Analyzing Young Adult Travelers’ Perception and Impacts of Carpooling on Traffic Sustainability

Abstract

Since young adults (i.e., aged between 18 and 30) are generally more flexible and conscious of economic and environmental concerns, it is essential to assess their travel perceptions and tendency to shift towards sustainable transportation modes. Therefore, this research aims to evaluate the acceptance of carpooling (CP) in the younger population to alleviate traffic congestion, fuel demand, and pollution. This study was conducted in Islamabad, a fast-growing city in Pakistan with a high percentage of young residents, to evaluate travelers’ perceptions, mode choice decisions, and potential acceptance of CP. An online questionnaire-based survey was developed and distributed to collect a total of 656 responses from the young adult population. Various factors influencing travel perceptions, mode choice, and tendency to carpool were categorized and analyzed using confirmatory factor analysis, stepwise, and mediated multiple regression analysis. Thereafter, the tangible impacts of CP, including reduced traffic congestion, fuel consumption, and emissions resulting from the potential adoption of CP by the younger population, were quantified. The findings imply that young travelers’ mode choice mediates the relationship between travelers’ perception and the tendency to carpool. The outcomes also affirm the impact of demographic variables, including gender and education levels, on the willingness of the young adult population to shift towards carpooling. Furthermore, results predicted that adopting CP, especially by younger adults, may reduce by about 33.6% the number of private vehicles in Islamabad. The findings of this study could offer helpful insights for transport planners, commercial carpool service operators, environmentalists, and policymakers to promote carpool systems, especially among young adults.

1. Introduction

Over the last few decades, problems like highly congested traffic and environmental pollution have grown alarmingly. The rapid expansion of the automotive industry has led to several adverse consequences, the most notable of which are increased travel costs and time, increased fuel consumption, deteriorated environment, elevated congestion levels, and decreased overall efficiency of the transportation systems [1,2]. Carpooling (CP) is a revolutionary transportation mode that has gained vast popularity worldwide in recent decades as an alternative to individual car ownership [3]. This mode of transportation results in a wide range of potential benefits, including reducing greenhouse gas emissions (GHG), lowering car ownership rates, addressing parking issues, enhancing mobility prospects, and mitigating traffic congestion [4,5,6]. In addition, CP positively influences various social factors, most notably by offering high-quality and well-balanced transportation alternatives to people who do not possess private vehicles [7,8,9,10,11].

The idea of CP originated in Europe in the 1940s and has since expanded globally to other parts of the world, including Australia, North America, and Asia [7]. In 2019, there were a total of 9818 registered vehicles across the continent of North America that provided services to a considerable user base of more than 378,000 people [8]. CP was introduced in 26 nations in 2010 to minimize traffic costs and environmental impact. This sustainable transportation concept encouraged commuters to share rides, which reduced traffic, energy use, and emissions, making transportation greener and cheaper [3,12]. Over the past few decades, CP has become an increasingly prominent and highly studied topic among urban planners. Both society and the environment have benefited from its widespread adoption in various cities across the globe. The increasing interest in CP can be attributed to its ability to reduce transportation difficulties and encourage environmentally friendly mobility options [13].

Realizing the significance of sustainable transportation modes, CP is being actively promoted by governments as well as commercial entities. CP is being encouraged through policy interventions and tax incentives [14]. Moreover, CP services are currently being facilitated through various applications worldwide, each with its operational guidelines. For instance, the Waze app for carpooling employs a peer-to-peer paradigm, where a higher number of passengers leads to lower costs per passenger. In addition, apps such as Getting-Together take a novel approach by repurposing taxis as minibusses in their transportation network. Numerous passengers traveling in the same general direction can share the same cab, making the ride more cost-effective and efficient for all parties [15,16]. All these strategies encourage the effective utilization of shared modes of transportation, making it a practical choice for urban travelers [17].

In developing countries like Pakistan, the rapid development of the urban population, increasing travel demand, and an exponential increase in vehicle ownership have created an unprecedented burden on the economy, environment, and overall living standards. Karachi, Lahore, Rawalpindi, and Islamabad, all major cities in Pakistan, are facing substantial traffic congestion issues. Mobility, air quality, and the overall livability of metropolitan areas are negatively impacted due to the increased number of vehicles on the roads, which, when paired with inadequate transportation infrastructure, have led to severe congestion challenges in many urban centers [18]. Residents of the metropolitan area, especially in developing countries, typically travel via private vehicles rather than rely on the insufficient and inefficient public transportation systems currently available [3]. The prevailing urbanization trends have put massive pressure on cities to expand infrastructure to meet current and future traffic demands and promote efficient city mobility. Nevertheless, infrastructure projects must be carefully planned and implemented after considering all the behavioral, spatial, and technical parameters [19].

According to the 2017 population census results in Pakistan, the capital city of Pakistan, Islamabad, exhibited a remarkable yearly population growth rate of 4.91% [20]. The city of Islamabad is experiencing a booming demand for automobiles due to the city’s continuously expanding population. More than 0.9 million automobiles were registered in 2019 to meet this growing demand, as stated in the data compiled by the Excise Office of Islamabad Capital Territory (ICT) [21]. Due to the ever-increasing number of cars on the road, the nation’s capital is experiencing extreme traffic congestion and environmental pollution. This scenario poses a significant challenge to decision-makers, planners, and other concerned authorities. Due to this issue, the nation sustains millions of dollars of economic damage annually. Conversely, creating a dynamic and efficient transport infrastructure can stimulate significant economic activity [4,18]. As a result, developing a solution to alleviate traffic congestion and environmental pollution has become an absolute necessity. It is also worth mentioning that providing vast public transportation networks, especially for struggling economies, is not generally viable. In this context, CP has the potential to become a sustainable transportation alternative for reducing road traffic, lowering fuel consumption, and decreasing carbon dioxide emissions. Even though this concept has been thoroughly investigated in developed nations, it still requires additional focus and comprehensive investigations in the developing world to adapt to their specific requirements and difficulties [13,22,23,24,25,26].

Implementing a CP program in Islamabad may be one of the potential solutions to the city’s problems with pollution and traffic congestion. The success of such a system is majorly dependent on the acceptability by the general public and market demand, which still needs to be analyzed comprehensively within the framework of the cultural attitude of Pakistanis. In order to induce a shift towards carpooling and other sustainable forms of transport, it is essential to examine different categories of residents and devise specific programs for them. In this regard, the younger adult population (i.e., aged between 18 and 30 years), generally more flexible and aware of economic and environmental concerns, may become the driving force in the attempt to change cultural attitudes and travel perceptions. University mates, office colleagues, and even strangers with similar characteristics in the time and space domain may be encouraged to join CP groups. CP may be the best option, especially for the young population, to achieve a sustainable transport system in Islamabad, where traffic congestion worsens daily, and most people drive their private automobiles.

