Handbook on Supply and Use

Tables and Input Output-Tables

with Extensions and Applications
 ST/ESA/STAT/SER.F/74/Rev.1

Department of Economic and Social Affairs

Statistics Division
Studies in Methods Series F No.74, Rev.1
Handbook of National Accounting

Handbook on Supply and Use Tables and

Input-Output Tables with Extensions and
Applications

Edited white cover version

United Nations
New York 2018
Department of Economic and Social Affairs
The Department of Economic and Social Affairs of the United Nations is a vital interface between
global policies in the economic, social and environmental spheres and national action. The
Department works in three main interlinked areas: (i) it compiles, generates and analyses a wide
range of economic, social and environmental data and information on which United Nations
Member States draw to review common problems and to take stock of policy options; (ii) it
facilitates the negotiations of Member States in many intergovernmental bodies on joint courses
of action to address ongoing or emerging global challenges; and (iii) it advises interested
Governments on the ways and means of translating policy frameworks developed in United
Nations conferences and summits into programmes at the country level and, through technical
assistance, helps build national capacities.

Note
The designations employed and the presentation of the material in the present publication
do not imply the expression of any opinion whatsoever on the part of the United Nations
concerning the legal status of any country or of its authorities, or the delimitations of its frontiers.
The term “country” as used in this report also refers, as appropriate, to territories or areas. The
designations of country groups are intended solely for statistical or analytical convenience and do
not necessarily express a judgement about the stage reached by a particular country, territory or
area in the development process. Mention of the names of firms and commercial products does not
imply endorsement by the United Nations. The symbols of United Nations documents are
composed of capital letters and numbers.

ST/ESA/STAT/SER.F/74/Rev.1
UNITED NATIONS PUBLICATION
Sales No.:
ISBN: 978-92-1-1
eISBN: 978-92-1-0
Copyright © United Nations, 2018
All rights reserved
 Preface and acknowledgements

The present Handbook on Supply and Use Tables and Input-Output Tables with Extensions
and Applications has been prepared as part of a series of handbooks on national accounting in
support of the implementation of the System of National Accounts 2008 (2008 SNA). The
objective of this Handbook is to provide step-by-step guidance for the compilation of supply and
use tables (SUTs) and input-output tables (IOTs) and an overview of the possible extensions of
SUTs and IOTs which increase their usefulness as analytical tools.

Preparation of the Handbook started as an update of the 1999 United Nations publication
entitled Handbook of National Accounting: Input-Output Table Compilation and Analysis, 1 with
the aim of incorporating changes in the underlying international economic accounting standards,
most notably the 2008 SNA, and classifications; extending the scope of the Handbook to provide
fuller coverage of SUTs; and providing practical compilation guidance for countries with advanced
and less advanced statistical systems. In this process, however, the Handbook has also evolved to
include an innovative approach to the compilation of SUTs and IOTs in the following three main
areas: first, the underlying use of an integrated approach to statistics; second, the use of a business
model for the compilation of SUTs and IOTs linking the various parts through the compilation
scheme known as the “H-Approach”; and, third, the mainstreaming of environmental
considerations.

The Handbook builds on the experience, practices and guidance available at national and
regional level, including the Eurostat Manual of Supply, Use and Input-Output Tables (Eurostat,
2008). It provides a consistent worked example of SUTs and IOTs, which runs throughout the
chapters (as far as practically possible) in order to facilitate understanding of the various
compilation steps. It also provides examples of best practices to illustrate certain aspects of the
compilation of SUTs, along with clear recommendations, principles and guidelines in order to
ensure best practice.
For the preparation and drafting of the Handbook, an editorial board was established in
May 2013, comprising 12 members and the United Nations Statistics Division. The editorial board
members were leading international experts, including members of the International Input-Output
Association, with decades of accumulated knowledge and experience from different regions and
from different institutions, such as national statistical offices, central banks, international
organizations and the academic community.

An editor (Sanjiv Mahajan, Office for National Statistics, United Kingdom) was appointed
to lead the work of the editorial board and coordinate the contributions of experts for the various

1
ST/ESA/STAT/SER.F/74, Sales No. E.99.XVII.9.

i
chapters. Initial drafts of the chapters were prepared by members of the editorial board, including
the editor. These were further refined and aligned by the editor in liaison with respective members
of the board and the United Nations Statistics Division into a coherent set of chapters. This was
achieved through many bilateral electronic communications between the editor and chapter
authors, a face-to-face meeting of all board members in New York in May 2014, and a final
editorial board review prior to a global consultation.

