Abstract

Parking management has become an increasingly important challenge in rapidly growing urban areas, yet evidence from intermediate Latin American cities remains scarce. This study analyzes parking choice behavior across Ecuadorian cities using a national stated preference (SP) experiment validated through revealed preference (RP) data and estimated via discrete choice models. Data were collected between 1 April and 3 May 2026 from 2150 active drivers across 12 cities spanning the Coast, Highlands, and Amazon regions. Multinomial Logit (MNL) and Mixed Logit (MIXL) models were estimated to evaluate the effects of parking cost, walking distance, search time, security, and surveillance on parking decisions. Results showed that all three cost-related attributes significantly reduced parking utility, while security improvements increased the probability of selecting formal parking alternatives. The MIXL model outperformed MNL specifications (ΔAIC = 211.84), revealing significant unobserved heterogeneity in cost (SD = 0.45) and security preferences (SD = 0.29). Willingness-to-pay estimates reached USD 0.64 per 100 m reduction in walking distance and USD 0.69 per 10-min reduction in search time, with substantial inter-city variability. Despite low current adoption (18.4%), willingness to use digital reservation systems exceeded 75% across income groups. The findings underscore the need for locally calibrated, context-sensitive parking policies and support differentiated smart parking strategies in developing urban systems.

IPC Classification

Keywords