Pedestrians’ Perceptions of Walkability Determinants: A Comparative Study of Bologna and Porto Through the Lens of the 15-Minute City

Abstract

The 15-minute city (15MC) promotes compact and accessible urban environments where residents can meet their daily needs through active mobility. Research on the 15MC has predominantly used objective assessments, with less attention to residents’ perceptions of the urban environment. Drawing on a questionnaire survey in Bologna and Porto (N = 1438), this study examines how pedestrians prioritize and evaluate walkability determinants and attributes, explores their relationship with walking behavior, and interprets selected attributes through the conceptual framework of the 15MC. Data were analyzed using descriptive statistics, chi-square tests, Mann–Whitney U tests, and Spearman’s rank correlations. The results reveal broadly similar perceptions across both cities, with pedestrian infrastructure and proximity to community facilities receiving the highest evaluations. Through the 15MC lens, these priorities are mainly associated with proximity and connectivity. Respondents from Bologna reported more frequent walking, whereas those from Porto assigned greater importance to urban ambiance and access to other transport modes. Perceived walkability was only weakly associated with walking behavior. The findings provide comparative evidence on perceptions of walkability attributes associated with selected 15MC principles in two European cities, revealing broadly similar priorities alongside differences in the importance assigned to specific attributes.

1. Introduction

This introductory outlines the challenges of urbanization and urban sprawl and their implications for sustainable mobility. It then reviews the principles and conceptual foundations of the 15-minute city (15MC). Finally, it identifies the research gaps addressed by this study and presents its objectives and contributions.

1.1. Urbanization, Urban Sprawl and Sustainability Challenges

The European Union (EU) is one of the most urbanized regions in the world. Of its 447 million inhabitants, approximately 75% live in urban areas, with a trend to exceed 80% by 2050 [1]. Rapid urbanization in many cases led to urban sprawl, as a result of uncontrolled expansion of cities into the surrounding rural areas. Urban sprawl has been increasing across all European countries, leading to heightened environmental, social, and economic problems. Specifically, Behnisch et al. [2] show that Europe is among the continents most affected by urban sprawl, which increased by 51% between 1990 and 2014.

The growth of urban sprawl, together with car-oriented urban development, has generated significant environmental and social negative externalities [3]. These include excessive land consumption, the reduction in urban green spaces, air pollution, longer travel times, poorer-quality public spaces, unequal access to services and opportunities, and increasingly sedentary lifestyles. In Europe, approximately 27% of greenhouse gas (GHG) emissions originate from the transportation sector, making it the second-largest emitting sector after energy production [4]. Around 72% of transportation-related emissions are generated by road transport, with private cars accounting for 61% of road transport emissions [5]. In response, the European Green Deal aims to reduce transport-related GHG emissions by 90% by 2050 compared to 1990 levels. Beyond their contribution to climate change, cars are also a major source of urban air pollution and noise. According to the World Health Organization, air and noise pollution are respectively the first and second most hazardous forms of pollution in urban areas, and they are associated with numerous adverse health impacts, including morbidity and premature mortality [6].

In response to these challenges, urban planning and mobility debates have increasingly emphasized the need for more sustainable, compact, inclusive, and people-oriented urban development models. Such urban planning models would be capable of reducing car dependency while improving accessibility, environmental quality, and social well-being [7,8,9]. Urban sprawl and the ecological footprints of cities can be mitigated through a variety of measures, particularly by promoting more compact and dense urban forms [2,10]. More recent approaches have progressively shifted attention from mobility-oriented paradigms toward accessibility and proximity-based models, emphasizing the reduction in travel demand through the spatial organization of urban functions [11,12]. In this context, proximity is increasingly understood not only as a spatial characteristic, but also as a social and environmental strategy aimed at improving quality of life, reducing transport-related emissions, and fostering healthier and more equitable urban environments [9,13]. The concept of the 15MC has emerged as one of the most influential contemporary paradigms of proximity-based urbanism and a key approach for achieving a net-zero urban future [2].

1.2. The 15-Minute City as a Proximity-Based Planning Concept

Popularized by Moreno et al. [14], the 15MC proposes that residents should be able to access essential daily needs, including employment, education, healthcare, commerce, and leisure, within a 15-minute (min) walk or bicycle ride from their homes [8,14,15]. The 15MC represents a significant conceptual shift in urban planning by prioritizing accessibility through proximity rather than through increasing travel speed or expanding transport infrastructure capacity [12]. In contrast to organizing cities around long-distance commuting patterns and functionally specialized urban zones, the model advocates for more compact, multifunctional, and decentralized urban structures capable of supporting everyday life within local neighborhoods. This proximity-oriented approach seeks to reduce dependence on private motorized vehicles while promoting active mobility modes such as walking and cycling [8,16].

Walking is the foundational mode of transport within the 15MC concept, as it represents the most universally accessible and environmentally sustainable form of mobility [15]. Consequently, some studies have framed concepts such as “15-min walkable city” [17] or “15-min walking environment” [18], emphasizing walking as the primary mode through which residents access daily services and urban opportunities. Compact urban morphologies characterized by dense street networks, mixed land uses, small urban blocks, and interconnected public spaces tend to facilitate pedestrian movement and enhance local accessibility [15,19]. Cycling further complements walking by enabling efficient medium-distance trips within urban areas [15]. From this perspective, sustainable mobility is achieved not only through improvements in active transport infrastructure, but also through the reconfiguration of urban form and the spatial distribution of opportunities, services, and amenities.

According to Moreno et al. [14], the 15MC is structured around four main principles: density, diversity, proximity, and digitalization. Density refers to the concentration of residents and activities required to sustain local services and urban vitality. Diversity concerns the coexistence of multiple land uses, social groups, and urban functions within neighborhoods. Proximity emphasizes short distances between residences and essential destinations, while digitalization relates to the use of information and communication technologies to facilitate access to services and optimize urban functioning. More recently, Khavarian-Garmsir et al. [20] expanded this framework to seven principles, by adding three other principles: human-scale urban design, connectivity, and flexibility. This framework emphasizes the importance of designing public spaces at the human scale, prioritizing walking and cycling, ensuring efficient connections between neighborhoods through active and public transport, and promoting flexible, multifunctional urban spaces capable of accommodating changing social and economic needs. Collectively, these principles provide a more comprehensive planning framework for developing walkable, accessible, and sustainable neighborhoods.

