Pre-registration
Cuba's life expectancy and infant mortality outcomes by 2000 were strong enough to rank competitively not just against Latin American peers, but against a broad ex-ante fixed pool of non-Latin-American market economies; if Cuba fails that harder comparison, the "socialist health-system superiority" story is more regional than universal.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
PRIMARY 1 (life expectancy rank, dispositive): within the fixed 21-country pool {CUB, ESP, PRT, GRC, TUR, ISR, JPN, KOR, THA, MYS, IDN, PHL, IND, LKA, PAK, BGD, MAR, TUN, DZA, EGY, JOR}, Cuba's life expectancy at birth in 2000 must rank in the top third of the pool (rank <= 7 of 21, rank 1 = best). PRIMARY 2 (infant mortality rank, dispositive): Cuba's infant mortality rate in 2000 must rank in the top quartile of the same pool (rank <= 5 of 21, lower = better). PRIMARY 3 (health-vs-income overperformance): Cuba's mean health rank across {life expectancy, infant mortality} in 2000 must beat its GDP-per-capita rank by at least 5 places. SUPPORTED if all three primaries hold. PARTIAL if exactly two hold. REFUTED if zero or one holds. INFORMATIVE: report 1960 ranks, 1991-2000 post-Soviet-subsidy sub-period trajectory, and separate Cuba's standing versus the advanced-market subgroup {ESP, PRT, GRC, ISR, JPN, KOR}. METHOD_VALID: requires coverage for Cuba plus at least 18 of the 20 non-Cuban comparators at the 2000 endpoint for life expectancy, infant mortality, and income-rank context.
formal test & threshold
test: cuba_non_latam_market_pool_rank_and_income_gap_2000 threshold: PRIMARY: LE_rank_CUB_2000 <= 7 of 21 AND IMR_rank_CUB_2000 <= 5 of 21 AND income_rank_CUB_2000 - mean(LE_rank_CUB_2000, IMR_rank_CUB_2000) >= 5
Method
- Template
descriptive- Sample
- 21 countries · 1960 – 2000
- Evidence type
- associational
Descriptive rank-table comparison of Cuba against a broad non-LatAm market comparator pool, 1960-2000. Primary statistics are 2000 endpoint ranks on life expectancy, infant mortality, and GDP per capita. The key diagnostic is whether Cuba's health rank materially outruns its income rank inside a pool that includes both developing and advanced market economies. Not causal and not a clean socialism-vs-capitalism estimator: Soviet subsidy, baseline advantage, sanctions, and official-data quality remain confounds. The point is to stress- test how far the Cuban health result travels outside a friendlier regional pool.
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 | level |
life_expectancy_who outcome | who_gho:life_expectancy_at_birthtier 1 | level |
cuba_socialist_indicator treatment | constructed:indicator = 1 for CUB; 0 for the non-LatAm market comparator pool.tier 5 | indicator |
log_gdp_per_capita_maddison control | owid:gdp-per-capita-maddison-2020tier 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
Cuba health outcomes vs non-LATAM market peers, 1960-2000
Verdict: PARTIAL — Cuba clears two of the three harder gates in the 21-country non-LATAM market-economy pool. Ranks: LE #6/21, IMR #7/21, income #18/21; mean-health-vs-income gap 11.5. Missed: IMR rank need <= 5.
Primary thresholds
- Life expectancy rank in 2000 must be <= 7 within the 21-country pool.
- Infant mortality rank in 2000 must be <= 5 within the 21-country pool.
- Cuba's mean health rank must beat its income rank by at least 5.0 places.
Cuba's standings
- 1960 ranks: LE #5, IMR #2.
- 2000 ranks: LE #6/21, IMR #7/21, income #18/21.
- Mean health rank in 2000: 6.5.
- Income minus health-rank gap: +11.5 places.
