Digital Service Substitution and Social Networks: Implications for Sustainable Urban Development

Abstract

Considering the rapid integration of digital services into daily life, it is crucial to analyze the impacts of the substitutability of physical services with digital alternatives. Limited studies have been conducted to investigate the relationship between service substitution and social networks and assess their impact on urban structure. Therefore, this study fills the gap by investigating how digital service substitution and social networks influence residential location choices and urban structure, aiming to support future sustainable urban modeling and planning tools. The study, through a comprehensive analysis incorporating cluster analysis, factor analysis, and binomial logistic regression on a web-based questionnaire survey (n = 6210), finds that socio-demographic factors significantly influence digital alternatives, and that digital service substitution and social networks impact sustainable urban structure. Younger individuals showed significantly higher adoption of digital alternatives, with age negatively associated with relocation likelihood. In urban areas, each additional year of age reduces the likelihood of relocation by approximately 4.4%, and individuals with high shopping substitution are 3.12 times more likely to consider relocation. These findings suggest that urban planners and policymakers to balancing physical and digital service provision to maintain a higher quality of life aligned with the SDGs and ensure sustainable urban development.

1. Introduction

The integration of cyberspace into contemporary life has dramatically changed different aspects of living environments, including cultural and socio-economic activities. These technological advancements have facilitated global engagement, allowing individuals from diverse regions and socio-economic characteristics to participate in a wide range of activities and services in cyberspace. Additionally, the rapid spread of COVID-19 and its announcement as a global pandemic [1] also impacted socio-economic activities and urban structures [2] and contributed to the acceleration of digital transformation. The lockdowns during pandemics made individuals adapt to a new environment, shifting daily activities and services from physical spaces to cyberspaces.

Although pandemics have historically affected different socioeconomic activities and urban structures [3], COVID-19 showcases how pandemics can accelerate the transition to digital alternatives where services are offered in cyberspace. The traveling restrictions during the COVID-19 pandemic led to drastic changes in urban mobility as governments enforced lockdowns and promoted remote work, resulting in a significant reduction in commuting and personal travel [4,5,6,7,8]. Similarly, public transportation saw a decline due to challenges associated with maintaining physical distance in crowded spaces [9,10,11]. Existing studies also show that the environmental impacts of these lifestyle shifts are significant. For instance, a study [12] shows global CO₂ emissions dropped by 17% during the pandemic, whereas a pollution reduction of around 30% was seen in heavily affected regions like Wuhan, Italy, Spain, and the United States [13,14,15]. Noise pollution in urban areas also decreased considerably [16,17,18].

The COVID-19 pandemic underscored the importance of access to information and services during crises, prompting research on analyzing the effects of COVID-19 during disaster evacuation [19] and the role of information accessibility during evacuations with its implications for future resilience planning [20,21]. The shift from physical services to digital alternatives, such as online shopping, remote work, and virtual meetings, reduced the need for physical travel and face-to-face interactions, thereby alleviating pressure on transportation systems and decreasing CO₂ emissions [22]. The shift toward online shopping, already gaining momentum before the COVID-19 pandemic, has seen an accelerated adoption in response to restrictions on mobility [4]. Similarly, the rise in e-learning, supported by Information and Communication Technologies (ICTs), has allowed many students to attend online classes from home [8]. The case is similar to teleworking, which has emerged as a critical component of modern work–life flexibility, reducing physical commuting and offering substantial environmental benefits by lowering CO₂ emissions [12]. With this transition, access to services began to rely more on ICT networks rather than traditional transportation systems, bringing new challenges and opportunities for researchers, urban planners, and policymakers.

The pandemic has profoundly altered socioeconomic activities [23] and daily lifestyles [24,25], with telecommuting and online conferencing emerging as viable alternatives to in-person interactions due to mobility restrictions during the COVID-19 pandemic. While ICT has enhanced the accessibility of people to services, leading to improvements in work–life balance [26] and cost and time savings [27], its impact on the quality and quantity of traffic in physical services has raised concerns. Reduced face-to-face interactions have negatively affected local economies and social engagement [28], with long-term implications for residential location choices and urban planning. Understanding the interplay between service substitutability, individual characteristics, social networks, and changes in urban structure and socio-economic environments is crucial in addressing these evolving challenges.

On the other hand, considering the increasing integration of technology into daily life and smart city development progress, coupled with the far-reaching impacts of pandemics such as COVID-19, more services are going to be delivered through both cyberspace and physical spaces to enhance the Quality of Life (QOL) of individuals [29]. Smart cities are specifically designed to deliver such services by leveraging advanced technologies [30,31], aiming to enhance the QOL of people [32]. Existing studies confirm that smart cities, due to relying on ICT, can reduce carbon emissions and contribute to the sustainable development of the cities [33,34]. Considering this digital transformation, it is essential to examine the substitutability of physical services with their digital alternatives and assess the implications for urban and regional structures. This study addresses this issue by analyzing data from a web-based questionnaire survey conducted in Japan and investigates how digital alternatives and social networks affect sustainable urban structure as a result of changes in residential location choice.