The diffusion of the 15MC concept crossed international borders and has stimulated numerous urban initiatives and policy experiments worldwide. Prominent examples include Paris’s “Ville du quart d’heure”, Barcelona’s Superblocks, Portland’s 20-min neighborhoods, Melbourne’s 20-min city, and China’s 15-min community life circles [12,21]. Many local governments have adopted the 15MC concept as a strategy to promote sustainable mobility, improve neighborhood accessibility, and strengthen local urban life [15]. At the same time, the growing popularity of the concept has generated an expanding body of academic research aimed at evaluating the accessibility conditions, urban forms, and policy implications associated with the 15MC [11,22,23].

Nevertheless, despite its rapid diffusion in both academic and policy debates, the 15MC has also been the subject of criticism and important conceptual discussions. Some authors question the feasibility of concentrating all daily activities within short travel distances, particularly in contemporary metropolitan regions characterized by complex labor markets, functional specialization, and highly differentiated mobility patterns [8,15]. Other critiques focus on the potential negative consequences associated with compact urban development, including excessive densification [2], rising property values and gentrification processes [8], as well as concerns that the 15MC may promote restrictive anti-car policies that limit individual mobility choices and personal freedom [24]. Moreover, the 15MC’s human-centered framing does not necessarily ensure equal accessibility for all population groups, particularly older people and people with disabilities or reduced mobility, whose diverse mobility needs may require specific adaptations of the built environment and pedestrian infrastructure [19,25]. Others argue that the mere existence of accessibility conditions does not necessarily lead residents to adopt active mobility practices [12]. For example, Birkenfeld et al. [26] found that even in compact urban environments, relatively few households perform all daily activities within short distances from their homes. Also Arifwidodo et al. [27] conclude that proximity to smaller parks in Bangkok, was not significantly associated with health outcomes despite these parks being within walking distance, suggesting that proximity alone does not necessarily translate into meaningful recreational use. Furthermore, socioeconomic characteristics such as gender, age, educational background, and mobility preferences may significantly influence the extent to which residents engage in local and active mobility practices [12].

1.3. Research Gaps, Objectives and Contributions of the Study

In recent years, the 15MC has gained significant attention in both research and planning practice, particularly across Europe [12,15]. Much of the existing literature on the 15MC has focused primarily on quantitative accessibility assessments based on Geographic Information System (GIS) analysis and spatial indicators measuring the proximity of services and amenities [11,22,23]. While these approaches provide valuable information regarding the spatial distribution of opportunities, accessibility is not solely determined by physical distance, but also by other dimensions, like the quality of the pedestrian and cycling infrastructure [28]. A neighborhood may perform well in terms of x-minute access, but few would walk or cycle, without pedestrian or bicycle-friendly infrastructure [29]. Therefore, applying the 15MC requires considering the quality of active mobility infrastructure, often overlooked in proximity analyses.

Moreover, the existing literature on the 15MC often overlooks the lived and subjective experiential dimensions of accessibility and proximity, resulting in insufficient evidence on how residents themselves perceive key principles of the 15MC [21]. Individuals have diverse priorities and needs and different perceptions about the influence of built environment on pedestrian safety and comfort. As highlighted by Guzman et al. [30], the role of these preferences and perceptions in the context of the 15MC has yet to be fully disentangled. Thus, while existing studies have predominantly focused on measuring accessibility mostly through metrics and GIS indexes [11], few have examined how residents evaluate these environments, which is crucial for assessing whether the 15MC aligns with their needs and preferences [12].

Europe is the region with the highest number of urban initiatives inspired by the 15MC model [31,32], which could be explained by the historical context of European cities characterized by dense networks of narrow streets, small blocks and mixed-use buildings where shops, offices and housing coexist [15]. However, there are limited research studies focusing on southern European cities. For example, in Portugal, scientific research on the 15MC remains scarce and fragmented, largely due to a delayed adoption of the concept, driven by the absence of national strategic guidelines [28,33].

The overall aim of this study is to compare Bologna and Porto in terms of the importance respondents assign to walkability determinants and attributes that can be conceptually related to selected principles of the 15MC, and to examine whether these perceptions are associated with self-reported walking behavior. This study is a continuation of our previous research based on the Smart Pedestrian Net (SPN) survey dataset, which identified and validated four perception-based walkability determinants through exploratory factor analysis [34]. Building upon that framework, the present study has three objectives: (i) to compare the importance attributed by respondents in Bologna and Porto to the identified walkability determinants and their associated built environment attributes; (ii) to examine the relationship between these evaluations and walking behavior; and (iii) to interpret the findings through the conceptual lens of the 15MC. None of these three aspects was explored in our previous study. By moving from the identification of walkability determinants in the pooled sample to a comparative analysis of Bologna and Porto and to an examination of their relationship with self-reported walking behavior, the present study provides new empirical evidence that was not addressed in the previous factor-analytic study. Beyond extending our earlier work, the present study contributes to the 15MC literature by examining residents’ perceptions of walkability attributes and interpreting these perceptions in relation to selected planning principles of the 15MC, an aspect that has received comparatively limited attention in the literature [12,30]. Furthermore, by focusing on Bologna and Porto, the study adds empirical evidence from Southern European cities, particularly Portugal, where research on the 15MC remains relatively scarce and fragmented [28,33]. To achieve this aim, the manuscript is guided by the following two research questions (RQ):

(RQ1) How do pedestrians in Bologna and Porto prioritize and evaluate walkability determinants and their associated built environment attributes that can be interpreted in relation to selected principles of the 15MC?

(RQ2) What is the relationship between the perceived importance of these walkability attributes and self-reported walking behavior?

2. Methods and Data

The following section elaborates on the methodology adopted in this study and it is organized into three subsections. Section 2.1 presents the conceptual framework and data source, Section 2.2 describes the data collection process, and Section 2.3 outlines the statistical procedures used to analyze the data and address the RQ.

