Dataset and Code for: The Industrial Inverted-U and Global Carbon Offshoring (1970–2024)

Overview This repository contains the comprehensive 55-year longitudinal dataset (1970–2024) and analytical code supporting the investigation of the Industrial Inverted-U of Embodied Carbon. The research tracks the spatial relocation and structural decoupling of the “Canonical Four” hard-to-abate industrial sectors: Cement, Iron & Steel, Chemicals, and Non-ferrous Metals. Through shifting the analytical focus from descriptive accounts of trade flows toward a generalizable theory of late-stage structural decoupling, this dataset provides the empirical foundation to understand the limits of global carbon offshoring. Scientific Context The architectural foundation of international climate governance is primarily built upon production-based accounting (PBA) of greenhouse gas emissions. While this framework allows high-income nations to report a successful “decoupling” of economic growth from domestic emissions, it frequently obscures the geographic dislocation of industrial activities. The redistribution of carbon-intensive production from the Global North to the Global South led to a historical surge within emissions embodied within international trade. This repository provides the harmonized data necessary to map this "Great Industrial Migration" and identify the precise macroeconomic conditions under which a nation’s reliance on imported embodied carbon begins to decline. Key Discoveries Supported through this Dataset The Dematerialization Threshold (τ): Empirical identification of the specific manufacturing value-added share of GDP where economies transition from an Offshoring Phase (rising reliance on imported CO2) to a Structural Decoupling regime. The Post-2015 Structural Plateau: Evidence confirming that major emerging markets (such as the BRICS group) are hitting their own structural turning points, ceasing to act as infinite sinks for carbon-intensive production. Structural Leapfrogging: Proof that modern emerging economies are reaching their dematerialization thresholds at significantly lower levels of industrial emission shares (≈18%) compared to the historical precedent established through early-industrializing G7 nations (≈28%). Data Structure & Contents The dataset integrates environmental and economic panel data across multiple decades, harmonized through Multi-Regional Input-Output (MRIO) frameworks. industrial_shares_1970_2024.csv: The primary panel dataset detailing territorial industrial emissions, manufacturing value-added shares, and net embodied carbon transfers (ECTit) across advanced and emerging economies. Sectoral Coverage: Cement, Iron & Steel, Chemicals, and Non-ferrous Metals. Temporal Coverage: 1970 to 2024. Replication Code The repository includes the Python and econometric scripts utilized to model the quadratic regression curves, calculate the τ thresholds, and generate the spatial relocation stacked area charts (Figure 4). Keywords: Carbon Offshoring, Emissions Embodied in Trade (EET), Structural Decoupling, Dematerialization Threshold, Heavy Industry, BRICS, Consumption-Based Accounting, Climate Economics