Skip to main content
European Commission logo print header

World Input-Output Database: Construction and Applications

Final Report Summary - WIOD (World Input-Output Database: construction and applications)

Executive summary:


If the amount of electronic products that German households import from China increases, what is the effect on the employment of low-skilled workers and CO2 emissions in Korea? Today's products and services are no longer produced within a single country. Whereas a product may list that it is "made in China", its key components are often produced also in other parts of the world. In the last couple of decades, production processes have been sliced up more and more into ever smaller parts, each of which is carried out by a different producer. In addition, this fragmentation more and more crosses the borders of countries. The common viewpoint today is that products and services are made in global value chains.

Project Context and Objectives:


WIOD in a nutshell

Does the increase in Chinese exports of textile products affect the employment of low-skilled workers in German retail trade? What would be the effects of a change in the European Union's agricultural policy on global CO2 emissions? And how strong is the trade-off between employment growth and environmental quality in Europe? Economic and environmental policies are designed at a very detailed level of industries and products.

Key objectives

The key objective in the first part of the project was to construct a global database of national SUTs and/or IOTs, which are linked through international trade flows, creating so-called inter-country SUTs and IOTs. An inter-country IOT, for example, contains information on, e.g. the sales from the Indian chemical industry to the German automobile industry.

Scientific and technical objectives

An important characteristic of the database is the integration of national SUTs and/or IOTs with data on bilateral international trade flows in so-called inter-country SUTs and IOTs. The following two examples indicate why this integration is of utmost importance. First, international trade plays an increasingly important role in the world's development.

The specific scientific and technical objectives of the project can be summarized as follows.
1. To build a time series of global inter-country input-output tables. The project constructed a time series of annual, global inter-country SUTs and IOTs (the so-called World Input-Output Tables, WIOTs) based on an integration of national SUTs with international trade statistics.
2. To build socio-economic and environmental satellite accounts. The project constructed satellite accounts for socio-economic and environmental indicators at the same level of industry detail and timeframe as for the WIOTs. Socio-economic indicators include numbers of employees and wages at various skill levels, and investments.
3. To measure and analyze the trends in trade, economic growth, technological change and environmental pressures. The new database was used to perform ex-ante and ex-post analyses of these trends, adopting IO techniques and/or econometric approaches.
4. To analyze changes in global trade structures and the effects of increasing global trade integration on European labour markets. The effects of outsourcing of activities from advanced to developing countries on specialization patterns and labour markets have been investigated using the new database.
5. To provide policy support to the European Commission on socio-economic and environmental issues.

Project Results:



Four types of "work" have been distinguished: the construction of the database (DATA), which is at the heart of the project; methodological research (METHODS); developing and carrying out applications such as IO and econometric techniques (APPLICATIONS); and building new models and adapting existing models for policy analysis (POLICY ANALYSIS).

WP1 Harmonization of supply and use tables


This work package provides time series of annual Supply and Use Tables (SUTs) for the period 1995-2006 on a country-by-country basis in nominal and real terms. The tables are internationally harmonized in terms of price and output concepts, and have a common industry- and product classification and detail.

Tasks within the project

The following tasks had to be carried out:
1. Collect national current price SUTs at lowest product and industry level possible for all countries. These will be available for some benchmark years, but not for all years. For some non-EU countries input-output (IO) tables (product-by-product or industry-by-industry) will be available, rather than SUTs (product-by-industry). These tables need to be "reverse-engineered", i.e. converted into SUTs using additional assumptions (see Beutel, 2005; Yamano and Ahmad, 2006). This process will differ from country to country depending on the information supplied.
2. Harmonize price concepts to basic price, and apply other adjustments needed to overcome differences in concepts such as treatment of financial intermediation services indirectly measured (FISIM) and c.i.f/f.o.b adjustments.
3. Benchmark current price SUTs on industry-level gross output, value added and intermediate inputs and final demand categories from the National Accounts (NA).
4. Derive full time-series of SUTs for the period 1995-2006 using interpolation techniques for years in between based on time series from the National Accounts (see Broersma and van Moergastel, 2007). For some major countries series back to 1980 will be attempted. Aggregate up to the WIOD product-level (60) and industry-level (30).
5. Deflate nominal SUTs on an industry-by-industry basis using gross output, intermediate input and import deflators from the National Accounts, possibly complemented by expenditure deflators and unit import values from trade statistics.

- D1.1: Report outlining detailed strategy for SUT data collection and processing
(due in month 1, delivered in month 5; beneficiaries 1, 3, 6 and 7 have carried out this task)
- D1.2: Preliminary harmonized nominal SUTs for 1995-2006
(due in month 6, delivered in month 18; beneficiaries 1, 3, 6 have carried out this task)
- D1.3: Preliminary harmonized constant national price SUTs for 1995-2006
(due in month 12, delivered in month 31; beneficiaries 1, 3, 6 have carried out this task)
- D1.4: Preliminary harmonized constant international price SUTs for 1995-2006
(due in month 16, delivered in month 31; mainly beneficiary 1has carried out this task)
- D1.5: Final set of harmonized nominal and constant price SUTs for 1995-2006
(due in month 36, delivered in month 36 mainly beneficiary 1has carried out this task)

Significant results

The most significant result in the first part of the project was reaching MILESTONE 1 Preliminary inter-country input-output tables in current prices in June 2010. In addition, a full set of annual SUTs for each country has been prepared (D1.2) based on the exploratory work in D1.1. As part of the methodological research concerning the interpolation of the benchmark SUTs, a new method has been developed, called SUT-RAS. This method has been used for updating applications as part of workpackage 6 and evolved into deliverable D6.3c.

WP2 International trade data


The objective of this workpackage was to provide a harmonized and consistent dataset of bilateral trade flows of goods and services between all countries in the WIOD-database covering the period 1995-2006. The time period covered has been changed to 1995-2009 for compatibility with the available national SUTs.