Currently, there are no formally incentivized, officially promoted, or systemized carpooling programs in Islamabad; hence, this research aims to evaluate the young adult population’s travel perceptions and the tendency to adopt CP. In order to study the influence of travelers’ perceptions on the adoption of carpooling as a mode of transportation, a model has been proposed that includes mode choice as a mediator on the relationship between traveler perception and tendency to carpool. Over the last few years, various models [4,27,28,29] in the context of Pakistan have been developed to study the influence of travelers’ perceptions on adopting carpooling as a mode of transportation. However, the novelty and contributions of this study lie in its focused examination of CP acceptance among young adults (aged 18 to 30 years) in Islamabad, Pakistan. This aspect has received limited attention in the existing literature on shared mobility modes, specifically in developing countries. This research provides comprehensive insights into the factors influencing young travel perceptions, mode choice, and tendency to carpool. Moreover, considering the dynamics of carpooling, ranging from behavioral aspects to technical limitations, the mediating effect of mode choice factors has been incorporated in the proposed model to investigate the relationship between traveler perception and tendency to carpool. In addition, the study aims to utilize the survey outcomes to quantify and predict the tangible impacts of CP on traffic sustainability, including factors like reductions in traffic congestion, fuel consumption, and emissions.

In order to achieve the outlined objectives, an online questionnaire-based survey was conducted, including several factors related to travel perception, mode choice decisions, and overall tendency to adopt carpooling by the young adult population of Islamabad. The data collected were used to determine the travel preferences of the participants and the level of their willingness to accept carpooling as an alternate mode of transportation. The proposed model and the developed hypotheses were evaluated based on confirmatory factor analysis, stepwise, and mediated multiple regression. Moreover, the obtained data regarding travel perception and willingness to carpool were further utilized to quantify the potential impacts of CP on traffic congestion, fuel consumption, and emissions. For this purpose, control data on Islamabad traffic were also collected to predict the influence on traffic congestion and the reduction in fuel consumption and pollution emissions.

The rest of this paper is organized as follows: Firstly, the literature review section gives an overview of existing studies. This is followed by a description of the proposed methodology, including research design, data collection, and model development approach. The succeeding section presents the descriptive and statistical analyses. Moreover, this section also includes quantifying the traffic sustainability factors, including reduced traffic congestion, fuel consumption, and emissions resulting from the potential adoption of CP by the younger population. Thereafter, the findings, study limitations, and future research directions are discussed. Finally, the paper concludes with a brief outline, recommendations, and implications of the study.

2. Literature Review

CP has been demonstrated in numerous studies to effectively cut greenhouse gas emissions within a range of 0.84 to 0.58 metric tons, depending on the study area [25,26]. Zong et al. [30] found that introducing an incentivized CP program effectively reduces traffic congestion and carbon emissions. After participating in a CP program, around 15- 29% of people in Canada, while 11–26% of people in the United States gave up ownership of their vehicles [7,8]. CP adoption is particularly beneficial for individuals who rely heavily on public transportation, resulting in decreased traffic congestion and lower demand for parking spots [9,13,22,23,24]. Many essential benefits of using CP include lower vehicle maintenance, reduced parking costs, and protected financial investments in automobiles [10,31,32,33]. CP initiatives promote environmentally friendly transportation modes like walking, biking, and public transit, improving overall health and reducing environmental pollutants [16]. Additionally, using electric vehicles in CP systems can reduce emissions and enhance ecological health [34,35,36,37,38,39].

Several researchers worldwide have investigated the behavioral aspects of CP adoption. A study by Yoon et al. [40] in Beijing, China, examined the patterns of one-way and two-way CP trips. According to the study’s conclusions, people with no personal vehicle were highly inclined towards CP for two-way trips. Julagasigorn et al. [41] proposed a conceptual framework to address the gap in psychological understanding of carpooling behavior by conducting a systematic literature review. It identified 18 psychological factors motivating drivers and passengers to carpool, categorizing them as typical or specific to each group. The Theory of Planned Behavior and Norm-Activation Model and other theories, such as Consumer Perceived Value and Social Capital, were proposed as suitable frameworks for carpooling research. Another study by Abutaleb et al. [42] investigated consumer intentions toward CP, utilizing the Theory of Planned Behavior (TPB) supplemented with personal norms and collaborative consumption motivators. A survey of 500 millennials in Egypt was analyzed using structural equation modeling. Subjective norms and attitudes significantly influenced intentions, followed by perceived behavioral control and personal norms. Additionally, economic benefits and sustainability positively impacted attitudes. This study contributed original insights into carpooling behavior and sustainability within the context of the sharing economy.

Moreover, Shaheen and Martin [43] analyzed citizens’ views toward CP. They concluded that younger individuals with greater education levels were more likely to accept and implement the CP approach. A study conducted in China evaluated the factors influencing college students’ attitudes toward CP. This study aimed to address the underutilization of CP despite its potential as a sustainable transportation option. Through an online survey at Shenzhen University with 514 participants, the study analyzed personal, travel, and attitude attributes using a multinomial logit model. Results indicated that safety concerns and high costs neutralize students’ attitudes, emphasizing the importance of comfort and targeted interventions for promoting carpooling among students [44]. Another research conducted by Liu et al. [45] examined the public transit behaviors of college students compared to other young adults in the Denver-Aurora region, Colorado, using 2015 survey data. The study explored the factors that influence students’ prioritization of environmentally sustainable transit options. Findings revealed that student transit users lived further from city centers and light rail stations. It was also observed that students in the study area often drove to stations and had different factors influencing their transit decisions, possibly due to housing affordability challenges in transit-rich areas.

Various studies and models in Pakistan have also been developed to study the influence of travelers’ perceptions on adopting carpooling as a mode of transportation. A study by Javid et al. [28] investigated carpooling policy in Lahore City to tackle traffic congestion in developing countries. The authors identified key motives and constraints using an online survey and analyzed data through factor analysis and structural modeling. Results highlighted social, environmental, and economic benefits as promoters, while concerns about privacy, security, and service constraints were hurdles in CP adoption. The findings stressed the importance of incentives, disincentives, and attitude modifications to promote carpooling effectively. Another study investigated the public perception and acceptance of carsharing systems in Lahore, Pakistan. Econometric models, including Multinomial Logit and Nested Logit, were employed to examine the influence of attributes like travel time, cost, and privacy alongside sociodemographic factors such as age, education, and income [29]. A similar study was also conducted to examine the acceptance of carsharing systems in Peshawar, Pakistan, through an online survey. 453 valid responses were analyzed using Multinomial Logit and Nested Logit models. Demographic characteristics, including gender, job, income, and service attributes, including travel time, travel cost, registration fees, and capital cost, were significant [4]. Finally, a study conducted by Ayaz et al. [27] investigated public acceptance of carpooling in Islamabad. A state preference questionnaire survey explored motives and constraints related to CP. Structural equation modeling was performed to analyze the collected data. Findings revealed the significance of demographic factors, i.e., gender. It was also found that high-occupancy vehicle (HOV) lanes had a positive impact, whereas unknown carpooling partners had a negative effect on the intention to shift to CP.

After a detailed literature review, it was concluded that there is a need to examine different categories of residents and devise specific CP programs for them to induce a shift towards carpooling and other sustainable forms of transport. This aspect has received limited attention in existing literature on shared mobility modes, specifically in developing countries. Therefore, the research was designed to assess the travel perceptions, mode choice decisions, and CP acceptance of the young adult population (aged 18 to 30 years) in Islamabad, Pakistan.