The Handbook is therefore the outcome of a collaborative team effort led by the editor in
liaison with the United Nations Statistics Division and the editorial board. This team comprises
the following:
• Sanjiv Mahajan, editor Office for National Statistics, United Kingdom
• Joerg Beutel Konstanz University of Applied Sciences, Germany
• Simon Guerrero Central Bank of Chile
• Satoshi Inomata Institute of Developing Economies, Japan External Trade
Organization
• Soren Larsen Statistics Denmark
• Brian Moyer Bureau of Economic Analysis, United States of America
• Isabelle Remond-Tiedrez European Commission, Eurostat
• José M. Rueda-Cantuche European Commission, Joint Research Centre
• Liv Hobbelstad Simpson Norway
• Bent Thage Denmark
• Catherine Van Rompaey Statistics Canada
• Piet Verbiest Statistics Netherlands
• Ilaria Di Matteo United Nations Statistics Division

The editorial board members contributed initial draft chapters and a detail review of all the
chapters in the various rounds of consultation. Substantive contributions on specific topics,
including initial draft chapters, were provided by the editorial board members as follows: Joerg
Beutel (transforming SUTs into IOTs, compiling physical SUTs (PSUTs) and environmentally
extended IOTs (EE-IOTs), extension of SUTs and IOTs and modelling applications of IOTs);
Simon Guerrero (examples of country practices); Satoshi Inomata (multi-country SUTs and IOTs);
Soren Larsen (compiling the use table); Brian Moyer (compiling the import use table and domestic
use table); José M. Rueda-Cantuche (transforming SUTs into IOTs and projecting SUTs and
IOTs); Liv Hobbelstad Simpson (guidance for countries with limited statistical resources and
examples of country practices); Bent Thage (classification of industries and products, compiling
the supply table, use table, valuation matrices, import use table and domestic use table, and
transforming SUTs into IOTs); Catherine Van Rompaey (regional SUTs); and Piet Verbiest
(compiling SUTs in volume terms and balancing). The editor also provided substantive
contributions to these topics, initial draft chapters and all other topics in the Handbook, and brought

ii
all the material together through numerous iterations with editorial board members reflecting many
changes and improvements.

The contributions by the editor and the members of the editorial board and their
commitment to the Handbook are very much acknowledged and appreciated. The following
specific contributions are also acknowledged: Joerg Beutel, in formatting and standardizing tables,
charts, boxes and figures throughout the Handbook; Ilaria Di Matteo, in reorganizing the chapters
and ensuring overall coherence and consistency of the Handbook; and Erwin Kolleritsch (Statistics
Austria), in kindly providing and checking much of the empirical data supporting the SUTs and
IOTs in parts two and three of the Handbook.

The Handbook also benefited from specific inputs provided by Issam Alsammak (Statistics
Canada), Gary Brown (Office for National Statistics, United Kingdom), Andrew Cadogan
(Australian Bureau of Statistics), Duncan Elliot (Office for National Statistics, United Kingdom),
Antonio F. Amores (European Commission Joint Research Centre), Ziad Ghanem (Statistics
Canada), Manfred Lenzen (University of Sydney, Australia), Bo Meng (Institute of Developing
Economies, Japan External Trade Organization), Louis de Mesnard (University of Bourgogne,
France), Carol Moylan and Tom Howells (Bureau of Economic Analysis, United States), Jan
Oosterhaven (University of Groningen, Netherlands), Ole Gravgard Pedersen (Statistics
Denmark), Xesús Pereira (University of Santiago de Compostela, Spain), Joao Rodrigues
(Technical University of Lisbon, Portugal), Jaroslav Sixta (Czech Statistical Office), Silke Stapel-
Weber (European Commission, Eurostat), Umed Temursho (European Commission Joint
Research Centre), Norihiko Yamano and Nadim Ahmad (OECD), and Herman Smith, Julian
Chow, Gulab Singh, Benson Sim and Alessandra Alfieri (United Nations Statistics Division).

Feedback was also received from participants at various meetings and conferences, most
notably the annual International Input-Output Association (2014, 2015 and 2016) and various
regional national accounts meetings. The Handbook has benefited greatly from the numerous
useful comments and suggestions made by national statistical offices, central banks, regional
commissions, academic associations and international organizations, and also by the
Intersecretariat Working Group on National Accounts during the global consultation in the period
August to October 2017.