Previous studies [35] have explored changes in online consumption behavior using surveys in Tokyo and Okayama, revealing that the frequency of in-person shopping declined during the pandemic. Findings suggest that women and younger individuals were more likely to engage in online shopping, with motivations differing based on city size. In the Tokyo metropolitan area, time savings were a key factor, while in rural areas, online shopping served as a complement to physical stores. Other researchers [36,37,38] have examined the impact of ICT on transportation behavior, work patterns, and urban structure. Existing studies have also examined the impact of ICT on QOL [39,40], and Mutahari [41] specifically investigated the substitution of physical activities with digital alternatives, focusing on QOL and decarbonization through a policy evaluation framework integrating transportation and ICT networks. However, these studies did not comprehensively explore how digital service substitution influences social networks and urban structure. To fill this gap, this study jointly analyzes digital substitution and social networks in the context of residential location choice. While prior research has provided insights into ICT usage, transportation behavior, and service access substitutability, few studies have addressed the interconnected effects of digital alternatives, social networks, and sustainable urban development.

This study aims to investigate how digital service substitution and social networks affect residential location choices and result in changes in urban structure. Specifically, the paper examines: (1) the behavioral patterns of digital alternative usage across different socio-demographic categories, (2) how the adoption of digital service alternatives, such as online shopping, teleworking, etc., affects social interaction patterns, and (3) the factors influencing the intention to relocate, including the effects of substitutability and social networks. By analyzing these relationships, the study contributes to a deeper understanding of digital substitution policy implications for sustainable urban development.

The study, using factor and cluster analyses, explores the relationships between individual attributes, residential location, and social networks to understand the broader implications for urban sustainability. A web-based questionnaire survey was conducted to examine the substitutability of access to daily life services through ICT networks, considering relationships within social networks. The analysis focuses on how sociodemographic factors such as age, gender, and place of residence, as well as the strength of social networks, influence these behaviors through cross-tabulation, factor analysis, and cluster analysis, ultimately identifying the factors affecting residential location choice through binomial logistic regression analysis. The overall objective of this paper is to fill the academic gap, providing urban planners and policymakers with insight into digital service usage and contributing to sustainable urban development by proposing a tradeoff between physical and cyberspace service usage and supply. It is believed that excessive usage of either will have negative impacts on social environments, well-being, and urban structure, and contribute to overall QOL reduction. To this end, the paper comprehensively studies the individual’s preferences and behaviors toward digital service adaptation and investigates the impacts on residential location choice. The result of this study suggests that urban structures will be affected by digital service substitution and urges urban planners and policymakers to carefully plan for future sustainable urban development through rules and regulations at the policy level; thus, the access to physical services and digital services shall be traded off and balanced based on individual needs.

Figure 1 presents the conceptual framework of the interrelationship between digital service substitution, social networks, and urban structure. In this study, digital service substitution refers to the replacement of physical activities or services in the real network that rely on transportation systems, such as shopping, working, etc., with digital activities or services in the virtual network that rely on ICT, such as online shopping, teleworking, etc. As can be seen in Figure 1, individuals, based on their preferences and characteristics, will choose either a transportation network or ICT network to reach a service. If the individual uses a transportation network to reach a service, that service is called a physical service. Whereas, if the individual uses an ICT network to reach a service, the service is called a digital service. The potential to replace physical services with digital services is called digital service substitutability. The study investigates the effects of this substitutability on social networks and residential relocation choice, that, as a result, affect the urban structure.

The remaining parts of this paper are as follows: After the introduction, Section 2 provides an overview of the data acquisition process, analyzes the survey, and presents the survey results. Section 3 and Section 4 detail the analysis methods, corresponding findings of a cross-tabulation analysis of behavioral substitution ratios, and human relationships affecting behavioral choices, respectively. Subsequently, Section 5 discloses the methodology and briefly discusses the residential relocation choice analysis. Section 6 discloses the main findings of this study, whereas Section 7 opens a discussion based on the findings. Finally, Section 8 offers conclusions and suggests directions for future research.

2. Materials and Methods

The study investigates the substitutability of physical services with digital alternatives using a web-based questionnaire survey conducted all over Japan. The survey aims to gather information about people’s preferences and behaviors while participating in activities and receiving services in both physical space and cyberspace. The web-based questionnaire survey by this study gathered information that is hard to obtain through other methods, such as how people access services through ICT networks and social networks, and how these factors affect their residential relocation choice. The ability to reach a wide range of respondents and collect reliable data efficiently across a large area are the main reasons that we used a web-based questionnaire survey in this study. The survey was conducted on 24 and 25 November 2023, and we received 6210 valid responses. The survey questions were divided into five main sections: personal characteristics, life behaviors, service access, use of information network services, social connections, and relocation choices.

The first section collected demographic information to understand how household characteristics affect access to both physical services and digital services. In this study, we considered 5 basic daily life activities such as shopping, dining, working, schooling, and healthcare visits, as shown in Figure 2. Dining refers to eating outside, such as in restaurants, where individuals use an actual mode of transportation, either walking, cycling, a private car, or public transport, to reach the facility and receive the services. Whereas, food delivery refers to ordering food from a restaurant and eating at a place of residence, where individuals use a mode of ICT to reach the virtual facility and receive the service. Food delivery is a digital alternative to dining in a restaurant. Similarly, online shopping, teleworking, e-learning, and e-medical consultation are digital alternatives for in-store shopping, in-office working, in-school learning, and in-hospital treatment, respectively.

These activities can be performed in physical space and cyberspace, and we named them physical activities and digital alternatives, respectively. Physical activities rely on transportation systems to reach the facilities or venues to perform the activity. For instance, to shop, eat outside, work, study, and visit a hospital, someone may use an actual mode of transportation to go to designated places to perform the activities. On the other hand, to perform activities in cyberspace (digital alternatives), people may use ICT to reach the virtual venue of activities to perform the activities and will not use the actual transportation systems. We define digital alternatives as an alternative to physical activities that use ICT rather than the transportation system. The second section of the questionnaire survey provided us with information on how frequently individuals perform physical activities or digital alternatives.