3. Methodology

3.1. Research Design

As mentioned in the Introduction section, currently, there are no formally incentivized, officially promoted, or systemized carpooling programs in Islamabad; hence, an online questionnaire-based survey including several factors related to travel perception, mode choice decisions, and overall tendency to adopt carpooling by the young adult population of Islamabad was conducted. The questionnaire was created in Google Forms and dispersed across Islamabad at various sectors and places. The data collected were used to determine the travel preferences of the participants and the level of their willingness to accept carpooling as an alternate mode of transportation in Islamabad, Pakistan. Moreover, the obtained data regarding travel perception and willingness to carpool were further utilized to quantify the impacts of CP on traffic congestion, fuel consumption, and emissions. The roadmap of the study is illustrated in Figure 1.

Numerous earlier studies have been carried out in a variety of locations all over the world, making use of online surveys as a method of research. In this scenario, people’s impressions of car and bike sharing were investigated with the help of an electronic questionnaire (survey) conducted with survey forms [46]. In addition, Wang et al. [47] conducted an online survey in China in 2017 to collect primary data on individual travelers’ acceptance of CP.

This study used lessons from studies conducted in developing countries to create a 36-question survey [27,28]. The survey consisted of three sections to collect specialized data: The first eight questions collected primary data and social demographics from travelers. Age, education, income, gender, location, vehicle ownership, and driving license status were covered. The second segment, consisting of six questions, gathered information regarding the awareness levels of respondents towards CP with graphics and definitions. This section measured participants’ knowledge regarding CP. The third section was based on twenty-two (22) questions. The section was a combination of questions aimed at collecting data regarding overall young adult travelers’ perceptions and mode preferences, specifically emphasizing attributes/factors that potentially influence the adoption of CP, e.g., ‘Which level of travel time reduction would encourage you to choose carpooling as the mode of your trip?’ with 5 options, i.e., 0, 5, 10, 15 and 20 min. Another example of a CP-related scenario is ‘How many co-travelers would you prefer in the ride if you choose carpooling as the mode of your trip?’ with 4 options, i.e., 1, 2, 3, and more than 3 co-travelers.

Respondents were asked about their preferred modes of transportation, travel perceptions, and other associated factors to gauge information regarding their tendency to adopt carpooling. Respondents expressed their preferences related to willingness to share their cars with local passengers, willingness to use carpooling in the presence of HOV lanes, whether they preferred being a CP passenger or driver, the purpose of CP, etc. This section also examined respondents’ perceptions regarding CP’s potential to reduce travel costs, distance, time, car parking issues, car parking costs, and safety and comfort concerns when traveling with strangers.

Table 1. Structure of the Survey.

Section	Number of Questions	Description
Section 1	8	Primary Data and Socio-Demographics: Age, education, income, gender, location, vehicle ownership, driving license status
Section 2	6	CP Awareness: Graphics and definitions, participants’ knowledge regarding carpooling
Section 3	22	Travelers’ Perceptions, Preferences, and CP Attributes: Preferred modes of transportation, willingness to share cars, role preference (driver or passenger), trip purpose, vehicle occupancy, presence of HOVs, perceptions of CP’s impact on travel costs, distance, time, car parking issues, car parking costs, and safety and comfort concerns

below shows the structure of the survey. Through these three sections, the survey sought to understand specifically younger adults’ views on CP as a transportation alternative.

3.2. Data Collection

According to the Capital Development Authority, Pakistan’s capital city, Islamabad, encompasses a total area of 906.5 km². As per the population census [48], there were around 1.015 million people in the city at the end of the year 2017. Moreover, the largest share of the population (approximately 25%) belongs to young adults, i.e., between 18 and 30 years old. Furthermore, the city’s total number of registered automobiles had topped 0.9 million by 2019 [21]. It is also important to mention that Islamabad is a planned city with a more well-developed road network than other cities in Pakistan. Figure 2 shows the study area selected for this research.

The vast majority of residents in Islamabad own private vehicles, such as cars and motorcycles. In contrast, the availability of public transit is substantially lower than expected, given the city’s population. This mismatch results in significant traffic congestion concerns in various areas throughout the city, thereby negatively impacting the city’s natural environment [27]. Islamabad residents’ considerable reliance on private cars is due to their desire to avoid public transit and the lack of public transport in their localities. This study used survey forms to deliver a questionnaire via social media and email to city inhabitants, mainly targeting students, professionals, drivers, workers, etc., in the age category of young adults. Convenience sampling technique [49] was utilized to collect survey data. All respondents understood the research purpose and questions to ensure transparency and to avoid ethical concerns. A total of 656 responses were collected from young adults residing in Islamabad. Sloven’s formula [50], as given in Equation (1), is generally used to determine the sample size for such research studies:

$$n=N/\left(1+N\times E^2\right)$$

(1)

Where n = sample size, N = population size, and E = margin of error indicating the acceptable amount by which the sample estimate might differ from the true population parameter. Considering a confidence Level of 95% with a 4% margin of error [49,51], a minimum sample size of 624 respondents was required to reflect the city’s young adult population accurately.

In order to accurately forecast the possible impact on traffic reduction, fuel consumption, and CO₂ emissions, control data were also required. The overall number of registered vehicles in Islamabad, including government and non-government vehicles, has continuously increased over the past several years. As shown in Figure 3, the number of registered cars in Islamabad exceeded 1.186 million between 1981 and 2020, as indicated by the statistics from the Excise and Taxation Office in Islamabad. A bar chart highlights the rise in the number of registered cars, with a particular emphasis on personal vehicles registered with the Excise and Taxation Office in Islamabad between 1981 and 2019. Figure 3 demonstrates a consistent increase in the number of private automobiles registered over this period. It was determined that the most critical element contributing to the increasingly severe traffic congestion difficulties in Islamabad was the considerable increase in the number of private automobiles operating on the city’s highways. The city’s transport infrastructure has been put under enormous pressure due to the consistent development in the volume of private cars, leading to increased traffic congestion and associated issues. The data presented serves as a visual representation of the increasing number of personal vehicles in Islamabad. It brings attention to the urgent need for efficient traffic management strategies and sustainable transportation solutions to address the growing congestion issues and ensure smoother mobility for residents and commuters.

3.3. Model Development

As discussed in previous sections, carpooling aids in lowering the amount of single-occupancy vehicles on the road [52]. Additionally, it enables travelers to enjoy flexibility and affordable travel. This transportation form reduces emissions and keeps society environmentally friendly [53]. Demographics, mode choices, and travelers’ perceptions are a few variables that may impact carpooling policies for these situations. Many researchers agree that people’s lifestyles and attitudes broadly impact travel perception [54,55,56]. The significant impact of varied travel perceptions regarding people’s travel behavior and travel demand management policies was also demonstrated by some additional studies [57,58,59]. More specifically, travel incentives and limits on personal automobiles are critical factors in the success of any travel demand management strategy [54]. According to a study by Sheldon and Heywood [60], the primary drivers of a successful carpooling service are lower costs, comfortable travel, and the availability of well-maintained facilities. If the carpooling service were supported with high-occupancy vehicle lanes, the tendency to carpool would increase depending on the mode choice people prefer [61]. The disposition of the travelers, carpooling incentives such as cost-saving, time-saving, and safety, the purpose of the trip, trip frequency, age, and car ownership disincentives may also be considered [62,63]. Additionally, it was discovered that people who are inclined towards public transportation and conscious of environmental issues are also more inclined towards carpooling [64].