The Handbook was prepared under the supervision of Herman Smith and the overall
responsibility of Ivo Havinga, both of the United Nations Statistics Division.

iii
iv
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Contents

PREFACE AND ACKNOWLEDGEMENTS ........................................................................................................... I

CONTENTS ................................................................................................................................................................ V
ABBREVIATIONS ............................................................................................................................................... XVII
PART ONE...................................................................................................................................................................1
CHAPTER 1. INTRODUCTION .........................................................................................................................3
A. BACKGROUND ...................................................................................................................................................3
B. USES OF SUTS AND IOTS ..................................................................................................................................5
C. SYSTEM OF NATIONAL ACCOUNTS ....................................................................................................................7
D. OBJECTIVES OF THIS HANDBOOK .................................................................................................................... 12
E. STRUCTURE OF THE HANDBOOK...................................................................................................................... 15
PART TWO ............................................................................................................................................................... 21
CHAPTER 2. OVERVIEW OF THE SUPPLY AND USE TABLES AND INPUT-OUTPUT TABLES ... 23
A. INTRODUCTION................................................................................................................................................ 23
B. OVERVIEW OF SUTS ....................................................................................................................................... 23
C. OVERVIEW OF IOTS ........................................................................................................................................ 35
D. STRUCTURE OF SUTS AND IOTS: BASIC ELEMENTS......................................................................................... 38
E. COMPILING SUTS AS AN INTEGRAL PART OF THE NATIONAL ACCOUNTS ......................................................... 57
CHAPTER 3. BUSINESS PROCESSES AND PRODUCTION STAGES ..................................................... 69
A. INTRODUCTION................................................................................................................................................ 69
B. INSTITUTIONAL ARRANGEMENTS..................................................................................................................... 70
C. OVERVIEW OF THE GENERIC STATISTICAL BUSINESS PROCESS MODEL .......................................................... 71
D. OVERALL STRATEGY FOR THE COMPILATION OF SUTS AND IOTS ................................................................... 75
ANNEX A TO CHAPTER 3: EXAMPLES OF INSTITUTIONAL ARRANGEMENTS IN COUNTRIES .... 90
A. CENTRALIZED PRODUCTION OF ECONOMIC STATISTICS: CANADA ................................................................... 90
B. CENTRALIZED PRODUCTION OF ECONOMIC STATISTICS: NORWAY .................................................................. 91
C. CENTRALIZED PRODUCTION OF ECONOMIC STATISTICS: UNITED KINGDOM .................................................... 92
D. DECENTRALIZED PRODUCTION OF ECONOMIC STATISTICS: CHILE ................................................................... 93
E. DECENTRALIZED PRODUCTION OF ECONOMIC STATISTICS: UNITED STATES OF AMERICA ............................... 95
CHAPTER 4. SPECIFY NEEDS, DESIGN, BUILD AND COLLECT STAGE ........................................... 99
A. INTRODUCTION................................................................................................................................................ 99
B. SPECIFY NEEDS, DESIGN AND BUILD PHASES ................................................................................................... 99
C. COLLECT PHASE ............................................................................................................................................ 120
PART THREE ......................................................................................................................................................... 129
CHAPTER 5. COMPILING THE SUPPLY TABLE .................................................................................... 131
A. INTRODUCTION.............................................................................................................................................. 131
B. STRUCTURE OF THE SUPPLY TABLE ............................................................................................................... 131
C. DOMESTIC OUTPUT ........................................................................................................................................ 135

v
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

D. IMPORTS OF GOODS AND SERVICES ................................................................................................................ 145