2.1. Conceptual Framework and Data Source

This study builds upon a previous investigation conducted by the authors in the cities of Bologna and Porto, which used a questionnaire survey to evaluate the perceptions of 1438 individuals regarding the importance of 19 built environment and streetscape attributes for walking [34]. Through exploratory factor analysis, the study found that 59% of the total variance could be explained by four latent dimensions comprising 13 attributes (

Table 1. Walkability determinants, associated built environment attributes, and their correspondence with selected 15MC planning principles.

Walkability Determinants	Built Environment Attributes	Primary 15MC Principle
Urban ambiance	Residential density Mixed land uses Shopping streets Street enclosure Transparency Architectural and landscape diversity	Density Diversity Diversity Human-scale urban design Human-scale urban design Human-scale urban design
Pedestrian infrastructure	Sidewalks in good condition Unobstructed sidewalks Wide sidewalks	Proximity Proximity Proximity
Connectivity and community facilities	Proximity to community facilities Street connectivity	Proximity Connectivity
Access to other modes of transport	Proximity to public transport Proximity to car parking	Proximity —

). These dimensions (urban ambiance, pedestrian infrastructure, connectivity and community facilities, and access to other modes of transport) were identified as the main determinants of respondents’ perceived importance of walkability attributes. The internal consistency of these four walkability determinants was assessed in the previous study [34], obtaining Cronbach’s alpha values of 0.85 for urban ambiance, 0.79 for pedestrian infrastructure, 0.70 for connectivity and community facilities, and 0.68 for access to other modes of transport. These values indicate generally acceptable internal consistency. To further assess the comparability of the factor structure across the two cities, the four-component solution identified in the pooled sample was examined separately for Bologna and Porto. Tucker’s coefficient of congruence was used to assess the similarity between the pooled and city-specific factor structures. The coefficients indicated high structural congruence for all four components in each city (above 0.850), indicating substantial similarity between the factor structures and supporting the use of the four determinants as a common analytical framework for the inter-city comparisons.

The four walkability determinants were therefore used as a common analytical framework to compare respondents from Bologna and Porto, examine the relationship between the perceived importance of walkability attributes and walking behavior within each city, and interpret the findings through the conceptual lens of the 15MC. The walkability attributes evaluated in this study can be interpreted through several planning principles commonly associated with the 15MC, including density, mixed land uses, proximity to services, active mobility, connectivity, multimodal accessibility, and human-scale urban design [35,36,37,38].The present study adopts the expanded 15MC framework proposed by Khavarian-Garmsir et al. [20], which extends the four original principles identified by Moreno et a. [14], by incorporating connectivity, human-scale urban design, and flexibility. As the SPN questionnaire was not originally designed to address these seven principles, the present study focuses on five principles (density, diversity, proximity, connectivity, and human-scale urban design) for which a meaningful correspondence with the available walkability attributes could be established. Digitalization and flexibility are not operationalized because the questionnaire did not include attributes that could adequately represent these principles.

The obtained mapping was developed by comparing the conceptual meaning of each attribute with the planning principles described in the 15MC literature. The resulting correspondence is summarized in Table 1. Although some attributes could reasonably relate to more than one principle, each was assigned to the principle that best reflects its predominant planning function according to the reviewed literature. Accordingly, residential density was associated with the density principle [19,39]; mixed land uses and shopping streets with diversity [15,19,40]; proximity to community facilities, proximity to public transport and pedestrian infrastructure with proximity [11,30,37,41]; street connectivity with the connectivity principle [13,19,23]; and street enclosure, transparency, and architectural and landscape diversity with human-scale urban design [36]. Proximity to car parking was retained as part of the original walkability framework and was therefore included in the attribute-level analyses, but it was not assigned to any of the selected 15MC principles because it does not correspond to the conceptualization of proximity adopted in the 15MC literature. The wording and conceptual meaning of all attributes are described in detail in the previous study [34].

As the SPN survey was not originally designed around the 15MC framework, this study does not directly measure or evaluate the implementation of 15MC principles. Instead, selected walkability attributes from the original questionnaire are interpreted in relation to concepts emphasized in the 15MC literature and should be understood as an interpretative framework rather than a formal operationalization of 15MC principles.

Bologna and Porto (Figure 1) were selected as case study cities because they served as pilot cities within the SPN project [42] and have been actively engaged in the development and implementation planning policies approaches aligned with the 15MC.

Both cities, particularly their historic centers, developed before the widespread adoption of the automobile and are characterized by compact urban fabrics, mixed-use neighborhoods, and relatively high proximity between residential areas and everyday destinations [19,40]. Their walkable street networks, traditional mixed-use buildings, and concentration of essential services create urban environments where many daily needs can be met within short walking distances, making them particularly suitable case studies for examining the principles of the 15MC. For those reasons, Bologna [11,40,43,44] and Porto [16,19,28,45] have also been examined in previous 15MC-related research, particularly through GIS-based assessments of accessibility, proximity, and walkability.

Building on this body of research, the present study adopts a complementary perspective by examining how pedestrians prioritize urban and walkability attributes and how these perceptions relate to walking behavior in the two cities, with selected attributes interpreted in relation to the principles of the 15MC.

2.2. Data Collection

This study uses data collected through a questionnaire administered in the cities of Bologna and Porto within the context of the SPN research project [42]. The survey was designed to investigate pedestrian travel habits, attitudes towards walking, and preferences regarding built environment and streetscape attributes.

Data were collected through an online questionnaire developed using Google Forms. The questionnaire comprised a combination of single-choice, multiple-choice, ranking, and open-ended questions and was organized into four main sections. The first section collected socio-demographic information, including gender, age, educational level, occupation, and pedestrian profile. The second section focused on travel behavior and walking habits, including walking frequency, walking time per trip, reasons for choosing or not choosing walking as a transport mode, and perceived barriers and incentives to walking. The third section evaluated the importance of built environment and streetscape attributes for walking using a five-point Likert scale ranging from 1 (“not important”) to 5 (“very important”). The fourth section explored additional attitudes and preferences related to walking, including route choice preferences and multimodal travel behavior.

