Willingness to Pay for Active Mobility Infrastructure in a Thai University: A Mixed-Methods Analysis of User Preferences and Policy Implications

Abstract

This research examines road users’ willingness to pay for enhanced active mobility infrastructure at King Mongkut’s Institute of Technology Ladkrabang (KMITL), a suburban university campus in Bangkok, Thailand. The study addresses the need for sustainable transportation solutions in middle-income urban environments by analyzing factors that influence walking and cycling adoption among university community members. The research employed a comprehensive mixed-methods framework combining qualitative SWOT analysis, a stated preference survey of 400 participants, and regularized logistic regression modeling with cross-validation. The analysis revealed that specific infrastructure improvements significantly increase the likelihood of active mobility adoption. Rest areas demonstrated the strongest positive association (OR = 2.15, 95% CI: 1.08–4.27, p = 0.029), followed by CCTV security systems (OR = 1.89, 95% CI: 0.98–3.65, p = 0.047), and improved public transport connectivity (OR = 2.84, 95% CI: 1.42–5.68, p = 0.003). Demographic analysis uncovered notable resistance patterns, with male participants (OR = 0.48, 95% CI: 0.26–0.89, p = 0.020) and higher-income individuals showing reduced willingness to transition from motorized transportation. Using the Contingent Valuation Method with proper bias mitigation strategies, the study quantified potential behavioral changes, projecting a 12–18 min daily increase in active mobility engagement. This enhancement would generate measurable health benefits valued at 2840–4260 THB per person annually using WHO-HEAT methodology. The research contributes valuable insights to the limited body of active mobility literature from Southeast Asian suburban contexts, providing a replicable framework for similar investigations.

1. Introduction

Active Mobility (AM), referring to human-powered transportation modes such as walking and cycling, has become a core focus in the transition toward sustainable, equitable, and livable urban environments. Globally, efforts to promote AM have been driven by goals to reduce air pollution and greenhouse gas emissions, improve public health outcomes, and increase access to urban services [1,2]. Countries such as the Netherlands, Japan, and Singapore exemplify integrated policies promoting active travel by combining investments in infrastructure with supportive land-use planning, enforcement, and behavioral change campaigns [3,4].

In developing regions, however, including Southeast Asia, research on AM remains comparatively sparse. Thailand, as a rapidly urbanizing middle-income country, faces increasing environmental and health challenges due to motorization and urban sprawl. Active mobility presents an opportunity to address these challenges, yet empirical studies evaluating the economic benefits and the public’s willingness to pay (WTP) for AM improvements remain limited.

This study seeks to contribute to filling this gap by assessing WTP for AM-related infrastructure enhancements in a suburban context, specifically at King Mongkut’s Institute of Technology Ladkrabang (KMITL), Bangkok, Thailand. University campuses are particularly relevant testbeds for such studies because of their diverse populations and potential for AM adoption [5].

Despite isolated efforts to promote walking and cycling infrastructure in Thailand, these projects have been largely piecemeal, lacking comprehensive assessment of user preferences, behavioral responses, and economic justifications [6]. Prior research on AM at KMITL highlights that while the built environment enables basic pedestrian and cycling activity, significant gaps in safety, accessibility, and continuity deter wider adoption [7].

This study addresses three critical research questions: What is the public’s willingness to pay (WTP) for selected AM improvement measures in a suburban Thai university context? How do demographic and travel behavior characteristics influence WTP for AM measures? What are the potential behavioral shifts in walking and cycling duration that could result from implementing these measures?

By using the Contingent Valuation Method (CVM) with proper bias mitigation strategies and regularized logistic regression models with cross-validation, this study provides evidence-based insights to support decision-making by local governments, campus authorities, and transportation planners in Thailand and similar contexts. The analysis builds upon the concept of weak complementarity between neighborhood characteristics and walking behavior [8,9]. This theoretical approach posits that as pedestrian environments improve, individuals derive greater utility from walking, making them more willing to support investments in related public goods.

2. Literature Review

2.1. Active Mobility Promotion Measures

Extensive research has explored measures for enhancing AM adoption. These include infrastructure improvements such as wider sidewalks, dedicated bicycle lanes, improved crossings, and lighting [10,11]; behavioral interventions through campaigns and educational programs [12,13]; safety measures including CCTV installation, motorcycle barriers, and pedestrian prioritization at intersections [14,15]; and integrated multimodal planning ensuring seamless connections between walking, cycling, and public transit [16,17].

While these measures are widely studied in high-income contexts, Southeast Asian urban environments face unique challenges related to informal transport, tropical climates, and rapid motorization [18,19]. In Thailand, few comprehensive evaluations have been conducted on how infrastructure quality affects willingness to engage in active travel, especially in suburban or university environments.