In the global context of increasing fuel costs and traffic congestion levels, adopting alternate transportation modes, including carpooling, is the need of the hour. Therefore, this research aims to evaluate the relationship between young adult travelers’ perception and the tendency to carpool in the presence of mode choice as a mediating variable. As identified in the literature review, various models [4,27,28,29] have been developed to study the influence of travelers’ perceptions on adopting carpooling as a mode of transportation. However, in this research, the mediating effect of mode choice has also been incorporated to study its impact on the relationship between young adult travelers’ perception and the tendency to carpool. Keeping in view the findings and the research gap identified in the existing research, the proposed research model, along with the formulated hypotheses, is illustrated in Figure 4 below:

The latent constructs in the above-proposed model incorporate various item scales collected via a questionnaire-based survey. The questionnaire was designed or structured to gather data related to three key construct variables: Travel Perception, Tendency to Carpool, and Mode Choice. Each construct variable included multiple-item scales to capture respondents’ behaviors and preferences comprehensively. For instance, the construct variable’ Travel Perception’ was assessed using a 10-item scale, with sample items such as ‘Influence of Carpooling on Parking Demand Reduction.’ Similarly, an example of an item for the ‘Mode Choice’ construct is ‘If you were struggling to reach your destination, what would be your preferred mode of transport?’. These item scales were designed to measure various aspects of each construct, providing a thorough understanding of respondents’ travel perceptions, mode choice decisions, and tendency to carpool. All the questionnaire items were designed to obtain accurate data from the target audience in relation to various aspects of carpooling. Items in the questionnaire were developed in accordance with the proposed hypotheses. The latent constructs and the respective item scales employed in this research are shown in

Table 2. Description of construct variables and item scales used for analysis.

Construct Variables	Item Scales Coding	Item Scales Description
Travel Perception	TP 1	Attraction of Carpooling
	TP 2	Influence of Carpooling on Air Quality
	TP 3	Influence of Carpooling on Parking Demand
	TP 4	Utility of Vehicle Ownership-I
	TP 5	Utility of Vehicle Ownership-II
	TP 6	Priority of Travel Characteristics
	TP 7	Traveling as a Driver with an Unknown Passenger
	TP 8	Traveling as a Passenger with an Unknown Driver
	TP 9	Utility of Carpooling for Daily Fixed Activities
	TP 10	Utility of Carpooling for Random Activities
Tendency to Carpool	TC 1	Awareness of Carpooling
	TC 2	Effect of HOV lane on Carpooling
	TC 3	Vehicle Occupancy for Comfortable Carpooling
	TC 4	Carpooling Purpose
	TC 5	Travel Cost Reduction to Prefer Carpooling
	TC 6	Travel Distance to Prefer Carpooling
	TC 7	Travel Time Reduction to Prefer Carpooling
	TC 8	Environmental Impacts to Prefer Carpooling
Mode Choice	MC 1	Impact of Vehicle Ownership on Mode Choice
	MC 2	Preferred Mode to Reach an Important Destination
	MC 3	Impact of Parking Issues on Mode Choice
	MC 4	Impact of Vehicle Occupancy on Mode Choice

below. As shown in Table 2 below, all these items were eventually linked to the relevant construct variables, i.e., travel perception, mode choice, and tendency to carpool for further modeling and analysis. After checking the model’s validity, confirmatory factor analysis (CFA) was performed to refine the item scales. A check for correlations among the latent constructs was also performed. Furthermore, stepwise multiple regression was performed with and without the presence of a mediator variable (i.e., Mode Choice) to examine the relationship between traveler perception and tendency to carpool. Firstly, stepwise multiple regression was performed to analyze the main effects of demographics, travel perception, and mode choice variables on the tendency to carpool. After that, mediated regression analysis examined the mediating role of mode choice on the relationship between travel perception and the tendency to carpool.

4. Results

Firstly, Cronbach’s Alpha, often called CRA, was utilized to determine whether the questionnaire was consistent and reliable. The values of the CRA can range anywhere from 0 to 1, with values closer to 1 indicating more reliability [27,28]. The computed CRA value for the data collected in this study was recorded as 0.74, indicating that the internal consistency of the questionnaire can be regarded as satisfactory [65]. It lends credence to the idea that the survey’s questions accurately assess the targeted dimensions and that they may be applied to subsequent analyses with a specific confidence level.

4.1. Descriptive Analysis

The demographic distribution of young adult respondents for this research is displayed in

Table 3. Carpooling Survey Demographic Distribution of Young Adult Respondents in Islamabad, Pakistan.

Demographics	Sample Distribution (%)
Gender	Male	72.7%
	Female	27.3%
Education	Undergraduate	75.5%
	Masters and Above	19.2%
	High School/Intermediate	5.3%
Driving License	Yes	27.6%
	No	72.4%
Job	Employee	12.5%
	Entrepreneur	4.4%
	Student	79.9%
	Unemployed	3.2%
Monthly Income	Less than PKR 20,000	74.4%
	PKR 20,000–40,000	12.5%
	PKR 40,000–60,000	6.7%
	PKR 60,000–100,000	4.3%
	More than PKR 100,000	2.1%

below. It is worth mentioning that the sociodemographic trends in this survey study were observed to be in line with the findings of a previous study conducted in Islamabad, Pakistan [27]. Out of 656 responses, approximately 72.7% of the respondents identified as male, while 27.3% identified as female. Analyzing the census data [48], it was revealed that the actual gender-wise distribution of the young adult population in Islamabad is 54% and 46% for males and females, respectively. However, the sample distribution pattern aligned with the local traveling trends, as females tend to travel less due to social and cultural constraints [27]. Moreover, Table 3 shows a breakdown of the respondents in terms of their educational backgrounds. It appears that 75.5% of the young adult participants specified their education at the undergraduate level and that 19.2% stated a higher level of education, such as a Master’s degree or beyond. Regarding job and monthly incomes, it was observed that a significant portion of the respondents belonging to the age bracket of 18 and 30 years were students and earned less than PKR 20,000 per month. These demographics were essential for comprehending the features of the study’s sample population, as they provided vital insights into the distribution of the survey respondents. However, it is worth mentioning that the survey was conducted via online platforms, and the convenience sampling technique was employed to gather responses from the young adult population of Islamabad; hence, the sample stratifications might not align with demographic patterns for specific categories. Nevertheless, it was ensured that the collected responses were more than the minimum sample size required to achieve reliable results for the young adult urban population of Islamabad. Hence, the collected data provided valuable insights into the young adult urban population’s travel perceptions and preferences related explicitly to CP.

