Pre-registration
Peru's 1990-1995 Fujimori shock-therapy package (price liberalisation, fiscal stabilisation under the August 1990 "Fujishock", Brady-style external debt restructuring 1996-1997, large-scale privatisation of SOEs, central-bank independence under the 1993 constitution, and trade liberalisation) produced a structural break in inflation and real-GDP per capita relative to Peru's 1985-1990 hyperinflation trajectory and relative to a Latin American peer pool that did not adopt comparable packages on the same timeline. The pre-registered claim is (a) inflation collapsed from >7000% (1990) to single digits by 1999, and (b) cumulative log real-GDP per capita 1990-2000 exceeds the Latin American median by at least 0.10 log-points. The mechanism is a fiscal+monetary regime change that broke the inflation-tax equilibrium and restored property-rights credibility for private investment.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
Not supported if any of (a) inflation does not break below 20% by 1995 and below 10% by 1999 (hyperinflation reversal channel), OR (b) cumulative log_gdp_pc_constant 1990-2000 (Peru minus synthetic peer) is < 0.10, OR (c) gross_capital_formation share of GDP does not rise at least 4 percentage points 1990-2000 relative to peer median.
formal test & threshold
test: synth_did_plus_inflation_break_plus_investment_share threshold: cumulative_log_gdp_pc(PER, 1990-2000) - cumulative_log_gdp_pc(synth, 1990-2000) >= 0.10 AND cpi_inflation_yoy(PER, 1999) < 10 AND delta_capital_formation_share(PER, 1990-2000) - delta_capital_formation_share(peer_median, 1990-2000) >= 4
Method
- Template
synth_did- Clustering
country- Sample
- 10 countries · 1985 – 2005
- Evidence type
- causal
Primary: synth_did with PER treated from 1990, Latin American peers as donor pool. Secondary: event-study on log GDP-pc relative to 1989 baseline. Tertiary: cointegration between credibility proxy (inflation level) and investment share to test whether the regime change had a credible-commitment signature.
Data
| Variable | Source | Transform |
|---|---|---|
log_gdp_pc_constant outcome | world_bank_wdi:NY.GDP.PCAP.KDtier 2 | log |
log_gdp_pc_ppp outcome | world_bank_wdi:NY.GDP.PCAP.PP.KDtier 2 | log |
cpi_inflation_yoy outcome | world_bank_wdi:FP.CPI.TOTL.ZGtier 2 | level |
gross_capital_formation_share_gdp outcome | world_bank_wdi:NE.GDI.TOTL.ZStier 2 | level |
fujimori_shock_indicator treatment | constructed:binary = 1 for PER from 1990 onwardtier 5 | binary |
terms_of_trade control | world_bank_wdi:TT.PRI.MRCH.XD.WDtier 2 | level |
copper_price control | fred:PCOPPUSDMtier 1 | log_level |
us_policy_rate control | fred:FEDFUNDStier 1 | level |
wgi_government_effectiveness control | wgi:GOV_WGI_GE.ESTtier 4 | level |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Result card — peru_fujimori_shock_therapy_1990_2000
Verdict: PARTIAL — mean_gap=-0.7927, |gap|/pre_sd=7.8, p_perm=0.4; claim direction ambiguous
Pre-registration
- Claim: Peru's 1990-1995 Fujimori shock-therapy package (price liberalisation, fiscal stabilisation under the August 1990 "Fujishock", Brady-style external debt restructuring 1996-1997, large-scale privatisation of SOEs, central-bank independence under the 1993 constitution, and trade liberalisation) produced a structural break in inflation and real-GDP per capita relative to Peru's 1985-1990 hyperinflation trajectory and relative to a Latin American peer pool that did not adopt comparable packages on the same timeline. The pre-registered claim is (a) inflation collapsed from >7000% (1990) to single digits by 1999, and (b) cumulative log real-GDP per capita 1990-2000 exceeds the Latin American median by at least 0.10 log-points. The mechanism is a fiscal+monetary regime change that broke the inflation-tax equilibrium and restored property-rights credibility for private investment.
- Falsification rule: Not supported if any of (a) inflation does not break below 20% by 1995 and below 10% by 1999 (hyperinflation reversal channel), OR (b) cumulative log_gdp_pc_constant 1990-2000 (Peru minus synthetic peer) is < 0.10, OR (c) gross_capital_formation share of GDP does not rise at least 4 percentage points 1990-2000 relative to peer median.
Synthetic-control estimate
- shape: synth_did
- treated_country: PER
- event_year: 1990
- n_donors: 9
- donor_weights (top): {'ARG': 0.6298, 'VEN': 0.3702, 'BRA': 0.0, 'CHL': 0.0, 'COL': 0.0}
- pre_rmse: 0.6828277998896773
- pre_period_sd: 0.10175383637722524
- mean_post_gap: -0.7927207350885753
- end_period_gap: -0.6740690895366956
- post_period_years: [1990, 2005]
- placebo_p_value: 0.4
- n_placebos: 9
- method: synthetic-control via NNLS, permutation inference
Variables resolved
world_bank_wdi:NY.GDP.PCAP.KD→ log_gdp_pc_constant (outcome, n=14131)world_bank_wdi:NY.GDP.PCAP.PP.KD→ log_gdp_pc_ppp (outcome, n=8360)world_bank_wdi:FP.CPI.TOTL.ZG→ cpi_inflation_yoy (outcome, n=9066)world_bank_wgi:GOV_WGI_GE.EST→ wgi_government_effectiveness (controls, n=5194)
Generated by scripts/run_synth_did.py at 2026-04-30T10:15:31+00:00
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 already on disk via WDI/WGI/FRED. Excludes the political-rights cost of the 1992 autogolpe; spec is narrowly economic. Channel- decomposition extension could test whether the property-rights channel (privatisation revenue 1991-2000) explains the investment- share rise.