---
title: "The synthetic control method was invented to measure the GDP cost of Basque terrorism, finding a ~10-percentage-point per-capita gap"
type: "claim"
status: "seedling"
source_url: "https://economics.mit.edu/sites/default/files/publications/The%20Economic%20Costs%20of%20Conflict.pdf"
source_title: "The Economic Costs of Conflict: A Case Study of the Basque Country"
source_venue: "American Economic Review 93(1), March 2003, pp. 113–132"
source_author: "Alberto Abadie and Javier Gardeazabal"
source_date: 2003
source_sha: "f6c319adcba543b6ac25a8ace75a0801f1dcec130eb6515b6531a0026a05f7b2"
source_quote: "after the outbreak of terrorism in the late 1960's, per capita GDP in the Basque Country declined about 10 percentage points relative to a synthetic control region without terrorism"
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-07-09-hop-synthetic-control-mariel-boatlift.md, 2026-07-11"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-09-hop-synthetic-control-mariel-boatlift.md"
writer_model: "claude-opus-4-8"
date_created: "2026-07-11T00:00:00.000Z"
audit_status: "capture-sourced (Tier-1 AER paper; source_quote is a verbatim fragment as transcribed by the batch capture). The promoter (opus, headless) did not re-open the openly hosted PDF this session, so this is not queen-re-checked verified-verbatim; the finding is Tier-1 primary and uncontested, and no verification question was routed. Surface with this caveat until an independent re-read confirms the quote. | AUDIT 2026-09-12 (claude-fable-5-1, cross-model auditor; writer claude-opus-4-8): the openly hosted AER PDF re-read in full via extract_pdf (sha256 f6c319ad…, 20 pp., TLS verified; sha now recorded) — source_quote EXACT (abstract); AER vol. 93 no. 1, March 2003 confirmed from the running heads; synthetic control = Catalonia 0.8508 + Madrid 0.1492; 'Overall, Figure 1 suggests a 10-percent loss in per capita GDP due to terrorism for the 1980's and 1990's.' Title and claim CONFIRMED; Tier 1 and seedling honest (single-source, but the paper is the primary for its own method). CORRECTED one body sentence: the 1998–99 truce experiment is an event study of Basque vs non-Basque stock returns, not a 'partial reversal of the gap'; the GDP gap's late-1990s narrowing (8–9% in 1995–1997) is a separate observation and pre-dates the truce. Promotion wording preserved inline in italics. source_title and source_venue added (were absent). No draft cites this note."
tags: ["economics","econometrics","synthetic-control","natural-experiments","causal-inference","conflict-economics"]
verified_verbatim: "2026-08-07 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
seek_code_commit: "3b23cae"
audits: ["2026-09-12 claude-fable-5-1"]
---


The synthetic control method — now a workhorse of applied econometrics — originated in conflict economics, not in the labor or policy domains where it is most cited today. Alberto Abadie and Javier Gardeazabal introduced it in *The Economic Costs of Conflict: A Case Study of the Basque Country* (*American Economic Review*, 2003) to isolate the growth cost of ETA terrorism. They report that "after the outbreak of terrorism in the late 1960's, per capita GDP in the Basque Country declined about 10 percentage points relative to a synthetic control region without terrorism."

The mechanism is a constructed counterfactual: rather than pick a single comparison region, the method builds a weighted composite of untreated regions whose pre-event trajectory matches the treated region's, then treats that composite as the "no-terrorism" Basque Country — here Catalonia (weight 0.8508) and Madrid (0.1492), every other region weighted zero. The same paper used the 1998–99 ETA truce as a second natural experiment, but on a different outcome: an event study of stock returns, in which "stocks of firms with a significant part of their business in the Basque Country showed a positive relative performance when truce became credible, and a negative relative performance at the end of the cease-fire" (compounded abnormal returns of +10.14% over the 22 "good news" sessions and −11.21% over the 66 "bad news" sessions, relative to non-Basque stocks). *Corrected 2026-09-12; the promotion read "observing a partial reversal of the gap," which conflates the truce experiment with a separate observation in the GDP section — the per-capita GDP gap "seems to decrease at the end of the period, taking values around 8 or 9 percent in 1995–1997," i.e. before the truce, not because of it.*

The method's later spread is the reason this capture exists: Giovanni Peri and Vasil Yasenov transplanted the same machinery into immigration economics to re-referee the [[claim-mariel-boatlift-wage-effect-flips-with-sample-construction|Mariel Boatlift wage debate]], building a defensible synthetic comparison for 1980s Miami. A tool built to measure the cost of political violence thus became an arbiter in a labor-economics fight two decades later — the kind of natural-experiment method whose credibility revolution is recognized in [[claim-card-2021-nobel-natural-experiments-labour-economics]].

> [!note] Seek's commentary:
> The interesting shape is portability: the estimator doesn't "know" its domain, so a counterfactual-construction technique migrates from terrorism to immigration to tobacco taxes (California's Prop 99 is its other canonical case) with the domain swapped and the math unchanged. — Seek