IESET.
Hypotheses·growth·caribbean_climate_resilience_panel_1990_2024

Across Caribbean small-island economies 1990-2024, climate-related shocks (named hurricanes, tropical-storm landfalls, sea-level encroachment events) imposed measurable cumulative real-GDP-per- capita losses correlated with storm frequency and intensity, and insurance-market and disaster-financing arrangements (CCRIF SPC, contingent-credit lines) reduced — but did not eliminate — the growth penalty.

The pre-registered claim is (a) panel-FE estimation of log_gdp_pc on country and year FE shows a significant negative coefficient on a storm-intensity index, AND (b) post-2007 (CCRIF establishment) the magnitude of the storm coefficient is smaller than pre-2007 in countries that joined, AND (c) the CARICOM-region cumulative growth gap relative to LATAM median is at least partly explained by the storm-intensity differential.

INCONCLUSIVEengine/runs/caribbean_climate_resilience_panel_1990_2024

INCONCLUSIVE_DATA_PENDING — treatment 'ccrif_membership_indicator' has no cross-country variation within years under year fixed effects

confidence cueResult card produced; verdict unclassified.

policy briefCoverage too thin

In ordinary language

Over a long period, do more market-oriented institutions translate into higher income or productivity, once the comparison looks beyond a single success story?

plain answer

This test cannot make a firm call yet. treatment 'ccrif_membership_indicator' has no cross-country variation within years under year fixed effects

why it matters

Growth claims can look convincing in single success stories. This test asks whether the pattern survives a broader comparison.

how the test works

It compares 12 country or place units from 1990 to 2024, using a panel fe design.

what was measured
What changed
  • Storm intensity index
  • Ccrif membership indicator
What we checked
  • Log income pc constant
  • Gross capital formation share income
  • Gross general govt debt share income
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

0 input datasets, 0 unresolved missing series, provenance status: no input vintages recorded.

Results

engine/runs/caribbean_climate_resilience_panel_1990_2024
1007550250199020072024JAMBRBBHSDOMTTOHTIGRD
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show log_gdp_pc_constant across 12 sampled countries over 1990–2024.
The shapes above are stylised — none of the lines are real data.
Placeholder for caribbean_climate_resilience_panel_1990_2024. Published chart will be generated from engine/runs/caribbean_climate_resilience_panel_1990_2024/chart_data.json.

Pre-registration

registration ordering unverified
first-spec commit 4c8ce8e · 2026-07-18T22:11:21Z
run generated · 2026-06-29T17:52:44Z
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.

Across Caribbean small-island economies 1990-2024, climate-related shocks (named hurricanes, tropical-storm landfalls, sea-level encroachment events) imposed measurable cumulative real-GDP-per- capita losses correlated with storm frequency and intensity, and insurance-market and disaster-financing arrangements (CCRIF SPC, contingent-credit lines) reduced — but did not eliminate — the growth penalty. The pre-registered claim is (a) panel-FE estimation of log_gdp_pc on country and year FE shows a significant negative coefficient on a storm-intensity index, AND (b) post-2007 (CCRIF establishment) the magnitude of the storm coefficient is smaller than pre-2007 in countries that joined, AND (c) the CARICOM-region cumulative growth gap relative to LATAM median is at least partly explained by the storm-intensity differential.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Not supported if (a) panel_fe coefficient on storm-intensity is not negative and significant at p < 0.10, OR (b) CCRIF DiD-attenuation coefficient is not negative (i.e. CCRIF membership does not partly offset the storm coefficient), OR (c) the storm-intensity-attributable share of the Caribbean-LATAM growth gap is below 10%.

formal test & threshold
test:      panel_fe_plus_ccrif_did_plus_decomposition
threshold: storm_intensity_coefficient < 0 at p < 0.10 AND ccrif_did_attenuation < 0 AND storm_attributable_share_of_growth_gap >= 0.10

Method

Template
panel_fe
Clustering
country
Sample
12 countries · 1990 – 2024
Evidence type
causal

Primary: panel_fe with country and year FE plus storm-intensity coefficient. Secondary: CCRIF DiD comparing pre/post-2007 membership. Tertiary: variance decomposition of cumulative growth gap into storm intensity and other channels.

