population of the city, and accessibility to earthquake evacuation areas, were chosen as indicators. For traffic-accident-related indicators, the number of traffic accident fatalities per 10,000 citizens within a year was selected.

For the fourth evaluation category of ④ “Local Economy”, indicators related to the service industry and real estate were chosen. To measure the service industry, the average sales per employee or establishment in the tertiary sector were selected as indicators, while for real-estate-related indicators, the average official land price of residential areas and the vacancy rate within the city were measured.

The fifth evaluation category of ⑤ “Municipal Governance” selected indicators related to urban management and tax revenue. The annual per capita expenditure and revenue, as well as the financial independence ratio of each local government, were chosen as indicators.

For the sixth evaluation category of ⑥ “Energy and Low Carbon”, indicators related to the energy aspect of the transportation sector were selected. In this study, the annual average CO₂ emissions per capita from automobiles were chosen as indicators (

Table 4. Established Evaluation Indicators and Calculation Methods.

Field No	Abbr No	Index		Unit	Data References	Year	Calculation
① Convenience of Daily Living	①-1	Proportion of Population within the Service Area * of Daily Life Services		%	Medical, Welfare, Commercial, and Transit Data presented below	-	(Population in Area where Walking Distance of Each Facility Overlaps) ÷ (Number of Citizen) × 100
	①-2	Population Density in Urban Areas		persons/km2	[11]	2020	(Population in Urban Area) ÷ (Area of Urban Areas)
	①-3	Population Density Within Walking Distance of Daily Life Service Facilities	Medical	%	[11,12]	2020	(Population within 800 m radius of Medical Facilities) ÷ (Number of Citizen) × 100
	①-4		Welfare		[11,12]	2020	(Population within 800 m radius of Welfare Facilities) ÷ (Number of Citizen) × 100
	①-5		Commerce		[11,12]	2020	(Population within 800 m radius of Commercial Facilities) ÷ (Number of Citizen) × 100
	①-6	Proportion of Population within the Service Area of Public Transportation		%	[11,13,14]	2020	(Population within 300/800 m radius of Bus Stop/Train Station) ÷ (Number of Citizen) × 100
	①-7	Population Density Within Walking Distance of Daily Life Service Facilities	Medical	persons/km2	Same as above Data	2020	(Population within 800 m radius of Medical Facilities) ÷ (Area of 800 m radius of Medical Facilities)
	①-8		Welfare			2020	(Population within 800 m radius of Welfare Facilities) ÷ (Area of 800 m radius of Welfare Facilities)
	①-9		Commerce			2020	(Population within 800 m radius of Commercial Facilities) ÷ (Area of 800 m radius of Commercial Facilities)
	①-10	Population Density within the Service Area of Public Transportation		persons/km2	Same as above Data	2020	(Population within Walking Distance of Public Transportation) ÷ (Area of 300/800 m radius of Bus Stop/Train Station)
	①-11	Modal Share of Public Transportation		%	[15]	2020	Use of Statistical Report Data
	①-12	Average Daily Mileage of a Private Car		km/day	[16]	2018	Use (Average Daily Mileage of a Private Car) from [Vehicle Mileage by Use, Type, Region]
② Health and Welfare	②-1	Proportion of Elderly Population in Service Area ** of Senior Center		%	[17]	2019	(Number of Elderly Population in Service Area of Senior Center) ÷ (Number of Elderly Population) × 100
	②-2	Proportion of Infant Population in Service Area of Kindergarten		%		2019	(Number of Infant Population in Service Area of Kindergarten) ÷ (Number of Infant Population) × 100
	②-3	Proportion of Vulnerable Population Outside Service Area of Hospital		%		2019	(Number of Vulnerable Population outside Service Area of Hospital) ÷ (Number of Vulnerable Population) × 100
	②-4	Proportion of Population in Service Area of Neighborhood Park		%		2019	(Number of Population in Service Area of Neighborhood Park) ÷ (Number of Citizen) × 100
	②-5	Proportion of Population in Service Area of Public Sports Facilities		%		2019	(Number of Population in Service Area of Public Sports Facilities) ÷ (Number of Citizen) × 100
	②-6	Walking Rate of the Elderly (Over 65 years old)		%	[18]	2020	(Number of People who Walked for more than 30 Minutes a Day and 5 Days in a week) ÷ (Number of Respondents) × 100
③ Safety	③-1	Proportion of Population in Disaster-Prone Areas		%	[19]	2020	(Population in Disaster-Prone Areas) ÷ (Number of Citizen) × 100
	③-2	Number of Traffic Accident Deaths per 10,000 Citizens		persons	[20]	2020	(Number of Traffic Accident Deaths) ÷ (Number of Citizen) × 10,000
	③-3	Accessibility to Earthquake Evacuation Shelter		km	[17]	2019	Distance to the Nearest Earthquake Evacuation Shelter
④ Local Economy	④-1	Sales of Tertiary Industries per Person		million-Won	[21]	2018	(Total Sales of Local Tertiary Industries) ÷ (Number of Employees)
	④-2	Average Sales of Local Retail Businesses		million-Won	[21]	2018	(Total Sales of Local Retail Business) ÷ (Number of Local Retailers)
	④-3	Average Official land Price of Residential Areas		thousand-Won/m2	[22]	2021	(Average Official Land Price of Residential Area)
	④-4	Proportion of Empty House		%	[23]	2019	(Number of Empty Houses) ÷ (Total Number of Houses) × 100
⑤ Municipal Governance	⑤-1	Expenditure per Person		million-Won	[24]	2019	(Total Amount of Annual Expenditures) ÷ (Number of Citizen)
	⑤-2	Local Government Fiscal Independence Rate		%	[25]	2020	(Local Government Income) ÷ (Local Government Budget) × 100
	⑤-3	Tax Revenue per Person		million-Won	[26]	2019	(Total amount of Annual Revenues) ÷ (Number of Citizen)
⑥ Energy and Low Carbon	⑥-1	Automobile CO2 Emissions per Person		t-CO2/ year	[27]	2019	(Mileage per Vehicle) × (Number of Vehicle Registrations) × (CO2 Emission Factor) ÷ (Number of Citizen)

* Walking Distance: 300 m Radius at Bus Stop, 800 m Radius at Medical, Welfare, Commercial Facilities, and Train Station. ** Service Area: 750 m Radius at Living Service Facilities.

