Skip to main content

GTAP Data Bases: GTAP 8 Data Base Issues

If you have encountered an issue with the GTAP 8 Data Base which you would like addressed, please email a detailed account to the Data Team. Below is an ongoing list of issues reported regarding GTAP 8.


Issues addressed in GTAP 8.1 Data Base (February 2013)

• Tariff Data fix: Firstly, CEPII revised the tariff estimates, by fixing widespread discrepancies in the unit values. Major changes/fixes have been found for tariffs imposed by China, Korea and to some extent, Switzerland. Secondly, small intra-EU tariffs that existed in GTAP 8 data Base, mainly because of non-zero tariffs imposed by overseas territories of France on its EU partners, were removed. Thirdly, we fixed a mix-up between Senegal and Singapore in GTAP 8 Data Base, because of faulty set definition in the tariff data pre-processing. Fourthly, the tariffs have been capped at 1000%, consistent with our practice in the past releases.


• OECD Domestic Support data for non-OECD members: With inputs from Joanna Komarowska and Hans Jensen, we incorporated this data for both 2004 and 2007. The countries included here are Brazil, China, Indonesia, Russia, Ukraine and South Africa. Secondly, the new headers introduced in GTAP 8 Data Base (PYRT, PAYD) on detailed domestic support information will no more be included in GTAP 8.1 Data Base. The reason is incompatibility of these values with the numbers in the data base. We shall provide them as part of GTAP Data Base documentation.


• Energy tax fixes for Australia: In GTAP 8 Data Base, Australian taxes on energy usage were wiped out completely. We have fixed this in the current release.


• Factor Tax fixes: A few minor bug-fixes have been incorporated in the treatment of income and factor taxes.


• I-O tables: Five new IO tables have been added (Benin contributed by Badri Narayanan, Burkina Faso, Guinea and Togo contributed by Lacina Balma and Rwanda contributed by Zekarias Hussein), bringing the tally of GTAP regions up to 134. Several I-O tables have also been updated/improved: Brazil (Joaquim Bento Ferreira-Filho), Belarus (Csilla Lakatos), Colombia (Gabriel Piraquive Galeano), Japan (Suzumu Suzuki), Korea (Jong-Hwan Ko), Nigeria (Khalid Siddig), Singapore (Nhi Tran), Taiwan (Hsing-Chun Lin), Malawi, Mozambique, Tanzania and Zambia (Zekarias Hussein).


• Minor changes to the data file: Long-name labels of a few headers on tax revenue and subsidy expenditure have been revised to improve clarity. For example, FBEP is labeled as Gross factor-based subsidies.


• Minor changes in Flexagg: The names of a few headers in the sets file in Flexagg output (DH1, DH2, etc.) have been changed to match those in GTAPAgg (OREG, OSEC, etc.).

• September 26, 2012: GTAP Data Base 8a release was meant to solve the issue of sparse storage, by including all input files in full storage format. We found that the data packages still generate the output files in sparse storage format, because the program files were all compiled with GEMPACK 11.
  Status: Patch file for GTAPAgg, to be unzipped to the folder where the aggregated files are stored. Please double-click on sprs2fullall.bat. This should work if you have not changed the names of the files. For Flexagg, we already have this batch file included in the main directory.



• September 26, 2012: Chinese tariffs, particularly on the "crp" sector of GTAP Data Base, are extremely high in 2007. Most of it arises from high tariffs on photographic films imports by China.
  Status: to be fixed in v8.1



• September 26, 2012: Tariffs on exports from Singapore and Senegal to all countries are wrong for 2007 version of GTAP 8 Data Base. This is due to a mapping problem, where these two countries got swapped in the mapping from the exporting countries to standard GTAP countries. Importing country mappings are still well-specified. So the tariffs on Singaporean exports appear in the final data base as those on Senegalese exports and vice-versa.
  Status: to be fixed in v8.1



• September 11, 2012: In the GTAP 8 Data Base, the new countries Oman (omn) and Israel (isr) share the same codenames with the sectors Other Mining (omn) and Insurance (isr), respectively. This causes a minor problem in GDX files used in GAMS: it rearranges the data such that Oman is the first country, while the rest of the data remains in the same order as in the original data.
  Status: Two alternative solutions are suggested to the users here. We may fix this in a future release by changing the names of a few or all set elements.



• September 11, 2012: Australia's taxes on consumption of petroleum and coal products (p_c) are negligible in both 2004 and 2007 releases of version 8 Data Base. This is due to a processing error in the energy data construction module.
  Status: To be fixed in v8.1.



• September 11, 2012: Malawi, Mauritius and Uganda have significant oil sector. In the construction process, we revise the energy sector to conform better to IEA data, which record data for these countries not individually, but as part of "rest of Africa". "Rest of Africa" now produces and exports oil in significant quantities. We apportion those flows, with other energy flows, among "rest of Africa" member countries according to GDP share. The problem with this approach is that the production and exports actually belong to Chad and Equatorial Guinea, which are not separately identified in our the data base, but end up reassigned to other countries like Uganda, which are.
  Status: For future releases, we hope to bring in data from other sources to improve our disaggregation of the IEA energy balance data countries and avoid this problem.