Additionally, the third section of the survey provided us with information on how people use digital platforms and what might prevent them from using these services. The social connections section of the survey asked about satisfaction with personal relationships, comparing in-person and online interactions to see how virtual engagement affects social relationships. Lastly, the survey investigated how access to both physical and digital services influences people’s decisions regarding residential relocation, along with the obstacles they may encounter in making such decisions. The survey could provide us with this information to study the substitutability of the services from physical to digital alternatives.

Table 1. Web-based questionnaire survey summary.

Items	Contents
Target area	All over Japan
Date	24 November to 26 November 2023
Number of samples	6210 samples
Survey items	1. Basic Attributes
	2. Questions related to lifestyle behavior
	3. Questions about information network use
	4. Questions related to human connections
	5. Questions about place of residence

represents an overview of the questionnaire survey.

The questionnaire survey was segmented and evenly distributed based on gender, age, and place of residence, based on the population distribution characteristics, as shown in Figure 3. This balanced distribution was carefully considered to improve the quality and reliability of the survey results and remove biases. Although we carefully looked into removing potential biases through survey segmentation, there might have been some bias within the elderly (70s and older) age group, who may not have been capable of using smart devices to answer the online questionnaire survey. However, this bias was insignificant in this study. As one of the objectives of this study, we discussed the future urban structure, where younger individuals are the target, who have enough capability to use smart devices and receive digital services. Therefore, the potential bias with elderly responses cannot be considered a limitation of this study.

Although we have gathered information about individuals’ income and employment type, as shown in Figure 4, we have not used these variables in our analysis. The focus of the current study is to consider major socio-demographic variables such as age, gender, and place of residence.

3. Cross-Tabulation Analysis of Behavioral Substitution Ratios

3.1. Substitution Ratio for Each Behavior

This section examines the substitutability of physical activities for digital alternatives across five daily life activities such as shopping, dining, commuting, schooling, and medical visits. For example, the section examines to what extent shopping in physical space can be substituted by online shopping, considering several factors such as age, residential location characteristics, and license ownership. The “substitution ratio” is calculated by comparing the frequency of performing physical activities with their digital alternatives.

Figure 5 presents the substitution ratio for these activities, highlighting tendencies toward digital alternatives. The survey analysis results disclose that shopping exhibits a higher substitution ratio toward digital alternatives, with only 25% of respondents reporting no adoption of digital alternatives, whereas 75% indicated varying levels of substitution. This result confirms that shopping has a higher substitution ratio toward its digital alternative than other activities. This trend suggests a transformation in consumer behavior, where digital network accessibility increasingly changes the shopping behavior of individuals. Considering the higher substitution ratio of shopping behavior and its direct relevance to transportation demand and urban land use, the study selects shopping behavior as a focal activity for substitution ratio analysis. This study mainly focuses on analyzing the factors influencing the choice of shopping digital alternatives and investigates the main cause of these behavioral shifts.

3.2. Impact of Age on Shopping Substitution Behavior

Figure 6 illustrates the percentage of shopping substitution across different age groups, highlighting key trends in the adoption of digital alternatives. The survey results disclose that the tendency of shopping digital alternative behavior decreases with age. It is found that younger generations (20s) are engaging more in online shopping compared to other age groups. However, even among the elderly (60+ years old), approximately 69% reported using online shopping as a substitute for traditional in-store purchases. A total of 80% of individuals belonging to the 20s age group reported online shopping to varying degrees. The trend is 78%, 77%, 76%, and 69% for age group 30s, 40s, 50s, and 60+, respectively. This finding emphasizes the widespread adoption of online shopping across all age groups, emphasizing the substitutability of access to physical shopping with its digital alternative, even among the elderly.

3.3. Impact of Residential Location on Shopping Substitution Behavior

Figure 7 demonstrates the relationship between residential location and the shopping substitution ratio. Residential characteristics were categorized according to responses to the question: ‘Please select the option that best describes your residential area’. Five residence categories, including city center, city center surrounding, suburbs, rural areas, and nature-rich areas, were listed hierarchically for the respondents.

The analysis results reveal that the average shopping substitution ratio, ranging from 0.1 to 0.2, was slightly higher among respondents living in the city center. This may suggest that residential environments may slightly influence shopping substitution behavior. However, as can be seen in Figure 7, no significant difference can be witnessed in online shopping frequency across various residential areas. In contrast, it can be witnessed that regardless of residential locations, online shopping has a higher substitution ratio. The survey results disclose that individuals tend to do online shopping at an average rate of 25.2%, regardless of their place of residence.

3.4. Impact of License Possession on Shopping Substitution Behavior

Figure 8 illustrates the relationship between license possession and the shopping substitution ratio. Although it was assumed that people without driving licenses might do more frequent online shopping compared to people with driving licenses due to convenience and delivery options, the result of this study shows otherwise. The survey results show that individuals with a driving license tend to do more online shopping. Almost 75% of individuals with a driving license engage in online shopping, whereas only 66% of individuals without a driving license engage in online shopping. However, the substitution ratio of 0.5–1.0 is higher by 2% among individuals without a driving license.