Tasks within the project

Task 2.1: provide a harmonized and consistent dataset of bilateral trade flows of goods
This task was concerned with collecting and preparing data for bilateral trade in goods data to be used in the construction of the international supply and use tables. For this, raw data was taken from the UN COMTRADE and downloaded at the HS 6-digit level. The trade database contains the 40 WIOD countries over the period 1995-2009 as reporter countries and all other countries as partner countries.

Task 2.2: provide a harmonized and consistent dataset of bilateral trade flows of services
The second task was concerned with constructing a set of bilateral trade flows in services. For this a database based on Balance of Payments (BOP) data on services trade by BOP-sector and covering services sectors for the period 1995-2009, covering all major countries in the world including the WIOD countries.


- D2.1 Report outlining detailed strategy for trade data collection and processing
(due in month 3; delivered in month 3; beneficiary 3 has carried out this task)
- D2.2 Preliminary harmonized and consistent data set of bilateral trade flows of goods and services
(due in month 9; delivered in month 10; beneficiary 3 has carried out this task).
- D2.3 Final harmonized and consistent data set of bilateral trade flows of goods and services
(due in month 36; delivered in month 36; beneficiary 3 has carried out this task).

Significant results

Within this workpackage, consistent sets of trade in goods and services statistics have been provided which afterwards served as inputs in the construction of the world-input-output database (WP3). The data have been successfully constructed for a longer time period than was planned in the Description of Work. The deliverable outlining the detailed strategy for trade data collection and processing has been delivered in time.

Deviations from the plans

This work package has successfully achieved its objectives in providing a harmonized and consistent set of trade in goods and services to the project and even for an extended period (1995-2009, instead of 1995-2006). These data have been sent in time to enable the construction of the international SUTs and WIOT.

WP3 Estimation of inter-country input-output tables


The objective of this workpackage was to provide inter-country input-output tables for 1995-2006 based on the harmonized supply and use tables from WP1 and the trade statistics from WP2. In addition, aggregated and integrated versions of these tables should be provided, required for various analyses in WP7 – WP9.

Tasks within the project

To derive inter-country input-output tables, harmonized national SUTs constructed in WP1 had to be merged with the bilateral trade statistics derived in WP2. This required completion of the following steps:
1. Breakdown of national use tables into domestic and import use tables, with an import use table for each trading partner. The output of this step was a set of time series of International SUTs for 1995-2009. The conversion of national currencies into US dollars was based on official market exchange rates.
2. Creation of full inter-country input-output tables. To this end, the International SUTs for all countries were first merged into a WorldSUT for every year.
3. By combining the bilateral trade statistics with deflated harmonized national SUTs (also constructed in WP1), inter-country input-output tables expressed in previous year's prices were constructed.


- D3.1 Set of preliminary harmonized inter-country SUTs for each country
(due in month 12; delivered in month 24; beneficiaries 1 and 3 have carried out this task).
- D3.2 Preliminary inter-country world input-output tables in current international prices
(due in month 12; delivered in month 20; beneficiary 1 has carried out this task).
- D3.3 Preliminary inter-country world input-output tables in constant international prices
(due in month 21; delivered in month 36; beneficiaries 1 and 3 have carried out this task).
- D3.4 Preliminary integrated and consolidated input-output tables for major regions
(due in month 21; delivered in month 36; beneficiary 1 has carried out this task).
- D3.5 Final set of harmonized inter-country SUTs for each country,

Significant results

This workpackage yielded the most prominent output of the project: consistent time series of inter-country input-output tables for the period 1995-2009, both in current prices and in previous year's prices.

Deviations from the plans

In comparison to the Description of Work, several relatively minor changes were needed (next to extending the time series from 1995-2006 to 1995-2009 as mentioned in the reports on WPs 1 and 2 already).

WP4 Satellite accounts: environmental indicators


The objective of this WP was to collect, process and organize a comprehensive set of satellite accounts of environmental indicators over the period 1995-2006 and that match the inter-country database developed in WP3. For each country (including the "rest of the world" region), the dataset contains at the WIOD sectoral breakdown (35 industries plus final demand) a time series (extended up to 2009) of the following indicators:
- Gross and emission-relevant energy use broken down by 27 energy entries, including different fossil fuels, major non-renewables (electricity, heat and waste), major renewables and losses;
- Water consumption;
- Emissions of greenhouse gases (CO2, N2O, CH4) and air pollutants relevant for acidification (SO2, NOX, NH3) and tropospheric ozone formation (NOX, NMVOC, CO, CH4).


Task 4.1: Collection of environmental data sources and organization of lower tier data set

With respect to methodology, this work package has extensively built on the protocols of the FP6 project EXIOPOL. However, as the EXIOPOL database is strictly limited to one single base year, the selection of the data origin has diverged in many cases from EXIOPOL, and made use of databases that contain yearly information for the full time series of WIOD, such as Eurostat, IEA, UNFCCC, FAO, and SERI.

Task 4.2: Transformation of the lower tier data set into a harmonized upper tier data set

This task has been the principal task of the work package, and it has implied addressing several methodological problems, for which solutions have been found and have been duly reported.

1) Attain coherence and consistency between the environmental satellites and the IO based economic data set. Correspondence links have been clearly defined between the industrial classification adopted in the IO tables (IOTs) with the definition of industries and of other entities by which the emission data sets, energy statistics and other environmental variables are recorded.

Task 4.3: Further data integration and transformation linkages with applied models

This task has integrated the environmental data with applied models, essentially a CGE model run by ZEW. This has required the development of transformation matrices including substitution elasticities and abatement cost functions for pollution control options.


D4.1 (month of delivery: 3): Technical report on the conceptual framework for environmental satellites integrated in the WIOD system of SUTs and IOTs
D4.2 (month of delivery: 12): Preliminary database of environmental satellite accounts
D4.3 (month of delivery: 16) Preliminary dataset of environmental information for integration in the prototype CGE model
D4.4 (month of delivery: 36): Final database of environmental satellite accounts and final dataset of environmental information for integration in the core CGE model
D4.5 (month of delivery: 36): Technical report on the compilation of the environmental satellites database.

