Road Safety as the Binding Constraint on Sustainable Mode Choice: Discrete Choice Experiment Estimated via Conditional Logit
11th International Conference on Civil, Structural and Transportation Engineering, ICCSTE 2026, Hybrid, Barcelona, Spain, 11 - 13 June 2026, vol.9, (Full Text)
- Publication Type: Conference Paper / Full Text
- Volume: 9
- Doi Number: 10.11159/iccste26.185
- City: Hybrid, Barcelona
- Country: Spain
- Keywords: Active travel, Last-mile accessibility, Model shift, Perceived safety, Shared mobility, Sustainable transport, Travel behavior, Travel demand management
- Open Archive Collection: AVESIS Open Access Collection
- Yıldız Technical University Affiliated: Yes
Abstract
Urban transport decarbonization and livability goals increasingly depend on shifting travelers toward public and active modes, yet such shifts may be constrained when these alternatives are perceived as unsafe; this study examines whether road safety and personal security act as binding constraints on sustainable mode choice when individuals face explicit trade-offs against travel time and monetary cost. Using a stated preference discrete choice experiment, primary data were collected from 380 urban travelers, each completing eight choice tasks comparing a status quo current-mode option with a sustainable alternative (transit + walk) under systematically varied attribute bundles. Choices were organized in long format and analyzed using a conditional logit model grounded in random utility theory, with standard errors clustered by respondent to account for repeated choices. The results show that generalized cost components operate as expected: longer travel time and higher monetary cost reduce the likelihood of choosing the sustainable alternative. In contrast, safety-related improvements substantially increase sustainable choice, with personal security improvements (e.g., better lighting and a safer waiting environment) exerting the strongest influence, followed by walk-quality improvements and road-safety improvements. To translate model estimates into policy-relevant terms, counterfactual prediction was used to compute average marginal effects, indicating that improving personal security increases the probability of choosing the sustainable alternative, holding other attributes constant. These findings indicate that safety, particularly personal security conditions around access and waiting environments, functions as a critical enabling condition for sustainable mode shift; therefore, safety-first intervention bundles targeting crossings, last-mile walkability, and well-lit secure waiting areas should accompany conventional time and pricing strategies to achieve meaningful uptake.