).

3.2. Current Status of the Target Cities

In this study, 15 cities within Gyeongsangbuk-do, experiencing population decline, have been set as the target cities for analysis. The population-related status of these target cities is as follows in

Table 5. Population-Related Status of the Target Cities (Source: [28]).

City (Abbr)	Population	Area	Population Density
Yeongju-si (a)	104,012	668.4 km2	156 persons/km2
Yeongcheon-si (b)	104,875	920.3 km2	114 persons/km2
Sangju-si (c)	98,138	1255 km2	78 persons/km2
Mungyeong-si (d)	72,067	911.7 km2	79 persons/km2
Gunwi-gun (e)	23,748	614.2 km2	39 persons/km2
Uiseong-gun (f)	52,294	1176 km2	44 persons/km2
Cheongsong-gun (g)	25,223	846 km2	30 persons/km2
Yeongyang-gun (h)	16,884	815.1 km2	21 persons/km2
Yeongdeok-gun (i)	37,249	741.1 km2	50 persons/km2
Cheongdo-gun (j)	43,229	696.5 km2	62 persons/km2
Goryeong-gun (k)	32,891	384 km2	86 persons/km2
Seongju-gun (l)	44,970	616.3 km2	73 persons/km2
Yecheon-gun (m)	56,165	660.7 km2	85 persons/km2
Bonghwa-gun (n)	31,812	1201 km2	26 persons/km2
Uljin-gun (o)	50,104	989 km2	51 persons/km2

.

4. Analysis of Urban Compactness in Domestic Small- and Medium-Sized Cities

The results of the urban structure evaluation for each target city, derived through the analysis method of urban compactness, are as presented in

Table 6. Evaluation Results of Target Cities’ Urban Compactness.

No	Unit	(a)	(b)	(c)	(d)	(e)	(f)	(g)	(h)	(i)	(j)	(k)	(l)	(m)	(n)	(o)
①-1	%	60.6	56.8	41.1	58.6	28.2	32.6	37.7	33.8	48.1	35.5	40.9	27.3	23.4	28.7	57
①-2	persons/km2	1585	1054	1065	1180	635	527	1053	525	1161	670	476	1146	1880	882	707
①-3	%	63	58.5	51	62	30.6	37.4	45.7	38.3	57	39.9	48.4	34.8	25.3	32.6	61.1
①-4		87.1	78.3	77.1	82.3	66.4	71.9	68	63.3	78	73.5	71.9	68.9	48.1	59.9	81
①-5		86.7	79.9	77.8	84.3	61.9	63.4	65.5	61.4	79.2	74.2	78.4	69.5	44.6	56.6	81.8
①-6	%	82.9	76.9	63.2	74.7	50.2	56.7	57.5	57	56.5	60.5	65.7	48.8	67.1	48.2	63.4
①-7	persons/km2	1067	937	494	817	216	352	257	277	472	348	421	445	260	279	602
①-8		279	205	146	195	58	68	66	53	112	103	142	102	71	56	167
①-9		250	174	129	144	61	84	62	47	116	89	98	86	80	52	149
①-10	persons/km2	812	416	338	506	203	247	257	159	458	369	419	312	321	177	397
①-11	%	6.2	9.5	6.2	6.1	5.4	2	3.3	4.4	3.1	6.8	12.2	7.1	3.8	3	4.7
①-12	km/day	33.8	35.9	35.6	35.8	38.4	37.7	39.5	37.8	37.6	35.6	37.1	37.3	39.5	38.6	38.5
②-1	%	80.94	73.23	74.13	75.99	68.81	73.91	63.24	35.16	76.44	76.34	76.37	66.53	70.44	52.82	73.6
②-2	%	37.6	18.49	41.69	55.14	24.22	18.91	12.25	45.24	35.3	35.6	30.28	30.67	40.46	32.07	19.76
②-3	%	81.21	99.48	100	97.01	100	93.92	99.75	87.3	100	92.82	94.63	91.52	97.77	96.8	99.04
②-4	%	45.57	35.41	29.5	34.08	12.33	6.71	3.51	9.31	25.48	15.21	17.51	19.32	26.69	16.48	27.01
②-5	%	5.98	12.9	65.88	10	10.31	10.68	10.13	12.29	12.49	25.56	3.95	13.88	19.01	19.31	23.87
②-6	%	45.9	22.5	26.2	35	43.7	30	33	20.5	47.5	29.3	30.4	43.7	30.5	30	45.8
③-1	%	0	0	2.21	1.98	0.48	0.47	1	0.41	0.38	1.59	0.54	0.85	1.04	2.24	0.78
③-2	km	15.88	4.45	4.72	5.54	12.32	4.12	6.43	6.94	2.96	4.07	11.89	4.16	5.08	6.53	23.14
③-3	persons	1.04	1.99	3.74	2.06	2.05	3.01	1.95	1.16	2.35	3.19	2.05	1.97	2.34	2.77	1.57
④-1	million-won	42.36	54.35	51.27	49.16	74.14	54.65	54.35	68.71	47.69	48.77	62.05	53.89	64.02	57.95	66.72
④-2	million-won	48.13	64.64	62.96	48.97	67.16	51.7	43.18	30.01	29.93	47.37	55.37	63.39	46.28	29.35	44.63
④-3	thousand-won	336	339	264	204	200	171	104	161	264	248	336	322	267	117	156
④-4	%	9.7	14.1	12.9	13	12.7	15.6	15.6	14.4	16.2	21	12.7	10.5	20.1	13.4	18.4
⑤-1	million-won	7.17	8.53	11.22	9.69	14.5	13.34	15.15	19.67	12.25	8.84	9.8	10.74	8.67	13.62	12.51
⑤-2	%	11.55	15.45	8.45	11.33	7.43	8.91	7.82	6.11	8.79	10.3	11.7	12.2	10.18	6.18	10.69
⑤-3	million-won	9.79	11.3	14.09	12.4	19.96	16.82	20.75	26.8	19.83	14.22	12.77	14.82	11.24	18.8	22.28
⑥-1	t-CO2/ year	1.45	2.45	1.83	1.26	1.99	1.95	1.52	1.49	1.19	1.83	2.11	3.04	1.49	1.36	1.23

