The Energy of Freedom? Solar Energy, Modern Slavery and the Just Transition

S-LCA
Supply Chains
Forced Labour
Energy
A Social Life Cycle Assessment estimating forced labour risk per kWh across the top 30 PV-producing countries — the first multi-country S-LCA in the renewable energy sector.
Published

July 1, 2022

Doi

10.5871/just-transitions-s-i/J-C

← Back to Projects

Supply Chains · Published 2022 · British Academy Just Transitions Programme

Overview

Is solar energy “the energy of freedom” — or does it, in fact, put that freedom at risk for the workers who produce it? Around 40–45 % of global polysilicon is manufactured in China’s Xinjiang Uyghur Autonomous Region under what multiple governments and independent tribunals have characterised as state-sponsored forced labour. A further 15–30 % of the cobalt used in lithium-ion battery storage originates from artisanal mines in the Democratic Republic of Congo where forced and child labour are prevalent.

This project — conducted for the British Academy’s Just Transitions within Sectors and Industries Globally programme — develops a new, scalable method for estimating forced labour risk across the photovoltaic (PV) supply chain, from raw silica extraction to consumer markets. We demonstrate how modern slavery risks can be quantified at the country-level production system, identifying where in the value chain risk accumulates and how sensitive the system is to changes in upstream conditions.

Project lead James Cockayne

Co-investigators Edgar Rodríguez-Huerta · Oana Burcu

Funder: British Academy — Just Transitions within Sectors and Industries Globally Period: November 2021 – March 2022 Host institution: Rights Lab, University of Nottingham

Method: Social Life Cycle Assessment of PV supply chains

This project applies a Social Life Cycle Assessment (S-LCA) framework to the global solar energy value chain — to our knowledge the first multi-country S-LCA in the renewable energy sector. Rather than measuring environmental impact per unit of product (as in conventional LCA), we measure forced labour risk as it cascades through the economic relationships that produce one kilowatt-hour of on-grid PV electricity.

Data integration

The method combines four independent datasets:

  • World Bank / IRENA — levelised cost of electricity (LCOE) by country and by supply-chain breakdown (PV module and inverter, balance-of-system hardware, installation, soft costs, O&M). LCOE provides the functional unit for disaggregating cost — and therefore risk — across the life cycle.
  • UN Comtrade — bilateral trade flows used to trace where each input into a country’s PV production system originates. This is the economic input–output component (EIO-LCA): large trade flows carry proportionally larger shares of the risk embedded in each source country.
  • PSILCA database — the Product Social Impact Life Cycle Assessment database, version 3, which provides 88 social impact indicators for every country-specific sector (CSS) combination available in the EORA Multi-Regional Input–Output (MRIO) database (187 countries; 15,909 sectors). Risk scores reflect 2018 data, updated for key CSS components using 2022 US Department of Labor data.
  • Walk Free Foundation Global Slavery Index — underpins the frequency-of-forced-labour indicator.

Forced labour indicators

Three indicators score each CSS component in the value chain:

Indicator Source What it captures
FFL — Frequency of Forced Labour Walk Free Foundation Global Slavery Index Estimated proportion of a country’s population in modern slavery
GFL — Goods produced by Forced Labour US Dept of Labor (ILAB) list Number of commodity classes in a sector produced in whole or in part by forced labour
TP — Trafficking in Persons US State Dept annual TIP Report tier rankings Country-level anti-trafficking enforcement

These combine into a composite Forced Labour Index (FLI):

\[\text{FLI}_i = 0.30 \times \text{FFL}_{\text{scaled},i} + 0.60 \times \text{GFL}_{\text{scaled},i} + 0.10 \times \text{TP}_{\text{scaled},i}\]

GFL receives the highest weight because it is most directly linked to specific goods in the supply chain; FFL captures broader population-level risk; TP proxies institutional enforcement capacity.

Output metrics

The FLI is translated into two final measures, applicable to any energy source:

  • FLR/kWh — forced labour risk in medium risk hour equivalents (mrh-eq) embodied in one kilowatt-hour of on-grid PV electricity, from cradle to gate.
  • FLR/USD LCOE — the same risk expressed per dollar of levelised electricity cost, useful for buyer and investor comparison.

Both metrics can be applied to other energy sources, opening a path towards forced labour risk integration into ESG benchmarking and energy market comparisons.

Key findings

Country-level results (top 30 producers, 96 % of global output)

  • China and India show the highest FLR/kWh scores on the GFL indicator, driven by the high volume of goods produced with forced labour in their PV hardware supply chains.
  • India and Ukraine show elevated FFL scores, reflecting broader national prevalence of forced labour across their economic sectors.
  • When risk is expressed as FLR/USD LCOE — the metric most relevant to buyers — India slightly exceeds China due to its higher solar LCOE.

Value-chain structure of risk

  • For GFL (goods-based risk), the primary driver is PV module and inverter hardware — particularly Chinese modules, which are both high-GFL and a large share of global LCOE.
  • For FFL (population-level risk), the highest mean arises in “Other soft costs” and O&M — sectors with large labour inputs across many countries.
  • Upstream components account for 55 % of FFL risk, 30 % of GFL risk, and 69 % of TP risk across the 30 countries studied.

Sensitivity analysis — system-level significance

A single CSS component — CN-Other electric machinery and equipment (the polysilicon production sector in China) — intervenes in 95 % of national PV value chains studied. Shifting its FFL risk score from “very low” (reflecting 2018 data, before Xinjiang allegations emerged) to “very high” produces a 15,385 % increase in mean FFL risk across all 30 countries (from 0.00026 to 0.04 mrh-eq/kWh). A change in one upstream node reverberates through nearly every solar energy system on the planet.

Policy implications

The study argues that a scalable S-LCA method could:

  1. Allow firms and investors to identify higher-risk CSS components for targeted due diligence, rather than resource-intensive audits across the entire supply chain.
  2. Enable integration of forced labour risk into ESG benchmarks, sustainability-linked finance, and energy procurement standards.
  3. Support policy actors in setting risk thresholds or milestones tied to forced labour metrics rather than arbitrary cut-off dates.
  4. Be replicated for battery supply chains (cobalt) and extended to other product systems beyond energy.

Interactive country risk profiles for all 30 countries are available via Tableau Public.

Outputs