ANNEX A TO CHAPTER 5: SAMPLE QUESTIONNAIRE COLLECTING SALES OF GOODS AND
SERVICES, INVENTORIES OF GOODS AND TRADE-RELATED DATA .................................................. 153
CHAPTER 6. COMPILING THE USE TABLE ............................................................................................ 157
A. INTRODUCTION.............................................................................................................................................. 157
B. STRUCTURE OF THE USE TABLE ..................................................................................................................... 157
C. INTERMEDIATE CONSUMPTION PART OF THE USE TABLE................................................................................ 162
D. GVA PART OF THE USE TABLE ....................................................................................................................... 170
E. FINAL CONSUMPTION EXPENDITURE PART OF THE USE TABLE ....................................................................... 173
F. GROSS CAPITAL FORMATION PART OF THE USE TABLE................................................................................... 182
G. EXPORTS ....................................................................................................................................................... 196
ANNEX A TO CHAPTER 6: SAMPLE QUESTIONNAIRE COLLECTING PURCHASES OF GOODS AND
SERVICES FOR INTERMEDIATE CONSUMPTION ...................................................................................... 198
ANNEX B TO CHAPTER 6: IMPACT OF CAPITALIZING THE COSTS OF RESEARCH AND
DEVELOPMENT IN SUTS AND IOTS ............................................................................................................... 201
A. RESEARCH AND DEVELOPMENT AS FIXED CAPITAL FORMATION .................................................................... 201
B. IMPLICATIONS OF VALUATION OF OUTPUT AS SUM OF COSTS......................................................................... 202
C. OWN-ACCOUNT RESEARCH AND DEVELOPMENT AS PRINCIPAL OR SECONDARY OUTPUT ............................... 203
D. BALANCING SUPPLY AND USE OF RESEARCH AND DEVELOPMENT SERVICES.................................................. 205
CHAPTER 7. COMPILING THE VALUATION MATRICES.................................................................... 207
A. INTRODUCTION.............................................................................................................................................. 207
B. VALUATION OF PRODUCT FLOWS................................................................................................................... 207
C. TRADE MARGINS ........................................................................................................................................... 216
D. TRANSPORT MARGINS ................................................................................................................................... 229
E. TAXES ON PRODUCTS AND SUBSIDIES ON PRODUCTS ..................................................................................... 236
ANNEX A TO CHAPTER 7. EXAMPLE FOR DERIVING TRADE MARGINS IN SUTS BASED ON
SURVEY DATA ...................................................................................................................................................... 241
A. SUPPLY TABLE............................................................................................................................................... 243
B. USE TABLE .................................................................................................................................................... 245
CHAPTER 8. COMPILING THE IMPORTS USE TABLE AND DOMESTIC USE TABLE ................. 249
A. INTRODUCTION.............................................................................................................................................. 249
B. STRUCTURE OF THE IMPORTS USE TABLE AND DOMESTIC USE TABLE ............................................................ 250
C. COMPILATION OF THE IMPORTS USE TABLE ................................................................................................... 255
D. SPECIFIC ISSUES IN THE COMPILATION OF THE IMPORTS USE TABLE .............................................................. 259
E. ENHANCEMENTS TO THE IMPORTS USE TABLE FOR ANALYTICAL USES .......................................................... 268
CHAPTER 9. COMPILING SUTS IN VOLUME TERMS .......................................................................... 271
A. INTRODUCTION.............................................................................................................................................. 271
B. RECOGNITION OF ALTERNATIVE APPROACHES............................................................................................... 272
C. OVERVIEW OF THE STEPS IN THE H-APPROACH WITH A FOCUS ON VOLUMES ................................................ 273
D. PRICE AND VOLUME INDICATORS IN THEORY................................................................................................. 284
E. PRICE AND VOLUME INDICATORS IN PRACTICE .............................................................................................. 285
F. INPUT-OUTPUT TABLES IN VOLUME TERMS .................................................................................................... 301