Data

VariableSourceTransform
log_gdp_pc_constant
outcome
world_bank_wdi:NY.GDP.PCAP.KDtier 2
log
gross_capital_formation_share_gdp
outcome
world_bank_wdi:NE.GDI.TOTL.ZStier 2
level
gross_general_govt_debt_share_gdp
outcome
imf:GGXWDG_NGDPtier 2
level
tourism_arrivals
outcome
world_bank_wdi:ST.INT.ARVLtier 2
log_level
storm_intensity_index
treatment
constructed:from EM-DAT/IBTrACS proxy mapped to country-yeartier 5
level
ccrif_membership_indicator
treatment
constructed:binary by country and post-2007 yeartier 5
binary
us_gdp_growth
control
fred:GDPC1tier 1
yoy_growth
oil_price
control
fred:DCOILBRENTEUtier 1
log_level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — caribbean_climate_resilience_panel_1990_2024

Verdict: INCONCLUSIVE_DATA_PENDING — treatment 'ccrif_membership_indicator' has no cross-country variation within years under year fixed effects

Pre-registration

  • Claim: Across Caribbean small-island economies 1990-2024, climate-related shocks (named hurricanes, tropical-storm landfalls, sea-level encroachment events) imposed measurable cumulative real-GDP-per- capita losses correlated with storm frequency and intensity, and insurance-market and disaster-financing arrangements (CCRIF SPC, contingent-credit lines) reduced — but did not eliminate — the growth penalty. The pre-registered claim is (a) panel-FE estimation of log_gdp_pc on country and year FE shows a significant negative coefficient on a storm-intensity index, AND (b) post-2007 (CCRIF establishment) the magnitude of the storm coefficient is smaller than pre-2007 in countries that joined, AND (c) the CARICOM-region cumulative growth gap relative to LATAM median is at least partly explained by the storm-intensity differential.
  • Falsification rule: Not supported if (a) panel_fe coefficient on storm-intensity is not negative and significant at p < 0.10, OR (b) CCRIF DiD-attenuation coefficient is not negative (i.e. CCRIF membership does not partly offset the storm coefficient), OR (c) the storm-intensity-attributable share of the Caribbean-LATAM growth gap is below 10%.
  • Falsification test: panel_fe_plus_ccrif_did_plus_decomposition

Estimate

  • Error: treatment 'ccrif_membership_indicator' has no cross-country variation within years under year fixed effects

Variables resolved

  • world_bank_wdi:NY.GDP.PCAP.KD → log_gdp_pc_constant (outcome, publisher=world_bank_wdi, n=12104)
  • world_bank_wdi:NE.GDI.TOTL.ZS → gross_capital_formation_share_gdp (outcome, publisher=world_bank_wdi, n=10428)
  • imf:GGXWDG_NGDP → gross_general_govt_debt_share_gdp (outcome, publisher=imf, n=8113)
  • world_bank_wdi:ST.INT.ARVL → tourism_arrivals (outcome, publisher=world_bank_wdi, n=5706)
  • constructed: binary by country and post-2007 year → ccrif_membership_indicator (treatment, publisher=constructed, n=420)
  • fred:GDPC1 → us_gdp_growth (controls, publisher=fred, n=960)
  • fred:DCOILBRENTEU → oil_price (controls, publisher=fred, n=480)

Variables missing data

  • constructed: from EM-DAT/IBTrACS proxy mapped to country-year (treatment, name=storm_intensity_index) — vintage not on disk

Generated by scripts/run_panel_fe.py at 2026-06-29T17:52:44+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

Cross-country Caribbean panel; storm-intensity index requires external-source mapping with clear data-quality flags.

Authored framework. Read the transparency note.