A new approach for vehicle-health system measurement by network data envelopment analysis and an application in the USA.

Public health and climate change mitigation are strongly related to the efficiency of a unified vehicle-health system. To date, however, the existing DEA studies have not incorporated a robust set of transportation and public health indicators to evaluate the performance of the vehicle-health system. Our approach maps the relationship between transportation and health indicators in a network DEA framework. In addition, this study develops a new methodological framework for identifying stage priorities for decision-makers, peer evaluation and an understanding of the effects of policy in a comprehensive vehicle-health production system. This study empirically evaluates indicators of vehicle use and their effects on public health for each US state. The main conclusions are summarized: First, the network DEA and cross-efficiency network DEA models have significant methodological differences, highlighting the importance of model selection for empirical analysis. Second, US states exhibit enormous disparities in their prior strategies. The relative weights of the vehicle use stage and health outcomes stage confirm differing levels of importance attached to either stage. Third, group heterogeneity and technology inequality are observed among US states. Both intra-group and inter-group inequalities drive technology inequality. Moreover, considerable heterogeneity exists in the intra-group decomposition of the overall efficiency Theil index among the three network DEA models. Finally, promoting a safe and integrated transportation network improves vehicle-health performance, implying that strategic transportation policies hold significant potential for improving public health. [ABSTRACT FROM AUTHOR]