Introduction

Carbon pricing is a key tool for addressing environmental externalities, yet the mechanisms of its success remain complex. This National Bureau of Economic Research (NBER) working paper develops a microfounded equilibrium model to quantify how firms and markets adjust to pollution pricing. It identifies three distinct adjustment margins: short-run input substitution, the reallocation of output from emissions-intensive to cleaner firms, and long-run adoption of clean technology.

Institutional setting and data

The study examines the first 17 years of the European Union Emissions Trading System (EU-ETS), the world’s largest carbon market established in 2005. The system covers power generation, manufacturing, and energy-intensive sectors. Researchers utilised a matched dataset linking European Union Transaction Log (EUTL) emissions records with Orbis firm-level financial data, covering 4,671 regulated firms across 24 countries from 2005 to 2021.

Model

The framework captures the complexity of production networks where many regulated sectors act as central input suppliers. The model treats emissions as an input alongside capital, labour, and materials. This allows for the estimation of short-run marginal abatement cost (MAC) curves while accounting for equilibrium reallocation across firms. Long-run dynamics are integrated through a law of motion where firms invest in clean technology in response to carbon price signals.

Estimation

Empirical estimation reveals that the input substitution margin is remarkably rigid. The elasticity of substitution between emissions and other productive inputs is essentially zero across all sectors, indicating that firms cannot easily reduce emissions in the short run without decreasing output. Within-sector demand elasticities were found to range from 4 in rubber and plastics to 12 in the refining sector, while the cross-sector demand elasticity is approximately 1.1.

Results

Reallocation of output is identified as the primary driver of the aggregate emissions response in the short run. This effect is most pronounced in sectors with high emissions heterogeneity, such as power and cement. Furthermore, clean technology adoption responds causally to permit price shocks; a 1% increase in the permit price raises clean technology by approximately 0.211% within five years. In carbon-intensive sectors like refining and power generation, this response increases to 0.318%.

Counterfactuals

Simulations of doubling the EU-ETS permit price show a 23% reduction in aggregate emissions in the short run, while output declines by 3.6%. After five years of induced clean technology adoption, emissions reductions deepen to 31%, while output losses ease to 2.5%. The study highlights that the economic burden falls significantly on downstream firms as carbon costs from upstream sectors, such as electricity, pass through the production network.

Conclusion

The findings suggest that carbon pricing delivers substantial emissions reductions with moderate aggregate economic costs. The stability of European output and employment despite falling emissions is explained by the dominance of market reallocation and endogenous technology adoption over within-firm substitution. Lasting and credible policy is emphasised as essential for securing the long-run technology dividend.