• August 28, 2012: Peter Minor reported that long names of the regions do not appear in GTAPAgg regional aggregation window in GTAP 8a Data Base for the year 2007. Mark Horridge identified the issue to be arising from mis-specified header "H1L" in the file sets.har.
  Status: We have prepared a patch file which needs to be unzipped to the GTAPAgg directory. Then the file "setpatch.bat" has to be double-clicked on to complete the patch.



• August 28, 2012: Glen Peters reported that the time-series trade data for Taiwan, Singapore and Netherlands for the years 2008 and 2009 are inconsistent with the rest of the years. There is a sharp fall of total exports and imports of these countries in these two years.
  Status: This is due to the lack of re-export margins data required for adjustments needed for these countries, to be consistent with those done for the years prior to 2008. This will be fixed in a future release.



• July 5, 2012: Although GTAP 8a Data Base provides all files in a non-sparse format that is completely compatible with all versions of GEMPACK, it may be possible that the finally aggregated files may get generated in the sparse storage format. Users with licenses older than GEMPACK 10 may then have an issue reading these files, much as they did with the final release of GTAP 8 Data Base.
  Status: We do not propose another bug-fix version for this, since this problem seems more random than the original one in the final release of GTAP 8 Data Base. Users facing this problem are advised to unzip this patch ZIP file into their flexagg directory and then run their aggregations.



• July 3, 3012: Kim Yong noted that the data on VST (exports of international transport service) from South Korea to all countries in total is very large particularly for OTP sector. This is strange given that Korea's export of international transportation service of Otp (road and others) should be very small as Korea is located in peninsula land and it is neighbored only by North Korea.
  Status: Unfortunately there is no data, so we had to fill. From the IMF, we have Korea's trade in transport services, but no breakdown between air, sea and land, or between passenger, freight, and other. We fill in the blanks by a procedure that, roughly speaking, should use cross-country average shares. However, the share of land freight transport exports in total transport exports is 11 per cent globally but 35 per cent for Korea. The filling-in program is not simply a matter of taking cross-country shares and it is possible that its close location to China may have something to do with the high filled share. We plan to look at it more closely when we redo the international transportation shares in v9.



• July 3, 3012: Dominique van der Mensbrugghe noted some large production subsidies in Belarus.
  Status: We have confirmed this is a problem with the contributed IO table. The IO contributor has been contacted and has agreed to submit a table with the correct tax data. To be fixed in version 8.1.

Issues addressed in the GTAP 8a Data Base (Bug Fix Version, June 2012)

• For users of Gempack 9 or earlier versions there was an issue with the data storage format. In this new version the data files are stored in "full storage format", as opposed to "sparse" format, so that the data base is compatible with all versions of Gempack.


• In the GTAP 8 Data Base a new header ("VTSS") containing the decomposition of tariffs into specific and ad valorem components was included in the GTAP Data Base for those interested in separating the two types of tariffs. This data was found to contain some defects and the definition was found to confuse users. These defects have now been fixed and the data re-defined as "tariff revenue decomposed into it specific and ad valorem components".


• The element "TOTAL" has been removed from the new domestic support headers PAYD and PYRT, to avoid showing two totals in Viewhar.


• The Flexagg2 package was revised to be able to aggregate the new emission file. We also added Viewhar.exe to the Flexagg2 package to assist the new users in reading the data more easily.

• June 18, 2012: Domestic support rates provided by Hans Jensen and Hsin Huang are not targeted in the data base construction process. As a result, the rates in the data base are different from those provided as inputs.
  Action: For version 8, we have provided the detailed information on OECD domestic support payments and rates in the headers PAYD and PYRT respectively. The original rates in total for any sector for any factor/input/output may be viewed by summing over the "decoupling" types. In future releases, we may target these rates explicitly in the data base construction process.



• May 15, 2012: Flexagg for GDyn 8 Data Base does not automatically produce header H0
  Action: fixed in the version 8a in June 2012



• May 15, 2012: Zero bilateral migration in GMig2 DB gives room to computational errors while modelling (for certain aggregations).
  Action: fixed in the version 8a in June 2012



• May 9, 2012: Data in the header VTTS is not usable, since it has some serious errors that cause its inconsistency with the core data base.
  Action: fixed in the version 8a in June 2012



• May 9, 2012: Flexagg for GTAP 8 Data Base does not automatically aggregate CO2 emissions data
  Action: fixed in the version 8a in June 2012



• May 9, 2012: Some users may not be able to read or use GTAP 8 Data Base, due to the sparse storage format of the data.
  Action: fixed in the version 8a in June 2012



• May 9, 2012: There are some small intra-EU tariffs on imports by France, while these should have been all zero. This is because of inclusion of small overseas territories of France in the GTAP region "FRA". Their inclusion, per se, is not the cause of the problem, however. The small tariffs arise due to an error in the tariff dataset for these territories, which are wrongly assumed to impose France's MFN tariff for all imports, including those from within the EU.
  Action: This will be fixed in a future release.



• May 9, 2012: There are some FTA partners that appear to impose tariff in some sectors on each other's exports. A detailed account on this and other tariff issues and the response from our contributor (Houssein Guimbard, CEPII) is available here
  Action: None suggested

Center for Global Trade Analysis
Department of Agricultural Economics
Purdue University
403 Mitch Daniels Blvd.
West Lafayette, Indiana 47907-2056 USA

© 2026, Purdue University. All rights reserved.
Purdue University is an equal access/equal opportunity university.
If you have trouble accessing this page due to disability, please contact contactgtap@purdue.edu.