For the present study, the analysis focuses on variables describing walking habits and travel behavior, together with built environment and streetscape attributes identified as determinants of walkability in our previous study (Table 1) [34]. These variables were selected because they capture respondents’ evaluations of urban attributes relevant to walkability, as well as their reported walking behavior. Selected attributes were subsequently interpreted in relation to the principles of the 15MC.

Following a pilot test and subsequent revisions to improve the reliability and clarity of the questionnaire, the survey was administered online in both cities. Respondents were recruited through municipal and university databases, social media channels, and the SPN project website. The questionnaire was distributed in Italian and in Portuguese. A total of 1438 responses were obtained, including 865 from Bologna and 573 from Porto.

2.3. Data Analysis

Building on the four walkability determinants identified in our previous paper [34], this study addresses the two RQ presented in Section 1.3. For this purpose, the attributes grouped into the four walkability determinants shown in Table 1 (urban ambiance, pedestrian infrastructure, connectivity and community facilities, and access to other modes of transport) were interpreted in relation to selected principles of the 15MC, based on the conceptual correspondence between the attributes and the planning principles described in the 15MC literature. The exception was proximity to car parking, which was retained in the original walkability framework but did not correspond to any of the selected 15MC principles.

Descriptive statistics were first used to characterize walking frequency and walking times, as well as to assess the importance assigned to each walkability attribute. Measures included mean values, standard deviations (SD), and ranking positions. Walking frequency was coded ordinally according to the reported frequency of walking: 1 (never), 2 (less than 3 times per week), 3 (3–6 times per week), and 4 (daily). Thus, higher scores indicate more frequent walking. Walking time per trip was measured using two categories: 1 (≤15 min), and 2 (>15 min), with higher scores indicating longer walking trips. Because most reported walking trips were utilitarian and destination-oriented, the 15-min threshold provides a relevant indication of short walking trips to everyday destinations. However, this variable should not be interpreted as a direct measure of adherence to the 15MC, as the 15MC encompasses broader dimensions of accessibility and proximity to multiple types of essential destinations. Determinant scores were then computed as the arithmetic mean of the importance ratings of the attributes composing each determinant, with equal weighting assigned to the constituent attributes. This approach preserved the previously established grouping of attributes into the four walkability determinants [34] without re-estimating factor loadings or deriving weighted factor scores. The resulting scores provided an overall measure of the importance respondents assigned to each walkability determinant in each city.

The analysis was subsequently extended using three inferential statistical tests to address the RQ. First, chi-square tests of independence were performed to assess the association between city and walking frequency, as well as between city and walking time. Second, given the ordinal nature of the Likert-scale data and the independence of the two samples, Mann–Whitney U tests were used to examine whether statistically significant differences existed between Bologna and Porto in the importance attributed to each determinant and its corresponding attributes. Third, Spearman’s rank-order correlations were used to examine bivariate associations between the perceived importance of walkability determinants and their associated attributes and self-reported walking behavior within each city. These analyses were intended to identify associations rather than estimate whether the perceived importance assigned to walkability attributes was independently associated with walking behavior after controlling for demographic or socioeconomic characteristics. To account for multiple comparisons, the Holm correction was applied separately to the family of 12 Mann–Whitney U tests assessing inter-city differences and to the family of 16 Spearman correlations, adjusting the corresponding p-values. All statistical analyses were conducted using a significance level of 0.05 and performed using the software Jamovi (version 2.7.34).

3. Results

The results are organized into three subsections addressing sample characteristics and walking habits, the perceived importance of walkability attributes associated with the 15MC, and their relationship with walking behavior in both cities.

3.1. Sample Description

The analysis is based on 1438 valid responses collected in Bologna (n = 865; 60.2%) and Porto (n = 573; 39.8%) as part of the SPN research project. The socio-demographic characteristics of the sample are summarized in

Table 2. Socio-demographic characteristics of respondents in Bologna and Porto.

Variable	Category	Bologna (n = 865)		Porto (n = 573)
		n	%	n	%
Gender	Female	507	58.6	341	59.5
	Male	358	41.4	232	40.5
Age	≤24 years	84	9.7	110	19.2
	25–44 years	266	30.8	236	41.2
	45–64 years	477	55.1	214	37.3
	≥65 years	38	4.4	13	2.3
Education	Undergraduate degree	562	64.9	308	53.8
	Graduate degree	303	35.1	265	46.2
Occupation	Student	111	12.8	155	27.0
	Employed	735	85.0	402	70.2
	Unemployed/Retired	19	2.2	16	2.8

.

As reported in our previous study [34], the sample is slightly biased towards female respondents and individuals in the 25–65 age groups, with a predominance of participants with higher education and full-time employment. While the sample broadly reflects the population structure of both cities in terms of gender distribution and place of residence, some deviations are observed, particularly the underrepresentation of older adults (65+) and the overrepresentation of highly educated and working-age individuals. This pattern is a common limitation of online survey-based data collection, as certain population groups, such as older adults and individuals with lower levels of education, are generally less digitally connected and may have more limited access to, or familiarity with, online questionnaires. Although these limitations should be considered when interpreting the findings, the sample enables a comparative examination of respondents’ perceptions of walkability determinants and the interpretation of selected attributes in relation to the 15MC framework.

3.2. Walking Habits and Perceptions of Walkability Determinants and Attributes Related with the 15MC

Table 3. Walking frequency and walking time among respondents in Bologna and Porto.