2000 rank table
| Country | LE rank | IMR rank | Income rank | Mean health rank | Income minus health | |---|---:|---:|---:|---:|---:| | JPN | 1 | 1 | 1 | 1.0 | +0.0 | | ESP | 2 | 2 | 2 | 2.0 | +0.0 | | GRC | 4 | 3 | 6 | 3.5 | +2.5 | | ISR | 3 | 5 | 3 | 4.0 | -1.0 | | PRT | 5 | 4 | 4 | 4.5 | -0.5 | | CUB | 6 | 7 | 18 | 6.5 | +11.5 | | KOR | 7 | 6 | 5 | 6.5 | -1.5 | | MYS | 8 | 8 | 7 | 8.0 | -1.0 | | TUN | 9 | 12 | 11 | 10.5 | +0.5 | | JOR | 11 | 11 | 10 | 11.0 | -1.0 | | THA | 12 | 10 | 9 | 11.0 | -2.0 | | LKA | 13 | 9 | 14 | 11.0 | +3.0 | | TUR | 10 | 14 | 8 | 12.0 | -4.0 | | PHL | 15 | 13 | 17 | 14.0 | +3.0 | | DZA | 14 | 15 | 12 | 14.5 | -2.5 | | EGY | 16 | 16 | 13 | 16.0 | -3.0 | | MAR | 17 | 18 | 16 | 17.5 | -1.5 | | IDN | 18 | 17 | 15 | 17.5 | -2.5 | | IND | 19 | 20 | 20 | 19.5 | +0.5 | | BGD | 20 | 19 | 21 | 19.5 | +1.5 | | PAK | 21 | 21 | 19 | 21.0 | -2.0 |
Advanced-market subgroup check (ESP, PRT, GRC, ISR, JPN, KOR, plus Cuba)
| Country | LE rank | IMR rank | Income rank | Mean health rank | Income minus health | |---|---:|---:|---:|---:|---:| | JPN | 1 | 1 | 1 | 1.0 | +0.0 | | ESP | 2 | 2 | 2 | 2.0 | +0.0 | | GRC | 4 | 3 | 6 | 3.5 | +2.5 | | ISR | 3 | 5 | 3 | 4.0 | -1.0 | | PRT | 5 | 4 | 4 | 4.5 | -0.5 | | CUB | 6 | 7 | 7 | 6.5 | +0.5 | | KOR | 7 | 6 | 5 | 6.5 | -1.5 |
Soviet-subsidy sub-period
- Cuba life expectancy: 73.8 in 1991 -> 75.9 in 2000.
- Cuba infant mortality: 10.3 in 1991 -> 6.8 in 2000.
Method
Descriptive rank-table test only. The dispositive question is whether Cuba remains genuinely competitive once the comparison pool expands beyond a friendlier Latin American frame. Life expectancy and infant mortality come from WDI; income-rank context comes from the OWID-packaged Maddison GDP-per-capita series because the WDI PPP-per-capita vintage does not report a usable Cuba 2000 endpoint. Method-validity requires Cuba plus at least 18 of the 20 non-Cuban comparators at the 2000 endpoint on life expectancy, infant mortality, and income.
Caveats
- WDI back-fills Cuba's 1960-2000 health series from Cuban official reporting; the usual official-data caveat still applies, especially for infant mortality classification practices and the lack of a full-window WHO-independent backfill.
- Income rank uses OWID's Maddison GDP-per-capita series rather than WDI PPP-per-capita because Cuba's WDI PPP endpoint is missing. That preserves cross-country comparability, but the exact income-rank gap should be read as an approximate context metric, not a causal control.
- A strong or weak rank in this pool does not by itself identify socialism as the cause: Soviet transfers, pre-revolution baseline, sanctions, public-health prioritisation, and state capacity remain bundled together.
Provenance
- world_bank_wdi:SP.DYN.LE00.IN
- world_bank_wdi:SP.DYN.IMRT.IN
- owid:gdp-per-capita-maddison-2020
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
Harder companion to `cuba_health_outcomes_vs_latam_peers`. The LATAM-only pool is informative but friendly: Cuba started ahead regionally and some of the best non-socialist health performers (Southern Europe, East Asia) sit outside that frame. This spec widens the test to a broad non-LatAm market-economy pool and asks where Cuba actually ranks, both in raw health outcomes and relative to its income rank. The question is not "did Cuba beat Japan or Spain on every metric," but whether Cuba's health ranking materially outran what its income position within a broad market comparator pool would predict.