Pre-registration
Argentina under Milei (December 2023 inauguration) executed a fiscal-surplus + monetary-contraction + de-facto-dollarisation programme that collapsed monthly CPI inflation from ~25% (Dec 2023) toward sub-3% by 2025-2026, eliminated the primary fiscal deficit within a single fiscal year, and compressed the parallel-market peso-USD gap (blue-dolar / official) from over 100% to near zero. The pre-registered claim is that across four monetary/fiscal metrics — monthly inflation, primary fiscal balance, peso-USD parallel-market gap, and BCRA real monetary base — the Milei programme produces extreme outcomes that diverge from prior Argentine stabilisation attempts (Convertibilidad 1991, Macri 2015-2019) and from Latin American disinflation comparators. The headline test is whether the inflation collapse holds without a recession of comparable magnitude to the disinflation gain (output decline less than 4% peak-to-trough).
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
Supported if 4 of 5 canonical metrics meet thresholds. Refuted if 2 or fewer metrics meet thresholds. Mixed verdict otherwise. The headline single-shot version: refuted if monthly inflation does NOT fall below 5% by end-2025, OR if real GDP peak-to-trough decline exceeds 6% (i.e., the disinflation costs more output than the Tequila-style stabilisation benchmark).
formal test & threshold
test: multi_metric_canonical_milei_stabilisation threshold: metrics_met >= 4 of 5
Method
- Template
multi_metric_checklist- Clustering
none- Sample
- 1 countries · 2020 – 2026
- Evidence type
- canonical_case_multi_metric
Single-country canonical case; no comparator panel. The 5-metric checklist substitutes for a single causal estimator. Comparison to historical Argentine stabilisations (Convertibilidad, Macri) is descriptive context, not gating.
Data
| Variable | Source | Transform |
|---|---|---|
monthly_cpi_inflation outcome | bcra:inflation_monthlytier 1 indec:ipc_monthlytier 2 | monthly_pct_change |
primary_fiscal_balance_pct_gdp outcome | imf:GGXCNL_NGDPtier 2 indec:resultado_primariotier 2 | annual_pct_gdp |
parallel_fx_gap outcome | bcra:fx_officialtier 1 dolartoday:blue_ratetier 3 | ratio_minus_one |
real_gdp_index outcome | indec:emaetier 2 | log_level |
poverty_rate_official outcome | indec:eph_pobrezatier 2 world_bank_wdi:SI.POV.DDAYtier 2 | level |
milei_era_indicator treatment | constructed:binary = 1 for ARG from 2023-12-10 onwardtier 5 | binary |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Result card — milei_dollarisation_inflation_collapse_2024_2026
Verdict: inconclusive (data gaps)
Reason: 0 metrics met, 5 pending; 4 more need resolution
Pre-registered rule: SUPPORT if >= 4 of 5 metrics met; REFUTE if <= 2 met (impossible to hit support).
Counts: 0 MET · 0 NOT_MET · 4 PENDING_DATA · 1 PENDING_EVAL
Primary country: ARG
Metric-by-metric
| # | Metric | Status | Observed | Threshold | Notes |
|---|---|:---:|---:|---|---|
| 1 | monthly_inflation_collapse | PENDING_DATA | | monthly inflation < 3% by Dec 2025; sustained < 4% for 6+ months | No usable vintage for: indec:ipc_monthly; Non-tidy (needs custom parser): bcra:inflation_monthly |
| 2 | primary_fiscal_surplus_achieved | PENDING_EVAL | 0.5 (2024) [max_in_window_fallback] | primary balance > 0% of GDP in 2024 AND 2025 | threshold expression unparseable by regex |
| 3 | parallel_fx_gap_compression | PENDING_DATA | | parallel_fx_gap < 0.20 by Dec 2025 | No usable vintage for: dolartoday:blue_rate; Non-tidy (needs custom parser): bcra:fx_official |
| 4 | output_recession_bounded | PENDING_DATA | | max peak_to_trough log_real_gdp decline < 0.04 | No usable vintage for: indec:emae |
| 5 | poverty_rate_recovery | PENDING_DATA | | INDEC poverty headcount < 42% by 2026-S1 | No usable vintage for: indec:eph_pobreza |
Claim
Argentina under Milei (December 2023 inauguration) executed a fiscal-surplus + monetary-contraction + de-facto-dollarisation programme that collapsed monthly CPI inflation from ~25% (Dec 2023) toward sub-3% by 2025-2026, eliminated the primary fiscal deficit within a single fiscal year, and compressed the parallel-market peso-USD gap (blue-dolar / official) from over 100% to near zero. The pre-registered claim is that across four monetary/fiscal metrics — monthly inflation, primary fiscal balance, peso-USD parallel-market gap, and BCRA real monetary base — the Milei programme produces extreme outcomes that diverge from prior Argentine stabilisation attempts (Convertibilidad 1991, Macri 2015-2019) and from Latin American disinflation comparators. The headline test is whether the inflation collapse holds without a recession of comparable magnitude to the disinflation gain (output decline less than 4% peak-to-trough).
Interpretation
Verdict is inconclusive (data gaps) — 4 metric(s) cannot be evaluated because the underlying data source is not yet in the vintages pipeline, and 1 metric(s) have data but a threshold expression the auto-evaluator does not recognise (complex conditions, discrete event counts, cross-country gaps). Close these gaps then re-run.
Steelman live concerns
See hypotheses/steelman/milei_dollarisation_inflation_collapse_2024_2026.md for the strongest opposing arguments. Canonical-case multi-metric evidence is a pattern match, not a causal identification — the result card should be read as 'outcome trajectory matches the predicted pattern to degree X' rather than 'policy P caused the outcome'.
Provenance
Vintages pinned in manifest.yaml. Full per-metric diagnostics in diagnostics.json. Machine-readable results in metric_results.parquet.
Strongest opposing argument
Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.
Notes
Multi-metric canonical-case design. Argentina's 2023-2026 stabilisation is a once-in-a-generation natural experiment combining a fiscal-shock-therapy package, central-bank-balance-sheet sterilisation, and incremental dollarisation steps (cepo relaxation, USD legal-tender expansion, BCRA reserve recomposition). The hypothesis treats it as a multi-metric pattern-match rather than a single estimator because the pre-period (2007-2023 INDEC falsification, 2018-2023 cepo regime) makes conventional panel/synth-control identification noisy. Output-trajectory and poverty-rate metrics are included to surface the social cost of the disinflation, not just the disinflation itself.