This article investigates the demand for demand responsive transit (DRT) services in shrinking rural areas, focusing on factors influencing adoption. As rural populations decline and traditional transit options become less viable, DRT offers a flexible and cost-effective alternative. Using an integrated choice and latent variable model, this study analyzes stated preference data from a survey conducted in South Korea’s rural regions experiencing population decline. It examines socio-demographic factors, service attributes, and latent traits such as sociability, tech-savviness, and punctuality. Results show that socio-demographic characteristics, including age, income, and children’s status, significantly affect mode choice, with younger individuals and higher-income households more likely to choose DRT. The study highlights that residents of severely shrinking areas are less likely to adopt DRT, suggesting population decline affects service viability. Additionally, latent variables like sociability, tech-savviness, and punctual mindset positively influence DRT adoption. Key attributes, including in-vehicle travel time, access time, egress time, cost, and detour time, significantly determine mode choice, with longer travel times and higher costs negatively affecting DRT preference. These findings highlight the importance of ensuring service efficiency, cost-effectiveness, and accessibility for the successful implementation of DRT in rural areas. The article offers policy recommendations, including targeted service design, infrastructure improvements, and financial incentives. It also suggests future research exploring induced demand and long-term user adaptation to DRT in rural settings.

This study is an investigation of the effect of general-purpose lane reduction policies—commonly referred to as “road diets”—on drivers’ travel behavior, focusing on changes in travel route, departure time, and mode choice. A stated preference–off–revealed preference survey targeting drivers residing in Seoul was conducted in November 2024, yielding 3,466 responses, of which 650 were identified as valid. The data were analyzed using an error component mixed logit model to relax the assumption of independence from irrelevant alternatives and to account for correlation across alternatives and repeated choices at the individual level. In the model, four options were considered: maintaining private car use without changes (status quo), changing travel route, changing departure time, and switching to public transportation. The results indicate that lane reductions significantly influence drivers’ decisions, with a stronger preference for altering travel routes and departure times than for maintaining original travel behavior. However, shifts to public transportation were less pronounced, as drivers more frequently chose to adjust their routes or departure times instead. These findings highlight the importance of accounting for behavioral responses when implementing lane reduction policies and suggest that complementary measures, such as improved public transportation services and targeted incentives, may be necessary to encourage more substantial and sustained shifts toward public transport use. Future research should be conducted to examine the long-term effects of lane reduction and the interaction between such policies and public transportation improvements.

New towns often experience a structural transit gap in early stages, where transport supply lags behind population growth. Demand-responsive transit (DRT) has emerged as a promising complementary solution; however, most studies rely on MNL-based ICLV models that do not account for error covariance across alternatives. This study applies an ICLV model, integrating an Error Component Mixed Logit kernel with latent variables, to analyze mode choice behavior in transit-poor new towns. Based on an SP-off-RP survey of 644 residents in new towns, 2576 observations were analyzed. The model incorporates five latent variables, including Transit Dissatisfaction, Convenience, Safety, Travel Time, and Travel Companion Sensitivity, and captures unobserved correlations through a two-level nesting structure. Results show that DRT has a significantly positive alternative-specific constant, indicating latent acceptance beyond observable attributes. DRT adoption is more common among transit-poor new town residents and highly educated individuals, but less common among car owners. Users are more sensitive to access and waiting time than to in-vehicle time. Convenience, Safety, and Travel Time significantly influence DRT utility, while Travel Companion Sensitivity reveals heterogeneous effects across modes. These findings provide behavioral insights for designing effective DRT strategies in transit-poor new towns.