According to the questionnaire findings, there is a connection between these sociodemographic variables and the awareness and perception of CP. According to the results, approximately 67.2% of respondents had some level of familiarity with CP, whereas 32.8% were clueless about this method of transportation. In addition, about 78.9% of participants believed that CP has the potential to positively impact air quality by lowering the number of individual cars on the roads. Additionally, more than half of the respondents stated that CP has the potential to reduce the demand for parking spots. Nearly half of the respondents indicated that they would be willing to postpone the purchase of a new vehicle if carpooling was able to meet the travel demands of their community sufficiently.

Similarly, roughly 40% of the participants were open to selling their vehicles if CP could efficiently cover their transportation needs. As seen in Figure 5, nearly one-third of the people who participated in the survey responded that they would prefer to carpool even if doing so would make it more difficult for them to reach their destination. Conversely, approximately 37% of people expressed that they would still favor driving their private automobiles as their primary form of transportation. Around 54.3% owned a vehicle, i.e., a car, a motorbike, or both, while 45.7% of participants did not own any car and relied on public transport or city taxis.

According to

Table 4. Carpooling Characteristics Distribution of Young Adult Respondents in Islamabad, Pakistan.

CP Variables	Distribution (%)
No. of persons in a carpool	With one person	22.70%
	With two persons	34.90%
	With three persons	31.40%
	More than three persons	11.00%
Share rides with strangers	Uncomfortable	40.90%
	Comfortable	59.10%
Receive rides from strangers	Uncomfortable	41.40%
	Comfortable	58.60%

, it is evident that respondents who showed their willingness to carpool had varying preferences in terms of the number of persons involved in the carpooling trip. Approximately 22.70% and 34.90% of people were willing to share their vehicles with one person and two persons, respectively. In comparison, approximately 31.4% of respondents would be happy sharing a transport with three people. Lastly, approximately 11% of respondents said they would be interested in riding in a vehicle with four or more other travelers.

Table 3 provides more insights, revealing that approximately 59.1% of the respondents expressed a level of comfort in offering a ride to a stranger. In comparison, approximately 58.6% of people felt comfortable sharing a ride with an unknown driver. Sharing rides with strangers is one of the perceived drawbacks of carpooling, particularly in developing countries where there is a potential risk to the safety of drivers or passengers. However, the survey findings imply that a sizeable section of the young adult population is receptive to sharing transportation with unfamiliar people despite the inherent risks connected with carpooling in specific locations or settings.

According to the survey findings, a substantial portion of the population, precisely 67.9%, expressed their intent to select carpooling as their preferred mode of travel when High Occupancy Vehicle (HOV) lanes are available. HOV lanes, also known as carpool or diamond lanes, are a kind of traffic management aimed to incentivize and encourage ride-sharing. This method has gained popularity, particularly among those commuting to work, university, or college. These findings prove the widespread acceptance and favorable reception of carpooling as a viable choice for addressing day-to-day transportation demands, mainly when supported by HOV lanes.

Additionally, the survey offers insight into the various elements that motivate and encourage travelers to adopt carpooling as their mode of travel. Around 31.9% of respondents stated that they would be more likely to carpool if there was a 10-min reduction in travel time. A potential reduction in journey time of 15 min motivated 16.7% of the total respondents, while 7% of persons found a decrease in travel time of 20 min more appealing. In addition, 20.5% of the total respondents claimed that even a reduction of 5 min in travel time would motivate them to use carpooling. In addition, interestingly, a sizeable percentage of respondents, i.e., 72.5%, stated that the motivation to select carpooling as a form of transportation comes from the desire to protect the environment and lower their carbon dioxide emissions. This finding shows the level of awareness and consciousness of the young adult population in Islamabad regarding the importance of environmental impact and the need to choose eco-friendly transportation alternatives.

Similarly, another key advantage of CP is the reduction in travel costs. As per the survey findings, more than half of the young adult respondents in Islamabad stated that a travel cost reduction of 50% would encourage them to opt for carpooling regularly. Furthermore, the survey also attempted to collect information regarding the trip distances for which young adult residents of Islamabad would prefer to carpool. In this context, the survey results indicated that CP is flexible and attractive for both short and long trips. 39.1% of respondents chose CP for 20–50 km journeys, while 24.2% preferred CP for more than 50 km trips. Overall, 83% of respondents showed their intent to carpool for longer journeys in an urban context, i.e., greater than 10 km, whereas 17% expressed their preference to carpool for shorter trips.

4.2. Control Variables

One-way analysis of variance (ANOVA) was carried out to investigate the variation in the willingness to carpool, taking into account various demographic factors. The findings revealed a significant gap in the likelihood of different groups, differentiated by gender and educational attainment, participating in carpooling. As evident from

Table 5. One-way ANOVA.

Demographics	Tendency to Carpool
	f Statistics	p-Value
Gender	2.235	0.000
Education	1.754	0.004
Job	0.853	0.724
Monthly income	1.075	0.352

, the most important demographic variables were gender and education. This implies that willingness to adopt carpooling varies significantly among gender classes and across people’s education levels. For gender groups, this pattern aligned with the local trends as females travel less overall, owing to social and cultural constraints. Interestingly, the educational classification of the young adult participants, among which the majority lie in the undergraduate category, significantly impacted the tendency to carpool. This implies that the level of education, even in the young adult population, is a crucial parameter. Conversely, the study did not identify any significant differences in the tendency to carpool among different classes of other demographic factors, i.e., employment and monthly income. As per Table 3 above, it was observed that a significant portion of the young adult respondents were current students and earned less than PKR 20,000 per month. However, the willingness to carpool among the classes of these demographic factors was the same.

4.3. Confirmatory Factor Analysis (CFA)

A strong statistical method, i.e., confirmatory factor analysis (CFA), was used to analyze the validity measures of the item scales. CFA helps to scrutinize the relationship between different variables by refining the number of scale items for each variable construct used in the research study. As mentioned in the previous section, Smart-PLS software version 4 was used for CFA. Figure 6 and Figure 7 each show the results of the CFA performed on level 1 and level 2, respectively. The relevant criteria for refining item scales are also described below. Moreover,

Table 6. Factor Item Loadings and Convergent Validity.

Construct	Items	Loadings	CR	AVE
Travel perception	TP 1	0.635
	TP 2	0.632
	TP 3	0.603
	TP 4	0.690
	TP 5	0.651
	TP 6	0.154
	TP 7	−0.413
	TP 8	−0.453
	TP 9	0.489
	TP 10	0.543	0.830	0.524
Tendency to carpool	TC1	0.625
	TC 2	0.700
	TC 3	−0.108
	TC 4	0.470
	TC 5	0.155
	TC 6	0.178
	TC 7	−0.149
	TC 8	0.589	0.736	0.583
Mode Choice	MC 1	0.342
	MC 2	0.696
	MC 3	0.606
	MC 4	0.373	0.686	0.523

indicates factor loadings of item scales and the convergent validity of each variable construct.