Findings on the relationship between license possession and online shopping may suggest that licensed individuals, while having the option to use private vehicles for investigating and checking the products in an actual store, may prefer online shopping due to convenience and lower prices. The result of this study emphasizes that mobility and digital alternatives can coexist, and the presence of a driving license does not necessarily reduce the use of digital alternatives. Additionally, the result emphasizes that online shopping may not necessarily reduce traffic congestion or private vehicle usage.

3.5. Impact of Social Relationships on Shopping Substitution Behavior

This section examines how personal relationships affect the adoption of digital alternatives. Respondents were asked to rate the importance of relationships across three categories: family, friends, and internet acquaintances, using a seven-point Likert scale, where one indicates “not at all important” and seven indicates “very important”. In this study, we divided the relationships into two categories: “neighbors” in real networks and “internet acquaintances” in virtual networks, such as social media, to investigate the relationship between social relationships and digital substitution. Figure 9 illustrates the relationship between the perceived importance of “Internet acquaintances” and the shopping substitution ratio. Respondents were grouped into three categories based on their relationship importance: “less important” (ratings 1–3), “neutral” (rating 4), and “important” (ratings 5–7).

The results show that the “important” group, which places a higher importance on relationships with internet acquaintances, had the highest shopping substitution ratio. This indicates that individuals who place more value on their online relationships are more likely to prefer digital alternatives, such as online shopping. Consequently, a stronger emphasis on online communication correlates with an increased tendency to adopt digital shopping behavior. This means people who are exposed to more online communication tend to do more online shopping and have a higher digital substitution ratio.

Figure 10 illustrates the relationship between personal relationships with neighbors and the adoption of digital shopping behavior. The results reveal that respondents who placed less importance on personal relationships with neighbors in physical space were more likely to have higher engagement with digital alternatives. The results of this study confirmed that people who stated that relationships with neighbors are less important shopped online more frequently compared to the other two groups who found relationships to be “neutral” or “important”. Individuals who responded that relationships and interactions with neighbors were important or remained neutral engaged in online shopping less, by 74.3% and 72%, respectively, compared to the other group, at 77.7%. The substitution ratio of 0.5–1.0 was higher among the less important category by 2% compared to the important category. Combining the results from Figure 9 and Figure 10, it can be suggested that social networks, either in real networks or virtual networks, affect the digital substitution ratio accordingly.

4. Analysis of Relationship Types Influencing Behavioral Choices

4.1. Factor Analysis of Relationship Types Influencing Behavioral Choices

To understand the significance of various relationships by respondents, an exploratory factor analysis was conducted. The maximum likelihood method was employed for factor extraction, with Promax rotation applied to accommodate correlations between factors. Factors were retained based on an eigenvalue of at least 1 and a factor loading of at least 0.40. This analysis identified five items and four distinct factors, which together accounted for approximately 62% of the total variance. The factor loadings and eigenvalues of the extracted factors are presented in

Table 2. Factor analysis results.

Item	Factor Loadings
	1	2	3	4
Relationships with family members	1.08	−0.07	−0.09	−0.24
Relationships with relatives	0.69	−0.12	0.19	0.02
Relationships at work or school	−0.12	1.02	0.07	−0.14
Relationships in the neighborhood or community	0.02	0.08	0.83	0.05
Relationships with acquaintances and friends on the internet	−0.12	−0.09	−0.05	0.73
Relationships with friends	0.39	0.26	0.11	0.08
Relationships with loved ones	0.35	0.20	−0.15	0.28
Eigenvalue	1.80	1.09	0.87	0.57
Factor contribution ratio	0.258	0.157	0.125	0.081
Cumulative contribution ratio	0.258	0.415	0.540	0.621

. The resulting factors were interpreted and labeled as “Family-Oriented Relationships” (emphasizing the importance of family and relatives), “Workplace/School Relationships” (emphasizing the significance of workplace and school connections), “Community Relationships” (emphasizing the neighborhood and community ties), and “Internet Relationships” (emphasizing the relationships formed in virtual networks).

4.2. Cluster Analysis of Respondents and Behavioral Characteristics

This section presents a cluster analysis using the K-means method, applied to the factor scores derived from the factor analysis represented in Table 2. K-means clustering was selected due to its computational efficiency and suitability for large datasets such as ours (n = 6210). Compared to hierarchical clustering, which can become computationally intensive and less scalable with larger samples, K-means can offer faster convergence and clearer interpretability especially when the number of clusters is predefined or empirically estimated [42]. We used the elbow method [43] to determine the optimal number of clusters, resulting in seven distinct clusters. Each cluster’s proportion of alternative behaviors, particularly those related to shopping, was calculated and standardized to serve as an explanatory variable. This clustering approach allowed us to capture meaningful behavioral profiles reflecting different levels of digital service usage and social interaction preferences across respondent groups. Figure 11 reveals the result of the cluster analysis.

The study further examined the characteristics and proportions of digital alternative behaviors within each cluster, providing insights into how respondents’ behavioral tendencies align with their factor scores and residential location attributes. This classification highlights digital alternative behavior across various respondent groups, offering a deeper understanding of the factors driving these behaviors. The results of cluster analysis and each cluster characteristic are described below.

Following the cluster analysis, respondents were grouped into distinct clusters based on their factor scores and the proportion of digital alternative behaviors, particularly in relation to shopping. The following clustering provided us with an in-depth explanation of the characteristics and behavioral patterns observed within each cluster, shedding light on how human relationships, preferences for digital alternatives, and shopping behaviors intersect across different respondent groups.