. As discussed earlier, the calculated evaluation figures are not absolute values for assessing urban compactness. Instead, these indicators are compared and analyzed with the same indicators of other target areas to evaluate the urban structure from a relative perspective, and to identify characteristics and challenges. The analysis results of urban compactness through the comparative analysis of indicators across evaluation fields can be summarized as follows (Figure 5).

In the case of Yeongju-si (a), all indicators within the domain of living convenience exhibit positive figures surpassing the average (50) of all target cities. This indicates a high accessibility to and ease of use of living services for its citizens. Moreover, the significantly lower average daily driving distance (30) compared to the average suggests a low dependency on private vehicles, allowing for easy pedestrian access to everyday living services. The health and welfare sector also shows high accessibility and density of exercise and welfare facilities. Yeongju-si can be assessed as having an overall well-concentrated urban structure.

In Sangju-si (c), the majority of indicators in domains such as living convenience, health and welfare, local economy, administrative operations, and energy and low carbon are at levels similar to the average. Notably, the proportion of the population within the service areas of public sports facilities (83) is significantly higher compared to other cities, indicating a widespread and even distribution across the city. Areas needing improvement in Sangju-si relate to safety concerns, as indicators in the safety domain show negative figures. The number of traffic accident fatalities per 10,000 citizens (70) and the proportion of the population in natural disaster risk zones (67) are notably negative, highlighting areas that require attention.

In Mungyeong-si (d), indicators across all sectors generally display positive figures, especially in the domain of living convenience, where all indicators show positive results surpassing the average. Other notably positive indicators include the proportion of young children within the kindergarten service areas (70) and the proportion of the population within the service areas of community parks (64). However, Mungyeong-si faces challenges with a negative figure in the proportion of the population living in natural disaster risk zones (64) and indicators in the local economy sector such as sales per worker in the tertiary sector (41) and the average official land price in residential areas (46) also show negative outcomes.

In Gunwi-gun (e), indicators related to the local economy sector are positive, especially those related to the service industry including the average production value of retail businesses (64) and sales per worker in the tertiary sector (70), which show very high figures, which seems to be attributed to Gunwi-gun’s geographical proximity to Daegu Metropolitan City. However, indicators in most other domains display negative figures, particularly those related to everyday living services, indicating significantly reduced accessibility to these services. Additionally, indicators in the administrative operations sector are also negatively evaluated, suggesting a need for improvements in the overall urban structure.

For Uiseong-gun (f), while most evaluation indicators are similar to the average, the indicators in the living convenience sector are rated somewhat lower than average. The very low public transportation modal share (37) and the long average daily driving distance (55) suggest low accessibility to everyday living services and high dependency on private vehicles. Furthermore, the higher number of traffic accident fatalities per 10,000 citizens (61) compared to the average indicates a need for improvements in the pedestrian environment.

In Cheongsong-gun (g), the indicators in the living convenience sector, including the population density within walking distance to living service facilities and the public transportation modal share (41), show lower results compared to other areas, suggesting that living service facilities are dispersed and there is a high dependency on private vehicles. Accessibility to health and welfare facilities for infants and the elderly is also low, and indicators such as the average official land price in residential areas and the vacancy rate indicate a sluggish real estate market.

The indicators in the administrative management sector also appear negative, predicting potential future difficulties in urban management. This necessitates policies aimed at consolidating unnecessary facilities to reduce management costs and improve urban management efficiency.

In Yeongyang-gun (h), the indicators in the health and welfare sector, including the proportion of young children within the kindergarten service areas and the vulnerable population outside hospital service areas, show positive figures, and the indicators in the safety sector are positively evaluated. Additionally, sales per worker in the tertiary sector are high compared to other areas. However, all indicators in the living convenience sector show negative figures, indicating reduced accessibility to everyday living services for citizens. Particularly, the indicators in the administrative operations sector are very negative compared to other areas, suggesting the need for efficient urban management strategies and overall urban consolidation.

Yeongdeok-gun (i)’s results show a high proportion of the population within walking distance to all living service facilities, and indicators related to walkability, such as the walking practice rate among the elderly population (65) and per capita automobile CO₂ emissions (39), present positive figures. However, the high vulnerable population outside hospital service areas (59) indicates a need to encourage the residency of vulnerable populations within service areas, and somewhat low figures in public-transportation-related indicators such as the population ratio within public transportation service areas (45) and the public transportation modal share (41) suggest the need for improvements in the public transportation system. Additionally, indicators related to the service industry, such as the average production value of retail businesses (35) and sales per worker in the tertiary sector (40), are low.

