Pre-registration
Argentine monthly CPI inflation declines from its late-2023 peak (around 25% month-on-month in December 2023 following the 54% peso devaluation) to below 5% month-on-month within 12 months of Milei's December 2023 inauguration, and below 3% month-on- month within 18 months. The decline is faster and deeper than in any prior Argentine stabilisation episode since the 1989-1991 Austral/Convertibility period (Menem), and faster than in any of the Latin American stabilisation episodes in the IMF's post- 1990 catalogue that did not adopt dollarisation, controlling for initial inflation level. The pre-registered primary outcome is monthly CPI inflation as published by INDEC.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
PRIMARY (dispositive, all three must hold for SUPPORTED): (1) median Argentine monthly CPI m/m inflation across the window [t+10, t+14] post-inauguration (Oct 2024 through Feb 2025) <= 5.0%; (2) median monthly CPI m/m inflation across the window [t+16, t+20] (Apr 2025 through Aug 2025) <= 3.0%; (3) months_to_first_sub_5 (first calendar month after t=0 whose m/m INDEC print falls below 5%) <= 10. Verdict aggregation: 3/3 hold = SUPPORTED; 0/3 = refuted; 1-2/3 = partial. METHOD_VALID gate: INDEC IPC Nacional monthly index covers Dec 2022 through Jun 2025 (t-12 through t+18) without gaps; failure emits inconclusive rather than refuted. INFORMATIVE colour: comparison to the IMF post-1990 non-dollarisation LatAm stabilisation catalogue (Brazil Real 1994, Mexico Tesobonos 1995, Peru Fujimori 1990, Uruguay 2002, Bolivia 1985) reported in diagnostics; colours the verdict but does not gate it.
formal test & threshold
test: event_study_monthly_inflation_thresholds threshold: PRIMARY: median(mom_pct, t in [10..14]) <= 5.0 AND median(mom_pct, t in [16..20]) <= 3.0 AND months_to_first_sub_5 <= 10
Method
- Template
event_study- Clustering
month- Sample
- 1 countries · 2018 – 2027
- Evidence type
- causal
Event-study presentation around the December 2023 inauguration date. Primary test: pre-registered magnitude of monthly inflation 12 and 18 months post-event. Secondary: comparison to a peer-Latam synthetic-control-style control panel using Abadie-Diamond-Hainmueller weights optimised on 2015-2023 pre-period monthly inflation dynamics. Alternative: local projection (Jordà 2005) on monthly inflation with Milei-era indicator and months-since-event interaction.
Data
| Variable | Source | Transform |
|---|---|---|
monthly_cpi_inflation_argentina outcome | imf:PCPIPCHtier 2 | monthly_percent_change |
monthly_cpi_inflation_annualised outcome | derived from monthly_cpi_inflation_argentina | ((1 + m/100)^12 - 1) * 100 |
milei_inauguration_indicator treatment | constructed:indicator = 1 from 2023-12-10 onwards; 0 prior.tier 5 | step_indicator |
milei_months_since_inauguration treatment | constructed:integer count of months since December 2023tier 5 | integer_count |
monthly_cpi_latam_peers control | countries BRA, CHL, COL, MEX, PER, URY | monthly_percent_change |
commodity_price_index control | world_bank_wdi:TT.PRI.MRCH.XD.WDtier 2 | monthly_log_change |
real_effective_exchange_rate_argentina control | bis:WS_EERtier 2 | monthly_log_change |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Milei reforms reduce Argentine inflation
Verdict: INCONCLUSIVE_DATA_PENDING - INDEC IPC Nacional vintage missing from local data/vintages; expected indec:148.3_INIVELNAL_DICI_M_26 (indec:148.3_INIVELNAL_DICI_M_26@*.parquet).
Data Gate
- Required local vintage:
data/vintages/indec/148.3_INIVELNAL_DICI_M_26@*.parquet. - The current workspace does not contain an
indecvintage directory or matching INDEC IPC file. - The prior supported result cannot be reproduced in this lane without the registered INDEC input.
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-gated: IMF IFS monthly CPI for ARG + LatAm peers must land on disk before run. If IFS coverage is delayed, the INDEC IPC series is a clean substitute via manual drop (Argentina-specific) combined with peer-country monthly CPI from each central bank via OWID / IMF mirrors. Entry #10 in mega-spec D.1 pre-registration list (monetary topic). First-commit timestamp is the authoritative pre- registration date for v1.