Variable	Category	Bologna (n = 865)	Porto (n = 573)	X2	p-Value
Walking frequency	Daily	566 (65.4%)	226 (39.4%)	135.38	<0.001 ***
	3–6 times/week	161 (18.6%)	102 (17.8%)
	<3 times/week	138 (16.0%)	245 (42.8%)
Walking time (per trip) 1	≤15 min	260 (35.8%)	138 (42.1%)	3.83	0.503
	>15 min	467 (64.2%)	190 (57.9%)

¹ Only regular pedestrians; *** p < 0.001.

presents the distribution of walking frequency and average walking time per trip among respondents from Bologna and Porto. Regarding walking behavior, respondents from Bologna reported more frequent walking habits than those from Porto. Nearly two-thirds of participants in Bologna (65.4%) indicated walking daily, compared with 39.4% in Porto. In turn, Porto presents a much larger proportion of occasional walkers (<3 times per week), representing almost 43% of respondents. Interestingly, the share of respondents walking 3–6 times per week is remarkably similar in both cities (around 18%). A chi-square test of independence revealed a statistically significant association between city and walking frequency (χ² = 135.38, p < 0.001), indicating that walking frequency patterns differed significantly between the two cities, with respondents from Bologna reporting more frequent walking than those from Porto.

Among regular pedestrians, defined as respondents who walked daily or 3–6 times per week, most participants in both cities reported typical walking trips lasting more than 15 min. In Bologna, 64.2% of regular pedestrians indicated walking for more than 15 min per trip, compared with 57.9% in Porto. Conversely, trips lasting up to 15 min were reported by 35.8% of respondents in Bologna and 42.1% in Porto. A chi-square test of independence did not indicate a statistically significant association between city and walking time class (χ²(1) = 3.83, p = 0.050), although respondents from Bologna were somewhat more likely to report trips exceeding 15 min.

The subsequent analysis examined how pedestrians in Bologna and Porto prioritized and evaluated the walkability determinants and their associated attributes. Although our previous study identified the four walkability determinants and their respective attributes through factor analysis, it did not examine whether the importance assigned to these determinants and attributes differed between the two cities. To address this gap, descriptive statistics and comparative analyses were performed to assess respondents’ evaluations of the four determinants and their associated attributes by city. The results are presented in

Table 4. Importance assigned to walkability determinants and attributes.

Determinants and Attributes	Bologna Mean (SD)	Porto Mean (SD)	Total Mean (SD)	Determinant Rank	Attribute Rank
Pedestrian infrastructure	4.38 (0.70)	4.32 (0.82)	4.36 (0.75)	1	-
Sidewalks in good condition	4.50 (0.79)	4.43 (0.89)	4.47 (0.83)		1
Unobstructed sidewalks	4.38 (0.88)	4.36 (0.93)	4.37 (0.90)		2
Wide sidewalks	4.25 (0.92)	4.18 (1.00)	4.22 (0.95)		3
Connectivity and community facilities	3.96 (0.94)	4.02 (0.94)	3.98 (0.94)	2	-
Proximity to community facilities	3.91 (1.08)	4.16 (1.01)	4.01 (1.06)		4
Street connectivity	4.00 (1.07)	3.87 (1.14)	3.95 (1.10)		5
Access to other modes of transport	2.85 (1.08)	3.71 (1.19)	3.19 (1.20)	3	-
Proximity to public transport	3.27 (1.28)	3.87 (1.31)	3.51 (1.33)		7
Proximity to car parking	2.43 (1.32)	3.54 (1.35)	2.88 (1.44)		11
Urban ambiance	2.76 (0.77)	3.29 (0.81)	2.97 (0.83)	4	-
Architectural and landscape diversity	3.38 (1.07)	3.84 (0.98)	3.57 (1.06)		6
Mixed land uses	2.98 (1.02)	3.55 (1.02)	3.21 (1.06)		8
Shopping streets	2.77 (1.16)	3.44 (1.11)	3.04 (1.19)		9
Transparency	2.89 (1.11)	3.03 (1.08)	2.94 (1.09)		10
Residential density	2.39 (0.98)	3.09 (1.08)	2.67 (1.08)		12
Street enclosure	2.15 (0.98)	2.79 (1.06)	2.41 (1.06)		13

,

Table 5. Mann–Whitney U test for inter-city differences in walkability determinants.

Determinants	U	p-Value	Holm-Adjusted p-Value	Effect Size
Urban ambiance	156,441	<0.001 ***	<0.001 ***	0.369
Pedestrian infrastructure	247,154	0.929	0.929	−0.003
Connectivity and community facilities	236,464	0.133	0.159	0.046
Access to other modes of transport	143,016	<0.001 ***	<0.001 ***	0.423

*** p < 0.001.

and

Table 6. Mann–Whitney U test for inter-city differences in walkability attributes.

Attributes	U	p-Value	Holm-Adjusted p-Value	Effect Size
Shopping streets	169,850	<0.001 ***	<0.001 ***	0.315
Residential density	159,591	<0.001 ***	<0.001 ***	0.356
Mixed land uses	171,381	<0.001 ***	<0.001 ***	0.309
Street enclosure	165,102	<0.001 ***	<0.001 ***	0.334
Arch. and landscape diversity	188,047	<0.001 ***	<0.001 ***	0.241
Transparency	232,712	0.040 *	0.119	0.061
Proximity to public transport	177,648	<0.001 ***	<0.001 ***	0.283
Proximity to car parking	140,033	<0.001 ***	<0.001 ***	0.435

* p < 0.05; *** p < 0.001.

.

According to the results presented in Table 4, the highest scores were generally assigned to attributes associated with pedestrian infrastructure, connectivity, and proximity to daily services. In particular, sidewalks in good condition, unobstructed sidewalks, wide sidewalks, proximity to community facilities, and street connectivity all achieved mean scores close to or above 4.0. In contrast, attributes related to urban ambiance, particularly residential density and street enclosure, received comparatively lower evaluations. The comparison between the two cities reveals several similarities. Pedestrian infrastructure was the highest-ranked determinant in both Bologna (4.38) and Porto (4.32), while urban ambiance received the lowest scores (2.76 and 3.29, respectively). Likewise, the three highest-ranked attributes in both cities were related to pedestrian infrastructure.

The Mann–Whitney U tests showed that statistically significant differences between the two cities were observed only for urban ambiance and access to other modes of transport (Table 5). Significant differences were found for urban ambiance (U = 156,441, p < 0.001) and access to other modes of transport (U = 143,016, p < 0.001), both showing moderate effect sizes (0.369 and 0.423, respectively). No statistically significant differences were observed for pedestrian infrastructure or connectivity and community facilities. Overall, these results indicate broadly similar priorities across the two cities, while suggesting that urban ambiance and access to other modes of transport may be perceived differently in the two urban contexts.