Figure 6 and Table 6 above show the calculated factor loadings for each item scale. According to the recommendations of several experts in the field of social science [66,67,68], the acceptance threshold for the factor loadings must be greater than 0.6 for each item scale. Level 2 of the confirmatory factor analysis is depicted in Figure 7. CFA level 2 contains refined item scales after eliminating the item scales that do not meet the specified factor loading criteria. It is evident from Figure 7 that the item scales in level 2 were reduced to a total of nine for the three construct variables, i.e., travel perception, mode choice, and tendency to carpool. Further analyses were conducted on the refined item scales.

4.4. Correlations

Correlation analysis determines the existence and magnitude of the relationship between the variables. The results of the correlation analysis are presented in

Table 7. Correlation Analysis.

Construct Variables	1	2	3
1. Travel perception	1
2. Tendency to carpool	0.430 **	1
3. Mode choice	0.170 **	0.264 **	1

n = 656, ** p < 0.01.

.

Table 7 indicates the degree of association between each variable of the study. The results suggested a significant and positive correlation existed between travel perception, mode choice, and tendency to carpool. All these variables had significantly low to moderate degrees of correlation among them. This analysis provides information on the strength and direction of the relationships between these three variables, which can help understand their patterns and potential interactions.

4.5. Stepwise Multiple Regression Analysis

Stepwise multiple regression analysis was performed to examine the main effects of different predictor variables on the outcome variable in the model. It involved stepwise inclusion and determination of statistically significant variables to develop a regression model. After stepwise iterations, only the statistically significant variables were kept in the final model. The findings of multiple regression analysis are presented in

Table 8. Stepwise Multiple Regression Analysis.

Predictors	Tendency to Carpool
	Β	R2	ΔR2
Step 1
Gender Education	0.200 *** 0.014	0.051
Step 2			0.190
Travel Perception	0.268 ***	0.241
Mode Choice	0.098 ***

n = 656, *** p < 0.001.

, indicating that travel perception significantly affected the tendency to carpool (β = 0.268, p < 0.001), which supported hypothesis H1. Similarly, the mode choice also substantially impacted the tendency to carpool (β = 0.098, p < 0.001), thereby supporting hypothesis H2. Moreover, among the demographic variables, only the gender variable was significant, whereas the other variable, i.e., educational level, was not significant; hence, it was excluded from the final model.

4.6. Mediated Regression Analysis

Mediated regression analysis is generally used to evaluate the role and magnitude of the mediator in the relationship between a dependent variable and an independent variable. In this research, mediated regression analysis was performed to check the mediating effect of mode choice on the relationship between travel perception and the tendency to carpool. This analysis provided valuable results for understanding the type of relationships between the model variables. The results presented in

Table 9. Mediation Regression Analysis.

	β	SE	T	p
TP → TC	0.2916	0.0239	12.1951	0.0000
TP → MC	0.1802	0.0408	4.4157	0.0000
MC → TC	0.1260	0.0224	5.6256	0.0000
TP → MC → TC	0.2689	0.0237	11.3385	0.0000
Bootstrap results in an indirect effect		Indirect Effect	LL 95% CI	UL 95% CI
		0.0227	0.0104	0.0389

n = 656, LL = Lower limit, UL = Upper limit, TP = Travel Perception, MC = Mode Choice, TC = Tendency to Carpool.

indicate that the direct effect of travel perception and the tendency to carpool was significant (β = 0.2916, p < 0.05). It also shows that travel perception was positively related to mode choice (β = 0.1802, p < 0.05), hence supporting hypothesis H3. Similarly, mode choice was positively associated with the tendency to carpool (β = 0.1260, p < 0.05). Moreover, the direct effect between travel perception and the tendency to carpool was more significant than the indirect effect between them through a mediator, i.e., mode choice (β = 0.2689, p < 0.05), so partial mediation exists, thereby supporting hypothesis H4.

4.7. Reduction Estimation of Traffic-Related Parameters

In this section, the impacts of CP on the reduction in vehicles, fuel consumption, and CO₂ emissions are quantified. Firstly, the reduction in vehicle numbers (CR) was determined using Equation (2). For this purpose, it is critical to obtain accurate data regarding the typical distance traveled by an individual daily, weekly, monthly, and annually to come to this conclusion. This information is essential for calculating the number of automobiles needed, which directly impacts the congestion on the roads, the demand for fuel, and levels of harmful CO₂ emissions. Consequently, Equation (3) calculates the typical amount of driving done in a month. Similarly, Equation (4) determines the typical distance traveled in a year. The particulars of the equations are presented below.

$$CR=PC\times {\displaystyle \frac{TCC }{100}}$$

(2)

Where TCC is the total number of cars in the city and PC is the percentage of residents, specifically young adults, who desire to share their vehicles with others (based on the survey findings).

$$AVTM=AWD\times AVTD$$

(3)

Where AVTM is the average vehicle travel per month, AWD stands for the average number of working days in one month, and AVTD stands for the average number of miles a car travels over one day.

$$AVTY=AVTM\times 12$$

(4)

Where AVTY is the average distance traveled by a vehicle in a year.

Table 10. Reducing car numbers and CO₂ and fuel-saving after carpooling.

Year	Total Cars	Car Reduction	After Carpooling	Fuel Saving per Year (Liter)	Total CO2 Emissions (mt)	Reduction in CO2 (Metric Tons)
2009	208,208	70,136	138,138	26,926,912	1206.111795	800.8582317
2010	236,521	79,471	157,050	30,516,886	1366.645114	907.4523557
2011	263,577	88,562	175,015	34,007,759	1522.977745	1011.257223
2012	305,641	102,695	202,946	39,435,024	1766.028299	1172.64279
2013	348,885	117,225	231,660	45,014,538	2015.897026	1338.555626
2014	384,564	129,214	255,350	49,617,986	2222.054328	1475.444073
2015	424,409	142,601	281,808	54,758,947	2452.283248	1628.316077
2016	467,049	156,928	310,121	60,260,530	2698.661995	1791.911565
2017	525,165	176,455	348,710	67,758,889	3034.462823	2014.883314
2018	592,019	198,918	393,101	76,384,659	3420.752803	2271.379861
2019	641,824	215,653	426,171	82,810,700	3708.531731	2462.46507
2020	673,796	226,395	447,401	86,935,855	3893.26957	2585.130992
2021	707,768	237,810	469,558	91,319,058	40898.56363	2715.470249
2022	745,740	250,569	495,171	96,218,358	4308.97014	2861.156175

displays the predicted values for each year from 2009 to 2022, presenting the annual decrease in the number of cars as computed from Equations (2) to (4). If an actively promoted and incentivized CP system had been introduced in Islamabad, it would have resulted in annual savings of 546,810 barrels of fuel, equivalent to approximately 3.42% of Pakistan’s total petrol consumption. The CP research findings indicated a sizeable drop in the total number of vehicles on the road. It was anticipated that an actively promoted CP program would have reduced the number of cars on the streets of Islamabad by around 0.25 million by 2022. This estimation was based on traveler perceptions and the tendency to adopt CP by the young adult population of Islamabad. These numbers imply that CP can significantly influence roadway congestion and foster the development of a more effective transport system.

