Dividends of AI-Driven Logistics Reform in Vietnam: A Recursive Dynamic CGE Analysis

Vietnam’s logistics costs are among the highest in Southeast Asia and a recognised bottleneck on export competitiveness. The government’s Logistics Development Strategy 2025–2035 mandates a reduction to 15% of GDP by 2030, but the technological adoption required to achieve this target has not been formally quantified. This paper employs a recursive dynamic Computable General Equilibrium (CGE) model with 40 sectors calibrated to Vietnam’s 2019 Input-Output Table to simulate three scenarios: a Business-as-Usual trajectory (S1), an empirically grounded AI-ITS adoption path drawing on recent literature (S2), and a reverse-closure targeted scenario that solves for the minimum annual TFP growth rate required to achieve the government’s price mandate (S3). Under the model and its calibration, the empirically plausible adoption trajectory raises projected real GDP by 0.75% and real wages by 0.62% above BAU by 2030 but yields only a 10% logistics price reduction — about five percentage points short of the government target. Closing the gap to a full 15% price reduction in the simulation requires a uniform TFP growth rate of 2.75% per annum across road freight, rail freight, and warehousing, associated with a 1.20% GDP gain. The residual gap between S3 and the government’s 15% cost-share target reflects inventory, administrative, and delay-related costs not captured by transport-sector TFP, suggesting complementary broader logistics ecosystem reforms.