The Environmental Smile Curve: Input-Output evidence on the Pollution Haven Hypothesis

This paper examines how the fragmentation of production across Global Value Chains (GVCs) generates both economic and environmental inequalities. Building on the "smile curve" framework (Mudambi, 2008; Meng et al., 2020), we show that developing countries specialize in low-value-added, high-emission production stages, while advanced economies capture high-value, low-emission activities like R&D and design (Riccio et al., 2025). Using OECD ICIO and CO2 emissions data, we demonstrate that GVC integration exacerbates a "double harm": production workers -particularly in middle-stage manufacturing- face wage suppression, while these same stages exhibit higher carbon intensity per unit of value added. This aligns with the Pollution Haven Hypothesis (Cole, 2004), as emissions are displaced to regions with weaker regulations. Our analysis reveals an environmental smile curve, where environmental and economic downgrading co-occur in middle segments of GVCs, reinforcing global inequalities. These disparities intensify with deeper GVC penetration, challenging the decoupling narrative of green growth. By integrating labour and emissions data, we provide novel evidence of how GVCs structurally embed unequal ecological and economic burdens.