Significant results

Halfway the project time (with 4 month's delay, month 20) a preliminary dataset of energy and air emission data was delivered.

Deviations from the plan

There have been no major deviations in this WP, only secondary adjustments.

A 4-month delay was experienced in the interim release of the first dataset draft, but this was largely compensated by regular updates before the final dataset release in month 36.

WP5 Satellite accounts: socio-economic indicators


To provide accounts of socio-economic indicators that will match the inter-country database developed in WP3. For each country, it will contain various indicators of employment, including hours worked by various types of skill and associated wage cost; various type of investment (including tangible and intangible assets) and shares of (affiliates of) foreign firms in total sales and investment. The data will match the industry-level detail of the WIOD-database.

Task 5.1: The update of the EU KLEMS database
The construction of EU KLEMS database was funded under the 6th framework program (see online). This database, grounded in national accounts statistics, provides industry-level measures of output, inputs and productivity for the European Union countries, Japan, Australia, Canada, South Korea and the US, covering the period 1970-2005.

Task 5.2: Construction of harmonized indicators of intangible investment
In addition to investment in software, which is already covered through the KLEMS database, two other major areas of intangible investment will be included: innovative property (mainly scientific and non-scientific R&D) and firm competencies (company spending on reputation, human and organizational capital). These statistics are collected under a proposed FP7 project called Competitiveness, Innovation and Intangible Investment in Europe (COINVEST) for European countries and the US.

Task 5.3: Construction of Foreign Direct Investment indicators (this task will mainly be done by partner 3)
Increasing Foreign Direct Investment is an important component of globalization not captured by the trade statistics to be constructed in WP2. The indicators will be mainly based on the OECD's Activities of Foreign Affiliates (AFA) and Foreign Affiliates Trade in Services (FATS) databases, which present detailed data on the performance of foreign affiliates in the manufacturing and services industry of OECD countries (inward and outward investment).

- D5.1: Update of EU KLEMS database
(due in month 12, delivered in month 12; mainly beneficiary 1and 3 have carried out this task)
- D5.2: Preliminary KLEMS database for WIOD countries not covered in EU KLEMS
(due in month 12, delivered in month 36; mainly beneficiary 1 has carried out this task)
- D5.3 Intangibles investment indicators by broad sector for WIOD countries
(due in month 24, delivered in month 36; mainly beneficiary 7 has carried out this task)
- D5.4 Foreign affiliates' indicators by broad sector for WIOD countries.
(due in month 24, delivered in month 28; mainly beneficiary 3 has carried out this task)
D5.5 Final satellite account: socio-economic indicators
(due in month 36, delivered in month 36; mainly beneficiary 1 and 3 have carried out this task)

Significant results

In the first period we reached an important mile stone: MILESTONE 3 Preliminary socio-economic satellite accounts. At the end of the project, MILESTONE 10 Final socio-economic satellite accounts has been reached. This part of the database is accessible for public use at the WIOD website from April 16, 2012 onwards.

Deviations from the plan

Deliverable 5.2 (extension of EU KLEMS to non-EU countries) was delayed in the first part of the project, due to diffifulties ingathering the necessary data. But in the second part of the project, this quickly ameliorated and in the end, the fnial database was delivered in time. As a result the intermediate deliverable 5.2 (preliminary KLEMS database) has never been formally submitted and is represented by the final database as described in D5.5

WP6 Methodological research related to the database


Constructing a large database as the one proposed in this project, involves a lot of estimation. This is due to the fact that some of the data are simply not available or incomplete. Another factor that plays a role is that some data are known with greater reliability than other data, and some are even known exactly (the so-called superior data). Also, even in an ideal world, IO tables cannot be measured but are a construct (i.e. a model) themselves.

Tasks within the project

Task 6.1: aspects related to the construction of national IO tables from SUTs
1. Analysis of the (dis)advantages of using product-by-product versus industry-by-industry type of IO tables.
2. Addressing the choice of method to use for constructing IO tables from SUTs. The most widely used methods are the product-technology model and the fixed product sales structure model, but a wide range of alternative models have been proposed. The merits and demerits of the available methods were investigated.

Task 6.2: aspects dealing with linking the tables across space and time, and making projections
3. Analysis of methods to harmonize IO tables, which requires using international prices. Next to the question whether the SUTs should be valuated in basic prices or in purchasers' prices (what are the advantages of each, what type to use when, how can they be linked?), the issue of making the prices internationally consistent was dealt with.
4. Analysis of methods to construct IO tables in constant prices. Various methods are available, of which "double deflation" is the best known. For some countries, tables in constant prices are readily available, but typically they are in prices of the previous year. The question in that case is how the information from the chained indexes can be best used.
5. Projection of the entire WIOD database up to the year 2009 (originally planned to update up to 2010). Many of the underlying annual statistics in this project become available with a relatively short time lag.

Task 6.3: fundamental aspects related to the use of the database for empirical analyses and model building
6. Sensitivity analysis with respect to types of IO tables used in empirical work. Although inter-country tables contain, in principle, more information than so-called multi-country tables, how much will be left of the advantage of using inter-country tables in the case of limited data availability?
7. Analysis of the effects of aggregation and disaggregation. In the WIOD database, a limited number of classifications for groups of countries with similar detail have been used when linking the SUTs. A consequence of this is that there were several options when models are built or when impact analyses are done. The question is: To what extent do the results depend on the chosen approach?
8. Sensitivity analysis with respect to the use of either current prices or constant prices. When comparing changes over time, one can take the table in constant prices as the starting point of an IO model or the table in current prices.