vi
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

CHAPTER 10. LINKING THE SUPPLY AND USE TABLES TO THE INSTITUTIONAL SECTOR
ACCOUNTS 303
A. INTRODUCTION.............................................................................................................................................. 303
B. INSTITUTIONAL SECTORS AND SUBSECTORS .................................................................................................. 304
C. TABLE LINKING SUTS AND INSTITUTIONAL SECTOR ACCOUNTS.................................................................... 307
D. COMPILATION METHODS ............................................................................................................................... 315
CHAPTER 11. BALANCING THE SUPPLY AND USE TABLES ............................................................... 319
A. INTRODUCTION.............................................................................................................................................. 319
B. OVERVIEW OF THE SYSTEM AND BASIC IDENTITIES ....................................................................................... 320
C. BALANCING ................................................................................................................................................... 324
D. STEP-BY-STEP PROCEDURE FOR SIMULTANEOUS BALANCING ........................................................................ 332
E. ALTERNATIVE BALANCING METHODS............................................................................................................ 336
F. EXTENDING THE BALANCING OF SUTS TO INCLUDE INSTITUTIONAL SECTOR ACCOUNTS, IOTS, PSUTS AND EE-
IOTS 338
G. PRACTICAL ASPECTS OF BALANCING ............................................................................................................. 342
ANNEX A TO CHAPTER 11. BALANCING SUPPLY AND USE TABLES ................................................... 350
CHAPTER 12. TRANSFORMING THE SUPPLY AND USE TABLES INTO INPUT-OUTPUT TABLES
369
A. INTRODUCTION.............................................................................................................................................. 369
B. OVERVIEW OF THE RELATIONSHIP BETWEEN IOTS AND SUTS ...................................................................... 369
C. CONVERSION OF SUTS TO IOTS .................................................................................................................... 374
D. INPUT-OUTPUT FRAMEWORK ......................................................................................................................... 379
E. EMPIRICAL APPLICATION OF THE TRANSFORMATION MODELS ....................................................................... 400
ANNEX A TO CHAPTER 12. MATHEMATICAL DERIVATION OF DIFFERENT IOTS ......................... 411
A. PRODUCT-BY-PRODUCT IOTS AND INDUSTRY-BY-INDUSTRY IOTS ............................................................... 411
B. PRODUCT-BY-PRODUCT IOTS........................................................................................................................ 412
C. INDUSTRY-BY-INDUSTRY IOTS ..................................................................................................................... 414
D. USE OF A HYBRID TECHNOLOGY ASSUMPTION FOR PRODUCT-BY-PRODUCT IOTS ......................................... 416
ANNEX B TO CHAPTER 12. CLASSICAL CAUSES AND TREATMENT OF NEGATIVE CELL ENTRIES
IN THE PRODUCT TECHNOLOGY ................................................................................................................... 419
A. CLASSICAL CAUSES OF NEGATIVE ELEMENTS IN THE PRODUCT TECHNOLOGY ............................................... 419
B. OVERALL STRATEGY FOR REMOVING NEGATIVES .......................................................................................... 421
C. SPECIFIC APPROACHES TO DEALING WITH NEGATIVES ................................................................................... 421
ANNEX C TO CHAPTER 12. EXAMPLES OF REVIEWS OF APPROACHES TO THE TREATMENT OF
SECONDARY PRODUCTS ................................................................................................................................... 427
CHAPTER 13. COMPILING PHYSICAL SUPPLY AND USE TABLES AND ENVIRONMENTALLY
EXTENDED INPUT-OUTPUT TABLES ............................................................................................................. 429
A. INTRODUCTION.............................................................................................................................................. 429
B. OVERVIEW OF PSUTS ................................................................................................................................... 430
C. COMPILATION OF PSUTS .............................................................................................................................. 442
D. ENVIRONMENTAL EXTENDED IOTS ............................................................................................................... 449
E. COMPILATION OF EE-IOTS ........................................................................................................................... 453
F. COUNTRY EXAMPLES .................................................................................................................................... 454

