Pre-registration
El Salvador's FDI inflow, real-GDP growth, tourism arrivals, and business-formation rate accelerated under the Bukele era (2019-2024) relative to a Central American peer-country donor pool (Honduras, Guatemala, Nicaragua, Costa Rica, Panama, Dominican Republic). The pre-registered claim is that under a synthetic-control design, Salvadoran cumulative real-GDP growth 2019Q2-2024Q4 and cumulative FDI-inflow-share-of-GDP exceed the synthetic counterfactual with permutation inference at p < 0.10, and that tourism arrivals growth 2022-2024 (once post-COVID recovery paths are netted out) is faster than the peer-pool weighted average. The intended mechanism is a combination of (a) the homicide-rate collapse reducing the country-risk premium and security-driven location avoidance, (b) the 2023 tech-incentive law and tax reform lowering effective entry costs, and (c) the Bukele-brand publicity effect (conferences, Bitcoin-tourism, 'CECOT-tour' narrative). The null counter-claim is that the Bukele era produced no material divergence from the Central American growth path once commodity-cycle and US-demand controls are included.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
Not supported if any of: (a) synth_did estimated CATT on log(real_gdp) over 2019Q2-2024Q4 is not positive at p_perm < 0.10 relative to the Central American donor pool, OR (b) CATT on fdi_inflow_share_gdp is not positive, OR (c) CATT on tourism_arrivals (post-2022 recovery window) is not positive, OR (d) the channel-decomposition test attributes less than 30% of the GDP gap to the homicide-rate mechanism (i.e., the security-channel story fails and the observed acceleration must be explained by other factors including remittances, commodity cycle, or post-COVID base effects), OR (e) the SLV real-GDP growth is statistically indistinguishable from the Central American peer-group weighted average.
formal test & threshold
test: synth_did_bundle_plus_channel_decomposition threshold: CATT_2019Q2_2024Q4(log_real_gdp) > 0 at p_perm < 0.10 AND CATT_2019Q2_2024Q4(fdi_share_gdp) > 0 at p_perm < 0.10 AND CATT_2022_2024(tourism_arrivals) > 0 AND homicide_channel_share_of_gdp_gap >= 0.30
Method
- Template
synth_did- Clustering
country- Sample
- 7 countries · 2010 – 2025
- Evidence type
- causal
Primary: synth_did with SLV treated from 2019Q2, Central American peers as donor pool. Secondary: Callaway-Sant'Anna DiD. Tertiary: local projections with SLV-indicator interacted with quarters-since-inauguration for dynamic effect. Channel decomposition: regress the post-treatment outcome gap on homicide-rate log (as mediator) to estimate the share of the GDP/FDI effect attributable to the security channel.
Data
| Variable | Source | Transform |
|---|---|---|
real_gdp outcome | world_bank_wdi:NY.GDP.MKTP.KDtier 2 sv_bcr:PIBtier 2 | log_level_cumulative |
fdi_inflow_share_gdp outcome | world_bank_wdi:BX.KLT.DINV.WD.GD.ZStier 2 unctad:Worldtier 2 | level |
tourism_arrivals outcome | world_bank_wdi:ST.INT.ARVLtier 2 sv_mitur:Estadtier 3 | log_level |
new_business_registration outcome | world_bank_enterprise:Businesstier 2 sv_cnr:Registrotier 2 | log_level |
bukele_era_indicator treatment | constructed:binary = 1 for SLV from 2019-06-01 onwardtier 5 | binary |
homicide_rate_log treatment | unodc:intentional_homicidetier 2 | log_level |
commodity_terms_of_trade control | world_bank_wdi:TT.PRI.MRCH.XD.WDtier 2 | level |
us_gdp_growth control | fred:GDPC1tier 1 | yoy_growth |
us_policy_rate control | fred:FEDFUNDStier 1 | level |
oil_price control | fred:DCOILBRENTEUtier 1 | log_level |
remittance_share_gdp control | world_bank_wdi:BX.TRF.PWKR.DT.GD.ZStier 2 | level |
wgi_rule_of_law control | wgi:GOV_WGI_RL.ESTtier 4 | level |
● ready · ● pending · ● reconstruct-needed
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 on UNCTAD FDI inflows (annual), BCR quarterly GDP, MITUR tourism arrivals, and CNR new-business-registration series. The channel-decomposition test (security mediator) is the load-bearing falsification criterion — a growth acceleration without the security-channel signal would indicate the acceleration is driven by factors the movement doctrine does not claim credit for.