- D6.1 Technical report on advantages and disadvantages of types of input-output tables product-by-product or industry-by-industry)
(due in month 6, delivered in month 9; mainly beneficiaries 2 and 6 have carried out this task)
- D6.2 Technical report on the construction of input-output tables (product-by-product or industry-by-industry)
(due in month 12, delivered in month 13; mainly beneficiaries 2 and 6 have carried out this task)
- D6.3 Technical report on: methods to harmonize input-output tables; methods to construct input-output tables at constant prices; projection of supply and use tables; and sensitivity of model results with respect to data availability
(due in month 18, delivered in month 19; beneficiaries 1, 2 and 6 have carried out this task)
- D6.4 Technical report on the multiplier bias from supply and use tables and the sensitivity with respect to using tables in constant or current prices
(due in month 24, delivered in month 27; beneficiaries 1 and 2 have carried out this task)

Significant results

As scheduled, WP6 has produced four deliverables. The deliverables D6.1 and D6.2 provide a scientific underpinning of two important choices that have been made for the construction of the database. Namely, to focus on industry by industry IO tables and to adopt the fixed product sales structure assumption when "translating" SUTs to IO tables.

Deviations from the plans

Item 5 in Task 6.2 originally covered the projection of the entire database up to 2010. The method to carry out the projections has been developed (i.e. SUT-RAS) and has been successfully applied (in WP1) for interpolation purposes. A crucial part of this method is that certain information is necessary for the margins of the tables. The reason that the implementation for the years up to 2010 could not be executed is that this necessary information was not available (or very incomplete).

WP7 Applications of the database: environmental aspects


This work package focused on designing and improving several types of models that attempt to explain recent changes in the environment and to predict future changes, at national, regional and global levels. The vast majority of improvements were possible due to the construction of the WIOD database. Its unique feature is that it contains harmonized data across time and space. In particular the time dimension allows for the use of (panel data) econometrics.

Tasks within the project

Task 7.1: Econometric approaches and structural decomposition analysis
1. Employing econometric techniques to study the linking of economic data and environmental indicators. Econometric panel data methods have been adopted to make full use of the rich dataset.
2. Focusing on the widely debated topic of the consequences of globalization and trade liberalization on the environment. The environmental implications of structural change have been studied in two consecutive steps.

Task 7.2: Environmental model building and related methodological aspects
3. This subtask dealt with the cost-efficiency assessment of environmental policies. The central objective was to come up with operational algorithms for integrating bottom-up estimates on abatement cost functions for selected pollutants in economy-wide models.
4. The second subtask aimed at improving the calibration of computable general equilibrium (CGE) models. Typically, parameters of CGE models are calibrated to benchmark data from IO tables in a certain year and are thus based on a single observation. One of the unique features of the WIOD database is that it provides a time series of IO tables. This offered opportunities (i) for investigating the robustness of model outcomes to the choice of calibration year, (ii) for estimating some of the elasticities required for calibration, and (iii) to study the sensitivity of the results with respect to the calibration method.
5. This subtask dealt with the research field of technological change in the context of environmental policy. In order to incorporate the significance of environmental innovations, the integration of induced technological progress into this work package is of high relevance. The WIOD database contains data on (implicit) prices as well as input requirements per unit of output, measured in constant prices.
6. The fourth subtask developed a prototype CGE model, aiming at a 100% fit with the WIOD data. Typically, a large variety of data sources is used in the construction of CGE models, while WIOD allowed for using a single source of harmonized data.


- D7.1 Report on linking economic data and environmental indicators using panel data econometrics
(due in month 12, preliminary version delivered in month 13 and final version in month 17; corresponds to task 7.1 subtask 1; mainly beneficiary 4 has carried out this task)
- D7.2 Report on the effects of structural change on the environment using a structural decomposition analysis and on the causes of structural change in selected industries using econometric techniques
(due in month 18, delivered in month 17; corresponds to task 7.1 subtask 2; mainly beneficiary 4 has carried out this task)
- D7.3 Report on the cost-efficiency assessment of environmental policies using algorithms for a bottom-up analysis of the associated abatement cost functions
(due in month 24, postponed until and delivered in month 36; corresponds to task 7.2 subtask 2; mainly beneficiary 5 carried out this task)
- D7.4 Report on calibration of CGE models using time series of IOTs
(due in month 18, delivered in month 18; corresponds to task 7.2 subtask 4; mainly beneficiary 4 has carried out this task)
- D7.5 Report on studying induced technological progress regarding energy use and pollution
(due in month 24, delivered in month 24; corresponds to task 7.2 subtask 3; mainly beneficiaries 4 and 5 have carried out this task)
- D7.6 The WIOD Prototype CGE Model
(due in month 30, delivered in month 24; corresponds to task 7.2 subtask 4; mainly beneficiary 4 has carried out this task)

Significant results

As scheduled, WP7 has produced six deliverables. D7.1 has tried to use as many environmental and other data from the WIOD database as possible. We have depicted carbon and energy intensities for the majority of the countries included in the database. Then we have applied an index decomposition analysis to identify determinants of the changing indicators.

Deviation from the plans

This work package has achieved all objectives and performed its tasks as outlined above in the required timeframe. It should be mentioned that no delays have occurred in the submission of the deliverables. The preliminary state of some deliverables was due to lack of important and reliable data that was necessary for the calculations within the deliverables. After consultation with the Management Team the delivery was postponed whenever necessary. It should be stressed, however, that the preliminary character of some of the deliveries has had no impact on other tasks, available resources, or planning of the project.

WP8 Applications of the database: socio-economic aspects


The central focus in this WP is the analysis of trade structures (trade in intermediate products in particular) and the effects of increasing trade integration with developing countries on labour markets. This topic has attracted more attention in the last decade when the effects of outsourcing to low-wage countries on labour markets and relative labour demand in the advanced countries have been discussed. It was argued that outsourcing of low-skilled intensive fragments of the production process to low wage countries has negative effects on relative demand and wages of the unskilled workers in advanced economies and thus lead to higher wage dispersion and inequality. In most of the studies outsourcing was measured either by using a benchmark IO table or SUT for one year or by using detailed trade data. Due to data availability most of the studies have been pursued for developed countries only (mostly the US).

Tasks within the project

Task 8.1: Studies of factor contents of trade
The first task was to study the "factor content of trade" allowing for trade in intermediate inputs using the WIOD-database. Each traded good can be seen as including a certain amount of the primary factors of production like labour (distinguished by skill type), capital, land, etc. These factor contents have been measured by use of IO tables, because one has to account for the indirect inputs as accounted for in the IO framework.

