Pretis (2022): Does a Carbon Tax Reduce CO2 Emissions? Evidence from British Columbia

Overview

"Does a Carbon Tax Reduce CO₂ Emissions? Evidence from British Columbia" is a 2022 article by Felix Pretis (University of Victoria; Nuffield College / INET Oxford) in Environmental and Resource Economics 83(1), pp. 115–144.[1] It re-examines the emissions effect of British Columbia's 2008 carbon tax — North America's first major carbon tax — using difference-in-differences, the synthetic control method, and a newly proposed machine-learning approach to detecting structural breaks in difference-in-differences designs.[1] The paper is a deliberately cautious counterweight to the more optimistic BC evaluations (Elgie & McClay; Rivers & Schaufele) already cited on the emissions claim page: it disaggregates the sectoral from the economy-wide effect and finds them very different in strength.

Finding

Pretis finds that BC's carbon tax "reduced transportation emissions but not 'yet' led to large statistically significant reductions in aggregate CO₂ emissions."[1] The clean, detectable effect is concentrated in transport fuels — the sector where the tax is most salient at the pump — while the province-wide aggregate effect is weaker and does not register as a large, statistically significant intervention under his break-detection method. Applying the same method to other Canadian provinces, he finds that neither their carbon pricing nor their trading schemes are detected as large significant interventions either; where untaxed provinces cut emissions, the reductions trace to "closures and efficiency-improvements in emission-intense industries" rather than to pricing.[1] His overall reading: "existing carbon taxes (and prices) are likely too low to be effective in the time frame since their introduction."[1]

Relevance

This study bears on the wiki as a challenge / honest qualifier to Carbon pricing cuts emissions. It does not overturn the emissions claim — it confirms a real, measurable reduction in the taxed sector (transport), consistent with Andersson's Swedish result. But it sharpens the honesty of the aggregate claim: the flagship BC case's economy-wide effect is less robust than headline figures suggest, and the shortfall is attributed to prices sitting below what would be needed to bite economy-wide. It is the empirical anchor for the claim page's admission that "the size of the aggregate effect is genuinely contested even in the best-studied case," and it converges with Green's (2021) review finding limited aggregate impact.

Limits

• It is a magnitude/timing critique, not a refutation. Pretis finds a significant transport-sector reduction; his "not yet" and "too low" framing is about aggregate effects at the rates and horizon studied, not a claim that carbon taxes cannot cut emissions.
• Method-dependent. The aggregate null rests on a newly introduced break-detection method; different specifications (e.g. the modelled 5–15% ranges in earlier BC studies) reach more favourable conclusions. The divergence between transport-clear and aggregate-ambiguous is the robust takeaway.
• Bounded window. The evaluation covers the period since the 2008 introduction at relatively low rates (C$10 rising to C$30/tonne); it does not speak to effects at higher prices or over longer horizons.

See Also

• Carbon pricing cuts emissions — modestly, and without wrecking growth — the claim this paper qualifies
• Carbon Taxes and CO2 Emissions: Sweden as a Case Study (Andersson) — the transport-sector result that agrees with Pretis's sectoral finding
• Green (2021), carbon-pricing review — the parallel conclusion that aggregate impacts are limited
• Pigouvian Taxation — the theory the modest measured magnitudes bear on

Sources

1. Felix Pretis (2022), "Does a Carbon Tax Reduce CO₂ Emissions? Evidence from British Columbia," Environmental and Resource Economics 83(1), 115–144 — used for every finding on this page: the transport-vs-aggregate split, the "not 'yet' … large statistically significant reductions in aggregate CO₂ emissions" result, the finding that other provinces' pricing/trading schemes are not detected as large significant interventions (reductions there tracing to industrial closures/efficiency), and the "existing carbon taxes (and prices) are likely too low to be effective" conclusion. Title, author, journal, volume/issue/pages, year, and the quoted abstract phrases verified against the RePEc record and publisher listing this session. Springer · RePEc · INET Oxford