The Impact of Cross-Border E-Commerce Pilot Zones on Regional Economic Growth: Evidence from the Yangtze and Pearl River Deltas, 2010-2023
Abstract
This study evaluates whether approval of China's Cross-Border E-Commerce (CBEC) comprehensive pilot zones changes city-level economic growth. The analysis uses a balanced panel of 39 cities in the Yangtze River Delta and Pearl River Delta from 2010 to 2023, comprising 546 city-year observations, 36 treated cities and three cities never treated by 2023. Because approval occurred in several cohorts, the main specification applies the Callaway-Sant'Anna group-time average treatment effect estimator with not-yet-treated controls and city-clustered standard errors. A joint test of event-time coefficients from five to two years before approval does not reject parallel trends (Wald = 6.540, p = 0.162). The estimated average treatment effect on treated cities is -0.0912 log points (SE = 0.0333, p = 0.006; 95% CI [-0.1565, -0.0259]), equivalent to approximately an 8.72% reduction in nominal GDP per capita. Effects are negative and significant in the approval year and the following year, providing reduced-form support for the revised short-run cost hypothesis: compliance, technological adaptation and resource-reallocation costs may temporarily outweigh initial gains. This interpretation is theoretical rather than a directly observed mediator. The estimate remains negative when Shanghai and Shenzhen are excluded, when the sample begins in 2014, and when 2020 is omitted, but it is imprecise when only the three never-treated cities are used as controls; a 500-draw cohort-randomisation placebo gives an empirical p-value of 0.094. On a common 2010-2021 horizon, both regional estimates are negative, while their direct difference is not significant (p = 0.161). An exploratory logistics analysis finds that policy exposure raises per-capita express-delivery revenue, but the activity-to-growth path and bootstrapped indirect effect are insignificant. The evidence therefore supports a conditional short- to medium-run negative effect and calls for longer evaluation horizons, transitional support and direct measurement of firm and worker adjustment.