As seen in Figure 8, the total number of vehicles registered in Islamabad was roughly 0.745 million for the year 2021–2022. However, if CP had been adopted, it was predicted that the number of vehicles would have reduced to approximately 0.49 million, suggesting a significant difference from the current situation. Hence, it was determined that an active CP program has the potential to reduce the number of registered cars on the road by around 34%. A reduction of this magnitude in the number of vehicles on the roadways suggested the potential significance that the introduction of CP in the nation’s capital may play in reducing future traffic congestion. The city would reap significant benefits, including reduced traffic congestion, improved traffic flow, and a lessened environmental impact. The anticipated reduction in the number of vehicles draws attention to the promising potential of carpooling as a viable approach to address traffic-related difficulties in Islamabad and to establish a transportation system that is more sustainable and efficient.

Furthermore, the impact of CP on the fuel demand was calculated using Equation (5):

$$TPRCRY=TCR\times PRY$$

(5)

Where TPRCRY is the total petrol required for the reduced number of cars per year, TCR is the total number of vehicles reduced on the road by CP, and PRY is the petrol necessary per year.

Figure 9 compares the annual petroleum usage with and without the impacts of the CP system. According to the findings, 1.63 million fuel barrels were required in the year 2021–2022 to supply all the registered cars in Islamabad without any CP program. Conversely, if a carpooling system had been implemented, it is predicted that this consumption would have dropped significantly to 1.08 million barrels, resulting in a significant difference of around 0.55 million barrels. This considerable reduction substantiates the hypothesis that CP can reduce the amount of petrol consumed by automobiles by approximately 33.6%. CP has the potential to successfully reduce the consumption of petroleum and contribute to the country’s economy. It also highlights the economic and environmental benefits of implementing a CP system, as it can lead to significant savings on gasoline and promote a more sustainable transportation alternative for the rapidly expanding cities of Pakistan, specifically Islamabad.

Like fuel consumption, the amount of CO₂ emissions can also be reduced in Islamabad by encouraging and incentivizing the CP system. According to research, consumption of one liter of gasoline produces roughly 2392.5 g of CO₂ [61]. As a consequence of this, a decrease in the amount of fuel consumption directly corresponds to a reduction in the amount of CO₂ emissions. Therefore, calculation data and outcomes from the survey were utilized to forecast the reduction in CO₂ emissions. The results revealed an apparent decrease if an active CP system had been introduced. These findings were directly linked to a decrease in CO₂ emissions. The calculations indicated that CP can cut carbon dioxide emissions by around 33.6%. It was calculated that all cars registered in Islamabad contributed 3893.3 tons of CO₂ to the atmosphere in 2021–2022. Notably, the city has no organized CP program, which significantly contributes to the negative influence on the environment. However, if the relevant authorities had implemented such a system, the CO₂ emissions would have been reduced to 2585.13 metric tons, a significant difference of 1308.2 metric tons. Figure 10 shows the reduction in CO₂ before and after the implementation of a CP system in Islamabad.

All the results presented in this section reflect the significance of developing and promoting CP systems. In addition to the role of individual travelers’ perceptions, mode choice decisions, and tendency to adopt CP, the quantified impact of CP on traffic sustainability factors like reduction in the number of vehicles, fuel consumption, and emissions imply the urgency of actively establishing CP policies and programs. As mentioned above, CP would have eliminated at least 0.25 million cars, i.e., 33.6% of Islamabad’s private vehicles in 2021–22. This reduction would have saved 546,810 barrels of fuel and reduced CO₂ emissions by 411,035 tons annually. Based on the findings, the effectiveness of CP as a sustainable transportation mode, specifically in developing countries, can be ascertained.

5. Discussion

Over the last few years, various models [4,27,28,29] in the context of Pakistan have been developed to study the influence of travelers’ perceptions on adopting carpooling as a mode of transportation. However, the novelty and contributions of this study lie in its focused examination of CP acceptance among young adults (aged 18 to 30 years) in Islamabad, Pakistan. This aspect has received limited attention in existing literature on shared mobility modes, specifically in developing countries. Therefore, this research provided comprehensive insights into the factors influencing young travelers’ perception, mode choice decisions, and tendency to carpool. Moreover, considering the dynamics of carpooling, ranging from behavioral aspects to technical limitations, the mediating effect of mode choice factors was incorporated in the proposed model to investigate the relationship between traveler perception and tendency to carpool. In addition, the study quantified and predicted the tangible impacts of CP on traffic sustainability.

This study was conducted in Islamabad, a fast-growing city in Pakistan with a high percentage of young residents, to assess the travelers’ perceptions, acceptance of CP, and the quantified effects of CP on traffic congestion, fuel consumption, and emissions. An online questionnaire-based survey investigated young adult travelers’ perceptions of CP. A total of 656 responses from the population segment of younger adults, which comprise almost 25% of Islamabad’s urban population, were collected. The survey findings confirmed that nearly 46% of young adult respondents did not own any vehicle and used other available modes for their daily trips. Moreover, most respondents were students or early career professionals with lower financial stability. These sociodemographic trends were observed to be in line with the findings of a previous study conducted in Islamabad, Pakistan [27]. Moreover, these statistics affirmed that the particular category of young adult residents could be targeted to induce an overall change in travel behavior within society.

This study employed various statistical modeling methods to analyze the relationship between young adult travelers’ perceptions and their tendency to adopt carpooling, with mode choice acting as a mediating variable. The proposed model and the developed hypotheses were evaluated based on CFA, stepwise, and mediated multiple regression. Overall, these statistical methods enabled a comprehensive analysis of the factors influencing the adoption of carpooling by the young adult population in Islamabad, which is essential for formulating targeted interventions and policies aimed at promoting sustainable transportation options. The use of CFA allowed for the validation of item scales used to measure different variables, ensuring the reliability and validity of the study’s constructs. CFA helped scrutinize the relationships between variables by refining the number of scale items for each construct. Moreover, stepwise multiple regression analyzed predictor variables’ primary effects, resulting in a final model with only significant variables. Finally, mediated regression analysis was conducted to evaluate the mediating effect of mode choice on the relationship between travel perception and the tendency to carpool.

The refined item scales in CFA reflected that CP awareness, impacts on the environment and parking demand, and availability of HOV lanes were essential factors that influenced the travel perception, mode choice, and tendency to carpool of young adult respondents. All these factors, especially the availability of HOV lanes and environmental factors related to CP, have been highlighted by previous researchers as well [27,28]. Further analyses were performed on the refined item scales, including stepwise and mediated multiple regression, to evaluate the proposed hypotheses. The results of stepwise regression analysis validated hypothesis H1 and hypothesis H2, indicating that both travel perception (β = 0.268, p < 0.001) and mode choice (β = 0.098, p < 0.001) significantly affected the tendency to carpool. The outcomes also affirmed the impact of demographic variables, among which only the gender variable was found to be significant on the willingness of the young adult population to shift towards carpooling. These results were similar to the findings of previous studies [4,27,28]. Moreover, the results of mediated regression analysis also validated hypothesis H3, which stated that travel perception was positively related to mode choice (β = 0.1802, p < 0.05). Additionally, the results also implied that the travelers’ mode choice mediated the relationship between travelers’ perception and tendency to carpool, thereby supporting hypothesis H4. Furthermore, the outcomes of stated CP perceptions and mode choice decisions of the young adult respondents were utilized to predict that CP has the potential to reduce the number of cars on the road by 0.25 million, constituting about 33.6% of the total private vehicles. Accordingly, a potential fuel saving of 546,810 barrels per year and a CO₂ emissions reduction of 411,035 tons can be achieved. All these results helped to understand the factors influencing the adoption of CP as a transportation alternative, which is essential for formulating targeted interventions and policies to promote sustainable transportation options.

Overall, this research provided a valuable in-depth analysis of various factors that influence the travel perception and tendency to carpool for a young adult population, specifically within the context of rapidly expanding cities of developing countries. Moreover, the quantification of the traffic sustainability factors related to the adoption of CP, including vehicle reduction, fuel consumption, and emissions, emphasized the importance of implementing and promoting alternative transportation modes.

Finally, it is essential to mention the study limitations and directions for future research. Although this study provided valuable results that can serve as a reference for future studies, large-scale studies representing the exact demographic patterns and stratifications may be conducted to enhance the proposed models’ outputs further. It is worth mentioning that the survey was shared via online platforms, and the convenience sampling technique was employed to gather responses from the young adult population of Islamabad; hence, the sample stratifications might not align with demographic patterns for different categories. Furthermore, a smaller proportion of female travelers in the survey sample might have influenced certain factors. Future studies could be conducted to investigate the perspectives of female travelers and their level of acceptance towards carpooling. Moreover, as the study focused on evaluating the travel perceptions of young adult travelers in connection with their tendency to adopt carpooling as a transportation mode, demographic characteristics such as higher education levels, unemployed/student status, and low monthly income levels may have influenced the underestimated effect of certain factors like tendency to share rides with strangers (i.e., TP7 and TP8 factors). As per the findings, the respondents showed a tendency to share rides with strangers despite the potential security risks. Approximately 59.1% of the respondents expressed a level of comfort in offering a ride to a stranger, while approximately 58.6% of people showed comfort in sharing a ride with an unknown driver. This conclusion implied that a sizeable section of the young adult population was receptive to sharing transportation with unfamiliar people despite the inherent risks connected with carpooling in specific locations or settings.

Moreover, the employed statistical methods might not capture all relevant factors influencing the tendency to adopt carpooling, such as unobserved variables or omitted variable bias. To address these limitations and enhance the robustness of the findings, future research could explore alternative advanced statistical analyses such as Bayesian methods, agent-based modeling, neural networks, or integration of spatial analysis techniques. These advanced techniques could enhance validity and generalizability. In addition to the CP attributes covered in this research, studies could also be conducted to investigate the challenges of CP, such as the need to adapt the schedule to accommodate another person, the vehicle’s functionality, and the likelihood of social incompatibilities. In this research, the mediating role of mode choice has been evaluated because of its practical significance in the carpooling scenario. Nevertheless, depending on the extent of collected data, other variables might also be included as mediators or moderators in the model to evaluate further the relationship between travel perception and the tendency to carpool. Apart from the young adult population, other categories of residents, e.g., older residents, corporate workers, government officers, etc., might also be examined with respect to the prospects of carpooling. Lastly, the adaptability and potential influence of technology on travel perception and the tendency to carpool can also be evaluated in future studies.

6. Conclusions

As a sustainable transportation mode, CP has excellent potential to reduce the adverse impacts of the urban transportation sector. In developing countries with insufficient transportation facilities, CP can play an even more critical role in efficiently managing traffic networks. Formally promoted and incentivized CP strategies can induce a shift in travel perception and mode choice decisions. In this regard, it is essential to examine different categories of residents and devise specific strategies for them. Therefore, it was identified in this research that the younger adult population segment (i.e., aged between 18 and 30 years), which is generally more flexible and aware of economic and environmental concerns, may prove to be the driving force in the attempt to change cultural attitude and travel perceptions [43]. This research evaluated the young adult population’s perception of travel and the tendency to adopt CP. In addition, the study aimed to quantify and predict the effects of CP on traffic sustainability, including factors like reductions in traffic congestion, fuel consumption, and emissions. Therefore, this is one of the first studies, particularly in the scenario of a developing country, to provide a comprehensive layout of all the factors influencing the adaptability of carpooling, specifically for the young adult population.

This study was conducted in Islamabad, a fast-growing city in Pakistan with a high percentage of young residents, to evaluate travelers’ perceptions, mode choice decisions, and potential acceptance of CP. An online questionnaire-based survey was developed and distributed to collect a total of 656 responses from the young adult population. In this study, various statistical modeling methods were employed to analyze the relationship between young adult travelers’ perceptions and their tendency to adopt carpooling, with mode choice acting as a mediating variable. Overall, these statistical methods enabled a comprehensive analysis of the factors influencing the adoption of carpooling by the young adult population in Islamabad. Based on the confirmatory factor analysis, stepwise, and mediated multiple regression analysis, all the hypotheses of this research were supported. The findings implied that young adult travelers’ perception and mode choice attributes had a positive and significant influence on the tendency to carpool. Additionally, it was also observed that young adult travelers’ mode choice mediated the relationship between travelers’ perceptions and the tendency of young adults to carpool. Moreover, the outcomes also affirmed the impact of a demographic variable, i.e., gender, on the willingness of the young adult population to shift towards carpooling. All these results reflected the motivations and considerations for developing and promoting CP systems, the role of individual behaviors, and their current mode choices in transitioning to a new mode of transportation. These results were further utilized to determine that the implementation of CP would have eliminated at least 0.25 million cars, i.e., 33.6% of Islamabad’s private vehicles, in 2021–2022. This reduction would have saved 546,810 barrels of fuel and reduced CO₂ emissions by 411,035 tons annually.

The above conclusions offer practical implications and helpful insights for transport planners, commercial carpool service operators, environmentalists, and policymakers to design and implement efficient, targeted CP systems specifically in the context of rapidly growing cities in developing countries. The identification and understanding of significant travel perception and mode choice factors related to CP can help transport planners develop effective and targeted CP programs that cater specifically to young adults, incorporating their preferences and addressing their concerns to enhance CP adoption. Moreover, authorities, policymakers, and environmentalists can promote CP as a sustainable transportation mode by recognizing the potential benefits of CP in lowering traffic congestion, fuel consumption, and CO₂ emissions. Implementing formal, incentivized, and targeted CP programs may promote sustainable urban mobility, specifically among the demographic group of young adult travelers. Policies could include financial incentives, dedicated CP parking lots, HOV lanes, and integration with public transportation systems to make carpooling a more attractive option. Moreover, the findings of this research can also assist commercial carpool service operators in developing marketing strategies and service models that appeal to young adults, emphasizing the economic and environmental benefits of carpooling. Hence, it can be concluded that the study’s findings provide a valuable guideline for creating CP systems that incorporate the needs and behaviors of young adult travelers, fostering a cultural shift towards more sustainable transportation modes.