Public Transport and the Scale of Perceived Neighbourhood Boundaries Among Older Adults in a Super-Aged Megacity: Evidence from Seoul, South Korea

Abstract

8 Neighbourhoods are central to residents' well-being, yet the conceptualisation and operationalisation of 9 their boundaries remain contested. Although prior studies have examined various determinants of 10 neighbourhood boundary scales, empirical evidence on the role of public transport—particularly service 11 satisfaction and inclusiveness for vulnerable groups—is limited, especially in relation to older adults 12 who often face mobility constraints.

13 Using the 2024 Seoul Survey and statistical modelling, this study examines associations between 14 public transport satisfaction, service inclusiveness, and perceived neighbourhood boundary scale, and 15 whether these differ between the young-old (65–74) and old-old (75+).

16 Transport inclusiveness for vulnerable groups was consistently and positively associated with 17 perceived neighbourhood scale across both age groups; overall service satisfaction was not. Older 18 adults in areas with high local accessibility tended to perceive smaller, more localised boundaries.

19 These findings indicate that targeted inclusiveness improvements—such as barrier-free design, step20 free vehicles, and priority seating—are more effective in expanding older adults' perceived 21 neighbourhood boundaries than general service quality enhancements.

23 Keywords: Public transport, neighbourhood boundaries, older adults

25 1. INTRODUCTION

26 Neighbourhoods constitute a fundamental spatial unit shaping everyday life and have long occupied a 27 central position across disciplines including urban planning, public health, and transport research 28 (Kallus and Law-Yone, 2000, Kawachi and Subramanian, 2007, Cerin et al., 2017). As the primary 29 arena in which residents conduct daily activities, access services, and form social ties, neighbourhoods 30 exert substantial influence on quality of life, health, and well-being (Hong et al., 2026, O’Campo et al., 31 2009). Consequently, neighbourhood improvement has been a persistent objective of urban policy 32 interventions worldwide. Despite broad agreement regarding their importance, however, the 33 conceptualisation and operationalisation of neighbourhood boundaries and scales remain theoretically 34 and methodologically contested.

35 A central challenge lies in the inherently subjective and socially constructed nature of neighbourhoods, 36 which are not fixed administrative units but lived spatial experiences. A growing body of research 37 demonstrates that perceived neighbourhood boundaries vary according to individual characteristics,

1 social position, and broader political and environmental contexts (Finlay et al., 2023, Baffoe, 2019). 2 This variation is not merely conceptual; it carries substantive implications, particularly for vulnerable 3 populations. For instance, older adults or pregnant women may experience more constrained activity 4 spaces in their daily lives. Although smaller perceived neighbourhoods are not inherently indicative of 5 social exclusion, empirical evidence suggests that larger perceived neighbourhoods are often associated 6 with greater walkability, more frequent physical activity, improved access to various destinations, and 7 stronger social integration (Finlay et al., 2023). Consequently, systematic differences in perceived 8 neighbourhood scale may contribute to inequalities in access and social participation, thereby wellbeing.

9 These issues become especially salient in ageing societies. Older adults consistently report smaller 10 perceived neighbourhood boundaries than younger adults (Charreire et al., 2016), partly reflecting age11 related declines in physical capacity, mobility, and social participation. In addition, older populations 12 are far from homogeneous. The determinants and consequences of neighbourhood scale may differ 13 substantially between the young-old (65–74) and the old-old (75+), reflecting disparities in health status, 14 mobility resources, and transport dependency. From a policy perspective, transport systems— 15 particularly public transport—may play a crucial role in expanding activity spaces by enabling older 16 adults to travel beyond their immediate residential surroundings. Yet despite extensive research on built17 environment correlates of neighbourhood perception, empirical evidence on the role of public transport 18 in shaping perceived neighbourhood scales among older adults remains limited, especially in super19 aged mega cities.

20 This study addresses these gaps using the 2024 Seoul Survey, a representative dataset sponsored by the 21 Seoul Metropolitan Government. Employing multivariate statistical models, this study examined how 22 two public transport–related dimensions—(1) overall satisfaction with the public transport system and 23 (2) perceived inclusiveness of vulnerable populations in the transport system—are associated with the 24 size of perceived neighbourhoods. Importantly, this study compared these relationships between young25 old and old-old adults, thereby capturing heterogeneity within the older population.

