Pre-registration
Cuban life expectancy and infant mortality outcomes 1960-2000 outperformed Latin American middle-income peers despite sanctions, demonstrating socialist health-system superiority.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
PRIMARY 1 (life expectancy, dispositive): Cuba's life expectancy at birth in 2000 minus the LATAM peer-pool mean in 2000 must be at least +1.0 year. PRIMARY 2 (infant mortality, dispositive): Cuba's IMR in 2000 must be at most 50% of the LATAM peer-pool mean (CUB / peer_mean <= 0.50). SUPPORTED if BOTH primaries hold AND Cuba's gap on at least one metric is more favourable in 2000 than in 1960 (so Cuba "pulled away" rather than just starting ahead). PARTIAL if exactly one primary holds. REFUTED if neither primary holds. INFORMATIVE: Cuba's rank within the 12-country pool at 2000 on each metric, plus 1991-2000 sub-period (post-Soviet-subsidy) trajectory. METHOD_VALID: requires WDI coverage for at least 8 of 11 peers at the 2000 endpoint; WHO GHO independent cross-check noted as unavailable for the full window (series begins ~2000).
formal test & threshold
test: cuba_vs_latam_peer_mean_endpoint_dual_threshold_2000 threshold: PRIMARY: (le_cub_2000 - le_peer_mean_2000) >= 1.0 AND (imr_cub_2000 / imr_peer_mean_2000) <= 0.50 AND ((le_gap_2000 > le_gap_1960) OR (imr_ratio_2000 < imr_ratio_1960))
Method
- Template
descriptive- Sample
- 12 countries · 1960 – 2000
- Evidence type
- associational
Descriptive cross-country comparison of CUB vs income-matched LATAM peers 1960-2000 on life expectancy and infant mortality. Plot trajectories with sanctions/embargo annotations; report rank-position of CUB within peer distribution at 5-year intervals. Cuban official data quality flagged as measurement caveat — sensitivity using WHO independent estimates. Not causal — does not separate health-system architecture from Soviet subsidies channel (subsidies ended 1991 makes 1991-2000 a useful sub-period).
Data
| Variable | Source | Transform |
|---|---|---|
life_expectancy_at_birth outcome | world_bank_wdi:SP.DYN.LE00.INtier 2 | level |
infant_mortality_per_1000 outcome | world_bank_wdi:SP.DYN.IMRT.INtier 2 | log |
life_expectancy_who outcome | who_gho:life_expectancy_at_birthtier 1 | level |
cuba_socialist_indicator treatment | constructed:indicator = 1 for CUB; 0 for income-matched LATAM peer pool.tier 5 | indicator |
log_gdp_per_capita control | world_bank_wdi:NY.GDP.PCAP.KDtier 2 | log |
log_health_expenditure_per_capita control | world_bank_wdi:SH.XPD.CHEX.PC.CDtier 2 | log |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Cuban health outcomes vs LATAM peers, 1960-2000
Verdict: SUPPORTED — Cuba's 2000 LE was 75.9y vs LATAM peer mean 72.8y (gap +3.1y, threshold +1.0y). Cuban IMR was 6.8/1k vs peer mean 21.1/1k (ratio 0.32, threshold ≤0.50). LE rank 3/12, IMR rank 1/12. IMR-ratio improved (0.36 → 0.32) but the LE gap NARROWED (+6.0y → +3.1y) — peers caught up on life expectancy.
Endpoint comparison
| Metric | Cuba 1960 | Cuba 2000 | Peer mean 1960 | Peer mean 2000 | 2000 gap / ratio | |---|---:|---:|---:|---:|---:| | Life expectancy (y) | 63.3 | 75.9 | 57.3 | 72.8 | +3.1y (need ≥ +1.0y) | | Infant mortality (per 1k) | 36.6 | 6.8 | 102.8 | 21.1 | 0.32 (need ≤ 0.50) |
At 2000 Cuba ranks #3/12 on life expectancy (higher = better) and #1/12 on infant mortality (lower = better) within the 12-country pool (Cuba + 11 peers).
Did Cuba pull away or just start ahead?
- LE gap 1960: +6.0y → 2000: +3.1y (change -2.9y).
- IMR ratio 1960: 0.36 → 2000: 0.32 (improvement +0.03).
- Pulled away on at least one metric: True.
Soviet-subsidy sub-period (1991-2000)
Soviet bloc collapse cut Cuban subsidies hard from 1991. If the gap held or widened in 1991-2000, that points more towards system-architecture; if the gap narrowed, subsidies were doing more of the work.
- LE gap 1991: +3.5y → 2000: +3.1y.
- IMR ratio 1991: 0.32 → 2000: 0.32.
Method
Descriptive endpoint comparison; no causal panel estimator. The thresholds come from a fair-reader interpretation of 'outperform' for a primary-care-emphasising system: at least one year better on LE at the endpoint, half-or-less infant-mortality vs the peer pool. The peer pool is the spec's 11 LATAM middle-income countries (MEX, BRA, ARG, CHL, COL, VEN, DOM, ECU, PER, URY, CRI). Method-validity gate: at least 8 of 11 peers must have data at the 2000 endpoint.
Caveats
- WDI back-fills the Cuban 1960-2000 series from Cuban government sources. WHO GHO independent life-expectancy estimates start around 2000, so a contemporaneous WHO cross-check across the full window is not possible from current vintages.
- This is descriptive only — it does not separate Cuban primary-care architecture from Soviet subsidy effects (1960-1991) or from primary-care diffusion across peer countries (Costa Rica's primary care expanded substantially over the same window and tracks Cuba closely).
- The 2000 ranking conditions on whichever peer countries had both an embargo-style shock and a non-Cuban political system; Venezuela's 1999 oil collapse and Argentina's 2001 crisis sit at the back end of this window.
Provenance
- world_bank_wdi:SP.DYN.LE00.IN
- world_bank_wdi:SP.DYN.IMRT.IN
See manifest.yaml for exact vintages. Reproduces from replication.py.
Strongest opposing argument
Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.
Notes
Stub seeded from a Marxist-Leninist school prediction about Cuban health-system superiority. Cuban official data quality is contested; needs human review of WHO/PAHO independent estimates and embargo controls.