Pre-registration
OECD economies 2010-2023 have absolutely decoupled territorial CO2 emissions from real GDP -- emissions fell while GDP rose -- AND the pace of that decoupling falls short of the rate consistent with a 1.5C-aligned linear pathway (50% reduction 2019-2030, ~7%/year). The joint claim holds the absolute- decoupling fact (which refutes growth-requires-emissions necessity) AND the insufficient-pace finding (which refutes optimistic green-growth sufficiency). Both legs must hold; the spec is structured to refuse a one-leg-only verdict.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
SUPPORTED iff BOTH (a) pooled OECD territorial CO2/cap 2023 / 2010 < 1.0 AND pooled OECD real GDP/cap 2023 / 2010 > 1.05 (absolute decoupling fact), AND (b) mean pace_gap_vs_15c across OECD-35 in 2023 > 0.05 (>= 5 percentage points above the 1.5C benchmark; pace insufficient). REFUTED if (a) fails (no decoupling) OR (b) fails by more than 5 pp (i.e. pace_gap < -0.05; pace actually faster than benchmark). PARTIAL if exactly one leg holds. INFORMATIVE: country-level dispersion, consumption-based cross- check, sub-period decomposition. METHOD_VALID: at least 28 of 35 OECD countries have both endpoints on territorial CO2pc and real GDPpc; OWID consumption-CO2 panel covers >= 25 OECD countries.
formal test & threshold
test: oecd_joint_decoupling_fact_and_pace_gap_2010_2023 threshold: PRIMARY: pooled co2pc_2023 / co2pc_2010 < 1.0 AND pooled gdppc_2023 / gdppc_2010 > 1.05 AND mean pace_gap_vs_15c_2023 > 0.05.
Method
- Template
descriptive- Clustering
country- Sample
- 35 countries · 2010 – 2023
- Evidence type
- descriptive
Joint two-leg test on the OECD-35 panel 2010-2023. Leg A (absolute-decoupling fact, holds the eco-socialist / PK-degrowth claim against simple green-growth sceptics): population-weighted mean territorial CO2 per capita 2023 vs 2010 declines AND population-weighted mean real GDP per capita 2023 vs 2010 rises. Leg B (insufficient-pace fact, holds against optimistic green-growth advocates): mean pace_gap_vs_15c is positive (i.e. observed 2023 emissions are above the 1.5C-consistent linear benchmark for 2023). Country- level dispersion reported (which countries are on/below pace, which are not). Consumption-based cross-check is reported as diagnostic; if territorial decoupled but consumption-based did not, the absolute-decoupling leg is annotated PARTIAL.
Data
| Variable | Source | Transform |
|---|---|---|
log_co2_per_capita outcome | owid:co-emissions-per-capitatier 2 | log |
log_real_gdp_per_capita outcome | world_bank_wdi:NY.GDP.PCAP.KDtier 2 | log |
log_consumption_co2_per_capita outcome | owid:consumption-co2-per-capitatier 2 | log |
pace_gap_vs_15c_pct outcome | derived: (observed_co2_2023 / observed_co2_2019) - benchmark_15c_2023 | percentage_points |
post_paris_indicator treatment | constructed:1 for years >= 2016 (Paris Agreement entry into force)tier 5 | indicator |
log_population control | world_bank_wdi:SP.POP.TOTLtier 2 | log |
log_oil_price control | imf_pcps:POILAPSPtier 1 | log |
services_share_gdp control | world_bank_wdi:NV.SRV.TOTL.ZStier 2 | level |
● ready · ● pending · ● reconstruct-needed
Notes
Distinct from eu_green_deal_decoupling_pace_vs_target which is an EU-16 panel 2019-2023 around the Green Deal as named treatment. This spec covers the broader OECD-35 panel 2010-2023 with no named policy treatment, joint-testing the decoupling-fact and pace-shortfall on the same observation. WDI + OWID inputs are on disk for OECD; consumption-CO2 OWID panel coverage for non-EU OECD members (KOR, JPN, CAN, AUS, USA, MEX, CHL, TUR) needs confirmation -- spec gates inconclusive on the consumption-side leg if missing.