• Cluster 1

Cluster 1 consists of individuals who assign minimal importance to human relationships, both in physical space and cyberspace. These respondents seem to prioritize their needs over daily social interactions and prefer online shopping. The behavioral tendencies of this group suggest that they rely on digital alternatives to a significant extent, potentially due to their low emphasis on social networks.

• Cluster 2

Similarly to Cluster 1, respondents in Cluster 2 also place little importance on social networks. However, in contrast to Cluster 1, this group exhibits a low proportion of shopping digital substitution. Instead, they show a preference for shopping at physical stores, which can potentially suggest that despite their reluctance to engage in social relationships, they still prefer to do their shopping in physical stores.

• Cluster 3

Cluster 3 is composed of individuals who place a relatively high value on social relationships, both in their physical space and cyberspace. Despite the emphasis on human interactions, they also actively engage in online shopping. This can suggest that respondents in this group are capable of efficiently performing both physical activities and their digital alternatives, as they can make strategic use of digital alternatives for shopping while maintaining social interactions across various networks.

• Cluster 4

In Cluster 4, although the respondents exhibit a low level of importance placed on social relationships, especially within the family, they tend to rely more on physical stores for shopping. This can suggest that there might be external factors (such as convenience, familiarity, or personal preferences) that influence their shopping choice behaviors. Additionally, we could witness this cluster’s strong tendency towards physical shopping rather than digital alternatives.

• Cluster 5

Although respondents in Cluster 5 valued social relationships to varying degrees, their shopping behaviors are not strongly influenced by giving importance to social relationships. Their low shopping substitution ratio indicates a preference for physical stores, potentially driven by the direct interaction they have with their environment and the people around them. This cluster may reflect individuals who engage in shopping activities as part of their social experience, valuing face-to-face interactions.

• Cluster 6

In Cluster 6, while respondents show a high level of importance placed on relationships, their shopping substitution rate remains low. This suggests that, like Cluster 5, these individuals prefer physical stores despite their value on social interactions. The emphasis on direct human connections may play a significant role in shaping their shopping preferences, where the in-person experience is preferred over digital alternatives.

• Cluster 7

Cluster 7 is characterized by individuals who do not place significant emphasis on any of their social relationships. These respondents have a moderate shopping substitution ratio, suggesting that they are somewhat indifferent toward both physical shopping and digital alternatives. This group may have a more passive approach toward shopping and social interactions, showing a general reluctance to engage in both social interactions and shopping digital alternatives.

Table 3. Characteristics of each cluster.

Item		Cluster 1	Cluster 2	Cluster 3	Cluster 4	Cluster 5	Cluster 6	Cluster 7
		n = 936 15.1%	n = 1292 20.8%	n = 492 7.9%	n = 1237 19.9%	n = 630 10.1%	n = 618 10.0%	n = 1004 16.2%
Gender	Men	15.2%	21.2%	7.5%	20.3%	10.0%	9.8%	15.9%
	Women	15.0%	20.4%	8.3%	19.5%	10.3%	10.1%	16.4%
Age	Average Age	48.5	51.1	47.5	50	49.2	49.5	48.9
	20s	17.4%	16.1%	10.0%	18.7%	10.4%	10.1%	17.1%
	30s	16.4%	19.9%	9.1%	20.8%	9.6%	8.9%	15.2%
	40s	14.9%	19.8%	8.5%	20.2%	10.3%	10.5%	15.7%
	50s	13.5%	22.8%	5.9%	19.1%	11.7%	11.6%	15.3%
	60 and over	14.7%	22.9%	7.3%	20.6%	9.4%	9.4%	15.6%
Driving License	Yes	15.0%	20.8%	7.9%	19.8%	10.6%	9.9%	16.0%
	Men	15.2%	21.2%	7.5%	20.3%	10.0%	9.8%	15.9%

provides an overview of cluster analysis by gender, age group, and driver’s license status. As can be seen in Table 3, clusters 2 and 4 have a higher percentage of men and women, and huge significant differences cannot be witnessed in the age distribution among the clusters. Regarding age distribution, clusters 1 and 3 have a higher concentration of respondents in their 20s and 30s, with respondents in these clusters showing a higher substitution ratio but varying importance placed on social networks, while Cluster 2 shows a greater proportion of older age groups, with respondents in this cluster ascribing higher importance to social networks and showing a lower substitution ratio. As previously shown in Figure 6, younger individuals are more likely to adopt digital alternative behaviors, such as online shopping, whereas older individuals tend to prefer physical shopping. However, Cluster 5 has a higher proportion of respondents with a driver’s license compared to those without. Since this cluster appears to favor physical shopping, we can infer that access to private vehicles may facilitate their preference for physical shopping.

5. Residential Relocation Choice Analysis

5.1. Overview of Residential Relocation Choice

This section investigates the respondents’ intentions to relocate from their current place of residence. The five residence categories described earlier are recategorized into urban and rural areas. Residences such as the city center and its surroundings fall into urban areas, whereas suburbs, rural, and nature-rich areas fall into rural areas. Figure 12 illustrates the relocation intention of the respondents obtained from the survey analysis. The results indicate that 35.9% of respondents in urban areas and 30.8% in rural areas expressed an intention to relocate, whereas 59.2% of respondents living in urban areas and 64.2% of respondents living in rural areas expressed no intention of relocation. However, about 5% of respondents living in both residence categories had already changed their place of residence. These findings confirm that the residential relocation intention is lower by 5% in rural areas compared to urban areas. The subsequent part of this paper will disclose what kinds of factors affect the residential relocation intention.