Task 8.2: Analyses of effects of trade in intermediates on labour markets
In the second task, the effects of trade in intermediates on labour markets will be studied in great detail. Changes over time, the effects of inclusion of services trade and a more detailed regional aspect will be addressed

Task 8.3: Modelling and simulating with WIOD times series data
The third task will entail an extensive modelling effort with full usage of the time-series dimension of the WIOD database focusing on the implementation of a dynamic trade matrix. This allows to study the relative importance of various phenomena (in particular the effects of outsourcing) on labour market outcomes and output specialization by means of simulations focusing on medium-run scenarios.


- D8.1 Report containing a literature review of studies on the factor contents of trade
(due in month 9; delivered in month 14; beneficiary 3 has carried out this task).
- D8.2 First results on the factor content of trade in the WIOD database
(due in month 16; delivered in month 19; beneficiary 3 has carried out this task).
- D8.3 Final report on the analysis of the factor contents of trade
(due in month 36; delivered in month 36; beneficiary 3 has carried out this task).
- D8.4 Trade in intermediate inputs and final goods - Descriptive analysis and decomposition
(due in month 18; delivered in month 20; beneficiary 3 has carried out this task).
- D8.5 Report with estimation results of the effect of outsourcing on labour markets, based on
preliminary WIOD data
(due in month 24; delivered in month 24; beneficiary 3 has carried out this task).
- D8.6 Final report with estimation results of the effects of outsourcing on labour markets and international specialization
(due in Month 36; delivered in Month 36; beneficiaries 3 and 5 have carried out this task).
- D8.7 Report on the construction of a new international interindustry econometric model with flexible and dynamic trade structures;
(due in month 36; delivered in Month 36; beneficiaries 3 and 5 have carried out this task).
- D8.8 Report on medium-run scenario studies using the new model, with a focus on the effects of increased trade openness and outsourcing on developed and developing countries;
(due in month 36; delivered in month 36; beneficiaries 3 and 5 have carried out this task).

Significant results

Within this work package WP8 it was planned to produce three deliverables in the first period. The first (D8.1) contains a review of the literature in factor contents trade including theoretical underpinnings and historical contributions with an emphasis on the more recent developments in this field. The second deliverable (D8.2) provides a description of results for the trade in value added and the various components both in a descriptive and more analytical way following the recent literature. This should lead to a final report on this issue by the end of the project (i.e. D8.3). The third deliverable (D8.4) provides a detailed account of trade statistics with respect to trade in intermediates versus trade in final goods using data from WP1 and WP2 including a comparison to the data then provided in the international SUTs and WIOT.

Deviations from the plans
None. The tasks in this work package heavily rely on the data to be made available in work packages WP1-WP3, notably the international supply and use tables, the additional data on national accounts and socio-economic indicators (also from WP5) and the underlying trade in goods and services data. Final results have therefore only become available at the end of the project, with no significant deviations from the plan.

WP9 Policy assessments making use of the new dataset


The purposes of this work package were a) to promote contacts between modellers and data providers in order to enhance the applicability of WIOD data for use in established policy-oriented models, b) to improve the models by using the new WIOD data and c) to exploit the WIOD datasets to better gear existing models to the actual assessment of specific policies and developments. Most of the models concerned (GEM-E3, Nemesis, PACE and WorldScan) are used in-house by the EC for policy analyses. Thus, the work package explored the potential benefits of the new datasets for policy assessments within the EC.

Tasks within the project

The tasks for all four modelling teams involved 1) to give feedback to the database construction part within WIOD -and to find additional data where necessary- such that the datasets become most valuable for the impact assessments envisaged, 2) to adapt the models to the new possibilities implied by the datasets, 3) to explore the usefulness and benefits of this approach by actually performing the impact assessments and 4) to deliver the adapted model versions to counterparts within the EC that use the models for policy analyses.

Task 9.1: Feedback on the data construction activities and extension of the datasets in view of the impact assessments envisaged. Comments and feedback on data construction from the modelling teams involved have been summarized and database extensions that were needed specifically for the policy assessments envisaged have been described and made available to the other partners.

In particular,
- The team of GEM-E3 checked the consistency of the dataset with the current model set-up, complemented the dataset with the additional data that the model requires and extended the calibration program to account for the new information that WIOD yielded on the productivity of production inputs;
- The Nemesis team used the new input-output tables and reorganized the data such that they became consistent with the accounting framework of the model;
- Similarly, the WorldScan-team performed checks to ensure that the new data remained consistent with the model set-up and collected (where necessary) additional data on skills, in particular with a view to represent the complementary of specific labour skills with Research and Development (R&D) capital stocks as inputs for industry-level production.

Task 9.2: Model adaptations
Model adaptations that a) came within reach with the new datasets from WIOD and b) were useful to implement with a view to the policy simulations envisaged have been conducted and documented.

Task 9.3: Policy analyses
Assessment of specific policies and developments have been conducted and documented. The exact list of policy questions has been defined at a later stage in order to match with the policy agenda of the EC and the exact availability of data within WIOD.
- The advanced model version of GEM-E3 has been used to study policy questions that are in the core of the Lisbon agenda and the recent EU strategic decisions on mitigating climate change.
- The new version of Nemesis has been used to assess policies related to the Lisbon Agenda. The main Lisbon policy applications are in the field of 'Knowledge Economics', reflecting policy objectives in the field of Research and Development (R&D), training and education
- The version of WorldScan adapted to reflect the formation and productivity of human capital focused on the impacts of rising human capital stocks in large, fast-growing countries under globalization and indicated the policy implications for the EU.
- The PACE model (adapted in WP 7) was used to address the effects of policy measures like the recent EU energy package, taking into account that environmental consequences of policy measures can be affected by induced innovation.

Task 9.4: Model delivery to the EC
The model versions used in Task 9.3 have been delivered to the respective model clients within the EC.


