Skip to main content
v2026.11,986 entries · CC-BY 4.0

Global Trade Alert (GTA): What It Records, Covers and Excludes

Global Trade Alert records government trade measures since 2008: inclusion tests, verification, coverage gaps, 2026 taxonomy change and CC BY-NC licence.

Written and maintained by CASRAI Editorial Board

Published

On this page

Global Trade Alert (GTA) is a public database of government measures that change how foreign firms can sell goods, supply services or invest in a market. It is run as an SGEPT initiative with support from the University of St. Gallen and the Schmidheiny Foundation, has recorded measures continuously since November 2008, and links every entry to the official document behind it. Researchers use it as a source of policy data on tariffs, subsidies, export restrictions and other interventions, alongside trade-flow sources such as UN Comtrade.

What is Global Trade Alert?

GTA describes itself as a public newsroom for trade and industrial policy. Where a trade-flow database tells you how much was traded, GTA records what governments did: a new import duty, an export ban, a subsidy scheme, a local-content rule, an investment screening step. Each record points back to the legal or official source.

Do not confuse it with other products that share the words. Law and consulting firms publish their own trade-advisory alerts, which are newsletters rather than a research dataset. The database discussed here is the one at globaltradealert.org, and its entries are what economists cite when they write that a country imposed a given number of trade-distorting measures in a period.

Who runs it and how far back does it go?

The coverage page names the St. Gallen-based SGEPT initiative as the operator, supported by the University of St. Gallen (UNISG) and the Schmidheiny Foundation. Coverage begins in November 2008 and has been maintained continuously since then, which is why the database is used for long-run counts of protectionist and liberalising measures rather than only for current news.

What counts as an intervention in GTA?

GTA does not record every government action. A measure enters the database only when it passes a relative-treatment test: it must change market access differently for foreign and domestic firms. A rule that treats everyone equally is not recorded. The coverage page lists seven instrument families: import barriers, export barriers, trade defence, subsidies, localisation requirements, FDI restrictions and export promotion. Each measure is also coded to six-digit Harmonised System products (across more than 5,000 tariff lines) and to three-digit CPC sectors.

Every record carries a direction of change on a three-colour scale: green for liberalising and non-discriminatory, amber for likely discriminatory, and red for almost certainly discriminatory. Counts of “harmful” measures in published analyses depend on which of these colours the author includes, so state the filter when you reproduce a number.

Inclusion and exclusion tests

Test Recorded Not recorded
Treatment of foreign firms Measures that change market access differently for foreign and domestic firms Measures that treat all firms equally
Regulatory scope The seven instrument families listed above SPS and TBT measures, UN Security Council sanctions, CITES measures
Origin of the rule Unilateral government measures Negotiated commitments such as bilateral or trade-agreement terms
Breadth Measures aimed at particular products, sectors or firms Economy-wide measures such as broad tax reform or general infrastructure spending
Subsidy size Financial support of USD 10 million or more (USD 100 million or more for SME-only schemes; for loans, the threshold applies to loan volume) Support below those thresholds
Status Measures with a credible path to implementation Announcements without one, and draft regulations still subject to approval

The practical consequence is that GTA is not a census of regulation. The exclusion of SPS and TBT measures means food-safety and technical-standard rules sit elsewhere: the coverage page itself points to the WTO ePing and I-TIP systems for those notifications. Because negotiated trade-agreement terms are also excluded, the page directs users to the World Bank Deep Trade Agreements database for that content (about 400 preferential agreements).

How are GTA entries verified?

Automated collection sweeps official gazettes, ministry sites and regulatory registers, but everything it finds is treated only as a lead. GTA describes a “six-eyes” principle across three desks: an analyst drafts the entry, a senior analyst checks it against the same source, and a publication desk publishes it, so that no entry reaches the public database on one person’s judgement.

More than 97% of entries cite an official government source. Where official publication is slow or scattered, analysts also use specialist legal and trade services. Where no official statement exists at all, a record requires at least three reputable, independent sources in agreement. Output is 20 to 30 new and updated records per business day.

A peer-reviewed description of the same pipeline appears in the COVID-19 trade-policy data paper by Evenett and co-authors, produced with the European University Institute and the World Bank. Its authors describe multilingual keyword searches over government agency sites, international organisation sites, online media and non-state sources such as law firms and research institutes, with every keyword hit becoming a lead. An analyst classifies each lead, a second analyst confirms, amends or rejects the classification, an official source is sought, a GTA analyst creates the entry using HS and ISIC codes, and a senior team member checks every entry before inclusion.

Which countries does GTA cover?

Sixty-three jurisdictions are on a standing basis, each with a named analyst: 36 on the priority tier and 27 on regular monitoring. The priority tier holds all G20 members, the European Union and the six largest Southeast Asian economies. Measures adopted at EU level bind all 27 member states.

Region What the coverage page says
G20, EU, six largest Southeast Asian economies Priority tier
South America Only Argentina, Brazil, Chile and Peru; Colombia and the smaller economies are not covered
Africa Six economies: South Africa, Nigeria, Kenya, Ghana, Egypt and Morocco; the rest of sub-Saharan Africa is not covered

For a study of a country outside the standing list, an absence of records is therefore not evidence that the government did nothing. Check the coverage page before interpreting zero counts.

