Network-specific design features are largely overlooked in the empirical literature, despite their potential role in strengthening urban rail transit as a net transport-sector decarbonizer. To address this gap, we examine the indirect contributions of these design features to the road-to-rail mode shift in China’s urban context, taking a structural equation modeling approach. In particular, we focus on three dimensions of physical network configurations—connectivity, accessibility, and efficiency. We hypothesize that the case specificity associated with these design features partly explains the mixed evidence regarding transit-led traffic-diversion effects. We find that, overall, metro rail service in China tends to significantly reduce vehicle use, with a net elasticity of -0.36. Path analysis shows that the road-to-rail traffic-diversion effects can be further strengthened by improved connectivity, optimized inter-station spacing, and enhanced transit accessibility. Conversely, the diversion effects are partly offset by rail-induced urban development and derived on-road traffic demand, while rail transit remains a net vehicular-traffic reducer. The overall net traffic diversion and the impact of network design features tend to be more pronounced and statistically significant in cities with larger populations, higher population densities, and longer periods of metro operation. This suggests that the case specificity underlying the existing mixed evidence is partly associated with network-specific factors and their interplay with the scale and density of a given city. From a policy perspective, these findings seem to provide a rationale for China’s continued investment in urban rail infrastructure, along with increased attention to network configuration features. Improvements in network connectivity and accessibility may be achieved through increased links between key stations, express rail services during rush hours, transit-oriented development, and complementary traffic management strategies.
Ou et al. (Mon,) studied this question.