- D9.1 Report on comments, suggestions and extensions needed in view of the policy assessments envisaged (due in month 16, delivered in month 25; beneficiaries 4, 9, 10 and 11 have carried out this task)
- D9.2 Report on model adaptations to GEM-E3 (due in month 24, delivered in month 37; beneficiary 10 has carried out this task)
- D9.3 Report on model adaptations to Nemesis (due in month 24, delivered in month 36; beneficiary 11 has carried out this task)
- D9.4 Report on model adaptations to WorldScan (due in month 24, delivered in month 36; beneficiary 9 has carried out this task)
- D9.5 Report on the impacts of selected policies for sustainable development with particular reference to the policies on the Lisbon agenda (due in month 36, delivered in month 37; beneficiary 10 has carried out this task)
- D9.6 Report on the impacts of selected policy objectives on the Lisbon agenda, in particular those aiming to enhance knowledge (due in month 36, delivered in month 33; beneficiary 11 has carried out this task)
- D9.7 Report on the economic role of human capital development, with special emphasis on the impacts of the rise of human capital in large, fast-growing countries under globalization (due in month 36, delivered in month 36; beneficiary 9 has carried out this task)
- D9.8 Report on the influence of induced technological progress on the environmental effects related to specific environmental taxes (due in month 36, delivered in month 37; beneficiary 4 has carried out this task)
- D9.9 The adapted version of GEM-E3 as described in D9.2 and used for the policy analysis of D9.5 (due in month 36, delivered in month 38; beneficiary 10 has carried out this task)
- D9.10 The adapted version of Nemesis as described in D9.3 and used for the policy analysis of D9.6 (due in month 36, delivered in month 36; beneficiary 11 has carried out this task)
- D9.11 The adapted version of WorldScan as described in D9.4 and used for the policy analysis of D9.7 (due in month 36, delivered in month 38; beneficiary 9 has carried out this task)
- D9.12 The new PACE model as described in D7.6 and used for the policy analysis of D9.8 (due in month 36, delivered in month 38; beneficiary 4 has carried out this task)

Significant results

The feedback of the modellers to the database developments has proven to be prerequisite for subsequent model development and policy simulations. Without this feedback and the adequate responses of WIOD-management and the WIOD-datateams the datasets would definitely have been less useful as an input to model development.
Different time-series from the datasets have been fruitfully used to improve the estimates of key parameters of the models.

Deviations from the plans

The tasks in this work package heavily relied on the new datasets made available via other work packages, notably the fully linked international analytical IO-tables and additional data on national accounts and socio-economic indicators. The delay in these datasets becoming available has carried over in delays in the appearance of deliverables D9.1 through D9.4. Apart from these delays, there have not been any significant changes from plan.

WP10 Dissemination of Results and Project Management

Consortium management tasks and achievements

With respect to the organization and administration of the project, the Project Board was in the centre. It consisted of all nine work package leaders, was the main decision-making body in the project, and set out the broad road-map of the project. In this, it was advised by a committee of external experts and a committee of stakeholders. To manage a project of this size a clear management structure is needed. Therefore, a Management Team was created to make sure that the day-to-day activities within the work packages were aligned. The management team was supported by a project administrator.

Project Board

Decisions regarding the project were made by the Project Board, which consisted of the members of the Management Team and the work package leaders. The partners that did not deliver work package leaders were represented by the leaders of the work packages they were involved in.

Management Team (MT)

The consortium consisted of a relatively small number of organizations. Nevertheless, a strong coordination structure was required to meet the objectives in time (and in accordance with the plans). The MT was created to make sure that the activities within the work packages (each of them led by one consortium member) were aligned. It was also be responsible for the scheduling and organization of consortium meetings every year.

Committee of External Experts

The Committee of External Experts offered its reflections on the progress of the project and choices to be made to the Project Board. They have been invited for several of the project meetings and have been contacted on a bilateral basis to discuss specific matters in their fields of expertise. The Committee of External Experts consisted of researchers who are widely regarded to be leading in fields relevant to the project.

Committee of Stakeholders

A Committee of Stakeholders was installed to ensure a good match between needs in the policy arena and the activities carried out by the WIOD consortium, as well as to promote wide dissemination of WIOD's methods and findings among policymakers. To maintain contacts with official statistical offices, Peter Ritzmann (EUROSTAT, later replaced by Isabelle Remond-Tiedrez) was involved. Two other members of the Committee of Stakeholders were Mark de Haan (chairman of the London Group on Environmental Accounting) and Joyashree Roy (coordinating lead author of a chapter in the IPCC's Fourth Assessment Report).


- D10.1 Launch of the project website
(due in month 1, delivered in month 1)
- D10.2 Organization and execution of data construction meeting and kick-off meeting in Amsterdam
(due in month 1, delivered in month 1; Dates: 14-15 May, 2009; Venue: Sheraton hotel, Schiphol Amsterdam)
- D10.3 Organization and execution of a data construction meeting for WP1-WP6 in Paris
(due in month 6, delivered in month 8; Dates: 3-4 December, 2009; Venue: OECD, Paris)
- D10.4 Organization and execution of first consortium meeting in Vienna
(due in month 12, delivered in month 13; Dates: 26 – 28 May, 2010; Venue: Technische Universitdt Wien, Vienna)
- D10.5 Organization and execution of Project Board meeting in Amsterdam
(due in month 18, delivered in month 19; Date: 4 November, 2010; Venue: Sheraton hotel, Schiphol Amsterdam)
- D10.6 Organization and execution of second consortium meeting, followed by a modeling meeting for WP7-WP9
(due in month 24, delivered in month 25; Dates: 25-27 May, 2011; Venue: IPTS, Seville, Spain)
- D10.7 Organization and execution of Project Board meeting
(due in month 30, delivered in month 31; Date: 28 November, 2011; Venue: ZEW, Mannheim, Germany)
- D10.8 Organization and execution of a final policy and dissemination meeting in Brussels
(due in month 36, delivered in month 36; Date: 16 April, 2012; Venue: Hotel Sofitel, Brussels, Belgium)
- D10.9 Organization and execution of an academic conference in Groningen
(due in monthe 36, delivered in month 36; Dates: 24-26 April, 2012; Venue: University of Groningen, The Netherlands)
- D10.10 Drafting and publication of progress reports to the European Commission
(months of delivery: several; delivered: several reports to the project officer and the Periodic Management Report on 17 December, 2010)
- D10.11 Drafting and publication of the Final Report
(due in month 36; delivered in month 38)
- D10.12 Drafting and publication of electronic WIOD Newsletters
(months of delivery: several; delivered in months 3, 9, 15, 20, 27)
- D10.13 Providing public-access facility to WIOD-database on dedicated website
(month of delivery: 36; delivered in month 36; the database was made public on 16 April, 2012)