5.2. Factors Influencing Residential Relocation Choice

To further analyze the determinants of residential relocation, a binomial logistic regression analysis was performed, with the intention to change residence as the dependent variable. Explanatory variables included gender, age, possession of a driver’s license, attachment to place of residence, and the percentage of shopping substitution. The analysis results are shown in

Table 4. Factors affecting residential relocation choice in urban areas.

Survey Item	Urban Areas
	Regression Coefficient	Standard Error	Odds Ratio	95% Confidence Interval	Significance
Age	−0.04536	0.38384	0.95565	0.94955–0.96169	1.60 × 10−44	***
Sex	−0.02446	0.00324	0.97584	0.80443–1.18366	0.803878
Driving license	0.19685	0.09845	1.21756	0.92918–1.60155	0.156044
Attachment to the place of residence	−0.34296	0.13877	0.70967	0.65198–0.77173	1.50 × 10−15	***
Shopping substitution rate	1.13927	0.26456	3.12450	1.86155–5.25563	1.66 × 10−5	***

Note: *** indicates statistical significance at p < 0.001.

for urban areas and

Table 5. Factors affecting residential relocation choice in rural areas.

Survey Item	Rural Areas
	Regression Coefficient	Standard Error	Odds Ratio	95% Confidence Interval	Significance
Age	−0.04632	0.00275	0.95474	0.94957–0.95984	6.95 × 10−64	***
Sex	0.00012	0.07725	1.00012	0.85955–1.16360	0.998740
Driver’s license	0.13059	0.13737	1.13950	0.87249–1.49557	0.341771
Attachment to the place of residence	−0.40555	0.03282	0.66661	0.62489–0.71069	4.38 × 10−35	***
Shopping substitution ratio	0.19790	0.21480	1.21884	0.79853–1.85411	0.356874

Note: *** indicates statistical significance at p < 0.001.

for rural areas.

The analysis result of this study reveals that age had a significant negative impact on the residential relocation choice of individuals in both urban and rural areas: as individuals age, their likelihood of relocating their residential place decreases. This may be attributed to older residents’ preference for maintaining their established living environments and avoiding relocation inconveniences.

Similarly, attachment to one’s residential place also showed a significant negative impact on the residential relocation choice of individuals in both urban and rural areas. The analysis results disclosed that a higher attachment to the current residential location is associated with a reduced intention to relocate, reflecting the importance of local community satisfaction and familiarity with the surrounding environment. However, the impact of attachment to residential places on residential relocation choice is stronger in rural areas, suggesting that community ties in rural settings play a more substantial role in relocation.

Additionally, the impact analysis of shopping substitution on residential location choices in this study disclosed some differences between urban and rural areas regarding the impact of shopping substitution on residential relocation choices. In urban areas, the odds ratio was 3.12, meaning the shopping substitution ratio has a higher impact on residential choice in urban areas, compared to the odds ratio of 1.21 in rural areas. It can be argued that the availability of alternative shopping options, such as online shopping, can increase residential retention in urban areas. This may suggest that improved accessibility to goods and services through ICT and transportation networks enhances the convenience of urban environments. In contrast, in rural areas, the odds ratio was 1.21, and the effect was not significant. This may suggest that in rural areas, shopping substitution has a limited impact on relocation choices. This may be due to established lifestyles that rely on car travel, where physical store accessibility remains an integral part of daily life.

6. Results

6.1. Behavioral Substitution Ratios

The analysis results of digital behavior substitutability in this study confirm the possibility of substitution of different physical activities with digital alternatives, especially shopping behavior. The substitution is reduced by orders from shopping, dining, schooling, working, and hospital visits. Only a 3% substitution was witnessed for the hospital visit digital alternative. Younger individuals show a significant substitution ratio in shopping digital alternatives, whereas the shopping substitution ratio reduces as people age. However, older adults and the elderly also demonstrate a considerable substitution ratio in shopping for digital alternatives. Furthermore, the study discloses that the impacts of residential location on the substitution ratio are insignificant. Respondents living in urban areas show an insignificant, only 0.33% higher substitution ratio compared to the respondents living in rural areas.

Assuming the individuals who have driving licenses can drive a private vehicle and their choice of transportation mode is greater, they will do more in-store shopping than online shopping. However, the study discloses that mobility, such as having a car, can influence digital alternative behavior, with individuals holding a driving license favoring online shopping, implying that access to transportation does not necessarily reduce the substitution ratio. In contrast, having a car enhances online shopping by giving individuals an opportunity to use their private vehicle to inspect products in a physical store and purchase products online at a lower price, which may increase traffic congestion and private vehicle usage.

Additionally, the study reveals that social relationships, particularly virtual network relationships, affect shopping digital behavior, and emphasizes the growing influence of virtual networks on online shopping. Furthermore, the result of this study confirms that digital substitution affects social and cultural environments, considering the tendency of people to favor digital alternatives but place less importance on face-to-face interaction. These findings show that digital alternatives affect social networks and urban infrastructures.

6.2. Relationship Types Influencing Behavioral Choices

The clustering analysis of this study, based on the K-means method and the elbow technique, identified seven distinct respondent groups, each characterized by different levels of emphasis on human relationships and shopping digital alternative behaviors. The results disclosed that younger individuals, particularly in clusters 1 and 3, are more inclined toward online shopping, while older respondents in cluster 2 exhibit a stronger preference for physical shopping. Additionally, it was found that respondents putting higher importance on social relationships, as seen in clusters 3 and 6, tend to have a higher and lower substitution ratio, respectively. This means that although clusters 3 and 6 value social relationships, their shopping behaviors are different. In contrast, respondents who place less importance on social relationships, such as Cluster 1, prefer more digital alternatives.

