IESET.
Hypotheses·energy·energy_crisis_2022_european_industrial_relocation

The 2022 European gas-price shock plus heterogeneous national policy responses (German Doppelwumms ~€200bn relief package, Spanish-Portuguese Iberian gas-price-cap mechanism, Italian energy-aid bridges, French EDF nationalisation + nuclear price cap) produced a divergent industrial- relocation pattern within Europe and across the Atlantic over 2022-2025.

Energy-intensive manufacturing share (chemicals, basic metals, fertiliser, glass, ceramics) fell more in DEU, ITA, and POL than in ESP, FRA, and USA; the lost share is partly structural (plant mothballing, capacity re- siting to Gulf Coast USA / Middle East / China) rather than cyclical. The intra-EU heterogeneity is large enough that the pooled-EU narrative obscures the policy-design lesson: countries that intervened on wholesale price formation (Iberian cap, French nuclear renationalisation) preserved more energy-intensive capacity than countries that intervened only on retail bills (Germany).

PARTIALengine/runs/energy_crisis_2022_european_industrial_relocation

PARTIAL — ATT=+0.1438, p=0.583, N=94, treated_countries=14 (above α=0.10)

confidence cueThe result is useful, but not decisive. Treat it as a clue, not a settled conclusion.

policy briefMixed or noisy

In ordinary language

In plain terms, this asks whether post 2021q4 gas shock dummy is actually linked to better or worse log energy intensive manufacturing va from 2018 to 2025.

plain answer

The evidence is suggestive but not decisive. ATT=+0.1438, p=0.583, N=94, treated_countries=14 (above α=0.10)

why it matters

This matters because energy claims should change belief only when they survive a pre-declared empirical test.

how the test works

It compares 14 country or place units from 2018 to 2025, using a did callaway santanna design, with fixed effects for country and year quarter.

what was measured
What changed
  • Post 2021q4 gas shock dummy
  • Country gas price intervention intensity
What we checked
  • Log energy intensive manufacturing va
  • Log industrial production index
  • Log chemical industry output
what this does not prove

A single test is not the whole truth. It narrows the claim under a specific sample, time period, and method. Strong policy conclusions need the pattern to survive nearby tests, alternative data, and serious objections.

verification

No evidence packet has been generated yet.

Results

engine/runs/energy_crisis_2022_european_industrial_relocation
1007550250201820222025DEUFRAITAESPNLDBELPOL
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show log_energy_intensive_manufacturing_va across 14 sampled countries over 2018–2025.
The shapes above are stylised — none of the lines are real data.
Placeholder for energy_crisis_2022_european_industrial_relocation. Published chart will be generated from engine/runs/energy_crisis_2022_european_industrial_relocation/chart_data.json.

Who has skin in the game — schools predicting on this

3 schools list this hypothesis as a test of their position. The chips below are school-level scoreboard outcomes, not a second hypothesis verdict.

hypothesis verdict vs scoreboard outcome

The banner verdict judges this hypothesis as written. The scoreboard asks whether each school's polarity-corrected prediction was right. Raw status is not a school win: SUPPORTED supports schools that needed SUPPORTED, but refutes schools that needed REFUTED.

Eco-Socialism / Red-Greenwins if SUPPORTEDpartial/no-callMarket Socialism (Lange-Lerner tradition)wins if MIXEDpartial/no-callModern Monetary Theorywins if MIXEDpartial/no-call

Pre-registration

registration ordering unverified
first-spec commit 4c8ce8e · 2026-07-18T22:11:21Z
run generated · 2026-05-04T08:18:31Z
Run timestamp predates this path's first git-add commit (rebase, rename, or pre-git local run). Spec hash is still the path's first-add commit — not repository HEAD — but ordering is not a clean pre-registration proof.

The 2022 European gas-price shock plus heterogeneous national policy responses (German Doppelwumms ~€200bn relief package, Spanish-Portuguese Iberian gas-price-cap mechanism, Italian energy-aid bridges, French EDF nationalisation + nuclear price cap) produced a divergent industrial- relocation pattern within Europe and across the Atlantic over 2022-2025. Energy-intensive manufacturing share (chemicals, basic metals, fertiliser, glass, ceramics) fell more in DEU, ITA, and POL than in ESP, FRA, and USA; the lost share is partly structural (plant mothballing, capacity re- siting to Gulf Coast USA / Middle East / China) rather than cyclical. The intra-EU heterogeneity is large enough that the pooled-EU narrative obscures the policy-design lesson: countries that intervened on wholesale price formation (Iberian cap, French nuclear renationalisation) preserved more energy-intensive capacity than countries that intervened only on retail bills (Germany).

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Not supported if (a) the post-2021Q4 × intervention-intensity interaction is zero or wrong-signed at p<0.10 (high-intervention countries should have smaller contraction), OR (b) intra-EU energy- intensive output gap (DEU+ITA vs FRA+ESP) does not widen by ≥5pp by 2024 relative to pre-shock baseline, OR (c) US Gulf Coast petrochemical capacity additions do not exceed pre-2022 trend, OR (d) the contraction reverses fully in 2024-2025 (cyclical not structural). If only the contemporaneous (2022-2023) effect is real and 2024-2025 shows recovery, classify as "cyclical shock without persistent relocation."

formal test & threshold
test:      european_relocation_did_panel
threshold: β_post×intervention positive on energy-intensive VA at p<0.10 AND DEU+ITA energy-intensive VA gap to FRA+ESP widens ≥ 5pp 2021-2024 AND US Gulf Coast petrochemical capacity exceeds 2018-2021 trend by ≥10% AND no full reversal by 2024-2025

Method

Template
did_callaway_santanna
Fixed effects
country, year_quarter
Clustering
country
Sample
14 countries · 2018 – 2025
Evidence type
causal

Primary specification: differences-in-differences with country-level treatment intensity = wholesale-intervention ordinal score interacted with the post-2021Q4 dummy. Outcome: log energy-intensive manufacturing VA + log chemical / steel / ammonia output. Test: does higher- intervention country show smaller post-2021Q4 contraction than lower- intervention country, after partialling out gas-price and oil-price? Second-stage: regress US Gulf-Coast petrochemical capacity additions on EU energy-intensive output decline to test the cross-Atlantic relocation channel — whether share lost in EU correlates with share gained in US Gulf Coast / SAU / ARE. Synthetic control for DEU and ITA on energy-intensive manufacturing VA using donor pool of FRA, ESP, USA, KOR, JPN. Known limitations: (1) Wholesale-intervention ordinal score is author-coded; sensitivity to alternative codings (binary, continuous-spend per GDP) reported. (2) Iberian cap was reformed and partially withdrawn 2023; treatment is time-varying. (3) Cyclical inventory destocking 2022-2023 is a confound for any 2024 reading; report 2024-2025 as the structural-persistence test. (4) Capacity re-siting decisions take 3-5 years; 2022-2025 window captures announcements + early plant mothballing but not full relocation.

Data

VariableSourceTransform
log_energy_intensive_manufacturing_va
outcome
constructed:OECD STAN VA — chemicals (C20) + basic metals (C24) + non-metallic minerals (C23) + paper (C17) + fertiliser. Cross-chectier 5
log
log_industrial_production_index
outcome
oecd:STSIND4_PRODMANtier 2
log
log_chemical_industry_output
outcome
world_bank_wdi:NV.IND.MANF.ZStier 2
log
log_steel_production
outcome
constructed:World Steel Association monthly crude-steel production by country. Manual-drop pending under data/manual/derived/.tier 5
log
log_ammonia_fertiliser_production
outcome
constructed:Fertilizers Europe / IFA monthly ammonia + nitrogen-fertiliser output by country. Manual-drop pending.tier 5
log
log_us_gulf_coast_petrochemical_capacity
outcome
constructed:ACC (American Chemistry Council) + EIA Gulf-Coast petrochemical capacity tracker. Cross-check eia: methanol / ethylene ptier 5
log
post_2021q4_gas_shock_dummy
treatment
constructed:indicator = 1 from 2021Q4 onwards (gas-price spike start).tier 5
indicator
country_gas_price_intervention_intensity
treatment
constructed:ordinal index of national wholesale-price interventions (0 = retail-only relief; 1 = retail + selective-industry rebate;tier 5
level
log_gas_price_ttf
control
constructed:Dutch TTF natural gas spot/forward, EUR/MWh, monthly. Cross-check fred:DHHNGSP (Henry Hub for USA). Manual-drop or eex ftier 5
log
log_brent_oil
control
imf_pcps:POILBREtier 1
log
log_real_gdp
control
world_bank_wdi:NY.GDP.MKTP.KDtier 2
log
trade_openness
control
world_bank_wdi:NE.TRD.GNFS.ZStier 2
level
eu_ets_price
control
eea:eu_ets_verified_emissionstier 2
level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — energy_crisis_2022_european_industrial_relocation

Verdict: PARTIAL — ATT=+0.1438, p=0.583, N=94, treated_countries=14 (above α=0.10)

Pre-registration

  • Claim: The 2022 European gas-price shock plus heterogeneous national policy responses (German Doppelwumms ~€200bn relief package, Spanish-Portuguese Iberian gas-price-cap mechanism, Italian energy-aid bridges, French EDF nationalisation + nuclear price cap) produced a divergent industrial- relocation pattern within Europe and across the Atlantic over 2022-2025. Energy-intensive manufacturing share (chemicals, basic metals, fertiliser, glass, ceramics) fell more in DEU, ITA, and POL than in ESP, FRA, and USA; the lost share is partly structural (plant mothballing, capacity re- siting to Gulf Coast USA / Middle East / China) rather than cyclical. The intra-EU heterogeneity is large enough that the pooled-EU narrative obscures the policy-design lesson: countries that intervened on wholesale price formation (Iberian cap, French nuclear renationalisation) preserved more energy-intensive capacity than countries that intervened only on retail bills (Germany).
  • Falsification rule: Not supported if (a) the post-2021Q4 × intervention-intensity interaction is zero or wrong-signed at p<0.10 (high-intervention countries should have smaller contraction), OR (b) intra-EU energy- intensive output gap (DEU+ITA vs FRA+ESP) does not widen by ≥5pp by 2024 relative to pre-shock baseline, OR (c) US Gulf Coast petrochemical capacity additions do not exceed pre-2022 trend, OR (d) the contraction reverses fully in 2024-2025 (cyclical not structural). If only the contemporaneous (2022-2023) effect is real and 2024-2025 shows recovery, classify as "cyclical shock without persistent relocation."

Estimate (Callaway-Sant'Anna staggered DiD, TWFE approximation)

  • coefficient: 0.14381729102378044
  • std_error: 0.26206389875517755
  • p_value: 0.5831515536149916
  • n_obs: 94
  • n_countries: 14
  • r_squared_within: 0.9857296543900552
  • fe_entity: True
  • fe_time: True
  • cluster: country
  • method: Callaway-Sant'Anna TWFE fallback (linearmodels failed: 'post')
  • n_treated_countries: 14
  • cohort_years: [2021]
  • dropped_controls_due_to_overlap: ['log_gas_price_ttf']

Variables resolved

  • world_bank_wdi:NV.IND.MANF.ZS → log_chemical_industry_output (outcome, n=9698)
  • constructed: indicator = 1 from 2021Q4 onwards (gas-price spike start). → post_2021q4_gas_shock_dummy (treatment, n=112)
  • constructed: ordinal index of national wholesale-price interventions (0 = retail-only relief; 1 = retail + selective-industry rebate; 2 = wholesale-price cap or renationalisation). DEU/POL = 1, ESP/PRT = 2, FRA = 2 (EDF), ITA/NLD = 1, USA = 0. → country_gas_price_intervention_intensity (treatment, n=112)
  • constructed: Dutch TTF natural gas spot/forward, EUR/MWh, monthly. Cross-check fred:DHHNGSP (Henry Hub for USA). Manual-drop or eex fetcher when shipped. → log_gas_price_ttf (controls, n=112)
  • imf_pcps:POILBRE → log_brent_oil (controls, n=518)
  • world_bank_wdi:NY.GDP.MKTP.KD → log_real_gdp (controls, n=14066)
  • world_bank_wdi:NE.TRD.GNFS.ZS → trade_openness (controls, n=10714)

Missing data

  • constructed: OECD STAN VA — chemicals (C20) + basic metals (C24) + non-metallic minerals (C23) + paper (C17) + fertiliser. Cross-check Eurostat sbs_na_ind_r2 sectoral turnover. (outcome)
  • oecd:STSIND4_PRODMAN (outcome)
  • constructed: World Steel Association monthly crude-steel production by country. Manual-drop pending under data/manual/derived/. (outcome)
  • constructed: Fertilizers Europe / IFA monthly ammonia + nitrogen-fertiliser output by country. Manual-drop pending. (outcome)
  • constructed: ACC (American Chemistry Council) + EIA Gulf-Coast petrochemical capacity tracker. Cross-check eia: methanol / ethylene production. Manual-drop pending. (outcome)
  • eea:eu_ets_verified_emissions (controls)

Generated by scripts/run_did_callaway_santanna.py at 2026-05-04T08:18:31+00:00

Strongest opposing argument

Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.

Notes

Data readiness: - OECD STAN sectoral VA: fetcher pending - OECD industrial production: ready - Eurostat sbs_na_ind_r2 / sts_inpr_m: ready - World Steel Association steel: manual-drop pending - Fertilizers Europe / IFA ammonia: manual-drop pending - ACC / EIA Gulf Coast petrochemical: manual-drop pending - TTF gas price: fetcher pending (eex) or manual-drop - EU ETS price: partial via eea fetcher - WDI / IMF PCPS: ready Run when OECD STAN + steel/ammonia manual drops + TTF series populate baseline.

Authored framework. Read the transparency note.