EXIOBASE 3 is the culmination of work in the FP7 DESIRE project (www.fp7desire.eu) and builds upon earlier work on EXIOBASE 2 in the FP7 CREEA projected (www.creea.eu) and the FP6 EXIOPOL project. 
The papers describing the work in EXIOBASE 3 are coming out at Journal of Industrial Ecology (https://onlinelibrary.wiley.com/journal/15309290) in June 2018. 
The database is available on www.exiobase.eu 

The data has aggregated energy and material accounts because of licensing issues with the IEA.

 See environmental footprint results.
The main data documentation is described here 

This corresponds to version 3.4 of the data for the time series 1995-2011.


There was an earlier release of the data for v3.3 – for some uses, e.g. energy, emissions and material accounting, the data is valid. 

However, care should be taken in the water and land use accounts. There were major differences that were corrected in v3.4, and a public release of data for v3.3 will not be made available. Depending on research questions, the data may be impacted.


We also have estimates based on trade and macro-economic data up to 2016 as well, however, a lot of care must be taken in use of this data. It is not suitable for analysing trends over time. The basic description of the process employed is in the relevant deliverable:


The data here has different end-dates for the environmental extensions (2011 energy, 2015 CO2 fuel combustion, 2013 material, 2011 most others, incl methane and N20, land, water), and is not part of the main public release. Some work is going on to update these, but other collaborative efforts are more than welcome.

Data quality/adherence

In terms of data quality for more recent years. Please be mindful of:

-2007 is the base year for the SUTs and the method of interpolation over time involves using data in different classifications and formats to understand structural change over time (i.e. intermediate and final demand flows by product and industry). Only output, trade and macroeconomic constraints (7 sectors) are direct inputs to final balancing in absolute terms. The database is under the NACE1 classification scheme.

-SUT and IOT data are only used to inform structural change

-Whilst the 2007 EXIOBASE2 table is the base, the National Account Main Aggregates is the main constraint. The main goal was temporal consistency, and not disaggregation of official tables in EXIO3. SUTs/IOTs do often not get updated going backwards, whilst National Account Main Aggregates data does. In general, National Account Main Aggregates are often 10-20% off the equivalent data points aggregated from the official SUTs/IOTs from 2007. 



More specific updates for each environmental extension will be provided in relevant publications.

Right now, our focus is on having a robust time-series to 2016. We are not currently working on updating data past 2016, although we do expect to continue the work in the future. The data past 2011 is currently being made available through Box. Unfortunately, the only method of access is manual addition to the shared folder – please email Konstantin or Richard. Also ensure that you have read the limitations above and have registered on exiobase.eu


In terms of financing future work, we of course are reliant on funding. We are updating parts of the dataset based on individual project requirements, and providing some internal funding for co-ordination across these efforts. We are of course welcome to participate in projects that help fund these updates – either central updates or peripheral work that has benefits for both a single end-user, but also that can be rolled into future public data releases.


We are continually updating the data behind EXIOBASE with a rolling release cycle. Hence more feedback that is given, the greater the chance of improvements in future updates!




Stay tuned for announcements:

Energy & emission accounts
Energy and Emission accounts are based on the IEA energy balances, but are not always fully consistent with the monetary IO tables (different data sources was used in the disaggregation of the energy accounts) – this does lead to some minor discrepancies in relationships between energy flows and monetary flows. Some efforts are ongoing to fix these. 

Household use of energy – this is available in disaggregated form (by EXIOBASE product), but is auxiliary to the main MRIO dataset, as it does not fit that data structure. Similar case for fuel combustion emissions.


For useful concordances, please see:
The final EXIOBASE2 database is available on www.exiobase.eu, but note that agricultural emissions are available at:
Final country specific tables of EXIOBASE2 are:
Source country tables of EXIOBASE2 are:

May 23, 2018

Leave a Reply