Potential Impact:


In this section it will be argued that the WIOD project not only meets the expected impacts listed in the work programme, but is more ambitious. The aim is that our database will become a (if not the) major point of reference in the EU for interindustry modeling and analyses of policy relevant issues. It should be stressed though that this takes time. During the WIOD project, the participants have been very active in promoting and explaining the WIOD project and (preliminary versions of) some of its applications. Several sessions have been organized at the annual international input-output conferences (that are also attended by representatives of national statistical institutes).

Before going into the details, it may be interesting to indicate some of the potential of this project by indicating the impact of the launch of the WIOD database. This event took place on 16 April, 2012, in Brussels and was organized by DG Trade in close co-operation with RUG (University of Groningen).

The article in the Wall Street Journal (heading "New Statistics Method To Shrink EU-China Trade Gap 36% -EU Official") writes about De Gucht's speech:

A new way to measure the value of goods traded between the European Union and the rest of the world would shrink the European Union-China gaping trade deficit significantly, a top EU official said Monday. Karel De Gucht, the EU commissioner for trade, said the new statistical method, which accounts for value of separate stages of production independently rather than just where the end product is assembled, would cut the EU's trade gap with China by 36%. "Our trade relationships with key partners are different from what we previously thought. For example, when we look at trade in value as opposed to traditional statistics, our trade deficit with China is reduced by 36%," De Gucht said speaking at a conference in Brussels. … De Gucht said that using the new method "China ... starts to look like less of a problem."

… The new trade statistical method is called the "World Input-Output Database" and was developed using EU budget funds. The World Trade Organization and the Organization for Economic Cooperation and Development were consulted to create it. … De Gucht said the change in the statistical accounting for trade had been developed to take into account a globalized assembly line where different parts of the same end product are produced in different parts of the world.

The EU trade commissioner gave the example of a Nokia smartphone: "It is listed as being made in China, but in reality 54% of its value comes from tasks that are carried out in Europe. Key components are produced in other parts of Asia and only the assembly itself actually happens in China." "Today, we measure trade by counting the total price of the good that is being exported or imported," De Gucht explained, "but because we do this both for components and for final products we get a distorted picture of what is really happening."

The following citations are taken from the speech by Alejandro Jara:

-I think we need to look at two developments in international trade that are at the core of the changes we are living through. The first of these relates to the rise of key emerging economies and the shift in economic realities that this implies. The second concerns the internationalization of production processes, leading to increased inter-dependency, expanded trade ties and a more deeply shared interest in a well-functioning trading system. The WIOD project, by providing the means to measure the strength and complexity of the internationalization of global supply chains, also allows us to better understand the complementarities and rivalries between the worlds of yesterday and today.

-Let me turn to growth of supply chains. This phenomenon is not entirely new, but has become increasingly pervasive and prominent, capturing more public policy attention. International supply chains are variously referred to as vertical integration, production sharing, outsourcing and offshoring. All these designations essentially refer to the same thing — the slicing up of production processes internationally. In several regions, more and more manufacturing activities and many services industries today are characterized by supply chain production, and nearly all supply chains embody an international dimension.

-A key challenge, of course, is measurement, and it is here that the results of the WIOD project are important. It is more difficult to measure trade in value-added terms than in gross terms. … The results obtained by WIOD add to the increasing stock of knowledge of several national and international initiatives. As often happens with statistics, new data answer old questions, but also raise new questions. And these new questions are not only important for researchers, but also for policy-makers: the G-20 meeting taking place in Mexico this week will look into the consequences of international supply chains and trade in value-added on the way that the international community looks at global governance.

-Several research questions suggest themselves when we think from a policy perspective. One is precisely how to decompose the complex elements of supply chain production into their component parts, especially in respect of services.

Relating the main elements in the Call to the WIOD project

According to the Work Programme of the call (SSH-2007-2.1.3) the following goals were to be reached in the project.
- The development of databases and accounting frameworks, including harmonized input-output tables
- The development corresponding analytical tools, models and consistent expert-systems with the aim to facilitate international comparisons and structured assessments of policies, thus enabling European and world projections of the future relationship between the main socio-economic, environmental and societal issues.

Therefore the project should have at least the following characteristics.
- Cover the European countries and the global level
- Being based on international accounting conventions
- Integrate the most advanced theories and methodologies
- Take into account the main cross-sector issues.

From the description of the project in the previous sections and the work packages in particular, it should be clear that the output of the project has met these goals.
- Work packages WP1–WP5 have constructed a set of databases and accounting frameworks: supply and use tables, input-output (IO) tables, and corresponding environmental and socio-economic satellite accounts. The latest methodologies in the field of IO analysis have been used for this (which was covered in WP6).
- The databases are fully harmonized across space using common product- and industry-classifications and are constructed according to the principles of the System of National Accounts (SNA) and its elaboration including environmental and economic accounts as laid out in the System of Integrated Environmental and Economic Accounts (SEEA). In addition to international harmonization, they are harmonized over time. The accounts cover all 27 EU countries and the major countries outside the EU, together responsible for at least 85% of world GDP.
- The constructed database has been applied for direct analysis and for model applications. In WP7 and WP8, IO structural decompositions and econometric techniques have been used. In WP7 – WP9, the database has been used for building new, flexible models and adapting existing models (most of which have been used in-house by the European Commission for policy assessments). Together these models constitute a multi-purpose expert system capable of addressing a wide range of policy-issues with respect to environmental issues and socio-economic aspects of growth. In particular, by including both environmental and socio-economic accounts main cross-sector trade-offs can be analyzed.
- The consortium included Europe's leading academics in the construction of international IO tables and the compilation of internationally comparable environmental and socio-economic accounts. In particular through the consortium members, the project was able to incorporate precious knowledge and expertise obtained from previous database projects. For example, the IO database of the OECD, the environmental database constructed in the large-scale EXIOPOL project (in which WIOD beneficiaries IPTS, ZEW and RUG were consortium members), and the socio-economic accounts compiled in the EU KLEMS project (in which WIOD beneficiaries RUG, WIIW, WIFO, HTWG and TCBE were consortium members).
- At the same time, the relevance of the database construction for European-wide policy analysis was ensured through the participation of IPTS, which is a Joint Research Centre of the European Commission that provides techno-economic analysis and policy support. Also, the consortium included the institutes responsible for the MODELS project.

