Title: Impact of the transportation organisation integration policy at Ningbo-Zhoushan Port using the double machine learning difference-in-differences model

Authors: Jiao Liu; Kaige Zhu; Yuanqiang Zhang; Pengjun Zheng

Addresses: Faculty of Maritime and Transportation, Ningbo University, Ningbo, China; Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, China; National Traffic Management Engineering and Technology Research Center Ningbo University Sub-Center, Ningbo, China ' Faculty of Maritime and Transportation, Ningbo University, Ningbo, China; Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, China; National Traffic Management Engineering and Technology Research Center Ningbo University Sub-Center, Ningbo, China ' Faculty of Maritime and Transportation, Ningbo University, Ningbo, China; Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, China; National Traffic Management Engineering and Technology Research Center Ningbo University Sub-Center, Ningbo, China ' Faculty of Maritime and Transportation, Ningbo University, Ningbo, China; Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, China; National Traffic Management Engineering and Technology Research Center Ningbo University Sub-Center, Ningbo, China

Abstract: In December 2019, Ningbo-Zhoushan Port introduced a transportation organisation integration policy to enhance navigation safety and operational efficiency. To assess the policy's impact on vessel navigation safety and efficiency, this paper employs the double machine learning difference-in-differences model, using weighted accidents, weighted incidents, and average waiting time as indicators, to conduct an empirical analysis on three ports within Zhejiang Province and 17 coastal ports across the country. Results show that the θ₀ coefficients for three indicators were -0.276, -0.314, and -0.523, respectively, all statistically significant at the 1% or 5% levels. These findings indicate that the policy significantly improved both navigational safety and efficiency, with the most pronounced effect observed in reducing vessel waiting time. This highlights the policy's stronger impact on efficiency, as reflected by the larger magnitude of the corresponding coefficient. The findings provide robust empirical support for integrated maritime traffic management and inform future policy development.

Keywords: impact analysis; transportation organisation integration policy; Ningbo-Zhoushan Port; difference-in-differences model; double machine learning; DML.

DOI: 10.1504/IJSTL.2026.150450

International Journal of Shipping and Transport Logistics, 2026 Vol.22 No.1, pp.82 - 117

Received: 17 Apr 2025
Accepted: 30 Jul 2025
Published online: 14 Dec 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article