The 2026 taxonomy change and what it does to time series

On 4 March 2026 GTA announced a refreshed intervention taxonomy. It separates price-related border instruments from standard tariffs and splits the broad residual category of import-related non-tariff measures (NTMs). Ten new intervention types were introduced: six under MAST chapters E, F and P, and four on market access under chapters C, H, J and K. Of 2,329 interventions examined, 2,044 were reclassified, covering November 2008 onward. The previous types affected were import tariff, export tax and import-related non-tariff measures, not elsewhere specified. Reclassification was done in two tranches, both relying entirely on expert validation: MAST chapters E, F and P in summer 2025, and import-related NTMs on 4 March 2026.

Group New intervention types
Price instruments Import price benchmark; Minimum import price; Other import charges; Voluntary export-restraint arrangements; Voluntary export-price restraints; Export price benchmark
Import access instruments Port restriction; Selective import channel restriction; Distribution restriction; Post-sales service restriction

The grouping in this table is an editorial arrangement of GTA’s own list, following the blog’s description of price instruments and import-access instruments. GTA points to the updated GTA Handbook on its methodology page for full definitions and coding guidance.

Editorial interpretation for data users: any series, replication package or chart built from an extraction made before March 2026 may use the older, coarser types. If you combine an old download with a new one, recode before comparing counts by intervention type, and record the extraction date alongside the data.

Worked example: why tracker counts differ

The 2022 data paper offers a rare side-by-side comparison of how many measures different trackers recorded for the same crisis. For 2 January to 9 October 2020, the GTA data set held 701 measures across 135 customs territories. The authors then removed measures announced in 2019, leaving 645. The arithmetic is 701 − 645 = 56 measures removed, and 645 ÷ 701 = 0.920, so 92.0% of the original set remained.

As of 9 October 2020 the International Trade Centre (ITC) repository listed 338 measures from 145 countries, customs unions and autonomous territories, and the WTO Secretariat’s verified list had 243 measures from 85 members and observers. The ratio of GTA to ITC is 701 ÷ 338 = 2.07, about twice as many measures, and GTA to the WTO verified list is 701 ÷ 243 = 2.88. The difference reflects scope and method, not an error: the paper notes that WTO notifications include standards and subsidy measures, and GTA records trade policy changes on medical goods and food products including export restrictions and import liberalisation, found through its own lead-and-validate process. The reading guide is to compare counts only after checking each tracker’s inclusion rules, period and unit of counting. The data table below sets the figures side by side.

How to use GTA in research

GTA versus UN Comtrade and other sources

GTA records policy, while UN Comtrade records flows. A typical design uses GTA to date and classify an intervention, then trade-flow data to measure what happened to imports or exports at the product level; the quantitative methods are covered in the trade policy analysis guide. GTA’s own coverage page recommends complementary sources: WTO ePing and I-TIP for technical and sanitary notifications (ePing covers notifications back to 1995), the IMF AREAER for exchange-rate regimes and capital-account rules, ITC Trade Map or Trade Data Monitor for trade flows (Trade Map covers more than 220 countries and territories), and Oxford Economics for country-risk scores and forecasts.

Treating GTA records as secondary data

GTA entries are compiled by analysts from official documents, so they are secondary data in the sense used in the secondary data analysis guide. Keep the measure identifier, the HS products affected, the direction colour and the date you extracted it. For regressions that use measure counts as a treatment variable, the panel data analysis guide covers fixed and random effects, and the extraction-date point above matters because the classification of past measures can be revised.

Licence and how to cite

Unless otherwise stated, GTA content is licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0), so commercial reuse is not covered by the default licence; see the CC BY versus CC BY-NC comparison for what the NonCommercial condition means. The coverage page gives no citation format of its own. Following the data citation principle, cite the database with its operator, the URL, the extraction date and the filters you applied. To cite the paper that documents the COVID-19 data set, an APA 7th-edition reference reads:

Evenett, S., Fiorini, M., Fritz, J., Hoekman, B., Lukaszuk, P., Rocha, N., Ruta, M., Santi, F., & Shingal, A. (2022). Trade policy responses to the COVID-19 pandemic crisis: Evidence from a new data set. The World Economy, 45(2), 342–364. https://doi.org/10.1111/twec.13119

Assumptions and common mistakes

  • Reading zero as no policy. The standing list has 63 jurisdictions, and South America and Africa are thinly covered.
  • Treating counts as intensity. A count of measures says nothing about trade value affected; one measure can cover a single tariff line or thousands.
  • Expecting everything. SPS and TBT measures, sanctions, CITES, negotiated commitments, economy-wide measures and small subsidies are excluded by design.
  • Mixing taxonomy vintages. Data downloaded before and after 4 March 2026 can classify the same intervention differently.
  • Ignoring the licence. The default licence is non-commercial.

Sources

Data: Trade-policy measures counted by COVID-19 trackers, 2 January to 9 October 2020

Counts of COVID-19-related trade measures and the jurisdictions they cover, as reported for three trackers (Global Trade Alert, ITC repository, WTO Secretariat verified list) in the 2022 data paper by Evenett and co-authors in The World Economy. Compiled into one table from the paper's text.

Period covered: 2020-01-02/2020-10-09.

Trade-policy measures counted by COVID-19 trackers, 2 January to 9 October 2020
Tracker or versionMeasures (count)Jurisdictions covered (count)
Global Trade Alert (full 2020 set)701135
Global Trade Alert (after removing 2019 announcements)645135
ITC repository (as of 9 October 2020)338145
WTO Secretariat verified list24385

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Global Trade Alert (GTA): What It Records, Covers and Excludes

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →