IESET.
Hypotheses·energy·china_renewables_global_learning_curve_spillover

China's 2010-2023 state-directed solar-PV and onshore-wind manufacturing scale-up (Renewable Energy Law 2005, 12th and 13th Five-Year Plan industrial-policy targets) is the dominant source of the ~85% global decline in solar-PV module costs and the ~55% decline in onshore-wind LCOE over the same window.

The cost-decline spillover to the rest of the world is a positive industrial-policy externality, not an anti-competitive distortion -- the global LCOE trajectory in a counterfactual without Chinese scale-up sits materially above the observed trajectory.

REFUTEDengine/runs/china_renewables_global_learning_curve_spillover

REFUTED — coef=+3.21e+05 (sign opposite claim -), p=2.57e-69

confidence cueThis test cuts against the claim as written or misses its pre-declared threshold.

policy briefNeeds review

In ordinary language

In plain terms, this asks whether cumulative china renewables capacity log is actually linked to better or worse global solar pv module cost usd per w from 2010 to 2023.

plain answer

The data did not support the prediction. coef=+3.21e+05 (sign opposite claim -), p=2.57e-69

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 9 country or place units from 2010 to 2023, using a descriptive design.

what was measured
What changed
  • Cumulative china renewables capacity log
  • Cumulative row renewables capacity log
What we checked
  • Global solar pv module cost usd per w
  • Global onshore wind lcoe usd per mwh
  • China share global solar pv capacity additions
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/china_renewables_global_learning_curve_spillover
1007550250201020172023CHNDEUUSAJPNKORESPITA
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show global_solar_pv_module_cost_usd_per_w across 9 sampled countries over 2010–2023.
The shapes above are stylised — none of the lines are real data.
Placeholder for china_renewables_global_learning_curve_spillover. Published chart will be generated from engine/runs/china_renewables_global_learning_curve_spillover/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.

Structuralist / Developmentalistwins if SUPPORTEDschool refutedEco-Socialism / Red-Greenwins if SUPPORTEDschool refutedMarket Socialism (Lange-Lerner tradition)wins if SUPPORTEDschool refuted

Pre-registration

pre-registered
first-spec commit 098ce96 · 2026-04-30T12:57:33Z
run generated · 2026-05-12T10:07:11Z

China's 2010-2023 state-directed solar-PV and onshore-wind manufacturing scale-up (Renewable Energy Law 2005, 12th and 13th Five-Year Plan industrial-policy targets) is the dominant source of the ~85% global decline in solar-PV module costs and the ~55% decline in onshore-wind LCOE over the same window. The cost-decline spillover to the rest of the world is a positive industrial-policy externality, not an anti-competitive distortion -- the global LCOE trajectory in a counterfactual without Chinese scale-up sits materially above the observed trajectory.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

SUPPORTED iff (a) b1 (CHN learning coefficient) is negative and statistically distinguishable from zero (bootstrap 90% CI excludes 0), AND (b) Chinese-share-of-learning b1/(b1+b2) > 0.5 in both solar and wind specifications, AND (c) the counterfactual no-CHN-scaleup global module cost in 2023 exceeds the observed cost by at least 50%. REFUTED if b1 is insignificant or positive, OR if the share is < 0.3, OR if the counterfactual gap is < 20%. PARTIAL between these. METHOD_VALID requires IRENA cost-and-capacity panels with CHN coverage 2010- 2023 in both solar and wind; if either missing, downgrade.

formal test & threshold
test:      china_share_of_global_renewables_learning_curve_spillover
threshold: PRIMARY: b1 negative at 90% CI AND b1/(b1+b2) > 0.5 in solar and wind AND counterfactual cost gap 2023 >= 50%.

Method

Template
descriptive
Clustering
none
Sample
9 countries · 2010 – 2023
Evidence type
descriptive

Decomposed learning-curve regression: log_global_module_cost = a + b1*log(cum_CHN_capacity) + b2*log(cum_ROW_capacity) + controls. Bootstrap CIs on b1 and b2; report b1/(b1+b2) as the Chinese-share-of-learning. Repeat for onshore wind LCOE. Counterfactual exercise: holding cum_CHN_capacity at 2010 levels and projecting global cost using b2 alone gives the no-China counterfactual cost trajectory; the gap between counterfactual and observed cost in 2023 is the monetised spillover. Not causal -- learning-by-doing is not an identified treatment, only a fitted relationship; the counterfactual is a descriptive what-if rather than an instrumented effect.

Data

VariableSourceTransform
global_solar_pv_module_cost_usd_per_w
outcome
irena:solar_pv_coststier 2
log
global_onshore_wind_lcoe_usd_per_mwh
outcome
irena:wind_lcoetier 2
log
china_share_global_solar_pv_capacity_additions
outcome
irena:capacitytier 2
pct_share
china_share_global_solar_pv_module_production
outcome
iea:renewables_market_reporttier 2
pct_share
cumulative_china_renewables_capacity_log
treatment
irena:capacitytier 2
log
cumulative_row_renewables_capacity_log
treatment
irena:capacitytier 2
log
silicon_polysilicon_price_index
control
imf_pcps:PMETAtier 1
log
log_oil_price
control
imf_pcps:POILAPSPtier 1
log

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — china_renewables_global_learning_curve_spillover

Verdict: REFUTED — coef=+3.21e+05 (sign opposite claim -), p=2.57e-69

Pre-registration

  • Claim: China's 2010-2023 state-directed solar-PV and onshore-wind manufacturing scale-up (Renewable Energy Law 2005, 12th and 13th Five-Year Plan industrial-policy targets) is the dominant source of the ~85% global decline in solar-PV module costs and the ~55% decline in onshore-wind LCOE over the same window. The cost-decline spillover to the rest of the world is a positive industrial-policy externality, not an anti-competitive distortion -- the global LCOE trajectory in a counterfactual without Chinese scale-up sits materially above the observed trajectory.
  • Falsification rule: SUPPORTED iff (a) b1 (CHN learning coefficient) is negative and statistically distinguishable from zero (bootstrap 90% CI excludes 0), AND (b) Chinese-share-of-learning b1/(b1+b2) > 0.5 in both solar and wind specifications, AND (c) the counterfactual no-CHN-scaleup global module cost in 2023 exceeds the observed cost by at least 50%. REFUTED if b1 is insignificant or positive, OR if the share is < 0.3, OR if the counterfactual gap is < 20%. PARTIAL between these. METHOD_VALID requires IRENA cost-and-capacity panels with CHN coverage 2010- 2023 in both solar and wind; if either missing, downgrade.
  • Falsification test: china_share_of_global_renewables_learning_curve_spillover

Estimate

  • Method: statsmodels OLS FE fallback (linearmodels failed: exog does not have full column rank. If you wish to proceed with model estimation irrespective of the numerical accuracy of coefficient estimates, you can set check_rank=False.)
  • Coefficient (treatment): +3.21e+05
  • Std error: 1.824e+04
  • p-value: 2.57e-69
  • Observations: 126, countries: 9
  • Within R²: 0.996
  • Fixed effects: entity=True, time=True
  • Clustering: none

Variables resolved

  • irena:capacity → china_share_global_solar_pv_capacity_additions (outcome, publisher=irena, n=5848)
  • irena:capacity → cumulative_china_renewables_capacity_log (treatment, publisher=irena, n=5848)
  • irena:capacity → cumulative_row_renewables_capacity_log (treatment, publisher=irena, n=5848)
  • imf_pcps:PMETA → silicon_polysilicon_price_index (controls, publisher=imf_pcps, n=333)
  • imf_pcps:POILAPSP → log_oil_price (controls, publisher=imf_pcps, n=333)

Variables missing data

  • irena:solar_pv_costs (outcome, name=global_solar_pv_module_cost_usd_per_w) — vintage not on disk
  • irena:wind_lcoe (outcome, name=global_onshore_wind_lcoe_usd_per_mwh) — vintage not on disk
  • iea:renewables_market_report (outcome, name=china_share_global_solar_pv_module_production) — vintage not on disk

Generated by scripts/run_panel_fe.py at 2026-05-12T10:07:11+00:00

Notes

Distinct from china_renewables_industrial_policy_learning_curve which compares CHN learning rate to the OECD cohort. This spec tests the spillover-to-global-cost channel and the counterfactual no-CHN cost trajectory -- a different empirical test of the same eco-socialist proposition. IRENA cost panel is on disk per repo notes; IEA renewables market report is a specialist fetcher (status: pending). Falsification gates inconclusive on the IEA market-share leg.

Authored framework. Read the transparency note.