Three levels of impacts

It is expected that this project will generate substantial impacts for three different user groups: academic researchers, statistical agencies, and policy analysts and makers at both the national and international level.

At the first level, methodological research has taken place and further research will continue to take place. Constructing large databases usually involves ample estimation, due to the fact that certain data are not available or incomplete. Another issue is the reliability of the data, ranging from reliable to highly unreliable. But even in an ideal world, IO tables are a construct (i.e. a model) themselves. Moreover, the choice of particular assumptions depends on the use of the tables. Methodological research on these issues is relevant for future evaluation of the possible procedures for constructing world IO tables.

At the second level, the public availability of the world input-output database will allow for a large variety of studies in a wide range of topics. On the one hand, there are groups of researchers and model builders who will use only certain parts of the database because they are interested in very specific topics. For example, the composition of the labour force, investments in ICT, bilateral trade flows, CO2 emissions or water consumption for water footprints. On the other hand, some groups of researchers will use the entire database (or large parts thereof) to build a model, or to analyze a set of related issues.

The third level is with respect to the applications that will be carried out in this project. These are (i) direct applications, using IO and/or econometric techniques; and (ii) building a new model that makes maximum use of the dynamic nature of the database and adapting existing models which have proven their usefulness in the policy arena. As for the "direct" applications in WP7 and WP8, IO tables and techniques continue to be used widely to analyze all sorts of economic and policy issues.

Future impacts, threats, and expansions

An important aim of this project is to make a lasting impact, i.e. one that goes far beyond the endpoint of this project. A crucial aspect for this is whether a successful institutionalization of the database and continuation is possible. Clearly, the consortium is committed to a long-term strategy which goes beyond the length of the proposed project. This follows readily from the set-up of the project which is firmly grounded in, and builds upon, the results of a large number of existing large-scale international research activities.

In particular, the beneficiaries have participated and co-operated in earlier successful framework programmes on topics that are closely related to this project:
- EU KLEMS project: RUG has co-ordinated this project in which WIIW, HTWG, TCBE and WIFO participated;
- EXIOPOL project: RUG, IPTS and ZEW were beneficiaries;
- MODELS project: ICCS co-ordinated this project in which CPB and CRSA participated.

The world input-output database will remain to be available and will be maintained through its website (see online) at the University of Groningen. The question whether it is possible to provide future updates for the database cannot be answered at the moment. Despite the fact that frequent updates are necessary to make a lasting impact, the funds are currently lacking. It should be stressed that updating requires a lot of additional work (and thus funds) because in the near future countries will adopt the "new SNA".

Next to updating, there are many possibilities for expanding the database in a meaningful way. In setting up this project, we have adopted a rather conservative scenario with respect to the construction of the database in terms of the countries covered. This was because we were aiming at a high quality of the data, that is, close to statistics from official sources and fully compatible with National Accounts data. For a large set of countries the consortium members had extensive experience with available statistical sources through various previous data projects (such as EU KLEMS, EXIOPOL and OECD work), and then on-going research activities (e.g. of RUG and TCBE in China and India). By this conscious limitation, we focused on providing at least the outputs that had been promised in the description of the project.

More years. For inter-temporal comparisons and the study of dynamic features of societal, economic and environmental issues it is desirable to have a time series of considerable length, that is fully comparable over time. In particular economic processes (such as technological changes and changes in the structure of production) evolve at a different pace in different countries. "Mature" economies typically exhibit smooth patterns, while some transition economies have shown periods of rapid changes. A longer time series allows to improve both the scope of the ex-post analyses and the quality of the dynamic aspects of the models with which ex-ante projections and evaluations of scenarios are made.

More countries. Due to the same reason, it may be expected that more and more countries outside Europe will start to publish supply and use tables (or IO tables). This will allow for a further expansion of the number of non-European countries in the database. In this project, the coverage is approximately 85% of world GDP. Still, an expansion would be very important from a socio-economic and environmental point of view. In terms of world population and questions regarding sustainable economic development, development aid, or migration issues, it seems crucial to include more developing countries, in particular in Africa and Asia. Also from an environmental and resource perspective it is crucial to expand the database with countries in the Middle East.

More information. An expansion of the database with new types of information is desirable. One aspect that will gain in importance is the treatment of disposals to nature, such as solid waste. This is no longer a local problem once it is transported from one place to another. In the literature, Kagawa et al. (2004) have presented an interregional make and use framework for "industrial waste" and have used a multi-regional IO account for Japan. The ideal case would be to develop an inter-country waste table at a global level. For example, it would allow for investigating how an increase in German private consumption of metal products affects the output of the French mining industry and its consequent "export" of waste to Spain.

More aspects. Adapting the database to "new" requirements in order to properly analyze certain phenomena. Outsourcing (or offshoring) – where certain parts of the manufacturing process or the provision of some services are done by low-wage countries instead of in the home country – is a phenomenon that has become important in international trade and is expected to expand further. Using standard IO tables, stimulating domestic production (e.g. by an increase in exports) affects domestic production in many other industries.

List of Websites: