This chapter provides advice on the dissemination of extended supply and use tables, extended input-output tables, and related indicators. It starts with general advice on the dissemination process, emphasising the importance of identifying the users and their specific needs. It then outlines different ways to reach users, ranging from the tables themselves and related analysis to graphs and videos. Special attention is given to confidential data, with guidance on how to handle this while still publishing as much relevant information as possible. The chapter concludes by proposing several indicators, both basic and more complex, that can be derived from extended supply and use tables and extended input-output tables.
Handbook on Extended Supply and Use Tables and Extended Input‑Output Tables
Abstract
Data dissemination involves the release of information obtained through a statistical activity. It consists of transmitting statistical data to the users through various media (paper publications, files available to specific users or for public use, official websites, social networks, public speeches, and press releases). Data dissemination is a crucial part of statistical activity as most official statistics, including extended supply and use tables (ESUTs) and extended input-output tables (EIOTs), are produced for public use. Dissemination provides visibility to the results and contributes to spreading statistical information and economic analysis to the public. Effective data dissemination requires good communication to make the information both accessible and clear to users. This process involves identifying user needs, determining what information will be released, communicating the dissemination schedule, ensuring timeliness and coherence among the disseminated data sets, maintaining statistical confidentiality, disseminating metadata and information on data quality, selecting formats (which are regularly reviewed for alignment with demand and standard practices), and means of dissemination.
Dissemination rules and good practices are largely similar to those described in the Handbook on Supply and Use Tables and Input-output Tables with Extensions and Applications (United Nations, 2018[1]). The purpose of this section is therefore not to provide detailed general principles of dissemination of statistics but to reiterate the main principles and to propose good practices of disseminating ESUTs to countries.
Identify users and their needs
Identifying the users and their needs is crucial when determining the appropriate level of information, explanations and messages and developing a good dissemination strategy. Promotion on professional social media can help to identify and reach more users and to understand their needs. Generally, there are two main types of users. First, expert users with analytical needs and regular data requirements for detailed variables, breakdowns and time series. Second, general users who need fewer details but more guidance on the interpretation and value added of statistics related to ESUTs and EIOTs. Good knowledge of users’ needs allows better prioritising the appropriate breakdowns in the supply and use tables (SUTs), providing more details on the behaviour of specific categories of enterprises that are of interest (e.g. by size, trading status and ownership).
The availability of a release calendar is an important element of the dissemination strategy that allows users, especially institutional users, to plan their activities. An advance release calendar with the first dissemination and revisions should be available at the beginning of each year, or at least well before the release date, on the websites of the national offices in charge of the dissemination. Punctuality (i.e. adherence to the schedule) is important too. It is one of the dimensions of quality in the Quality Framework and Guidelines for OECD Statistical Activities (OECD, 2011[2]).
Presentation of the ESUTs, EIOTs and related analysis
While ESUTs and EIOTs can address policy questions regarding the role of certain types of enterprises, they remain complex. Thus, it is crucial to present and explain the ESUTs, EIOTs and related indicators in the most user-friendly manner possible. It is recommended to make ESUTs, EIOTs, the analysis and all related publications available together, as they complement each other. This enables users to find all relevant information about this topic quickly. For the first release, it may be appropriate to organise specific media briefings or a conference for specific users to present insightful information derived from the findings. This may also be appropriate when there are substantial changes in the methodology or interesting key results.
The following elements can support the dissemination process as well.
Provide clear and detailed metadata. Metadata may be defined as “everything one needs to know to understand the results”. They describe the complete statistical production process. They ease sharing, querying, understanding and using statistical data at their various levels of aggregation. Metadata encompasses the data or data collection, concepts, definitions, classifications, thresholds, and main statistical processing before the compilation, dissemination and storage of the results in a database. Metadata are a very important tool for understanding the statistical process, how the data should or should not be used, together with the nature of the information they contain. For ESUTs, it is important to detail all the microdata linking and the sources used for the extension of the SUTs, as well as the methods and sources that were used to delineate the different enterprise characteristics.
Use different ways to disseminate the tables and related analyses. The ESUTs’ data can be disseminated in various forms and formats, for example in Excel, with several worksheets covering different parts of the ESUT framework for a specific year to enhance clarity. Many statistical offices have developed a large and detailed open database as well as an API (application programming interface) allowing users to select the specific data they need, view the data as graphs or download the data for use.
It remains a challenge to digest the main points from the data because of the possible large size of the tables. Although this is desirable for a specific audience that is interested in detailed information, this makes it difficult to analyse and visualise for the general public. Therefore, it is recommended to present aggregated information as well, for example, the main indicators for each type of enterprise in the table or the main indicators for manufacturing, services, possibly another industry grouping depending on data availability, by type of enterprise. Alternatively, focus on the larger industries, or the larger industries that are policy relevant, and provide the indicators by enterprise type.
Graphs can make the information more accessible and effectively highlight some specific points related to the specific ESUT/EIOT. Several countries use graphic presentations to complement their ESUTs with basic explanations of the main results. Such initiatives can be helpful to complement the ESUTs, illustrating the main global messages. For example, Sallusti and Cuicchio (2023[3]) compiled three types of ESUTs for Italy. These tables detail industries for three types of groupings: size, trading status and ownership. Examples of their work are described below.
Note: MNE: multinational enterprise.
Source: Sallusti and Cuicchio (2023[3]).
From Figure 8.1, the authors derived four key messages: MNEs employ 20% of total workers, there is a productivity gap between MNEs and domestic enterprises (over EUR 94 000 vs. less than EUR 73 000 per worker respectively), per capita compensation is slightly higher for domestic enterprises, and MNEs show a higher mark‑up on labour costs (measured as value added divided by labour costs).
Source: Sallusti and Cuicchio (2023[3]).
Three key messages from Figure 8.2 are that 66% of overall value added is generated by enterprises that neither import nor export; two-way traders account for 25% of value added; and two-way traders are relatively large in manufacturing, trade and transportation.
Source: Sallusti and Cuicchio (2023[3]).
Three key messages from Figure 8.3 are that 51% of value added is generated by microenterprises (1 to 9 persons employed); 21% of value added is generated by large enterprises (250 persons employed or more); and large enterprises are relatively more prominent in mining and quarrying, energy, and transportation.
INEGI (2023[4]) has implemented a similar framework for Mexico. It presented the main results using visuals that are easy to understand, while emphasising the main points of interest. For example, Figure 8.4 clearly shows that exporters in Mexico account for 35.4% of domestic production. Exporting foreign-owned large enterprises account for 6.9% of domestic production.
Note: S: small enterprises, M: medium-sized enterprises, L: large enterprises.
Source: INEGI (2023[4])
A short pedagogic video can present ESUTs and EIOTs, their uses, and meaning through appropriate examples. For example, Israel (CBS Israel, 2022[5]) and Singapore (Singapore Department of Statistics, 2020[6]) have each designed a short clip that illustrates the function of an SUT through concrete examples (olive oil production for Israel and oil refining industry for Singapore). The European Commission (2021[7]) provided a short pedagogic video to explain FIGARO (Full International and Global Accounts for Research in Input-Output analysis) and its possible uses. Such user-friendly videos can be an attractive way to make ESUTs, EIOTs and related indicators accessible to a larger audience. They can help explain why it is so important to introduce additional dimensions such as the size of the enterprise or its involvement in the globalisation process (trading status, ownership) and highlight insights offered by the new data.
Compiling ESUTs or EIOTs allow new indicators to be derived. These may relate to the type of enterprise in the domestic economy, links in the domestic economy or links with foreign economies. The links with foreign economies provide new insights into the participation of different types of enterprises in global value chains. It is worth noting that some indicators are very similar to existing indicators from regular SUTs and input-output tables, or to Trade in Value Added analysis (TiVA) (Martins Guilhoto, Webb and Yamano, 2022[8]), while others are new.
As an illustration, this section presents indicators for the case where industries are broken down by size class. Of course, all indicators can be calculated in a similar way when the industries are broken down by another category. Furthermore, note that one can calculate the indicators for the total economy or by industry as well, not using a breakdown by enterprise type. As indicated in Chapter 1, this will lead to higher quality estimates when heterogeneity within industries is high and disaggregated information is of high quality.
Basic indicators about the domestic economy by size class
Using ESUTs
production
value added
gross operating surplus
compensation of employees
value added as a share of production
compensation of employees as a share of value added.
The first four indicators show the share of each size class in well-known macroeconomic totals. These can be obtained using business statistics as well. However, such estimates can be inconsistent, do not entail illegal activities and are not comparable to macroeconomic aggregates such as gross domestic product. Value added as a share of production indicates whether production is relatively in-house or that inputs of goods and services are sourced mainly outside the enterprise, from domestic or foreign suppliers. It tells something about the cost structure. Compensation of employees as a share of value added shows how much value added goes to the providers of labour and how much to the providers of capital. This indicator is similar to the labour share. Its inverse, value added divided by compensation of employees, can be used to calculate the mark-up on labour costs, which is 100%*(this inverse minus 1).
Using EIOTs
Just as in regular SUTs, ESUTs show which products are imported and exported, but not who is importing or exporting. This information can be obtained from an input-output table or an EIOT. Combining information about imports and exports with production and value added yields several indicators describing globalisation. These can be by size class or by size class and industry:
Imports.
Exports.
Import intensity, defined as imports divided by production. This shows the importance of own/direct imports for a size class, how much it directly depends on foreign countries for its inputs.
Export intensity, defined as exports divided by production. This shows the importance of own/direct exports for a size class, how much it directly depends on foreign countries for its sales.
Value added due to direct exports, also known as direct domestic value-added content of exports, defined as export intensity times value added. This shows how much of its value added is directly related to foreign sales. This value added can be further split into compensation of employees and gross operating surplus.
The three latter indicators are examples of OECD’s Trade in Value Added (TiVA) indicators (Martins Guilhoto, Webb and Yamano, 2022[8]) and the Eurostat’s Macroeconomic Globalisation Indicators (Eurostat, 2024[9]).
Indicators using extended input-output tables by size class
Using an EIOT can yield estimates for indirect trade and the corresponding value added, for example, how much small and medium-sized enterprises (SMEs) depend on large enterprises to provide them access to emerging markets, or to what extent large manufacturing enterprises depend on SMEs in the services industries. Combining information about indirect trade with that of direct trade shows foreign dependence of a size class at total level.
To facilitate computations with the EIOT, it may be advisable to group the different size classes together instead of grouping the different industries together. This is depicted in Table 8.1.
Note: SME: small and medium-sized enterprise. C: final consumption expenditure; GCF gross capital formation; X: exports
Standard input-output analysis yields a table that shows how much value added one type of enterprise has due to production that is ultimately used by a type of enterprise to produce for domestic final use (such as household consumption) or exports. For example, the value added of SMEs due to production for the exports by large enterprises. This input-output analysis can be carried out as follows:
Set f1 and f2 as the vectors of domestic final use and exports, respectively.
Set x as the vector of production and Int as the matrix of intermediate supplies.
The technical coefficient matrix A is defined as usual; each of the columns of Int is divided by the column’s total (an element of x).
The Leontief matrix L is also defined as usual, namely L = (I-A)-1, where I is the identity matrix of the appropriate dimension. Then x = L·(f1 + f2).
Set v as the vector of value added per unit of production.
For a vector c, diag(c) is the matrix with vector c on the diagonal and 0 elsewhere.
Set B = diag(v)·L·diag(f2).
The matrix B is a who-to-whom matrix; a matrix element bij represents the value added at (industry by size class) i due to exports of (industry by size class) j. Or, interpreting the data the other way around, given exports of (industry by size class) j, how much value added is at (industry by size class) i. Direct exports are defined as the production of a size class that is ultimately used in its own exports. These are the diagonal elements of L·diag(f2). Define indirect exports as production of one size class that is ultimately used in exports of another size class. These are the off-diagonal elements of L·diag(f2).
Table 8.2 depicts a who-to-whom table. The numbers in the table refer to parts of the who-to-whom table that will be used in later calculations.
Note: SME: small and medium-sized enterprise.
Indicators
Value added due to indirect exports, when one produces goods and services that are ultimately embodied in exports by others in other size classes. For SMEs, this is equal to 3+4+5+8+9+10.
Value added due to exports is calculated by summing value added due to direct and indirect exports.
Value added due to exports relative to total value added. This indicator shows how important foreign sales are for this particular size class.
Value added due to indirect exports by size class and exporting industry (e.g. manufacturing, services and other). This is a further breakdown of the total value added due to indirect exports. It shows how size class A is dependent on the exports of which industries in size class B. For example, SMEs have value added due to direct exports of large manufacturers equal to 3+8.
Value added of others supplying for one’s exports, by size class and supplying industry (e.g. manufacturing, services and other). It shows how the exports of size class A enable its suppliers in size class B, and in which industries, to reach foreign markets indirectly. For example, exports of SMEs embody 11+12 value added by large manufacturers.
Domestic value added embodied in exports, by size class, is calculated as value added due to direct exports plus value added at other size classes supplying for the exports of this size class. For example, the exports of large enterprises in services embody 4+9+14+19+24 domestic value added.
Domestic value added embodied in exports by size class as a share of exports by this size class. This shows how much of the exports are locally produced. One minus this share is the foreign content of exports, showing the dependency of the size class on foreign inputs.
Value added generated by domestic suppliers for one unit of domestic value added due to direct exports, by size class, shows a multiplier effect. For example, the multiplier effect of SMEs in services is (2+12+17+22)/7.
Similar indicators can be calculated for value added embodied in production for final domestic use and for value added embodied in production for total final use. While these indicators are not that common, they may be useful for large economies such as the People’s Republic of China or the United States which rely relatively more on domestic consumption than smaller open economies such as Belgium and the Netherlands, which rely more on foreign demand.
The previous indicators are related to exports, providing information about the supply to foreign markets, by size class. Note that many indicators related to imports can be calculated in a similar way. That would provide information about the use of foreign inputs, by size class. Combining them with indicators about domestic inputs from other size classes and the own value added would provide insights into the input structure of industries by size class.
Country experiences
Statistics Finland regularly produces and publishes numbers based on linked trade and business statistics in combination with national accounts. Currently, seven tables are publicly available (Statistics Finland, 2024[10]). They contain TiVA indicators where industries are split by size class and group relation, ownership, or trading status, for example, the direct value added of gross exports, or the total domestic value added in gross exports of SMEs. To put this in a broader perspective, gross imports, gross exports and value added are provided too. This allows gauging the importance of exports for a specific type of enterprise in a specific industry. Similar information is available for employment related to exports. The data are available yearly for the period 2017-22. OECD and Statistics Finland (2020[11]) used a previous version to present granular insights on Finland’s integration into global value chains.
Figure 8.5 presents information for 2022, differentiating enterprises by size and group relation. An enterprise is said to be dependent when it has intra-enterprise relations with larger entities and independent when it does not. Dependent enterprises attribute more of their value added to foreign demand than independent enterprises do. For example, dependent enterprises with 50-249 persons employed have 28% of their value added due to their own (direct) exports, compared to 17% for independent enterprises of the same size. That is because a larger share of the turnover of dependent enterprises consists of exports. Differences between dependent and independent enterprises in the share of value added attributed to indirect exports (producing goods and services in the supply chain of an exporter) are substantial as well.
Source: Calculations based on Statistics Finland (2025[12]).