26 This study makes three primary contributions. First, it advances neighbourhood research by shifting 27 attention from static built-environment indicators to mobility-enabling institutional factors, highlighting 28 public transport as a key mechanism shaping perceived spatial size of neighbourhoods. Second, it 29 introduces an intra-generational ageing perspective by systematically comparing young-old and old-old 30 groups, moving beyond the conventional treatment of older adults as a homogeneous category. Third, 31 it provides robust empirical evidence based on a large, representative metropolitan sample, thereby 32 strengthening the external validity of findings in a rapidly ageing, high-density urban context.

34 2. DATA AND METHOD 35 2.1. Data

36 The study area is Seoul, South Korea (population ~9.3 million), which operates an extensive public 37 transport system comprising 308 subway stations and approximately 6,600 bus stops. In 2024, daily 38 public transport ridership averaged 9.4 million passengers, representing 64.4% of total modal share. 39 The Seoul Metropolitan Government additionally provides free urban railway services for residents 40 aged 65 and over.

41 Data were drawn from the 2024 Seoul Survey, an annually conducted survey funded by the Seoul 42 Metropolitan Government since 2003, covering quality of life, public policy impacts, socio43 demographics, health, and neighbourhood conditions.

1 The dependent variable in this study is the scale of perceived neighbourhood boundaries among older 2 adults. The survey asks respondents: “How do you define the spatial boundary of your neighbourhood?” 3 Responses were measured using four categories: (1) the same apartment building (a residential area 4 within approximately a 15-minute walking distance); (2) Dong, the smallest administrative district in 5 Seoul (there are 426 Dongs, with an average size of 1.42 km²); (3) Gu, the district-level administrative 6 unit (there are 25 Gus in Seoul, with an average size of 24.2 km²); and (4) Seoul as a whole. For the 7 analyses, the variable was recategorised into three groups by combining the Gu and Seoul as a whole 8 categories due to the relatively small sample size in the latter.

9 To measure perceived neighbourhood environment, four survey items were used and validated through 10 confirmatory factor analysis (CFA). Respondents were asked: “To what extent do you agree or disagree 11 with the following statements about the environment within a 1 km radius (approximately a 15-minute 12 walk) of your home?” The four statements were: (1) My neighbourhood is suitable for physical activities 13 such as walking, running, and strolling; (2) My neighbourhood is convenient for grocery shopping and 14 buying daily necessities; (3) My neighbourhood has adequate public facilities; and (4) My 15 neighbourhood is safe. Responses were measured on a five-point Likert scale, ranging from strongly 16 disagree to strongly agree.

17 Public transport factors were captured through two dimensions. First, public transport satisfaction was 18 measured for the two main modes in Seoul—bus and subway. The survey asked respondents about their 19 level of satisfaction with different transport modes in Seoul, including bus, subway, taxi, and intercity 20 and express buses. Responses were recorded on a five-point Likert scale, ranging from very dissatisfied 21 to very satisfied. Only responses related to bus and subway services were included in this study, as these 22 represent the primary public transport modes for intra-city daily travel. A CFA was conducted to 23 construct a latent variable capturing overall public transport satisfaction. Second, public transport 24 inclusiveness was measured using the question: “Do you think that transport modes in Seoul adequately 25 accommodate vulnerable groups (such as children, older adults, people with disabilities, and pregnant 26 women)?” Responses were recorded on a five-point Likert scale ranging from strongly disagree to 27 strongly agree. For the analysis, responses were recategorised into three groups (i.e., disagree, neutral, 28 and agree). In addition, a range of socio-demographic factors were included as control variables in the 29 analysis.

30 Descriptive statistics (Table 1) indicate that the young-old (65–74) exhibit more favourable 31 socioeconomic status, better health, and larger household sizes than the old-old (75+), consistent with 32 expected ageing effects. Both groups report similar levels of transport satisfaction and inclusiveness, 33 though the young-old express marginally higher subway satisfaction.