To further investigate these inter-city differences, a second Mann–Whitney U test was conducted at the attribute level for the determinants that exhibited significant differences (urban ambiance and access to other modes of transport). The results are shown in Table 6. Within urban ambiance, all six attributes differed significantly between Bologna and Porto. The largest effect sizes were observed for residential density (0.356), street enclosure (0.334), shopping streets (0.315), and mixed land uses (0.309), whereas transparency showed only a small effect size (0.061) and its initially observed statistical difference did not remain significant after Holm correction. Within access to other modes of transport, both proximity to public transport and proximity to car parking differed significantly between the two cities (p < 0.001), with proximity to car parking exhibiting the largest effect size among all attributes analyzed (0.435).

Finally, as described in Section 2.3, the selected walkability attributes were interpreted in relation to five principles of the 15MC framework: density, diversity, proximity, connectivity, and human-scale urban design. Overall, respondents in both Bologna and Porto assigned the highest importance to attributes interpreted in relation to the proximity, particularly pedestrian infrastructure, and proximity to community facilities, and connectivity principles. The two attributes associated with the diversity principle (mixed land uses and shopping streets) occupied intermediate positions in the overall ranking. By contrast, attributes associated with density, and human-scale urban design generally received lower importance scores.

3.3. Relationship Between Walkability Determinants and Walking Behavior

Spearman’s rank correlation coefficients (ρ) were computed to examine the relationship between walkability determinants and walking behavior, measured through walking frequency and walking time per trip, separately for Bologna and Porto (

Table 7. Spearman correlations between walkability determinants and walking behavior.

Determinants	Bologna		Porto
	Walking Frequency (ρ)	Walking Time (ρ)	Walking Frequency (ρ)	Walking Time (ρ)
Urban ambiance	0.057	0.019	0.086 *	−0.023
Pedestrian infrastructure	0.156 ***	0.012	0.055	0.082
Connectivity and community facilities	0.084 *	−0.046	0.010	0.009
Access to other modes of transport	−0.106 **	−0.185 ***	−0.146 ***	−0.150 **

Significance levels are based on Holm-adjusted p-values; * p < 0.05; ** p < 0.01; *** p < 0.001.

).

In terms of walking frequency, different patterns were observed between the two cities. In Bologna, walking frequency was positively associated with pedestrian infrastructure (ρ = 0.156, p < 0.001) and, more weakly, with connectivity and community facilities (ρ = 0.084, p < 0.05). Urban ambiance was not significantly associated with walking frequency (ρ = 0.057), while access to other modes of transport showed a weak negative association (ρ = −0.106, p < 0.01). In Porto, walking frequency was very weakly and positively associated with urban ambiance (ρ = 0.086, p < 0.05) and weakly and negatively associated with access to other modes of transport (ρ = −0.146, p < 0.001). No significant associations were observed for pedestrian infrastructure (ρ = 0.055) or connectivity and community facilities (ρ = 0.010).

Walking time showed limited associations with the perceived importance of walkability determinants in both cities. Only access to other modes of transport was significantly and negatively associated with walking time in Bologna (ρ = −0.185, p < 0.001) and Porto (ρ = −0.150, p < 0.01).

Overall, the correlations were weak, indicating limited associations between the perceived importance assigned to walkability determinants and self-reported walking behavior. The patterns for walking frequency nevertheless differed between the two cities: the largest positive association in Bologna was observed for pedestrian infrastructure, whereas in Porto the only significant positive association was with urban ambiance.

4. Discussion

This study examined how respondents in Bologna and Porto perceive and prioritize urban attributes relevant to walkability, whether these perceptions differ between the two cities, and how they relate to walking behavior. The findings were interpreted through the conceptual framework of the 15MC. Four main findings emerge from the analysis. First, respondents in both cities assigned broadly similar levels of importance to the walkability attributes, with attributes interpreted in relation to proximity and connectivity receiving the highest importance scores. Second, although overall patterns were comparable, significant inter-city differences were observed for urban ambiance and access to other transport modes. Third, the perceived importance assigned to walkability determinants showed only weak associations with walking behavior, particularly walking time. Fourth, highly valued attributes did not consistently show stronger associations with walking frequency or walking time, indicating a limited correspondence between the importance assigned to walkability attributes and reported walking behavior.

With this in mind, and in relation to the first RQ, the findings indicate that respondents in Bologna and Porto exhibit remarkably similar priorities regarding the walkability determinants and attributes. Despite differences in planning traditions and mobility dynamics, and some similarities in their consolidated and historically rich urban fabrics, the overall ranking of walkability determinants and their respective attributes is broadly comparable. Across both cities, attributes reflecting the proximity principle received the highest evaluations, particularly pedestrian infrastructure, proximity to community facilities and proximity to public transport. These results corroborate recent studies indicating that the quality of pedestrian infrastructure should be analyzed within concept of the 15MC [30] and is crucial for promoting inclusive 15MC [46]. Attributes associated with diversity, namely mixed land uses and shopping streets, received importance scores slightly above 3, while density and two attributes associated with human-scale urban design (street enclosure and transparency) received scores below 3. These findings nevertheless suggest that respondents recognized the relevance of diverse land uses and commercial streets for walking. This is consistent with previous analyses of Bologna and Porto showing that shopping is a purposes of daily walking trips and that commercial streets and diverse land uses can facilitate access to essential goods and services within local neighborhoods [34]. These findings are also consistent with broader literature highlighting the importance of diversity for promoting the 15MC and supporting active modes of transport [14,20], as well as with recent policies undertaken during and after the pandemic to expand and improve pedestrian areas [47].

