As a core component of high-standard opening-up, institutional opening-up seeks to reduce institutional transaction costs, deepen domestic reforms, stimulate market vitality, and foster inclusive growth. Within this framework, the Free Trade Zone (FTZ) and the Cross-border E-commerce Comprehensive Pilot Zone (CECZ) serve as two major policy experiments, prompting the following questions: Do they generate a synergistic effect in narrowing the income gap? What is the underlying mechanism? And what are the heterogeneous policy effects?

This paper decomposes the urban-rural income gap into two dimensions: inter-regional disparity and intra-regional disparity, measured respectively by the population-weighted coefficient of variation and the Theil index. Using panel data from 285 prefecture-level cities in China from 2011 to 2021, it employs a double machine learning framework to account for high-dimensional confounders and complex nonlinear relationships in causal inference. In addition, an intensity DID model is constructed, in which policy intensity is jointly defined by the import-export scale and e-commerce transaction volume of cities in the previous year, enabling a more precise identification of heterogeneous policy effects.

The empirical results indicate that, first, the “dual pilot” policy significantly narrows inter-regional disparity, while its effect on intra-regional disparity is limited and may even lead to widening. Second, the policy operates mainly through two channels: improvements in digital infrastructure and expansion of employment opportunities. Third, the policy effects exhibit pronounced heterogeneity. The convergence effect of inter-regional disparity is stronger in northeast China, central cities, regions with moderate income imbalance, and Belt and Road Initiative node cities; whereas the reduction effect of intra-regional disparity is more evident in western and inland cities. Fourth, the synergistic effect of the “dual pilot” policy is substantially stronger than that of either pilot alone, and an implementation sequence in which the FTZ precedes the CECZ yields more favorable outcomes.

The academic values of this paper are threefold: First, it situates the “dual pilot” policy within the broader literature on institutional opening-up and income distribution, shedding light on the distributional consequences of policy synergy and institutional coupling. Second, by decomposing income disparities and integrating double machine learning with intensity DID, it offers a more robust identification strategy for evaluating the causal effects of complex policy combinations. Third, it provides policy-relevant evidence for optimizing the spatial layout and sequencing of institutional opening-up initiatives, offering valuable insights for promoting common prosperity through high-standard opening-up.