IESET.
Hypotheses·growth·milei_shock_therapy_output_recovery_trajectory

Argentine quarterly real GDP, having contracted in 2024H1 under the Milei stabilisation shock, recovers along a trajectory that by Q4 2025 closes at least 50% of the peak-to-trough output gap observed during the shock's worst quarter, and by Q4 2026 returns to or exceeds the pre-Milei (Q4 2023) real-GDP level.

The trajectory is faster than the LatAm peer donor pool (Brazil, Chile, Colombia, Peru, Mexico, Uruguay) weighted average for an equivalent shock absorbed without reform content. The hypothesis is consistent with the Austrian / monetarist reading that post- stabilisation recoveries are front-loaded once the fiscal- dominance distortion is removed and with empirical findings on sharp-stabilisation recovery patterns (e.g., Peru 1990-1993, Bolivia 1985-1989).

PARTIALengine/runs/milei_shock_therapy_output_recovery_trajectory

partial - Level-recovery legs hold (2025 = +3.0%, 2026 = +6.4%) but ARG cumulative 2024-2026 log-deviation +8.1% does NOT exceed peer mean +14.1% — recovery, but not faster than the LatAm peer pool. Note: 2 of 3 post-period years (2025, 2026) are WEO projections (last realised year = 2024).

confidence cueThe result is useful, but not decisive. Treat it as a clue, not a settled conclusion.

policy briefMixed or noisy

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

The evidence is suggestive but not decisive. Level-recovery legs hold (2025 = +3.0%, 2026 = +6.4%) but ARG cumulative 2024-2026 log-deviation +8.1% does NOT exceed peer mean +14.1% - recovery, but not faster than the LatAm peer pool.

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 7 country or place units from 2015 to 2027, using a synth did design.

what was measured
What changed
  • Milei era indicator
  • Quarters since inauguration
What we checked
  • Real income quarterly
  • Real income deviation from trend
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

No evidence packet has been generated yet.

Results

engine/runs/milei_shock_therapy_output_recovery_trajectory
Loading chart…

Who has skin in the game — schools predicting on this

2 schools list this hypothesis as a test of their position. The chips below are school-level scoreboard outcomes, not a second hypothesis verdict.

hypothesis verdict vs scoreboard outcome

The banner verdict judges this hypothesis as written. The scoreboard asks whether each school's polarity-corrected prediction was right. Raw status is not a school win: SUPPORTED supports schools that needed SUPPORTED, but refutes schools that needed REFUTED.

Austrian Business Cycle Theory (ABCT)wins if SUPPORTEDpartial/no-callChicago School Monetarismwins if SUPPORTEDpartial/no-call

Pre-registration

pre-registered
first-spec commit 4c8ce8e · 2026-07-18T22:11:21Z

Argentine quarterly real GDP, having contracted in 2024H1 under the Milei stabilisation shock, recovers along a trajectory that by Q4 2025 closes at least 50% of the peak-to-trough output gap observed during the shock's worst quarter, and by Q4 2026 returns to or exceeds the pre-Milei (Q4 2023) real-GDP level. The trajectory is faster than the LatAm peer donor pool (Brazil, Chile, Colombia, Peru, Mexico, Uruguay) weighted average for an equivalent shock absorbed without reform content. The hypothesis is consistent with the Austrian / monetarist reading that post- stabilisation recoveries are front-loaded once the fiscal- dominance distortion is removed and with empirical findings on sharp-stabilisation recovery patterns (e.g., Peru 1990-1993, Bolivia 1985-1989).

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

PRIMARY (dispositive, three legs — all must hold for SUPPORTED): (A) ARG annual real-GDP log-level at 2025 minus log-level at 2023 is at least -0.04 (i.e. by 2025 the contraction has not exceeded ~4 log points relative to the pre-Milei baseline). (B) ARG annual real-GDP log-level at 2026 is at least the 2023 baseline (level recovery). (C) ARG cumulative log-deviation across 2024-2026 EXCEEDS the unweighted mean cumulative log-deviation of the LatAm peer pool (BRA, CHL, COL, MEX, PER, URY) over the same window — the spec's "faster than peers" claim. PARTIAL if A and B hold but not C, or if B holds but A misses by less than half. REFUTED if B fails. METHOD_VALID downgrade to inconclusive when >=2/3 of 2024-2026 are IMF WEO projections rather than realised outturns AND the headline result would be SUPPORTED on those projections. Translated from the spec's quarterly wording (2025Q4, 2026Q4) to annual under the spec.notes-authorised quarterly-to-annual fallback because quarterly real-GDP series for the LatAm peer pool are not on disk.

formal test & threshold
test:      annual_log_index_three_leg_with_peer_comparison
threshold: log(GDP_ARG_2025) - log(GDP_ARG_2023) >= -0.04 AND log(GDP_ARG_2026) - log(GDP_ARG_2023) >= 0.00 AND sum_y in {2024,2025,2026} log_idx_ARG_y > mean over peers in {BRA,CHL,COL,MEX,PER,URY} of sum_y log_idx_peer_y

Method

Template
synth_did
Clustering
country
Sample
7 countries · 2015 – 2027
Evidence type
causal

Primary specification: synthetic DiD (Arkhangelsky, Athey, Hirshberg, Imbens, Wager 2021) with ARG as treated unit and LatAm peer pool as donor set. Pre-period 2015-2023 defines the synthetic counterfactual weights. Treatment effect is the cumulative and quarterly gap between observed ARG real GDP and synthetic ARG from 2023Q4 onwards. Secondary specification: Callaway-Sant'Anna DiD treating the LatAm peer pool as never-treated control and ARG as a single treated unit at 2023Q4, with dynamic event-time coefficients through Q+12. Third specification: local projections with ARG indicator interacted with quarters-since-inauguration for a robustness check.

Data

VariableSourceTransform
real_gdp_quarterly
outcome
imf:NGDP_RPCHtier 2
world_bank_wdi:NY.GDP.MKTP.KDtier 2
log_level_quarterly
real_gdp_deviation_from_trend
outcome
derived from real_gdp_quarterlylog_deviation_from_2015_2019_linear_trend
milei_era_indicator
treatment
constructed:binary = 1 for ARG from 2023Q4 onward; 0 otherwisetier 5
binary
quarters_since_inauguration
treatment
constructed:integer count of quarters since 2023Q4tier 5
integer_count
commodity_terms_of_trade
control
world_bank_wdi:TT.PRI.MRCH.XD.WDtier 2
quarterly_log_change
global_financial_conditions
control
fred:NFCItier 1
level
us_policy_rate
control
fred:FEDFUNDStier 1
level
pre_treatment_output_gap
control
oecd:OutputGaptier 2
level_at_2023Q3

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Milei shock-therapy output recovery trajectory

Verdict: partial - Level-recovery legs hold (2025 = +3.0%, 2026 = +6.4%) but ARG cumulative 2024-2026 log-deviation +8.1% does NOT exceed peer mean +14.1% — recovery, but not faster than the LatAm peer pool. Note: 2 of 3 post-period years (2025, 2026) are WEO projections (last realised year = 2024).

Summary

  • ARG real-GDP log-level index, 2024 = -1.3%, 2025 = +3.0%, 2026 = +6.4% (2023 = 0% baseline).
  • Spec leg A (2025 floor >= -4%): PASS.
  • Spec leg B (2026 level recovery >= 2023): PASS.
  • Spec leg C (ARG cum 2024-2026 > LatAm peer mean): ARG = +8.1%, peer mean = +14.1% — FAIL.
  • Peer-by-peer cumulative log-deviation 2024-2026: BRA=+16.5%, CHL=+14.6%, COL=+11.9%, MEX=+7.0%, PER=+19.8%, URY=+15.1%.

Method

Dispositive primary test in three legs translated from the spec's quarterly thresholds to annual under the spec.notes-authorised annual fallback (quarterly real-GDP series for the LatAm peer pool are not on disk; the spec explicitly anticipates this).

Annual real-GDP percent change (IMF NGDP_RPCH) is converted to a log-level index pinned to 0 at 2023 (the pre-Milei baseline year the spec uses) by cumulating log(1 + g_t/100). The three thresholds are then evaluated directly:

  1. ARG 2025 log-gap vs 2023 must be >= -0.04 (4 log points).
  2. ARG 2026 log-gap vs 2023 must be >= 0.00 (level recovery).
  3. ARG cumulative log-deviation 2024-2026 must EXCEED the unweighted mean of the LatAm peer pool over the same window.

Caveats

  • IMF WEO projection content. As of run date 2026-04-27, the last unambiguously realised year in NGDP_RPCH is 2024. Years 2025, 2026 are WEO projections, not outturns. The verdict is therefore the joint hypothesis (claim is correct AND IMF projects it correctly). When the realised-data window extends, re-run with the same script.
  • Annual fallback was used in lieu of quarterly GDP per spec.notes. Synthetic-DiD with permutation inference (the spec's primary estimator) is not run because the annual sample (one ARG observation per post-treatment year, six donors) is too thin for permutation p-values to be informative; the level-and-peer comparison above is the legitimate annualised analogue.

Data

  • imf:NGDP_RPCH (annual real-GDP % change, includes WEO projections)

Strongest opposing argument

Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.

Notes

Synthetic control donor pool choice: LatAm peers with comparable institutional-quality baseline and commodity- exposure profile. Brazil included despite macro-size differences because pre-2023 trajectory similarity is strong. If the quarterly GDP data for any peer country is thinner than Argentina's own, the panel reverts to annual GDP for pre-period weighting with quarterly GDP only for post-2023Q4 event window. Data-gated: national-accounts quarterly GDP for ARG and peer pool must land on disk by run date. INDEC (ARG), IBGE (BRA), INE CHL, DANE (COL), INEGI (MEX), INEI (PER), INE URY are the primary sources via OWID or IMF IFS mirror.

Authored framework. Read the transparency note.