In Cheongdo-gun (g), the indicators in the sectors of living convenience, health and welfare, administrative operations, and energy and low carbon generally show results similar to the average. However, in the safety sector, while accessibility to earthquake emergency shelters (43) appears good, the number of traffic accident fatalities per 10,000 citizens (63) and the proportion of the population residing in natural disaster risk zones (59) are significantly high, indicating a need for improvements in living environment safety. In the local economy sector, sales per worker in the tertiary sector (41) are low compared to other areas, and the vacancy rate (70) is the highest among all areas, highlighting the need for a vacant house management strategy.

In Goryeong-gun (k), the public transportation modal share (75) is significantly high, which appears to be due to its geographical proximity to Daegu Metropolitan City and the availability of various public transportation options such as rural buses. Other indicators in the local economy and administrative operation sectors also show positive outcomes, suggesting stable urban management. However, the population density within urban areas is very low compared to other locations, indicating a need for consolidation through residential and urban function inducement. Apart from the population ratio within public sports facilities service areas (41), other indicators are generally at levels similar to the average.

Seongju-gun (l) exhibits positive figures in the local economy sector indicators, such as the average official land price in residential areas (61), local retail business production value (61), and vacancy rate (38), suggesting a more vibrant local economy compared to other cities. Additionally, the walking practice rate among the elderly population (65 years and above) (61) is higher than the average. However, the lower population ratio within walking distance to medical facilities (42) and public transportation service areas (37) indicates a need for expansion of medical facilities and public transport routes. Also, the high per capita automobile CO₂ emissions (75) suggest the necessity for strategies to reduce car dependency and support for eco-friendly vehicles.

In Yecheon-gun (m), the population density within urban areas (73) is significantly higher than in other locations, but other indicators show negative figures. The population ratio within walking distance to living service facilities (29) is the lowest among the compared groups for all three types of facilities, indicating a lack of various living service facilities. This suggests a need for strategies to utilize vacant houses and expand various living services due to reduced accessibility to everyday living services and a high average daily driving distance, with the vacancy rate (67) also being high, second only to Yeongdeok.

Bonghwa-gun (n)’s evaluation indicators across all sectors are generally similar to or slightly negative compared to the average. The only positive indicator is the per capita automobile CO₂ emissions (43) in the energy and environmental sector, which is significantly lower than the average. Most indicators in the living convenience sector are negative, showing very low accessibility to everyday living services and a high dependency on private vehicles. Moreover, the low elderly population ratio within senior center service areas (24) suggests a need for improvement in everyday living services through the expansion of various social overhead capital (SOC) facilities.

In Uljin-gun (o), indicators related to the accessibility of everyday living services appear positive, especially the population ratio within everyday living services areas (63) being significantly higher than the average, indicating well-concentrated living service facilities. Indicators needing improvement include the higher average daily driving distance (61) and the low young children population ratio within kindergarten service areas (40). Additionally, the very low accessibility to earthquake emergency shelters (77) indicates a need for disaster management strategies related to earthquakes. The higher vacancy rate (60) compared to other cities also suggests the necessity for a response strategy.

5. Discussion of the Analysis of Urban Structure Types in Small- and Medium-Sized Cities in Gyeongsangbuk-do Province

In this chapter, we aim to classify cities with similar characteristics based on the previously derived indicators of urban structure evaluation and to propose urban planning directions for each type. To achieve this, we first conducted Principal Component Analysis (PCA) using SPSS statistical software (version 28), employing the Varimax orthogonal rotation method to extract key variables for cluster analysis. The principal component scores for each city were then calculated and used as input variables for hierarchical clustering analysis.

For the clustering process, we applied the widely used hierarchical clustering method with Ward’s linkage, which minimizes the sum of squared deviations within clusters to ensure compact and homogeneous group formation. Euclidean distance was used to measure similarities between objects. The dendrogram generated from this process determined the optimal number of clusters, classifying cities into five distinct types. Each cluster was analyzed by comparing the average scores of the principal components and other relevant indicators to interpret their characteristics and identify common patterns.

This study utilized all 29 indicators introduced in the previous chapter to analyze the aggregation of urban structure comprehensively. However, using all indicators as variables for cluster analysis posed challenges in identifying distinctive characteristics and naming clusters. Therefore, PCA was employed to consolidate indicators with similar properties into a smaller number of principal components, enabling a more efficient and interpretable clustering analysis.

5.1. Principal Component Analysis Results

In Section 3, a principal component analysis (PCA) was conducted using the urban structure aggregation evaluation results derived from the studied target sites as variables. Through Varimax orthogonal rotation, four principal components with eigenvalues greater than 2.0 were extracted. Eigenvalues represent the explanatory power of each factor, with larger eigenvalues indicating that the factor effectively explains the variance (information) of the variables. The loading values of the principal components account for 74.184% of the total variance of the input variables, signifying that they explain a significant portion of the data. Statistically, when the loading values of extracted factors exceed 65%, they are considered to have high explanatory power (

Table 7. Eigenvalues and Variances of Principal Component Analysis Results by Factor.

	Eigenvalues	%Variance	%Cumulative
component 1	12.406	42.780	42.780
component 2	3.743	12.907	55.687
component 3	3.093	10.664	66.351
component 4	2.271	7.832	74.184
component 5	1.733	5.977	80.161
component 6	1.331	4.589	84.750
component 7	1.193	4.115	88.865
component 8	0.983	3.390	92.256
~	~	~	~

).