Increasing population density also contributes to fostering the 15MC, as a larger proportion of the population will be able to reach the most basic offerings of a city [48]. However, in our study, respondents did not perceive this attribute as very relevant for walking. This finding deserves particular attention because residential density is widely regarded as a fundamental prerequisite for implementing the 15MC, namely for supporting local services and reducing travel distances [14,20,49]. Our results suggest that pedestrians may not directly associate higher residential densities with better walkability. Respondents appear to value the functional outcomes that density can support, such as the proximity of services, and to public transport, rather than density itself. In this sense, density may act as an enabling condition that facilitates other walkability attributes rather than being perceived as a direct determinant of walking. This finding could also reflect the perception that excessively high residential densities can reduce moving speed, pedestrian comfort and infrastructure efficiency [50]. Previous studies have also shown that human-scale urban design can contribute to pleasant pedestrian environments and support walkability [20,38,51]. For example, Hassan and Elkhateeb [38] argue that urban design qualities can influence walkability and contribute to walkers’ psychological contentment. However, our findings do not confirm a similarly high perceived importance of human-scale design attributes. Instead, respondents assigned greater importance to attributes directly related to accessibility and everyday mobility. One possible explanation is that both Bologna and Porto are characterized by consolidated urban fabrics and historically rich built environments, particularly within their central areas, where high-quality urban design and architectural heritage are well established. Both cities also contain UNESCO World Heritage sites, reflecting the historical and architectural significance of their urban environments [52]. As a result, attributes such as architectural diversity and street enclosure may be perceived as inherent characteristics of the urban environment rather than as factors that actively influence everyday walking decisions.

Although the overall importance assigned to the four walkability determinants was broadly similar between the two cities, the descriptive results showed some differences in mean scores. These differences were statistically significant only for urban ambiance and access to other modes of transport, both of which received higher importance scores in Porto. No statistically significant differences were observed for pedestrian infrastructure or connectivity and community facilities. These differences may partly reflect the distinct urban and mobility contexts of the two cities and the way residents experience them. Bologna has a well-developed public transport system, comprising buses, trolleybuses, and suburban rail services, which is generally well integrated with the city’s compact urban structure and pedestrian environment [40]. This context may partly explain the lower importance assigned to access to other modes of transport by respondents in Bologna. By contrast, Porto’s steeper topography [28] and previous evidence highlighting the need to improve public transport coverage, network expansion and intermodal integration [53] provide a possible contextual explanation for the greater importance assigned to access to other modes of transport. Differences in modal split may provide additional context. Private car use accounts for a larger share of daily trips in Porto (54%) [54] than in Bologna (42%) [55], which may indicate differences in mobility patterns between the two cities. These interpretations should nevertheless be treated with caution, as the present study did not directly examine how objective characteristics of public transport provision or intermodal accessibility influence respondents’ evaluations.

When interpreted through the conceptual framework of the 15MC, the findings indicate that several of the highly valued attributes can be conceptually associated with proximity and connectivity, whereas attributes associated with density and human-scale urban design generally received lower importance scores. Because the analysis is based on subjective perceptions rather than objective measures of the built environment, these findings reflect how respondents perceive and value specific urban attributes, which may also be influenced by their socioeconomic characteristics, travel habits, and previous experiences. Consequently, differences in perceived importance between Bologna and Porto should not necessarily be interpreted as evidence of corresponding differences in the objective characteristics of their built environments.

Regarding the second RQ, respondents from Bologna report significantly higher walking frequency and a higher proportion of trips lasting more than 15 min than respondents from Porto, indicating differences in walking behavior between the two urban contexts. These differences may reflect the combined influence of urban form, mobility conditions, and travel culture. Bologna, particularly its historic center, is widely recognized as a highly walkable city, characterized by a compact urban layout, a dense network of narrow streets, and its extensive system of porticoes that provide shade and shelter [40,56,57]. By contrast, Porto’s pedestrian environment is constrained in some areas by its steep topography [19,58], rainy weather, particularly during the winter [56,57], and the limited number of highly walkable streets in the city center [59]. Together with the higher share of private car use reported for Porto [54], these characteristics may help explain the lower levels of walking reported by its respondents. These interpretations should nevertheless be treated with caution, as the present study did not directly examine the effects of these contextual factors on walking behavior.

Within each city, the correlation analysis revealed generally weak associations between the perceived importance of walkability determinants and walking behavior. After correction for multiple comparisons, significant associations were concentrated on access to other modes of transport, while pedestrian infrastructure was also significantly associated with walking frequency in Bologna. The remaining associations did not reach the corrected significance threshold. In Bologna, walking frequency was weakly and positively associated with pedestrian infrastructure, consistent with previous evidence that adequate sidewalks and comfortable walking conditions can support everyday walking [51]. In both cities, walking frequency was negatively associated with the perceived importance of access to other modes of transport. This association may indicate that respondents who walk more frequently place less importance on proximity to alternative transport modes, although the analysis does not establish whether this reflects lower reliance on public transport or private cars. In Porto, walking frequency also showed a statistically significant but very weak positive association with urban ambiance (ρ = 0.086, p < 0.05), which should be interpreted cautiously given its small magnitude.

Overall, the findings indicate limited correspondence between the importance respondents assign to walkability attributes and their reported walking behavior. Attributes considered important for supporting walking did not consistently show stronger associations with walking frequency or walking time. This distinction may reflect the fact that recognizing the importance of an attribute does not necessarily indicate that it is adequately provided or experienced in the respondent’s environment. Walking behavior is likely to result from the interaction of multiple built-environment, mobility, individual, and socioeconomic factors, many of which were not directly examined in the present analysis. Previous research likewise suggests that socio-demographic characteristics may play an important role in shaping adherence to the 15-min lifestyle [12]. The findings also differ from those of De Vos et al. [60], who examined perceived walkability of the surrounding environment and found positive associations with walking frequency and duration. However, their correlations were also consistently small, suggesting that perceived walkability alone may have limited explanatory power. Taken together, these findings reinforce the importance of distinguishing between the perceived importance of specific walkability attributes and the actual experience of the built environment when interpreting their relationship with walking behavior.