Statistics Netherlands had several projects examining the role and interplay of SMEs, large enterprises and multinationals. One such example is Onat et al. (2018[13]). For 48 industries, a Dutch input-output table was split into five categories: 1) foreign multinational; 2) Dutch multinationals large enterprises; 3) Dutch multinational SMEs; 4) non-multinational large enterprises; and 5) non-multinational SMEs. Enterprises in 17 industries, such as agriculture, financial services and government services, were not split into the five categories but were classified under another category “other”. Subsequently, different indicators of the mutual interdependencies were calculated (Exel et al., 2018[14]). Table 8.3 shows value added in a type of enterprise due to exports of a type of enterprise.
Non-multinational SMEs have more value added due to indirect exports than due to direct exports. The other types of enterprises serve as a channel through which non-multinational SMEs reach foreign markets. In contrast, the value added due to exports for multinationals is mostly due to their own (direct) exports.
Confidentiality should remain a core principle in dissemination, with specific attention paid to the confidentiality rules to avoid any possible direct or indirect identification. In ESUTs, attention should be paid to the adequate degree of granularity for dissemination. Chapter 3 discussed this problem. General guidance on preserving confidentiality is provided in Chapter 6 in the Handbook on Integrating Business and Trade Statistics (United Nations, forthcoming[15]).
Disaggregating data leads to greater granularity, which is bound to lead to confidentiality issues and the necessity to safeguard that confidentiality. Confidentiality problems arise when the sample of remaining enterprises producing or using a given product is too small, or when data about an individual enterprise might be deduced, for example when there is a dominating enterprise. In general, national statistical offices are not allowed to publish such data by law. Furthermore, Principle 6 of the Fundamental Principles of Official Statistics stipulates that data collected by statistical agencies for statistical compilation about individual entities are to be kept strictly confidential. Use is allowed only for statistical purposes, regardless of whether the entity is a natural or legal person. Preserving confidentiality is necessary to gain and maintain the trust of survey respondents.
There are several ways to address this issue in an ESUT setting. Two of the methods, namely aggregating industries and aggregating the extension, are described in the country experiences below. The Handbook on Supply, Use and Input-output Tables with Extensions and Applications (United Nations, 2018[1]) mentions a third option. In the case with one or two dominant producers in an industry, it recommends, “when necessary, that specific permission is sought from the business when their data are publicly available from other public sources, for example published company annual reports and accounts”.
Country experiences
This section present case studies from Mexico and the Netherlands. Note that generally it is advisable to work at a level as detailed as possible to allow for flexibility in aggregation later. If the aggregation step occurs at the beginning, this is no longer possible.
Mexico’s ESUTs (INEGI, 2023[4]) were developed at the 822 industries level. Each industry in the regular SUT was expanded according to the different focuses. However, it was clear that publishing the data at this level of disaggregation would not be possible due to the likely breaches of confidentiality.
The Mexican solution was to aggregate the 822 industries into the 20 main sectors of the North American Industry Classification System (NAICS). In that way, the ESUTs were able to provide sufficient granularity – as each industry was disaggregated – without risking the identification of particular entities. A more resource-intensive solution would have been to analyse each cell under the 822 industries in the ESUTs and compare it with the individual records from the Economic Censuses. For example, if, in a particular industry (say 325130 – Synthetic Dye and Pigment Manufacturing), the Economic Censuses (source) data show that there are only three exporters, then such an industry could be aggregated with another one (close to it) under the export-oriented focus. However, if the same industry shows ten small enterprises, then it can be disaggregated under the size of enterprise focus. The advantage of this approach is that it provides a maximum amount of detail. The disadvantage is that it might lead to discrepancies when comparing between extensions. However, those discrepancies could be addressed by comparing the focuses according to the least disaggregated one.
The Dutch ESUTs (Chong et al., 2019[16]) were developed at 128 industries. Of these, 115 industries were split by size and 13 (e.g. agriculture, financial intermediation, government services) were not. The published tables contained 56 industries split by size and 12 industries that were not. Even after this aggregation step, some confidentiality issues appeared. The guiding principle to solve them was to leave as much as possible relevant information for the user. One can choose to aggregate industries or to aggregate size classes.
