Econometric Estimation
GTA's product-level scope (HS-6, the six-digit customs product code) and the exact date of each recorded measure let a policy event enter an estimating equation as a treatment or regressor: a gravity model, a difference-in-differences panel, or a structural model of misallocation. GTA supplies the dated, scoped policy event; trade, FDI, or productivity data supply the outcome.
Descriptive Evidence
Here GTA supplies the counts, shares, and time trends that document a phenomenon without entering an estimating equation: a chart or a summary statistic, the fastest use and the most common first reach for the data.
- Sheng, Zhu and Zhu (2026), Global Resurgence of Industrial Policy: Evidence from 2023 GTA-NIPO Data, Economy of Regions. Documents 2,580 industrial-policy measures recorded in 2023 by instrument and region, using GTA-NIPO, the joint GTA/IMF database.
- Airaudo, de Soyres, Gaillard and Santacreu (2025), Recent Evolutions in the Global Trade System: From Integration to Strategic Realignment, Federal Reserve Bank of St. Louis Working Paper 2025-027. Draws on GTA data to document the rise of industrial policy and discriminatory measures across the trading system.
Text as Data
The most methodologically ambitious use treats GTA as a corpus rather than a finished variable. Each entry carries a human-written description of the measure; a classifier trained on a hand-labelled subset is then deployed across the full database to produce a new count at the country-sector-year level.
- Juhász, Lane, Oehlsen and Perez (2025), Measuring Industrial Policy: A Text-Based Approach, NBER Working Paper 33895. A supervised classifier trained on annotated GTA entries identifies industrial policies across the full database, 2010–2022.
- Goldberg, Juhász, Lane, Lo Forte and Thurk (2024), Industrial Policy in the Global Semiconductor Sector, NBER Working Paper 32651. Text analysis of GTA descriptions isolates semiconductor industrial policies, then combines them with a structural model of the industry.
Production-Network Exposure
When a measure on one product matters for sectors upstream or downstream, GTA records are joined to input-output tables so policy can be traced through supply chains. GTA supplies the located, dated measure; the input-output table supplies the linkages.
Real-Time Indices and Nowcasting
GTA updates monthly as measures are announced, so its flow of new measures can drive a high-frequency index tracking policy intensity close to real time.
Benchmarking and Validation
Because GTA is the established public inventory of commercial-policy measures, researchers building alternative collections measure themselves against it to show added coverage or confirm completeness.
Institutional Monitoring
International organisations use GTA for situational monitoring: a recurring account of what governments are doing, tabulated for a professional audience.