Investigating the rotated component matrix of principal component analysis, we define each principal component through the indicators corresponding to them. Typically, in the component matrix, factors with an absolute value of factor loadings exceeding 0.5 are defined as significant factors for each principal component, and the indicators corresponding to each principal component are as follows (

Table 8. Rotated Component Matrix of Principal Component Analysis.

Urban Structure Assessment Indicators	Principal Component 1	Principal Component 2	Principal Component 3	Principal Component 4
Indicator ①-1	0.939	0.109	−0.051	−0.183
Indicator ①-2	−0.030	0.871	−0.080	−0.115
Indicator ①-3	0.955	0.070	−0.025	−0.101
Indicator ①-4	0.958	−0.064	0.123	−0.011
Indicator ①-5	0.949	−0.023	0.203	0.003
Indicator ①-6	0.620	0.542	0.185	−0.204
Indicator ①-7	0.800	0.394	0.237	−0.246
Indicator ①-8	0.853	0.380	0.244	−0.210
Indicator ①-9	0.815	0.426	0.178	−0.238
Indicator ①-10	0.721	0.570	0.119	−0.292
Indicator ①-11	0.299	0.002	0.780	−0.064
Indicator ①-12	−0.640	−0.322	−0.445	−0.057
Indicator ②-1	0.646	0.272	0.315	−0.080
Indicator ②-2	0.113	0.583	−0.218	0.233
Indicator ②-3	−0.117	−0.286	−0.024	0.497
Indicator ②-4	0.631	0.646	0.146	−0.118
Indicator ②-5	0.131	0.039	−0.036	0.777
Indicator ②-6	0.277	0.027	−0.136	−0.400
Indicator ③-1	−0.074	0.146	−0.279	0.792
Indicator ③-2	0.409	−0.265	−0.120	−0.514
Indicator ③-3	−0.177	−0.007	0.086	0.917
Indicator ④-1	−0.502	−0.535	−0.025	−0.262
Indicator ④-2	0.107	−0.065	0.843	0.120
Indicator ④-3	−0.278	0.008	−0.283	0.242
Indicator ④-4	0.235	0.474	0.736	−0.139
Indicator ⑤-1	−0.364	−0.669	−0.493	−0.081
Indicator ⑤-2	0.417	0.327	0.692	−0.167
Indicator ⑤-3	−0.203	−0.732	−0.563	−0.119
Indicator ⑥-1	−0.236	−0.153	0.891	0.039

).

In the principal component analysis results, Principal Component 1 is characterized by a significant loading of indicators related to everyday life services, ranging from the population ratio within the service area of daily life services to the population density within the public transportation service area. Therefore, it is named “Accessibility of Daily Life Services”.

Principal Component 2, on the other hand, is characterized by a significant loading of indicators related to local-industry-related economic indicators such as “④-1 Sales of Tertiary Industries per Person” and “⑤-1 Expenditure per Person”, “⑤-3 Tax Revenue per Person”. Additionally, it includes indicators related to administrative operation and local finance; thus, it is named “Local Economic Stability”.

Principal Component 3, with a significant loading of indicators such as “①-11 Modal Share of Public Transportation” and “⑥-1 Automobile CO₂ Emissions per Person”, is named “Automobile Dependency” due to its association with the dependency on automobiles in the target area.

Lastly, Principal Component 4, characterized by significant loadings of indicators related to safety aspects such as “③-1 Proportion of Population in Disaster-Prone Areas” and “③-3 Accessibility to Earthquake Evacuation Shelter ”, as well as indicators related to safety in the transportation sector like “③-2 Number of Traffic Accident Deaths per 10,000 Citizens ”, is named “Safety of Living Environment” (

Table 9. Urban Structure Evaluation Indicators by Principal Component.

Principal Component	Urban Structure Evaluation Indicators
Principal Component 1 (Accessibility of Daily Life Services)	①-1 Proportion of Population within the Service Area of Daily Life Services ①-3 Population Density within Walking Distance of Medical Facilities ①-4 Population Density within Walking Distance of Welfare Facilities ①-5 Population Density within Walking Distance of Commercial Facilities ①-6 Proportion of Population within the Service Area of Public Transportation ①-7 Population Density within Walking Distance of Medical Facilities ①-8 Population Density within Walking Distance of Welfare Facilities ①-9 Population Density within Walking Distance of Commercial Facilities ①-10 Population Density within the Service Area of Public Transportation ②-1 Proportion of Elderly Population within the Service Area of Senior Center
Principal Component 2 (Local Economic Stability)	④-1 Sales of Tertiary Industries per Person ⑤-1 Expenditure per Person ⑤-3 Tax Revenue per Person
Principal Component 3 (Automobile Dependency)	①-11 Modal Share of Public Transportation ⑥-1 Automobile CO2 Emissions per Person
Principal Component 4 (Safety of Living Environment)	③-1 Proportion of Population in Disaster-Prone Areas ③-2 Number of Traffic Accident Deaths per 10,000 Citizens ③-3 Accessibility to Earthquake Evacuation Shelter

).

5.2. Cluster Analysis Results

In this study, cluster analysis was conducted to typify the study areas based on the principal component scores of urban structure evaluation indicators deemed significant, extracted from four principal components. For the method of cluster analysis, the Ward’s method, which merges clusters based on the sum of squared deviations within clusters, was utilized using IBM’s SPSS Statistics program. The measurement of distances between objects was conducted using Euclidean distance (Figure 6).

The clustering analysis resulted in the classification of target cities into five clusters. The cities belonging to each cluster and the average scores of principal components for each cluster are presented in

Table 10. Average Scores of Principal Components by Cluster Type.