An example of the industry aggregation is information services and IT services. In these industries, there is a lot of heterogeneity when it concerns size. For instance, IT services can range from a one-person enterprise to a large conglomerate. It seems better to keep the detailed information about size and aggregate the two industries instead. The Handbook on Supply, Use and Input-output Tables with Extensions and Applications (United Nations, 2018[1]) also recommends that “The aggregation of industries and products with non-disclosive industries and products should, however, be avoided, as this results in the loss of useful details for non-disclosive industries and products.”
An example of aggregation by size class is mining of oil and gas. In economic reality, this industry is not operated by small enterprises. Therefore, it was not split by size. Similarly, veterinary services, predominantly comprising small enterprises, were also not split by size.
References
[5] CBS Israel (2022), What is an Input-output Table?, https://www.youtube.com/watch?v=W5wwja5_Wrg.
[16] Chong, S. et al. (2019), “The role of small-and medium-sized enterprises in the Dutch economy: An analysis using an extended supply and use table”, Journal of Economic Structures, Vol. 8, pp. 1-24, https://doi.org/10.1186/s40008-019-0139-1.
[7] European Commission (2021), Eurostat FIGARO, https://www.youtube.com/watch?v=oRtP5J_7V1Q.
[9] Eurostat (2024), Macroeconomic globalisation indicators based on FIGARO. Insights into the measurement of value added and employment in the EU - 2024 edition, Eurostat’s Statistical Working Papers Series, Luxembourg: Publications Office of the European Union, 2024, https://ec.europa.eu/eurostat/web/esa-supply-use-input-tables/database.
[14] Exel, J. et al. (2018), Effecten in de waardeketen, Statistics Netherlands, https://www.cbs.nl/-/media/_excel/2018/41/tabellenset-multinationals-effecten-in-de-waardeketen.xlsx.
[4] INEGI (2023), Extended Supply and Use Tables: Base Year 2018, National Institute of Statistics and Geography, Mexico, https://en.www.inegi.org.mx/programas/coue/2018.
[8] Martins Guilhoto, J., C. Webb and N. Yamano (2022), “Guide to OECD TiVA Indicators, 2021 edition”, OECD Science, Technology and Industry Working Papers, No. 2022/02, OECD Publishing, Paris, https://doi.org/10.1787/58aa22b1-en.
[2] OECD (2011), Quality Framework and Guidelines for OECD Statistical Activities, OECD, Paris, https://one.oecd.org/document/STD/QFS(2011)1/en/pdf.
[11] OECD and Statistics Finland (2020), Globalisation in Finland: granular insights into the impact on businesses and employment, OECD and Statistics Finland, https://doi.org/10.13140/RG.2.2.22456.16640.
[13] Onat, E. et al. (2018), Multinationals en niet-multinationals in de Nederlandse economie, Statistics Netherlands.
[3] Sallusti, F. and S. Cuicchio (2023), Towards the Compilation of eSUT for Italian Economy, Paper presented at the 2023 IIOA Conference in Alghero, https://www.iioa.org/conferences/29th/papers/files/4861_sallusti_cuicchio_ISTAT_eSUT.pdf.
[6] Singapore Department of Statistics (2020), What Are Supply and Use Tables?, https://www.youtube.com/watch?v=vFNqVHs_j4E.
[12] Statistics Finland (2025), Experimental Statistics, trade in value added, Statistics Finland’s free-of-charge statistical databases, https://pxdata.stat.fi/PxWeb/pxweb/en/Kokeelliset_tilastot/Kokeelliset_tilastot__tiva/koeti_tiva_pxt_12t1.px/.
[10] Statistics Finland (2024), Experimental Statistics, trade in value added, https://pxdata.stat.fi/PXWeb/pxweb/en/Kokeelliset_tilastot/Kokeelliset_tilastot__tiva/.
[1] United Nations (2018), Handbook on Supply, Use and Input-output Tables with Extensions and Applications, United Nations, New York, NY, https://doi.org/10.18356/9789213582794.
[15] United Nations (forthcoming), Handbook on Integrating Business and Trade Statistics, United Nations, forthcoming.