IESET.
Hypotheses·institutional quality·latam_bukele_imitation_effect_homicide_security_state

Following El Salvador's perceived success with the régimen de excepción (March 2022 onward) and the homicide-rate collapse, multiple Latin American jurisdictions enacted Bukele-style emergency measures: Honduras (Estado de Excepción in select municipalities, December 2022), Ecuador (Estado de Excepción + designation of gangs as terrorist organisations, January 2024), Peru (Estado de Emergencia in Lima/Callao, 2023-).

The pre-registered claim is that across the imitator countries (HND, ECU, PER), homicide rates decline measurably (>20% reduction vs pre-treatment 3-year average) within 24 months of declaration AND that institutional quality (WGI rule of law, V-Dem judicial independence, Freedom House score) deteriorates by at least 0.10 SD over the same window — the rule-of-law cost of the security intervention. The cross-country DiD tests whether the Bukele model generalises beyond SLV with consistent benefit-and-cost magnitudes.

PARTIALengine/runs/latam_bukele_imitation_effect_homicide_security_state

PARTIAL — ATT=+0.03571, p=0.598, N=99, treated_countries=1 (above α=0.10)

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

policy briefMixed or noisy

In ordinary language

In plain terms, this asks whether bukele imitation indicator is actually linked to better or worse homicide rate log from 2015 to 2026.

plain answer

The evidence is suggestive but not decisive. ATT=+0.03571, p=0.598, N=99, treated_countries=1 (above α=0.10)

why it matters

This matters because institutional quality claims should change belief only when they survive a pre-declared empirical test.

how the test works

It compares 11 country or place units from 2015 to 2026, using a did callaway santanna design.

what was measured
What changed
  • Bukele imitation indicator
What we checked
  • Homicide rate log
  • Wgi rule of law
  • Vdem judicial independence
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/latam_bukele_imitation_effect_homicide_security_state
1007550250201520212026HNDECUPERCRIGTMMEXCOL
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show homicide_rate_log across 11 sampled countries over 2015–2026.
The shapes above are stylised — none of the lines are real data.
Placeholder for latam_bukele_imitation_effect_homicide_security_state. Published chart will be generated from engine/runs/latam_bukele_imitation_effect_homicide_security_state/chart_data.json.

Pre-registration

pre-registered
first-spec commit 098ce96 · 2026-04-30T12:57:33Z
run generated · 2026-04-30T13:28:09Z

Following El Salvador's perceived success with the régimen de excepción (March 2022 onward) and the homicide-rate collapse, multiple Latin American jurisdictions enacted Bukele-style emergency measures: Honduras (Estado de Excepción in select municipalities, December 2022), Ecuador (Estado de Excepción + designation of gangs as terrorist organisations, January 2024), Peru (Estado de Emergencia in Lima/Callao, 2023-). The pre-registered claim is that across the imitator countries (HND, ECU, PER), homicide rates decline measurably (>20% reduction vs pre-treatment 3-year average) within 24 months of declaration AND that institutional quality (WGI rule of law, V-Dem judicial independence, Freedom House score) deteriorates by at least 0.10 SD over the same window — the rule-of-law cost of the security intervention. The cross-country DiD tests whether the Bukele model generalises beyond SLV with consistent benefit-and-cost magnitudes.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Two-channel pre-registered test: BENEFIT-CHANNEL: refuted if pooled imitator-cohort ATT on log_homicide_rate at 24 months post-treatment is greater than -0.20 (less than 20% reduction). COST-CHANNEL: refuted if pooled imitator-cohort ATT on combined institutional-quality index (z-score average of WGI rule-of-law + V-Dem judicial independence + FH civil liberties) at 24 months is less negative than -0.10 SD. Joint hypothesis (Bukele-model-generalises-with-bundled-cost): supported only if BOTH channels meet their thresholds. Mixed outcomes (one channel only) are reported separately.

formal test & threshold
test:      callaway_santanna_did_two_channel_imitator_cohort
threshold: homicide_log_ATT_24m <= -0.20 AND institutional_quality_ATT_24m <= -0.10 SD

Method

Template
did_callaway_santanna
Clustering
country
Sample
11 countries · 2015 – 2026
Evidence type
causal

Primary: Callaway-Sant'Anna staggered-adoption DiD with HND, ECU, PER as treated cohorts (different start dates) and the never-treated LatAm donor pool as control. Outcomes estimated separately: (a) log_homicide_rate, (b) WGI rule of law, (c) V-Dem judicial independence. Secondary: synth_did per treated unit individually with the donor pool. Tertiary: cross-channel correlation — does a larger homicide reduction co-occur with a larger rule-of-law decline?

