The Impact of the China–ASEAN Free Trade Area on the Quality of China’s Forest Product Exports: Causal Evidence from Double Machine Learning

Free trade agreements are widely used to promote trade liberalization, yet their implications for export quality in resource-based sectors remain poorly understood. This study examines whether the China–ASEAN Free Trade Area (CAFTA) changed the quality of China’s forest product exports. Using HS6 product-level bilateral trade data for 1995–2023, we estimate export quality through a demand-side residual approach and apply double machine learning (DML) to identify the causal effect while controlling for product, destination, and year heterogeneity. The results show that CAFTA significantly reduces the average quality of China’s forest product exports, a finding robust to outlier treatment, alternative algorithms and sample splits, and alternative substitution elasticities, and that survives dropping potentially endogenous controls, excluding COVID-19 and other crisis periods, clustering standard errors by importing country, and excluding each ASEAN partner. Mechanism tests indicate that the quality-downgrading effect operates mainly through the trade cost effect and the standard harmonization effect; both channels lower entry barriers and weaken incentives for quality differentiation. The negative effect is concentrated in the ASEAN-6, whereas the ASEAN-4 exhibits a small positive effect, and resource-intensive products experience the largest quality decline. Policy should combine CAFTA 3.0 cooperation with stronger quality systems, traceability, and technology-oriented upgrading in Chinese forestry firms.