It was also found that mobility factors also influence these digital alternative behaviors, with Cluster 5 preferring physical shopping and having a higher proportion of respondents with a driving license, indicating that access to a private vehicle supports physical shopping for some individual groups. These findings highlight the diversity in shopping digital alternative behavior shaped by social interactions, social network usage, and mobility, providing valuable insights for businesses, policymakers, and urban planners.

Cluster analysis of this study, which categorized respondents into distinct groups based on their digital alternative behavioral tendencies and factor scores, reveals different levels of digital substitution ratio. Some clusters show a strong preference for shopping digital alternatives, often due to minimal emphasis on social relationships, while others prefer physical stores despite being socially active. The findings of this study disclose that social relationships, mobility, and age significantly influence the digital substitution ratio. This classification provides valuable insights into understanding how different respondent groups navigate their activity paths in both real and virtual networks. These findings can contribute to discussions on the integration of digital technologies in daily life, substitutability of access, and proposing a tradeoff between physical activities and digital alternatives to enhance overall QOL.

6.3. Residential Relocation Choice

The analysis result of this study discloses that residential relocation intentions are shaped by multiple factors such as age, attachment to the place of residence, and the availability of shopping alternatives. Younger individuals and those with weaker community ties are more inclined to relocate, and the impact of shopping substitution differs between urban and rural areas. In urban areas, digital shopping alternatives enhance residential convenience, and they affect the residential location choice. Whereas, in rural areas, traditional mobility patterns and local attachments remain dominant determinants of relocation choice behavior, and digital shopping alternatives have less impact on residential choice. The result of this study confirms that digital substitution affects sustainable urban structures, considering residential relocation, changes in land use and transportation, and population distribution. These changes will require new facilities to meet the people’s demands. Therefore, the result of this study suggests that sustainable urban planning requires a tradeoff between services offered in physical space and cyberspace to contribute to the development of sustainable urban structures, considering the needs of each individual and their well-being.

7. Discussion

The result of this study reveals that urban structure, including land use and social interactions, will be affected by digital service substitution. The findings of the study suggest that individuals who have a higher digital service substitution ratio are less likely to give importance to actual human relationships formed in physical space, such as those between families or relatives; in contrast, they are more likely to place higher importance on acquaintanceships and friendships formed in cyberspace, such as social media. The increase in digital service substitution will weaken in-person relationships and can potentially affect the strength, stability, and connectedness of communities over time. Individuals with a higher digital service substitution ratio tend to be more comfortable and active in digital environments. This digital lifestyle naturally extends to social behavior, where online communication becomes not just a supplement, but a primary mode of social interaction. Due to this, individuals valuing virtual relationships more, are more likely to have a higher digital service substitution ratio. These findings redefine social networks and emphasize long-term implications for community cohesion by addressing potential challenges related to disaster response and evacuation, social inclusion and mental health, public safety and crime prevention, and overall sustainable behavior and shared resources.

Additionally, in line with existing studies [44,45] that have argued that online shopping reduces traffic congestion and greenhouse gas emissions, our study reveals an unexpected finding that individuals with driving licenses and cars are approximately 9% more likely to engage in online shopping compared to those who do not have cars. The reason behind this could be that having a car eases goods inspection for individuals when purchasing, while enjoying the convenience and lower prices of online shopping. For instance, one of the factors that could diminish online shopping service value is in-person product inspection [41]. In physical stores, individuals have the opportunity to actually look closely at the product, but in online shopping, such opportunity is limited. However, online shopping is said to be convenient and have a lower price. We believe these factors are associated with the shopping choice behavior of individuals, and the results suggest that online shopping may not necessarily reduce traffic congestion or private vehicle usage. These insights have policy relevance for sustainable urban development and smart cities.

The findings of this study also highlight important digital inequalities. Older adults and rural residents were found to engage less in digital service substitution, which may reflect barriers such as limited digital literacy, lower access to high-speed internet, or reduced exposure to online platforms. These disparities can lead to unequal access to public digital services, digital commercial goods, and virtual social interaction opportunities. In the context of sustainable and inclusive urban development, especially targeting SDG 11 (Sustainable Cities and Communities), it is essential that policymakers address these gaps by investing in digital infrastructure, education, and support systems that ensure all populations—regardless of age or geography—can benefit from the digital transformation of services and social life to have a higher gross regional happiness.

Considering the findings of this study, the authors of this paper argue that physical services need not be substituted for digital alternatives, completely considering their consequences. However, we believe that digital services should be complementary to physical services, which would enhance service access and result in user satisfaction and a higher overall QOL, ensuring equity and sustainability. Additionally, the study suggests making a tradeoff between physical and digital alternative service usage. Excessive usage of either of them will negatively impact the sustainable urban structure, social environment, and overall QOL. For instance, excessive usage of digital services will contribute to more social isolation and reduce face-to-face communication and mobility, which will affect physical and mental health. On the other hand, there is a need for digital services for different reasons, such as increasing accessibility for all, including disabled people, demand and preference of individuals, and opportunities for businesses to innovate, expand their reach, and generate revenue in an increasingly digital economy. In contrast, physical activities generate more CO₂, due to relying on transportation, and consume more resources, e.g., facility construction, maintenance, etc., but they enhance face-to-face social and cultural interactions, promote individual mobility, and diminish the negative effects of digital service usage.

