Linking Local Government Road Safety Efforts to Road-User Behavior

Abstract

Korea’s Traffic Culture Index (TCI) combines local-government administrative effort, road-user behavior, and crash outcomes within a single evaluation framework. This study examines whether administrative effort is empirically associated with safer resident behavior using TCI sub-indicators for all 229 Korean municipalities from 2023 to 2025. A balanced panel of 687 observations was analyzed using municipality and year fixed-effects models, one-year lag models, and adjacent-year change models, with multiple-comparison corrections and measurement-stability checks. Administrative effort increased over the period, whereas behavioral change was small. No consistent association was found between the general administrative-effort composite and safety behavior across contemporaneous, lagged, or change-based analyses. Estimates for policy implementation were less precise and were positive for driving behavior across several specifications, leaving the component-specific evidence inconclusive. Equivalence testing rejected effects as large as ±0.5 behavior-score points per administrative-effort point for driving behavior, but not for pedestrian behavior. No domain-specific association survived false-discovery-rate correction. Several behavioral indicators also showed low year-to-year repeatability, which was systematically lower for survey-based than for field-observed measures. The findings suggest that the TCI’s administrative-input and behavioral components should be interpreted as distinct stages and that intermediate indicators are needed to track how administrative inputs translate into actual implementation and behavioral change.