Cluster	City Name	Principal Component 1 (Lifestyle Services)	Principal Component 2 (Local Economy)	Principal Component 3 (Transportation)	Principal Component 4 (Safety)
Type A	Mungyeong-si, Ulji-gun, Goryeong-gun	5.85	−3.26	−0.35	−11.45
Type B	Yeongju-si, Yeongcheon-si	6.47	−2.625	−1.34	−9.43
Type C	Sangju-si, Cheongdo-gun, Bonghwa-gun	4.78	−3.55	−1.62	−13.54
Type D	Gunwi-gun, Cheongsong-gun, Yeongyang-gun, Yeongdeok-gun	4.68	−4.59	−1.72	−10.36
Type E	Uiseong-gun, Seongju-gun, Yecheon-gun	4.29	−3.17	−3.15	−11.11

.

The identified clusters were categorized based on key variables that reflect the unique characteristics of each city. For example, “Type A” represents cities with high accessibility to daily services, highlighting the importance of policies that enhance and maintain this accessibility. Similarly, other types reflect varying levels of urban compactness, economic stability, and safety, which provide insights into tailored urban planning strategies for each cluster (

Table 11. Principal Component Scores and Auxiliary Indicators by Type.

Cluster	Population (Persons)	Area (km2)	Population Density (Persons/km2)	Aging Rate (%)	Principal Component 1	Principal Component 2	Principal Component 3	Principal Component 4
Type A	51,687	761.6	72	30	5.85	−3.26	−0.35	−11.45
Type B	104,444	794.4	135	27.8	6.47	−2.625	−1.34	−9.43
Type C	57,726	1050.8	55	35.3	4.78	−3.55	−1.62	−13.54
Type D	25,776	754.1	35	38.4	4.68	−4.59	−1.72	−10.36
Type E	51,143	817.7	67	34.7	4.29	−3.17	−3.15	−11.11

).

For Type A areas, the urban structure evaluation indicators commonly show that indicators in the convenience of living sector are more positive than average, while indicators in other sectors are generally at a similar level to the average. Thus, Type A is named “Concentrated Living Services Type”.

Type B areas have the highest total population and population density, with the lowest aging rate. In terms of principal component scores, these areas measure highest in accessibility to living services, regional economic stability, and safety of the living environment, while showing the lowest dependency on cars. Urban structure evaluations for cities within Type B also show more positive figures in all aspects compared to cities of other types. Accordingly, Type B is named “Urban Function Stability Type”.

Type C areas show a larger average urban area compared to other types, with lower accessibility to general services facilities, including living services. Specifically, accessibility to safety-related evacuation facilities is lower than in the other types, and urban structure evaluation indicators related to safety, such as population density in natural disaster risk areas and the number of traffic accident fatalities per 10,000 citizens, show more negative figures than the average. Therefore, Type C is named “Reduced Living Environment Safety Type”.

Type D areas have the lowest average population and population density among the types, with the highest aging rate, indicating a smaller working-age population (ages 15−64). The primary industries in cities within Type D are centered around primary industries like agriculture and fishing, or secondary industries such as manufacturing. With a small population, high aging rate, and an industry structure focused on agriculture, fishing, and manufacturing, Type D scored low in the regional economy part of the principal components and showed negative figures in overall indicators, including administrative operations and regional economy, on the urban structure evaluation indicators. Thus, Type D is named “Primary-Industry-Centered Type (Figure 7).

5.3. Analysis of Characteristics by Type

When examining the urban structure evaluation indicators for the Concentrated Living Services Type, it can be observed that the indicators in the convenience of living sector generally appear higher than average. The population density and the proportion of the population within walking distance to living service facilities (medical, welfare, commercial facilities) are higher than the average, and the population density and proportion within the public transportation service area are also above average, indicating high accessibility to everyday living services (

Table 12. Average Values of Urban Structure Evaluation Indicators by Type.

Indicator		Type A	Type B	Type C	Type D
Convenience of Living	①	59.3	64.5	45.7	44.5
	②	45.3	58.5	47.7	47
	③	59	62.5	46.3	47.8
	④	56.3	60.5	48.3	46.8
	⑤	58.7	60.5	49	46.8
	⑥	56	67.5	45.3	43.5
	⑦	55.3	69.5	45.7	43.3
	⑧	57	68	47	43
	⑨	54.3	69	46.7	43.3
	⑩	55	65.5	46.3	44.5
	⑪	58	58.5	49	44.3
	⑫	50.3	36	48.3	58
Health and Welfare	⑬	56	58.5	45	45.8
	⑭	52.7	46.5	54.3	47
	⑮	53	40	52.7	52.8
	⑯	54.3	66	48.7	42.3
	⑰	47	45	63.3	46
	⑱	53	50	46.3	52.5
Safety	⑲	52	38	64.3	45.3
	⑳	60	38	45	48.5
	㉑	46	54	63.7	45.8
Regional Economy	㉒	52.7	40	45.3	55.3
	㉓	50.7	40.5	47.7	44.8
	㉔	50	41.5	53.3	51
	㉕	49.7	63	47.3	43.8
Administrative Operations	㉖	46.7	38	48.3	61.8
	㉗	56	65	44.3	40.8
	㉘	49	38	48.3	61
Energy Low Carbon	㉙	46	54	49	46.25

and

Table 13. Strengths and Weaknesses by Type through Urban Structure Evaluation Indicators.