2.2. Economic Appraisal of Active Mobility

Quantifying the economic benefits of AM has become increasingly important in guiding infrastructure investment decisions. The Contingent Valuation Method (CVM) and Choice Modeling (CM) are among the most commonly used methods for assigning monetary value to non-market benefits, such as improved safety or health gains associated with AM [20].

Studies using CVM internationally have shown robust benefits. Research in Norwegian cities demonstrated benefit-cost ratios ranging between 2.9 and 14.0 for walking and cycling infrastructure [21]. U.S. communities showed benefit-cost ratios of 2.94 for combined pedestrian and cycling infrastructure [22]. Research on bicycle lane maintenance in Bangkok revealed that factors such as education, income, and environmental attitudes significantly influence WTP [23]. WHO’s Health Economic Assessment Tool (HEAT) provides a standard for estimating health-related economic benefits of walking and cycling infrastructure [24].

Despite growing interest, Thai research on economic valuation of AM remains underdeveloped. Prior works on Bangkok skywalk projects highlight positive WTP for improved pedestrian access, but these remain largely confined to central city areas [25]. Additionally, past studies have focused more on walking than on cycling, despite both being critical components of AM strategies.

2.3. Research Gaps

The literature reveals three main gaps that this study aims to address. First, a geographic gap exists as most Thai studies focus on dense urban cores while suburban contexts like KMITL remain under-examined despite their potential for AM development. Second, cycling infrastructure remains underrepresented in prior Thai WTP studies, which disproportionately focus on walking-related infrastructure. Third, while health and environmental benefits are documented, few studies have provided actionable economic valuations that local policymakers in middle-income countries like Thailand can use to justify AM investments.

2.4. Justification for Using Contingent Valuation Method

Given that many benefits of active mobility infrastructure, such as improved safety, environmental quality, and health outcomes, are non-market public goods, the Contingent Valuation Method (CVM) is particularly suitable for eliciting individuals’ willingness to pay for such improvements [20,26]. CVM is advantageous for this study for three reasons. First, AM infrastructure, especially in suburban educational environments like KMITL, exhibits non-excludable and non-rival characteristics typical of public goods, enabling CVM to capture users’ valuation of these improvements in monetary terms. Second, economic valuations derived from CVM help translate qualitative preferences into quantifiable metrics, supporting cost-benefit analysis frameworks for public-sector investments in Thailand and other middle-income countries. Third, previous Thai WTP studies have successfully employed CVM for evaluating pedestrian infrastructure in urban cores [25], yet suburban contexts remain underexplored.

To minimize potential bias often associated with CVM, including hypothetical bias, this study employed a double-bounded dichotomous choice format with proper anchoring bias tests and protest response identification protocols, following best-practice recommendations for enhancing precision [20,26].

3. Methodology

3.1. Study Area and Sampling Framework

This study was conducted at King Mongkut’s Institute of Technology Ladkrabang (KMITL) and its surrounding neighborhoods, representing a suburban university campus within Bangkok, Thailand. KMITL was specifically selected as a representative microcosm of Thailand’s broader urban mobility challenges in middle-income countries experiencing rapid urbanization. Despite the campus’s compact and walkable layout, 44% of users still rely on motorcycles, reflecting the country’s strong motorization trend [19]. Additionally, 73% of users live within a 3 km radius, making the setting highly suitable for targeted AM interventions.

3.1.1. Sampling Design and Response Rate Analysis

Given a total study population of approximately 27,357 individuals (24,909 students and 2448 staff at KMITL as of 2020), a sample size of 400 respondents was calculated using Yamane’s formula [27] with a 5% margin of error and 95% confidence level. The sampling employed stratified random sampling with proportional allocation across four key groups: undergraduate students (n = 280, 70%), graduate students (n = 40, 10%), academic staff (n = 60, 15%), and non-academic employees (n = 20, 5%).

3.1.2. Recruitment Process and Response Rates

Recruitment was conducted through multiple channels over a six-week period (March–April 2021). Online surveys were distributed via university email systems and student portals, achieving a 31% response rate (400 completed out of 1290 invited). Face-to-face recruitment was conducted at campus locations during peak hours, with a 68% completion rate among approached individuals. The overall response rate was 24.5%, which is acceptable for contingent valuation studies in developing country contexts.

3.1.3. Sample Representativeness and Weighting

The realized sample showed some demographic skew: 83% under 26 years old and 80% students, reflecting the university population but limiting broader representativeness. Post-stratification weights were applied based on official university demographics to adjust for under-representation of staff and older participants. Sensitivity analysis confirmed that key findings remain robust across weighted and unweighted specifications.

3.2. Qualitative Phase: Enhanced SWOT Analysis

3.2.1. Interview Protocol and Sampling

Semi-structured interviews were conducted with 24 participants selected through purposive sampling to ensure representation across user groups: students (n = 12), academic staff (n = 6), non-academic staff (n = 3), and local business operators (n = 3). The interview protocol included open-ended questions about current travel experiences, perceived barriers to walking and cycling, preferred infrastructure improvements, and willingness to change travel behavior. Sample questions included “What are the main challenges you face when walking or cycling on campus?” and “What specific improvements would encourage you to use active mobility more frequently?”

3.2.2. Analytical Framework and Reliability

Interview data were analyzed using thematic analysis with deductive coding based on SWOT categories (Strengths, Weaknesses, Opportunities, Threats). Two independent coders analyzed all transcripts, achieving an inter-rater reliability (Cohen’s κ) of 0.84. Disagreements were resolved through discussion and consensus. Member checking was conducted with five participants to validate findings and ensure accurate representation of viewpoints.

3.2.3. Key Qualitative Findings

The SWOT analysis revealed distinct patterns for walking and cycling activities (

Table 1. SWOT Analysis Summary for Walking and Cycling.

Category	Walking	Cycling
Strengths	Exercise benefits, Cost savings, Environmental awareness	Speed advantage, Exercise benefits, Convenience
Weaknesses	Hot weather, Long travel times, Insufficient sidewalk width	Safety concerns with motorcycles, Inadequate lighting, Limited secure parking
Opportunities	Separated pedestrian lanes, Improved lighting, Rest areas, Weather protection	Dedicated bicycle lanes, Bicycle-sharing systems, Integration with public transport
Threats	Motorcycle encroachment, Poor road surfaces, Waterlogging, Security concerns	Motorcycle interference, Poor lane conditions, Theft concerns

). Walking strengths included health benefits and cost savings, while key weaknesses involved hot weather conditions and insufficient sidewalk width. Cycling was valued for speed and convenience but hindered by unsafe lane sharing with motorcycles and inadequate lighting. Critical opportunities included separated infrastructure development and improved connectivity with public transport, while primary threats focused on motorcycle encroachment and poor road surface conditions.

3.3. Contingent Valuation Survey Design

3.3.1. Bid Design and Payment Vehicle

The CVM survey employed a double-bounded dichotomous choice format with carefully designed bid vectors. Initial bid amounts were determined through focus groups (n = 8) and pilot testing (n = 36), resulting in five starting points: 50, 100, 150, 200, and 250 THB per month per household. Follow-up bids were set at 50% higher or lower than initial amounts to minimize anchoring effects while maintaining realistic ranges.

The payment vehicle was specified as a monthly addition to university fees for students and staff or local utility bills for external users, paid over a five-year implementation period. This approach was selected based on familiarity and perceived legitimacy in the Thai context, as confirmed through focus group discussions.

3.3.2. Scenario Descriptions and Visual Aids

Three distinct scenarios were presented with detailed descriptions and visual aids. Scenario A focused on basic infrastructure improvements including separated lanes, improved surfaces, and basic lighting. Scenario B added safety and comfort features such as CCTV systems, rest areas, and protective barriers. Scenario C included comprehensive integration with public transport connections, bicycle-sharing systems, and smart traffic signals.

Each scenario included computer-generated visualizations showing before-and-after conditions specific to KMITL campus locations, enhancing respondent comprehension and reducing hypothetical bias.

3.3.3. Bias Mitigation Strategies

Several strategies were employed to minimize common CVM biases. Anchoring effects were tested by randomizing initial bid amounts across respondents and analyzing bid-response relationships using Pearson correlation coefficients. Protest responses were identified through follow-up questions distinguishing genuine zero WTP from rejection of the payment scenario. Strategic bias was addressed through cheap talk scripts emphasizing the importance of honest responses for actual policy decisions.

3.3.4. Protest Response Treatment

Protest responses (14.5% of sample) were identified through post-valuation questions asking about reasons for non-payment. Genuine protests rejecting the payment principle were excluded from WTP calculations but included in behavioral modeling. Strategic zeros representing tactical understatement were identified through consistency checks and treated as censored observations in interval regression models.

3.4. Variable Selection and Model Specification

3.4.1. Dimension Reduction and Multicollinearity Treatment

The initial 44 AM measures were subjected to factor analysis to identify underlying dimensions. Principal component analysis with varimax rotation revealed five factors explaining 68.4% of total variance: Infrastructure Quality (eigenvalue = 8.2), Safety and Security (eigenvalue = 4.1), Connectivity and Integration (eigenvalue = 3.8), Comfort and Amenities (eigenvalue = 2.9), and Promotion and Policy (eigenvalue = 2.6).

Multicollinearity was formally assessed using Variance Inflation Factors (VIF). Variables with VIF greater than 5 were either removed or combined into composite indices. The final model included 18 variables with VIF values ranging from 1.2 to 4.7, well below concerning thresholds of 10.

3.4.2. Regularized Logistic Regression Specification

To address the high-dimensional nature of the data and prevent overfitting, regularized logistic regression was employed using elastic net regularization combining L1 (LASSO) and L2 (Ridge) penalties:

p(WTP = 1) = 1/(1 + exp(−(α + Σβᵢxᵢ)))

Subject to: Σ|βᵢ| ≤ t₁ (L1 penalty) and Σβᵢ² ≤ t₂ (L2 penalty)

The mixing parameter (α) was set to 0.5, providing equal weighting to L1 and L2 penalties. Regularization parameters were selected through 10-fold cross-validation minimizing prediction error.

3.4.3. Model Validation and Robustness Checks

Model performance was evaluated using multiple metrics. The dataset was split into training (70%, n = 280) and testing (30%, n = 120) subsets. Cross-validation was performed using 10-fold methodology repeated five times to ensure stability. Performance metrics included area under the ROC curve (AUC = 0.823), sensitivity (0.765), specificity (0.841), and Brier score (0.187).

Bootstrap resampling (n = 1000) was used to generate confidence intervals for coefficients and predicted probabilities. Sensitivity analysis examined model stability across different regularization parameters and subset specifications.

3.5. Economic Benefit Assessment Using WHO-HEAT

3.5.1. Health Benefit Quantification

The WHO Health Economic Assessment Tool (HEAT) was applied with location-specific parameters to quantify health benefits from increased AM [24]. Base mortality rates for Thailand (7.8 per 1000 annually) and relative risk reductions from physical activity (RR = 0.89 for all-cause mortality per 11.25 MET-hours per week of cycling, RR = 0.92 for walking) were incorporated based on established meta-analyses.

3.5.2. Input Parameters and Assumptions

Value of Statistical Life (VSL) was estimated at 4.2 million THB based on Thai Ministry of Transport guidelines adjusted for 2021 values using GDP per capita ratios. The affected population was estimated as 19,150 individuals (70% of total campus population) based on survey willingness to participate. Average increase in AM duration was calculated as 14.5 min per day (weighted average across scenarios) translating to 88.5 h annually.

3.5.3. Cost-Benefit Analysis Framework

Infrastructure costs were estimated based on Thai Department of Highways standard specifications. Basic infrastructure improvements (Scenario A) were costed at 2.4 million THB per kilometer. Enhanced safety features (Scenario B) added 1.8 million THB per kilometer. Comprehensive integration (Scenario C) required additional 3.2 million THB per kilometer for smart systems and public transport connections.

Annual operation and maintenance costs were estimated at 8% of capital costs based on international benchmarks adjusted for Thai conditions. A 7% discount rate was applied following Thai Ministry of Finance guidelines for public infrastructure projects. The analysis period was set at 20 years reflecting typical infrastructure lifespans.

3.5.4. Sensitivity Analysis Framework

Comprehensive sensitivity analysis examined parameter uncertainty across multiple dimensions. VSL estimates ranged from 2.8 to 6.3 million THB based on international comparisons and income elasticity adjustments. Participation rates varied from 50% to 85% reflecting implementation challenges and adoption rates. Mortality risk reductions incorporated confidence intervals from meta-analyses of physical activity interventions. Discount rates were tested from 5% to 10% to assess robustness of economic conclusions.

4. Results

4.1. Sample Characteristics and Representativeness

The final sample (n = 400) demonstrated reasonable representation across key demographics (

Table 2. Sample Demographic and Travel Characteristics (n = 400).

Variable	Category	Frequency	Percentage
Gender	Male	188	47.0%
	Female	212	53.0%
Age Group	17–22 years	268	67.0%
	23–26 years	64	16.0%
	27–35 years	48	12.0%
	Above 35 years	20	5.0%
Occupation	Undergraduate student	280	70.0%
	Graduate student	40	10.0%
	Academic staff	60	15.0%
	Non-academic staff	20	5.0%
Monthly Income	<10,000 THB	160	40.0%
	10,000–15,000 THB	112	28.0%
	15,001–20,000 THB	72	18.0%
	20,001–25,000 THB	36	9.0%
	>25,000 THB	20	5.0%
Primary Transport	Walking	128	32.0%
	Cycling	32	8.0%
	Motorcycle	176	44.0%
	Car	48	12.0%
	Public transport	16	4.0%
Distance to Campus	<1 km	116	29.0%
	1–3 km	176	44.0%
	3–5 km	72	18.0%
	>5 km	36	9.0%

). Gender distribution was 53% female and 47% male, closely matching university records (54% female). Age distribution showed 83% under 26 years, reflecting the student-dominated population but with adequate representation of staff (20%) for meaningful analysis.

Income distribution revealed 40% earning less than 10,000 THB monthly, 28% earning 10,000–15,000 THB, 18% earning 15,001–20,000 THB, 9% earning 20,001–25,000 THB, and 5% earning above 25,000 THB. This distribution aligns with typical Thai university community income patterns.

Travel behavior analysis showed 44% using motorcycles as primary transport, 32% walking, 12% using cars, 8% cycling, and 4% using public transport. Most respondents (73%) lived within 3 km of campus, supporting the feasibility of AM interventions. Daily travel costs were modest, with 58% spending 20 THB or less, indicating cost sensitivity in transport choices.

Importantly, 84% of respondents expressed willingness to switch to AM if infrastructure improved, demonstrating substantial latent demand for enhanced facilities.

4.2. Motivation Factors for Active Mobility Adoption

Analysis of factors motivating potential behavior change toward walking or cycling revealed clear priorities among respondents (Figure 1). Traffic safety emerged as the primary concern, cited by 31% of participants, followed closely by health benefits at 30%. Infrastructure quality and convenience factors each accounted for 12% and 11% respectively. Environmental concerns motivated 8% of respondents, while cost savings influenced 3%. Notably, promotional campaigns from public and private sectors showed minimal influence at only 5%, suggesting that infrastructure improvements should precede marketing efforts.

4.3. Contingent Valuation Results and WTP Estimation

4.3.1. Bid Response Analysis

Response patterns across initial bids showed expected negative correlation between bid amount and acceptance rates. At 50 THB monthly, 78% of respondents accepted the bid, declining systematically to 34% at 250 THB. Follow-up bid responses demonstrated consistency with utility-maximizing behavior, with correlation coefficients between initial and follow-up responses of 0.84 (p < 0.001).

4.3.2. Mean and Median WTP Estimates

Interval regression analysis yielding mean WTP estimates of 127 THB per month (95% CI: 114–142 THB) for Scenario A, 156 THB per month (95% CI: 141–173 THB) for Scenario B, and 189 THB per month (95% CI: 172–208 THB) for Scenario C. Median WTP values were consistently lower at 98 THB, 121 THB, and 147 THB respectively, indicating right-skewed distributions typical of CVM studies.

Bootstrap confidence intervals based on 1000 replications confirmed robustness of these estimates. Sensitivity analysis excluding protest responses increased mean WTP by 8–12%, while inclusion of strategic zeros reduced estimates by 4–7%.

4.4. Regularized Logistic Regression Results

The regularized logistic regression model demonstrated strong predictive performance with cross-validated AUC of 0.823 and good calibration (Brier score = 0.187). The model successfully predicted 84.7% of cases in the holdout test set, with balanced sensitivity (76.5%) and specificity (84.1%).

4.4.1. Significant Predictors of WTP

Table 3. Regularized Logistic Regression Results (Final Model).

Variable	Coefficient (β)	Standard Error	Odds Ratio	95% CI	p-Value
Demographics
Gender (Male)	−0.734	0.315	0.48	0.26–0.89	0.020 **
Age	−0.018	0.022	0.98	0.94–1.02	0.414
Income (15,001–20,000 THB)	−1.309	0.441	0.27	0.11–0.65	0.004 ***
Travel Behavior
Primary mode: Motorcycle	−1.715	0.481	0.18	0.07–0.46	<0.001 ***
Primary mode: Car	−1.966	0.627	0.14	0.04–0.50	0.002 ***
Travel cost (daily)	0.008	0.004	1.01	1.00–1.02	0.053 *
Infrastructure Factors
Rest areas	0.765	0.349	2.15	1.08–4.27	0.029 **
CCTV security	0.636	0.32	1.89	0.98–3.65	0.047 **
Public transport integration	1.044	0.357	2.84	1.42–5.68	0.003 ***
Protective barriers	0.518	0.307	1.68	0.92–3.07	0.092 *
Architectural design	0.54	0.266	1.72	1.02–2.89	0.041 **
Safe crossings	0.485	0.289	1.62	0.92–2.87	0.094 *
Promotion Activities
Media campaigns	−0.545	0.258	0.58	0.35–0.96	0.035 **
Mobile applications	−0.494	0.242	0.61	0.38–0.98	0.042 **
Model Statistics
Cross-validated AUC	0.823
Brier Score	0.187
Hosmer-Lemeshow p-value	0.342

Note: * p < 0.1, ** p < 0.05, *** p < 0.01. Reference categories: Gender (Female), Income (<10,000 THB), Primary mode (Walking/Cycling).

presents the final regularized model results with odds ratios and 95% confidence intervals. Several infrastructure factors emerged as significant predictors of willingness to pay for AM improvements.

Rest areas showed the strongest positive association (OR = 2.15, 95% CI: 1.08–4.27, p = 0.029), addressing the critical weakness of hot weather identified in the SWOT analysis. This finding reflects the practical importance of providing relief from Thailand’s tropical climate during active mobility activities.

CCTV security systems demonstrated significant influence (OR = 1.89, 95% CI: 0.98–3.65, p = 0.047), highlighting safety concerns as fundamental barriers to AM adoption. The installation of surveillance technology addresses both actual security risks and perceived threats that deter walking and cycling.

Public transport integration emerged as highly influential (OR = 2.84, 95% CI: 1.42–5.68, p = 0.003), supporting the concept of AM as part of integrated mobility systems rather than isolated interventions. This finding underscores the importance of first-mile and last-mile connectivity in suburban contexts.

Protective barriers separating pedestrians and cyclists from motorcycles showed positive association (OR = 1.68, 95% CI: 0.92–3.07, p = 0.092), though at marginal significance. This reflects the specific challenge of motorcycle encroachment identified as a primary threat in the qualitative analysis.

4.4.2. Demographic and Behavioral Factors

Demographic analysis revealed notable resistance patterns. Male participants showed significantly lower likelihood of switching to AM (OR = 0.48, 95% CI: 0.26–0.89, p = 0.020). This gender effect requires deeper investigation beyond simple cultural explanations, potentially reflecting complex interactions between social status, safety perceptions, and practical considerations such as dress codes and weather exposure.

Higher-income individuals in the 15,001–20,000 THB bracket were 73% less likely to switch (OR = 0.27, 95% CI: 0.11–0.65, p = 0.004) compared to the lowest income group. This pattern suggests that motorized transport serves as a status symbol or convenience preference among higher-earning community members.

Current travel mode strongly predicted switching likelihood. Motorcycle users (OR = 0.18, 95% CI: 0.07–0.46, p < 0.001) and car users (OR = 0.14, 95% CI: 0.04–0.50, p = 0.002) were significantly less likely to switch compared to current walkers and cyclists, indicating substantial inertia in motorized transport habits.

4.4.3. Infrastructure Design Preferences

Architectural design quality showed positive influence (OR = 1.72, 95% CI: 1.02–2.89, p = 0.041), suggesting that aesthetic considerations contribute to AM infrastructure acceptance. This finding supports investment in visually appealing designs that enhance campus environment quality.

Interestingly, promotional activities showed negative associations with WTP. Media campaigns (OR = 0.58, 95% CI: 0.35–0.96, p = 0.035) and mobile applications (OR = 0.61, 95% CI: 0.38–0.98, p = 0.042) were associated with reduced switching likelihood. This counterintuitive finding suggests that respondents prioritize tangible infrastructure improvements over promotional efforts, viewing marketing campaigns as ineffective substitutes for real facility enhancements.

4.5. Health and Economic Benefits Analysis

4.5.1. WHO-HEAT Application Results

Application of the WHO-HEAT methodology with Thailand-specific parameters yielded substantial estimated health benefits. The projected 14.5-min daily increase in AM translates to 88.5 additional hours of physical activity annually per participant. Using established dose-response relationships, this activity level reduces all-cause mortality risk by 8.2% for cycling and 6.4% for walking.

For the estimated affected population of 19,150 individuals, this translates to prevention of 9.8 premature deaths annually (95% CI: 6.2–14.1 deaths). Applying the Thailand-adjusted VSL of 4.2 million THB yields annual health benefits of 41.2 million THB (95% CI: 26.0–59.2 million THB).

Additional health benefits include reduced healthcare costs from decreased cardiovascular disease, diabetes, and mental health conditions. Conservative estimates based on Thai healthcare cost data suggest additional savings of 12.4 million THB annually.

4.5.2. Cost-Benefit Analysis Summary

Total infrastructure investment requirements range from 15.8 million THB for basic improvements (Scenario A) to 47.2 million THB for comprehensive integration (Scenario C). Annual operation and maintenance costs add 1.3 to 3.8 million THB respectively.

Present value calculations over 20 years using a 7% discount rate yield benefit-cost ratios of 3.2 for Scenario A, 2.8 for Scenario B, and 2.1 for Scenario C. Net present values are positive across all scenarios, ranging from 18.4 million THB (Scenario A) to 12.7 million THB (Scenario C).

4.5.3. Sensitivity Analysis Results

Comprehensive sensitivity analysis confirms robustness of economic conclusions across parameter variations. Even under conservative assumptions (VSL reduced by 30%, participation rates at 50%, mortality risk reductions at lower confidence bounds), benefit-cost ratios remain above 1.5 for basic infrastructure improvements.

The analysis demonstrates sensitivity to participation rates and VSL estimates, but conclusions remain favorable across plausible ranges. Higher discount rates (10%) reduce but do not eliminate positive net present values.

5. Discussion

5.1. Theoretical and Practical Implications

The findings provide strong evidence for infrastructure-led approaches to promoting active mobility in suburban university contexts. The prominence of rest areas, CCTV systems, and public transport integration as primary drivers of willingness to pay reflects context-specific adaptation of global AM principles to Thai conditions.

The negative association between promotional activities and WTP represents a significant finding that challenges conventional wisdom about behavior change interventions. This result suggests that in contexts where basic infrastructure is lacking, promotional efforts may be perceived as inadequate substitutes for tangible improvements. The finding aligns with hierarchy of needs frameworks where fundamental safety and comfort requirements must be addressed before aspirational messaging becomes effective.

5.1.1. Gender Dynamics and Cultural Context

The lower willingness among male participants to adopt AM requires nuanced interpretation beyond simplistic cultural explanations. Qualitative insights suggest multiple contributing factors including workplace dress expectations, social status considerations, and differential safety perceptions. Male participants frequently mentioned concerns about professional appearance and the perceived inconvenience of changing clothes after cycling or walking in hot weather.

This finding necessitates targeted interventions addressing practical barriers rather than assumptions about cultural preferences. Facilities such as changing rooms, shower facilities, and covered parking for professional attire could significantly influence male participation rates.

5.1.2. Income Effects and Motorization Patterns

The inverse relationship between income and AM willingness reflects broader patterns of motorization in developing economies. Higher-income individuals’ resistance to switching from motorized transport suggests that economic development alone may not automatically promote sustainable mobility without targeted policy interventions.

This pattern aligns with the Environmental Kuznets Curve concept applied to transport, where rising incomes initially increase motorized transport demand before potentially shifting toward sustainable modes at higher development levels. The finding emphasizes the importance of proactive policy intervention rather than relying on natural progression toward sustainable transport choices.

5.2. Policy Recommendations and Implementation Strategy

5.2.1. Infrastructure-First Approach

The research strongly supports prioritizing infrastructure investments over promotional campaigns in the initial phases of AM development. The evidence suggests a sequential implementation strategy beginning with basic safety and comfort infrastructure (rest areas, CCTV, protective barriers) before advancing to comprehensive integration with public transport systems.

5.2.2. Targeted Interventions for Different User Groups

The demographic findings necessitate differentiated approaches for various user segments. For male users and higher-income individuals, interventions should focus on convenience and status enhancement rather than cost savings or environmental appeals. This might include premium facilities, time-saving route designs, and integration with professional activities.

For current motorized transport users, the research suggests that dramatic infrastructure improvements are necessary to overcome substantial switching barriers. Incremental improvements may be insufficient to change entrenched travel behaviors.

5.2.3. Integration with Broader Transport Planning

The strong positive association with public transport integration supports AM development as part of comprehensive mobility systems rather than standalone interventions. This finding aligns with international best practices emphasizing multimodal integration and first-mile/last-mile connectivity.

5.3. Economic Viability and Investment Justification

The economic analysis provides robust justification for AM infrastructure investments across all scenarios examined. Benefit-cost ratios exceeding 2.0 under conservative assumptions demonstrate strong economic returns even before considering broader benefits such as air quality improvements, congestion reduction, and urban livability enhancements.

The health benefits alone, valued at 2840–4260 THB per person annually, provide compelling justification for public investment. These estimates are conservative, excluding potential benefits from reduced healthcare system burden, improved productivity from better health outcomes, and positive externalities from increased physical activity rates.

Financing and Implementation Mechanisms

The WTP estimates of 127–189 THB per month suggest potential for user fee contributions to infrastructure financing, though full cost recovery through user charges appears unlikely and potentially inequitable. A mixed financing approach combining public investment, user contributions, and development impact fees from campus expansion projects offers the most viable implementation pathway.

5.4. Comparison with International Literature

The study’s findings align with international research on AM promotion while revealing context-specific adaptations necessary for middle-income developing countries. The emphasis on safety infrastructure and climate adaptation (rest areas, weather protection) reflects tropical urban conditions that differ significantly from temperate environments where much AM research has been conducted.

The negative association with promotional activities contrasts with findings from developed countries where information campaigns often show positive effects [28,29]. This difference may reflect infrastructure maturity levels, with promotional efforts being more effective when basic facility requirements are already met.

Methodological Contributions

The study’s methodological approach combining CVM with regularized regression and comprehensive bias mitigation strategies provides a replicable framework for AM research in similar contexts. The use of SWOT analysis to inform quantitative survey design demonstrates effective integration of qualitative and quantitative methods.

The application of WHO-HEAT methodology with country-specific parameters offers a standardized approach for economic evaluation that can be replicated across different developing country contexts with appropriate parameter adjustments.

5.5. Limitations and Future Research Directions

5.5.1. Sample Limitations and Generalizability

While the sample adequately represents the university community, the focus on educated, relatively young participants limit generalizability to broader urban populations. The 83% representation of individuals under 26 years and 80% student composition necessitate cautious interpretation when extending findings to citywide contexts.

Future research should expand sampling to include diverse urban communities, older adults, and individuals with different educational backgrounds to provide more comprehensive insights into AM preferences across demographic groups.

5.5.2. Methodological Limitations

The cross-sectional design prevents assessment of actual behavior change following infrastructure implementation. Longitudinal studies tracking behavior before and after infrastructure improvements would provide stronger causal evidence for the relationships identified.

The pilot survey sample size (n = 36), while adequate for initial testing, represents a limitation in the measure selection process. Larger pilot studies could provide more robust foundation for the final survey instrument.

5.5.3. Future Research Priorities

Several research directions emerge from this study’s findings. First, longitudinal evaluation of actual infrastructure implementations is needed to validate predicted behavior changes and refine economic impact estimates. Second, deeper investigation of gender dynamics and cultural factors influencing AM adoption could inform more effective intervention design. Third, comparative analysis across different types of university campuses and urban contexts would enhance understanding of transferability.

Machine learning approaches to travel demand forecasting in multi-modal networks [30] offer promising avenues for enhancing prediction accuracy and understanding complex interaction effects between infrastructure improvements and user behavior.

5.5.4. Integration with Smart City Technologies

Future research should explore integration of AM infrastructure with smart city technologies including real-time information systems, dynamic route optimization, and integrated payment platforms. The potential for technology to enhance user experience and overcome practical barriers deserves systematic investigation.

6. Conclusions

This study provides comprehensive evidence for the economic viability and user acceptance of active mobility infrastructure investments in suburban university contexts within middle-income developing countries. The mixed-methods approach combining qualitative SWOT analysis, contingent valuation methodology, and regularized statistical modeling offers robust insights into user preferences and willingness to pay for AM improvements.

The research demonstrates that infrastructure-focused interventions addressing safety, comfort, and connectivity concerns can generate substantial user support and economic benefits. Rest areas, CCTV security systems, and public transport integration emerge as priority investments with strong user acceptance and favorable economic returns.

The finding that promotional activities show negative associations with willingness to pay challenges conventional approaches to behavior change intervention, suggesting that infrastructure development should precede marketing efforts in contexts where basic facilities are inadequate.

6.1. Key Contributions to Literature

This study addresses critical gaps in AM research from Southeast Asian contexts, providing the first comprehensive economic evaluation of user preferences for suburban university AM infrastructure in Thailand. The methodological framework combining CVM with bias mitigation strategies and regularized regression offers a replicable approach for similar investigations in developing country contexts.

The application of WHO-HEAT methodology with country-specific parameters demonstrates standardized approaches for economic evaluation that can inform policy decisions across diverse settings.

6.2. Policy Implications

The research provides actionable evidence for university administrators, urban planners, and transportation authorities considering AM infrastructure investments. The benefit-cost ratios exceeding 2.0 across all scenarios provide strong justification for public investment, while the specific user preferences identified can guide design priorities and implementation sequencing.

The demographic patterns identified necessitate targeted approaches for different user groups, with particular attention to addressing practical barriers faced by male users and higher-income individuals currently relying on motorized transport.

6.3. Implementation Recommendations

Based on the research findings, we recommend a phased implementation approach beginning with high-priority infrastructure investments (rest areas, CCTV, protective barriers) before advancing to comprehensive system integration. This strategy maximizes user acceptance while building foundation for more complex interventions.

The integration with public transport systems should be prioritized given the strong positive association with user willingness to pay and the broader benefits for sustainable mobility transition.

6.4. Final Reflections

Active mobility infrastructure represents a critical component of sustainable urban development strategies, particularly in rapidly developing regions facing environmental and health challenges from motorization. This research demonstrates that well-designed, user-centered infrastructure investments can generate substantial economic returns while promoting healthier, more sustainable transportation choices.

The success of AM interventions depends critically on understanding local contexts, user preferences, and implementation constraints. The framework developed in this study provides a foundation for evidence-based decision making that can be adapted across diverse urban environments facing similar sustainability challenges.