35 Table 1 Descriptive statistics

Young_old (65-74) Old_old (75+) Mean SD Min Max Mean SD Min Max

Socio-demographic characteristics Age 68.88 2.79 65 74 78.85 3.77 75 97 Sex (male=1) 0.48 0.5 0 1 0.45 0.5 0 1 Social status (1: very low-6 very high) 2.95 0.89 1 6 2.69 0.93 1 6 Health 6.37 1.5 1 10 5.79 1.65 1 10

Household size 1.89 0.71 1 7 1.75 0.82 1 8 Perceived neighbourhood environment (1: Strongly disagree – 5: Strongly agree) Good for exercise 3.46 0.85 1 5 3.4 0.82 1 5 Convenient for grocery shopping 3.4 0.88 1 5 3.35 0.89 1 5 Sufficient public facilities 3.23 0.9 1 5 3.13 0.91 1 5 Safety 3.5 0.76 1 5 3.4 0.77 1 5 Public transport factors (1: Strongly disagree – 5: Strongly agree) Bus satisfaction 3.93 0.78 1 5 3.93 0.78 1 5 Subway satisfaction 3.82 0.91 1 5 3.72 0.97 1 5 Public transport inclusiveness 2.5 0.66 1 3 2.49 0.67 1 3 Sample size 6067 1816

2 2.2. Analytical model

3 To examine the associations between public transport satisfaction, inclusiveness of vulnerable groups, 4 and the scale of perceived neighbourhood boundaries, multinomial logistic regression models were 5 employed. The probability that individual i chooses neighbourhood boundary scale m, conditional on a 6 vector of explanatory variables 𝑥𝑖 is defined as: exp (𝛽𝑚 𝑥𝑖 ) 7 Pr( 𝑦𝑖 = 𝑚|𝑥𝑖 ) = (1) ∑𝐽𝑗=1 exp (𝛽𝑗 𝑥𝑖 )

8 where 𝑦𝑖 denotes the scale of perceived neighbourhood boundaries with three categories: (1) the same 9 apartment building—a residential area within approximately a 15-minute walking distance (reference 10 category), (2) Dong, and (3) Gu or the entire city of Seoul. The independent variables 𝑥𝑖 includes socio11 demographic characteristics, perceived neighbourhood environment, and public transport factors 12 (satisfaction and inclusiveness). Figure 1 shows the analytical framework.

15 Figure 1 Framework

2 To account for potential heterogeneity within the older population, the analyses were conducted 3 separately for two age groups: the young-old (65–74) and the old-old (75+).

5 3. RESULTS AND DISCUSSION

6 Model fit for the full sample, as well as for the young-old and old-old subsamples, was assessed using 7 standard fit indices (CFI, TLI, RMSEA, and SRMR), all of which indicate a good fit to the data. All 8 indicators load significantly on their respective latent constructs (p < 0.001).

9 Factor loadings for perceived neighbourhood environment are positive and range from 0.61 to 0.73, 10 suggesting satisfactory construct validity. This indicates that higher values of the latent variable 11 represent neighbourhoods with favourable environmental conditions, including supportive 12 infrastructure for physical activity, good access to local facilities, and a safe environment.

13 Public transport satisfaction is strongly represented by satisfaction with subway services (0.81) and bus 14 services (0.61). Interestingly, the factor loading for subway satisfaction is higher among the young-old 15 group than the old-old group, whereas the opposite pattern is observed for bus satisfaction. This pattern 16 may reflect accessibility differences between transport modes. Subway stations are typically located 17 underground and often require substantial vertical movement, which may present barriers for older 18 adults with reduced mobility.

20 Table 2 Confirmative factor analyses (CFAs)

All Young-old Old-old Estimate P-value Estimate P-value Estimate P-value Perceived neighbourhood environment Good for exercise 0.70 0.00 0.71 0.00 0.64 0.00 Convenient for 0.61 0.00 0.63 0.00 0.57 0.00 grocery shopping Sufficient public 0.73 0.00 0.73 0.00 0.73 0.00 facilities Safety 0.69 0.00 0.70 0.00 0.67 0.00 Public Transport satisfaction Bus satisfaction 0.61 0.00 0.60 0.00 0.64 0.00 Subway 0.81 0.00 0.82 0.00 0.77 0.00 satisfaction Model fit CFI 0.997 0.998 0.994 TLI 0.995 0.996 0.990 RMSEA 0.035 0.032 0.045 SRMR 0.021 0.019 0.029