Type		Evaluation
Concentrated Living Services Type	Strengths	①-1 Proportion of Population within the Service Area of Daily Life Services (59.33) ①-3 Population Density within Walking Distance of Medical Facilities (59) ①-4 Population Density within Walking Distance of Welfare Facilities (56.33) ①-5 Population Density within Walking Distance of Commercial Facilities (58.67) ①-6 Proportion of Population within the Service Area of Public Transportation (56) ①-8 Population Density within Walking Distance of Welfare Facilities (57) ①-10 Population Density within the Service Area of Public Transportation (55) ②-1 Proportion of Elderly Population within the Service Area of Senior Center (56) ⑤-2 Local Government Fiscal Independence Rate (56)
	Weaknesses	-
Urban Function Stability Type	Strengths	①-1 Proportion of Population within the Service Area of Daily Life Services (64.5) ①-2 Population Density in Urban Areas (58.5) ①-3 Population Density within Walking Distance of Medical Facilities (62.5) ①-4 Population Density within Walking Distance of Welfare Facilities (60.5) ①-5 Population Density within Walking Distance of Commercial Facilities (60.5) ①-6 Proportion of Population within the Service Area of Public Transportation (67.5) ①-7 Population Density within Walking Distance of Medical Facilities (69.5) ①-8 Population Density within Walking Distance of Welfare Facilities (68) ①-9 Population Density within Walking Distance of Commercial Facilities (69) ①-10 Population Density within the Service Area of Public Transportation (65.5) ①-11 Modal Share of Public Transportation (58.5) ①-12 Average Daily Mileage of a Private Car (36) ②-1 Proportion of Elderly Population within the Service Area of Senior Center (58.5) ②-4 Proportion of Population in Service Area of Neighborhood Park (66) ③-1 Proportion of Population in Disaster-Prone Areas (38) ③-3 Accessibility to Earthquake Evacuation Shelter (40) ④-3 Average Official land Price of Residential Areas (41.5) ④-4 Proportion of Empty House (63) ⑤-1 Expenditure per Person (38) ⑤-2 Local Government Fiscal Independence Rate (65) ⑤-3 Tax Revenue per Person (38)
	Weaknesses	②-5 Proportion of Population in Service Area of Public Sports Facilities (45) ④-1 Sales of Tertiary Industries per Person (40.5)
Reduced Living Environment Safety Type	Strengths	②-5 Proportion of Population in Service Area of Public Sports Facilities (63.3)
	Weaknesses	②-6 Walking Rate of the Elderly (Over 65 years old) (43.3) ③-1 Proportion of Population in Disaster-Prone Areas (64.33) ③-3 Accessibility to Earthquake Evacuation Shelter (63.67)
Primary-Industry-Centered Type	Strengths	-
	Weaknesses	①-1 Proportion of Population within the Service Area of Daily Life Services (44.5) ①-6 Proportion of Population within the Service Area of Public Transportation (43.5) ①-7 Population Density within Walking Distance of Medical Facilities (43.25) ①-8 Population Density within Walking Distance of Welfare Facilities (43) ①-9 Population Density within Walking Distance of Commercial Facilities (43.25) ①-11 Modal Share of Public Transportation (44.25) ①-12 Average Daily Mileage of a Private Car (58) ⑤-1 Expenditure per Person (61.75) ⑤-2 Local Government Fiscal Independence Rate (40.75) ⑤-3 Tax Revenue per Person (61)
Urban Dispersion Type	Strengths	③-2 Number of Traffic Accident Deaths per 10,000 Citizens (43.67)
	Weaknesses	①-1 Proportion of Population within the Service Area of Daily Life Services (39.67) ①-3 Population Density within Walking Distance of Medical Facilities (39.67) ①-4 Population Density within Walking Distance of Welfare Facilities (41) ①-5 Population Density within Walking Distance of Commercial Facilities (40) ①-8 Population Density within Walking Distance of Welfare Facilities (43.67) ⑥-1 Automobile CO2 Emissions per Person (58)

).

In the fields of health and welfare, safety, regional economy, and energy and low carbon, there were no indicators with values significantly high or low enough to be considered strengths or weaknesses. The population size, urban area, population density, and aging rate were also similar to the average of the types, showing no significant characteristics other than accessibility to living services.

The characteristic feature of the urban structure evaluation indicators for the Urban Function Stability Type is that overall, all sectors of the urban structure evaluation indicators produced positive results. While the percentage of population within the public sports facility service area (45) and sales per person employed in the tertiary industry (40.5) were negative, these are not common figures across the cities of this type, so it is difficult to regard them as definitive indicators of the type’s characteristics.

Consequently, for the Urban Function Stability Type, it can be interpreted that the overall urban structure is well consolidated and the urban functions are likely to be maintained stably.

The Reduced Living Environment Safety Type does not have indicators that can be considered strengths, except for the percentage of the population within the public sports facility service area (63.3). However, this indicator is only characteristic of Sangju-Si among the cities of this type, making it difficult to consider it a strength of the type. Indicators that could be seen as weaknesses include the percentage of the population living in natural disaster risk areas (64.33) and the number of traffic accident fatalities per 10,000 citizens (63.67), both of which are safety indicators that consistently show negative figures.

This indicates that the cities of this type have lower safety levels in their living environments compared to other types. Thus, it seems necessary to develop measures to create an environment where citizens can live safely and securely.

The urban structure evaluation indicators for the Primary-Industry-Centered Type generally appear average or negative, lacking indicators that could be considered strengths. This type is characterized by cities whose main industries are centered around primary industries like agriculture and fishing or manufacturing, with a population size of 20,000 to 30,000, which is very small even among small- and medium-sized cities with a population of less than 100,000. It has the lowest population density (35) and the highest aging rate (38.4) compared to other types.