From a planning perspective, the comparison between Bologna and Porto provides comparative evidence on how walkability attributes associated with selected 15MC principles are perceived in two Southern European cities with broadly comparable urban and historical contexts. The findings reveal broadly similar priorities alongside differences in the importance assigned to specific attributes, suggesting that the relevance of individual walkability attributes may vary even across relatively comparable urban settings. Although respondents in both Bologna and Porto generally assigned importance to pedestrian infrastructure, proximity to community facilities, and connectivity, the differences observed between the two cities indicate that these attributes are not perceived identically across contexts. Such differences may partly reflect the distinct urban and mobility conditions of the two cities. In Bologna, where walking is already well established and supported by a compact urban structure, future interventions should focus on maintaining the quality and continuity of the pedestrian environment while preserving the mixed-use character of neighborhoods. In Porto, where the steeper topography and the higher share of private car use reported in statistic data provide a different mobility context, improving pedestrian accessibility could be complemented by strengthening public transport integration, enhancing first- and last-mile connections, and expanding the provision of local services within walking distance. More broadly, the findings suggest that promoting the 15MC should extend beyond increasing density or introducing new urban design features. Instead, planners should prioritize improving the everyday functionality of neighborhoods through high-quality pedestrian infrastructure, accessible public facilities, mixed land uses, and integrated multimodal transport systems. These recommendations are likely to be relevant for Bologna and Porto and may also inspire other medium-sized European cities facing similar challenges of balancing compact urban development, sustainable mobility, and high-quality walking spaces.

5. Conclusions

Over recent years, the 15MC has become an influential framework for promoting sustainable urban development by encouraging active mobility, improving proximity to daily services, and reducing reliance on private motorized transport. Within this context, examining how residents perceive and prioritize urban attributes relevant to walkability, and interpreting these perceptions through the 15MC framework, can provide useful insights for planning interventions.

This study examined how pedestrians in Bologna and Porto prioritize and evaluate walkability determinants and attributes, and whether these perceptions are related to self-reported walking behavior. Selected attributes were interpreted in relation to the conceptual framework of the 15MC. By adopting a comparative perspective, the study contributes to the growing body of literature on the 15MC by exploring how its underlying principles are reflected in residents’ perceptions of the urban environment in two Southern European cities. Overall, the findings reveal remarkable similarities between Bologna and Porto regarding the importance assigned to walkability determinants and attributes. Despite differences in reported walking frequency and duration, respondents in both cities consistently attributed greater importance to pedestrian infrastructure, proximity to community facilities and public transport, and street connectivity than to residential density or human-scale urban design attributes. When interpreted through the conceptual framework of the 15MC, these findings indicate that several of the most highly valued walkability attributes can be associated with the proximity and connectivity principles, whereas attributes related to density and human-scale urban design generally received lower importance scores. From a planning perspective, the findings highlight pedestrian infrastructure and accessibility to daily services and public transport as common areas of importance across the two cities, while suggesting that these priorities should be considered alongside other built-environment and mobility factors that may shape walking behavior.

Despite its contributions, this study has some limitations that should be acknowledged. First, the questionnaire was originally designed to examine the perceived importance of built-environment attributes for walking rather than to directly measure the principles of the 15MC. Consequently, the interpretation of the findings through the 15MC framework is conceptual rather than a direct assessment of its principles. For that reason, not all surveyed attributes correspond directly to the planning principles of the 15MC. In particular, proximity to car parking was retained as part of the original walkability framework but was not interpreted as representing the proximity principle of the 15MC. Second, the analysis is based on stated perceptions and self-reported walking behavior rather than objective measures of accessibility or mobility. This self-reported evaluation may contain inconsistencies between reported preferences and individual behaviors and could be influenced by socioeconomic characteristics, such as age and gender, travel habits, previous experiences, and personal preferences. Therefore, the observed inter-city differences should not be interpreted as direct evidence of differences in urban form, but rather as differences in how residents perceive and value specific urban attributes. Third, the study relied on an online questionnaire, which may have introduced self-selection bias and contributed to the underrepresentation of certain population groups, particularly older adults and individuals with lower digital literacy, owing to the inherent difficulties of reaching these groups through a web-based survey. This limitation may have influenced the findings. In particular, older adults are generally more sensitive to barriers and deficiencies in the pedestrian environment; therefore, their underrepresentation may have affected the relative importance attributed to walkability determinants and attributes, especially those related to pedestrian infrastructure and accessibility. Consequently, the findings should be interpreted as representative of the surveyed sample rather than of the entire populations of both cities. Fourth, the analysis of walking behavior was based on bivariate Spearman correlations and therefore did not control for potential confounding factors such as age, gender, or educational level. Fifth, the Bologna and Porto samples also differed in their demographic composition, which may contribute to differences in the perceived importance assigned to walkability attributes and in walking behavior. The observed associations should consequently not be interpreted as independent effects of the perceived importance assigned to walkability attributes on walking behavior. Sixth, walking duration was collected using two categories (≤15 min and >15 min), which entails some information loss compared with a more detailed or continuous measure and may reduce the sensitivity of the correlation analysis. Nevertheless, because most reported trips were utilitarian and destination-oriented, the ≤15-min category provides a meaningful indication of short walking trips to everyday destinations. Finally, the study is based on two case-study cities, Bologna and Porto, and the findings should not be generalized directly to other urban contexts. Both cities share several characteristics, including consolidated urban fabrics and a strong historical heritage, but they also differ in their urban form, topography, mobility patterns, and planning contexts. The results therefore provide context-specific evidence rather than a representative assessment of Southern European cities as a whole.

To overcome these limitations, future studies could combine subjective evaluations with objective built environment indicators, such as density, land-use mix, accessibility, and street connectivity, to provide a more comprehensive understanding of walkability within the 15MC framework. Multivariate models, such as ordinal regression for walking frequency, could also be employed to assess whether the perceived importance assigned to walkability determinants remains associated with walking behavior after accounting for demographic, socioeconomic, and mobility-related characteristics. Comparisons with additional cities representing different urban forms, population structures, and planning and mobility contexts would be necessary to assess the extent to which the observed patterns are transferable to other settings. Further studies specifically designed around the principles of the 15MC would also provide a more comprehensive understanding of how residents perceive, experience, and respond to this increasingly influential planning framework.