The author of this paper urges urban planners and policymakers to address both the negative and positive impacts of digital service substitution for future sustainable urban planning and smart city developments. This paper suggests making a tradeoff between physical and digital service usage and should be considered at policy- and strategy-making levels to have a larger positive socio-economic impact aligned with the United Nations Sustainable Development Goals (SDGs) and reduce the negative impact of either of them.

8. Conclusions

8.1. Summary

This study initially explored trends in the frequency of digital alternative usage, examining how these behaviors correlated with factors such as gender, age, and driving license status. The analysis result confirmed that socio-demographic factors significantly influence the adoption of digital alternatives. A cross-tabulation in this study further investigated the relationship between the emphasis placed on interpersonal relationships and the frequency of digital alternative usage. The analysis results suggest that the value placed on human relationships in real space will negatively affect digital alternative behavior choices.

Subsequently, factor analysis was performed to investigate the importance respondents placed on different relationships, identifying key factors and their influence on digital alternative behaviors. Using these results, respondents were classified into distinct clusters, with specific attributes and digital alternative behavior patterns explained through the percentage of alternative behaviors within each cluster. The analysis results disclosed that individuals who frequently use digital alternatives place less importance on human relationships and face-to-face interaction, and confirmed that digital alternative substitution has impacts on social and cultural environments.

A binomial logistic regression analysis of age and residential attachment in this study confirmed that both factors significantly affect residential relocation choices. The impact of attachment was especially strong in rural areas, where personal ties to the community are more pronounced. Additionally, the study disclosed that in urban areas, digital alternative substitution affects residential relocation choice, while in rural areas, it has no clear impact, likely reflecting the differences in lifestyle and mobility across residential locations. The result of this study suggests that sustainable urban structure, land use, transportation, and population distribution are affected by digital alternative substitution to some extent.

This study contributes to several Sustainable Development Goals, particularly SDG 11 (Sustainable Cities and Communities), by informing digital-age urban planning strategies. It also supports SDG 9 (Industry, Innovation, and Infrastructure) and SDG 10 (Reduced Inequalities) through its insights on digital access, behavioral shifts, and inclusive service design. Furthermore, the environmental implications of reduced physical mobility align with the objectives of SDG 13 (Climate Action).

8.2. Policy Recommendations

Based on the study findings, four actionable recommendations to balance digital innovation with inclusive, sustainable, and human-centered urban design are suggested below for urban planners and policymakers, which can guide sustainable urban planning in the digital era:

• Design well-balanced city development that integrates both digital and physical infrastructure to support diverse lifestyles. Urban planners should ensure that digital services complement, rather than replace, essential in-person services such as healthcare, tourism, and community centers.
• Promote digital service substitution only as a means of improving efficiency, reducing environmental impacts, and enhancing service accessibility, particularly for populations with mobility constraints. However, adoption should be encouraged based on user needs and digital readiness.
• Recognize the limits of digital substitution. While digital tools enable flexibility and convenience, real-life human networks remain critical for quality of life, mental health, social equity, and disaster resilience. Urban policies should continue to invest in social infrastructure that fosters community cohesion.
• Pursue the development of sustainable compact, well-designed, and inclusive urban forms, where digital services are embedded into daily life, but where walkable neighborhoods, public spaces, and physical connectivity are preserved and enhanced. A sustainable compact well-designed smart city model can promote sustainability while supporting social diversity and digital inclusion.

These recommendations can help urban planners navigate the tradeoffs of digital transformation and ensure that smart city development remains equitable, resilient, and centered on human well-being.

8.3. Limitations and Future Research Prospects

In the current study’s analyses, we did not include variables such as income, occupation type, or education due to the relatively low variance in these attributes in Japan. This might limit the generalizability of our findings and could be considered a limitation of the study. Future research can expand the study target area to other countries and can incorporate these variables in more diverse socio-economic contexts or cross-country comparative studies to better understand their influence on digital service substitution and residential decision-making. In addition, although we distributed a questionnaire survey that was segmented by age, gender, and place of residence and could obtain enough responses (n = 6210), there might have been some potential bias with the self-reporting survey. However, a combination of stated preference and revealed preference studies can enhance this study.

Furthermore, another study should be conducted to develop a quantitative tradeoff model between physical and digital service usage to balance digital service substitution while considering the impacts of digital service substitution on sustainable urban structures, social environments, and the overall QOL of individuals. Additionally, a study is required to delve deeper into the role of social networks and the barriers to relocation, especially regarding residential location choices concerning digital service substitution, and explore the longitudinal impacts of digital service substitution on the urban form to investigate the temporal changes in the residential location choices, land use patterns, and transportation demand. Moreover, the development of a comprehensive model that evaluates the impact of digital service substitution within a multi-layer network is crucial for understanding its broader implications on an individual’s daily life behavior, urban mobility, and residential location choice while considering the different socio-demographic characteristics of individuals. Incorporating a multi-layer network approach and social dynamic simulations to consider temporal changes, as demonstrated by Sugiki [46] and tested effectively by Mutahari [47], would provide an innovative and powerful tool for urban planners and decision-makers. This approach can guide sustainable urban development strategies that align with the United Nations’ SDGs, ensuring sustainable and inclusive urban environments with a higher QOL for everyone.