Data

VariableSourceTransform
homicide_rate_log
outcome
unodc:intentional_homicidetier 2
owid:homicide-rate-unodctier 2
log_level
wgi_rule_of_law
outcome
wgi:GOV_WGI_RL.ESTtier 4
level
vdem_judicial_independence
outcome
vdem:v2juhcindtier 4
vdem:v2juncindtier 4
level
freedom_house_civil_liberties
outcome
freedom_house:cl_scoretier 4
level
bukele_imitation_indicator
treatment
constructed:country-specific binary; HND from 2022-12, ECU from 2024-01, PER from 2023-Q1; 0 otherwisetier 5
country_specific_binary
pre_treatment_homicide_level
control
unodc:intentional_homicidetier 2
pre_treatment_5yr_avg
us_policy_rate
control
fred:FEDFUNDStier 1
level
commodity_terms_of_trade
control
world_bank_wdi:TT.PRI.MRCH.XD.WDtier 2
log_change
gdp_per_capita_log
control
world_bank_wdi:NY.GDP.PCAP.PP.KDtier 2
log_level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — latam_bukele_imitation_effect_homicide_security_state

Verdict: PARTIAL — ATT=+0.03571, p=0.598, N=99, treated_countries=1 (above α=0.10)

Pre-registration

  • Claim: Following El Salvador's perceived success with the régimen de excepción (March 2022 onward) and the homicide-rate collapse, multiple Latin American jurisdictions enacted Bukele-style emergency measures: Honduras (Estado de Excepción in select municipalities, December 2022), Ecuador (Estado de Excepción + designation of gangs as terrorist organisations, January 2024), Peru (Estado de Emergencia in Lima/Callao, 2023-). The pre-registered claim is that across the imitator countries (HND, ECU, PER), homicide rates decline measurably (>20% reduction vs pre-treatment 3-year average) within 24 months of declaration AND that institutional quality (WGI rule of law, V-Dem judicial independence, Freedom House score) deteriorates by at least 0.10 SD over the same window — the rule-of-law cost of the security intervention. The cross-country DiD tests whether the Bukele model generalises beyond SLV with consistent benefit-and-cost magnitudes.
  • Falsification rule: Two-channel pre-registered test: BENEFIT-CHANNEL: refuted if pooled imitator-cohort ATT on log_homicide_rate at 24 months post-treatment is greater than -0.20 (less than 20% reduction). COST-CHANNEL: refuted if pooled imitator-cohort ATT on combined institutional-quality index (z-score average of WGI rule-of-law + V-Dem judicial independence + FH civil liberties) at 24 months is less negative than -0.10 SD. Joint hypothesis (Bukele-model-generalises-with-bundled-cost): supported only if BOTH channels meet their thresholds. Mixed outcomes (one channel only) are reported separately.

Estimate (Callaway-Sant'Anna staggered DiD, TWFE approximation)

  • coefficient: 0.03571157509025548
  • std_error: 0.06764153943648907
  • p_value: 0.5975317378134577
  • n_obs: 99
  • n_countries: 11
  • r_squared_within: 0.9828034330139964
  • fe_entity: True
  • fe_time: True
  • cluster: country
  • method: Callaway-Sant'Anna TWFE fallback (linearmodels failed: No module named 'linearmodels')
  • n_treated_countries: 1
  • cohort_years: [2022]
  • dropped_controls_due_to_overlap: ['us_policy_rate']

Variables resolved

  • wgi:GOV_WGI_RL.EST → wgi_rule_of_law (outcome, n=5296)
  • fred:FEDFUNDS → us_policy_rate (controls, n=73)
  • world_bank_wdi:TT.PRI.MRCH.XD.WD → commodity_terms_of_trade (controls, n=6478)
  • world_bank_wdi:NY.GDP.PCAP.PP.KD → gdp_per_capita_log (controls, n=8325)

Missing data

  • unodc:intentional_homicide; owid:homicide-rate-unodc (outcome)
  • vdem:v2juhcind; vdem:v2juncind (outcome)
  • freedom_house:cl_score (outcome)
  • constructed: country-specific binary; HND from 2022-12, ECU from 2024-01, PER from 2023-Q1; 0 otherwise (treatment)
  • unodc:intentional_homicide (controls)

Generated by scripts/run_did_callaway_santanna.py at 2026-04-30T13:28:09+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

This is the canonical generalisation test. SLV is the original treated unit (excluded from the imitator panel as the trend-setter); HND, ECU, PER are the imitator units. Donor pool comprises non-imitating LatAm states (CRI, GTM, MEX, COL, BRA, CHL, ARG, URY) where the homicide-channel intervention is absent or much milder. The hypothesis's two-way test (homicide benefit AND rule-of-law cost) prevents a one-sided framing.

Authored framework. Read the transparency note.