Behavioral mechanisms of traffic congestion in motorized-dominated cities: Insights from street-level observations in Ho Chi Minh City

Urban traffic congestion in motorcycle-dependent cities such as Ho Chi Minh City (HCMC) is often examined through infrastructure capacity, traffic volume, or intersection performance. However, these approaches can overlook the behavioral and spatial dynamics embedded in everyday use of street space. This study adopts a micro-behavioral lens to examine how informal socio-economic activities and concurrent human vehicle interactions contribute to recurring congestion. Drawing on field observations and structured video analysis across representative road segments in HCMC, the research develops a behavior-coding scheme and identifies five behavioral groups (A–E). Spatial temporal intensity mapping and co-occurrence analysis show that congestion rarely stems from a single behavior; rather, it emerges from recurring combinations, particularly those involving traffic participation (A), street vending (B), logistics activity (C), and school or workplace-related activity (D). Based on these patterns, the study identifies four dominant congestion mechanisms: spatial encroachment, flow interruption, traffic weaving and conflict, and temporal density surges. These mechanisms reflect context-specific interactions shaped by street width, land use, and daily urban rhythms. The findings underscore the need to move beyond infrastructure-centered traffic management toward integrated behavioral and spatial approaches. Policy implications include context-sensitive curb-space management, time-based regulation of street activities, and closer coordination between transport planning and land use policy. Although grounded in HCMC, the proposed framework offers transferable insights for traffic governance in other motorcycle-reliant cities across the Global South.