Examining the urban structure evaluation indicators for this type, the population density within the service area of everyday living services and facilities overall appears lower than the average, similar to the city’s population size and density. The indicators in the health and welfare, safety, regional economy, and energy and low-carbon sectors generally show average or negative figures. In terms of administrative operations, the amount of tax revenue and expenditure is very high compared to the average, while the fiscal self-reliance ratio is very low at 40.8. This suggests a need for efficient administrative management strategies and measures to revitalize the local economy for stable urban management.

The Urban Dispersion Type has the lowest average score for accessibility to living services among the types in this study, at 4.29, indicating the least accessibility to living services. Moreover, its average score for car dependency is −3.15, the highest dependency on cars among the types. In contrast, the Concentrated Living Services Type, similar in population and urban size to the Urban Dispersion Type, shows a higher average score for accessibility to living services at 5.85, indicating better accessibility, and the lowest car dependency score among the types at −0.35.

Urban structure evaluation indicators also show that cities of the Urban Dispersion Type display weaknesses in indicators related to the convenience of living. From the percentage of the population within the service area of everyday living services to the population density and proportion within walking distance to living service facilities, indicators related to accessibility to living services are lower than the average for the target areas. Conversely, indicators related to car dependency, such as the average daily driving distance per car (45.33) and CO₂ emissions per citizen from cars (58), are higher than the average for the target areas.

Indicators in fields other than convenience of living, energy and low carbon show figures similar to the average for the target areas. Such Urban Dispersion Type cities may need to focus on creating regional hubs to encourage residential and urban functions, thereby consolidating the dispersed urban areas.

6. Conclusions

This study was conducted under the premise that forming a compact urban structure should be aimed at as a solution to urban problems caused by population decline. It sought to examine the current status and necessity of forming such structures in domestic regional small- and medium-sized cities. To this end, the study utilized the urban structure evaluation system of Japan, which is pursuing urban regeneration policies in the face of similar population decline phenomena and applied it to domestic regional small- and medium-sized cities to review their current status and the necessity for compact urban structure formation. Furthermore, based on the results of the urban structure compactness evaluation, typification of the target areas and analysis of characteristics by type were conducted to propose the direction and implications of urban planning for each type.

The comprehensive results of the urban structure compactness evaluation indicate that the urban structures of Yeongju-Si, Yeongcheon-Si, and Mungyeong-Si are generally well compacted. In contrast, Cheongsong-gun, Yeongyang-gun, and Bonghwa-gun appear to require overall improvements to their urban structures and consideration for the formation of compact urban structures.

The results can be summarized as follows (

Table 14. Relative Comparison of Urban Structural Compactness among Target Cities.

City	Field
	Convenience of Daily Living	Health and Welfare	Safety	Local Economy	Municipal Governance	Energy and Low Carbon
Yeongju-si	◎	◎	◎	◎	◎	○
Yeongcheon-si	◎	○	◎	◎	◎	●
Sangju-si	○	○	●	○	○	○
Mungyeong-si	◎	○	○	○	◎	◎
Gunwi-gun	●	●	○	◎	●	○
Uiseong-gun	●	●	○	○	○	○
Cheongsong-gun	●	●	○	●	●	○
Yeongyang-gun	●	●	◎	●	●	○
Yeongdeok-gun	○	○	◎	●	○	◎
Cheongdo-gun	○	○	●	●	○	○
Goryeong-gun	○	○	○	◎	◎	●
Seongju-gun	○	○	○	◎	○	●
Yecheon-gun	●	○	○	○	○	○
Bonghwa-gun	●	●	●	●	●	◎
Uljin-gun	○	○	○	●	○	◎

◎ Strong/○ Average/● Weak.

).

Based on the analysis of urban structure compactness, this study classified domestic regional small- and medium-sized cities into five types (Concentrated Living Services Type, Urban Function Stability Type, Reduced Living Environment Safety Type, Primary-Industry-Centered Type, Urban Dispersion Type) through principal component analysis and cluster analysis. The results of the type-specific urban structure compactness analysis, using principal component scores and auxiliary indicators such as population size, population density, and aging rate, were utilized to analyze the characteristics of each type and derive future directions for urban planning. In terms of directions for urban planning by type, it was observed that for the Concentrated Living Services Type and Urban Function Stability Type, the overall urban structure, including the area of convenience of living, is well concentrated. Although there does not seem to be a necessity for forming compact urban structures, preventive measures against densification are needed in anticipation of continued population decline. For the Reduced Living Environment Safety Type, indicators related to safety and principal component scores showed negative figures, indicating the need for measures such as strengthening disaster prevention plans, improving pedestrian environments, and encouraging residency in safer areas. For the Primary-Industry-Centered Type, exploring measures to revitalize the local economy to attract and support new higher-order industries and considering overall urban structure concentration in light of low population density and high aging rates were seen as necessary. Lastly, for the Urban Dispersion Type, low accessibility to everyday living services due to dispersed population distribution was observed. Thus, instead of expanding living service facilities, setting zones to induce residential and urban functions to concentrate the urban areas might be needed.

This research is significant in that it provides the first objective and quantitative evaluation of the current state of urban structures in South Korea’s regional small- and medium-sized cities. This can facilitate the setting of common goals and consensus among various local stakeholders, including citizens, business operators, and administrators, within South Korea. Additionally, it can serve as foundational data for efficiently deriving future directions when establishing urban plans for the target cities.

Future research should aim to conduct further studies that analyze the characteristics of the target cities outlined in this study in more detail. Expanding the scope of analysis from being limited to Gyeongsangbuk-do to nationwide, enabling comparative analyses with similar-sized cities across the country, could yield interesting findings. Through such additional research, the urgent discussion on what specific models of urban compactness are needed for South Korea’s regional small- and medium-sized cities should be conducted.