vii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

CHAPTER 14. SUPPLY AND USE TABLES AND QUARTERLY NATIONAL ACCOUNTS ................ 467
A. INTRODUCTION.............................................................................................................................................. 467
B. QUARTERLY NATIONAL ACCOUNTS ............................................................................................................... 468
C. SUTS AND QUARTERLY NATIONAL ACCOUNTS .............................................................................................. 474
CHAPTER 15. DISSEMINATING SUPPLY, USE AND INPUT-OUTPUT TABLES ................................ 487
A. INTRODUCTION.............................................................................................................................................. 487
B. USER IDENTIFICATION ................................................................................................................................... 487
C. DISSEMINATION STRATEGY ........................................................................................................................... 488
D. COMMUNICATIONS OF SUTS AND IOTS WITH USERS .................................................................................... 493
E. DISSEMINATION FORMAT FOR SUTS AND IOTS............................................................................................. 494
F. STATISTICAL DATA AND METADATA EXCHANGE INITIATIVE ........................................................................ 497
PART FOUR ............................................................................................................................................................ 499
CHAPTER 16. REGIONAL SUPPLY AND USE TABLES ........................................................................... 501
A. INTRODUCTION.............................................................................................................................................. 501
B. ISSUES ARISING IN AND METHODS FOR THE COMPILATION OF REGIONAL SUTS AND IOTS ............................ 501
C. EXAMPLE OF BOTTOM-UP METHODS FOR REGIONAL SUTS: CANADIAN EXPERIENCE .................................... 504
CHAPTER 17. MULTI-COUNTRY SUPPLY AND USE TABLES AND INPUT-OUTPUT TABLES ..... 521
A. INTRODUCTION.............................................................................................................................................. 521
B. OVERVIEW OF MULTI-COUNTRY SUTS AND IOTS AND MAIN COMPILATION ISSUES ...................................... 522
C. COMPILATION PROCEDURE ............................................................................................................................ 530
D. MULTI-COUNTRY INPUT-OUTPUT DATABASE INITIATIVES ............................................................................. 541
E. WAY AHEAD .................................................................................................................................................. 543
CHAPTER 18. PROJECTING SUPPLY, USE AND INPUT-OUTPUT TABLES ....................................... 549
A. INTRODUCTION.............................................................................................................................................. 549
B. SITUATIONS NEEDING PROJECTION METHODS ................................................................................................ 549
C. GENERAL APPROACHES TO PROJECTION FROM A HISTORICAL PERSPECTIVE .................................................. 551
D. NUMERICAL EXAMPLES ................................................................................................................................. 567
E. CRITERIA TO CONSIDER WHEN CHOOSING A METHOD .................................................................................... 580
CHAPTER 19. EXTENSIONS OF SUTS AND IOTS AS PART OF SATELLITE SYSTEMS .................. 583
A. INTRODUCTION.............................................................................................................................................. 583
B. OVERVIEW OF POSSIBLE EXTENSIONS ............................................................................................................ 584
C. SOCIAL ACCOUNTING MATRIX ....................................................................................................................... 591
D. EXTENDED INPUT-OUTPUT TABLES ................................................................................................................ 597
E. OTHER EXAMPLES OF SATELLITE SYSTEMS.................................................................................................... 601
CHAPTER 20. MODELLING APPLICATIONS OF IOTS ........................................................................... 603
A. INTRODUCTION.............................................................................................................................................. 603
B. NUMERICAL EXAMPLE OF IOTS AS A STARTING POINT .................................................................................. 604
C. DISTINCTION BETWEEN PRICE, VOLUME, QUANTITY, QUALITY AND PHYSICAL UNITS .................................... 605
D. INPUT COEFFICIENTS ..................................................................................................................................... 609
E. OUTPUT COEFFICIENTS .................................................................................................................................. 611
F. QUANTITY MODEL OF INPUT-OUTPUT ANALYSIS ........................................................................................... 612
G. PRICE MODEL OF INPUT-OUTPUT ANALYSIS ................................................................................................... 616

viii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

H. INPUT-OUTPUT MODELS WITH INPUT AND OUTPUT COEFFICIENTS ................................................................. 622

I. CENTRAL MODEL OF INPUT-OUTPUT ANALYSIS ............................................................................................. 623
J. INDICATORS .................................................................................................................................................. 626
K. MULTIPLIERS................................................................................................................................................. 629
L. INTER-INDUSTRIAL LINKAGE ANALYSIS ........................................................................................................ 636
CHAPTER 21. EXAMPLES OF COMPILATION PRACTICES .................................................................. 641
A. INTRODUCTION.............................................................................................................................................. 641
B. BASIC CONSIDERATIONS FOR THE COMPILATION OF NATIONAL ACCOUNTS AND SUTS ................................. 642
C. EFFECT ON GDP OF INTEGRATING SUTS IN THE NATIONAL ACCOUNTS FOR MALAWI................................... 655
D. DEVELOPMENT OF THE APPLICATION OF THE INPUT-OUTPUT FRAMEWORK IN THE CZECH REPUBLIC ............ 659
E. CONTINUAL CHANGE, DEVELOPMENT AND IMPROVEMENT IN CHILE ............................................................. 663
REFERENCES ........................................................................................................................................................ 675
ADDITIONAL READING ..................................................................................................................................... 695

Box 1.1 Evolution of the SUTs and IOTs within the national accounts ....................................... 13
Box 2.1 Numerical example of the SUTs system ......................................................................... 28
Box 2.2 Numerical example showing a use table split between consumption of domestic
production and imports ................................................................................................................. 29
Box 2.3 SUTs and product-by-product IOTs ................................................................................ 37
Box 2.4 SUTs and industry-by-industry IOTs .............................................................................. 38
Box 2.5 Three approaches to measuring GDP .............................................................................. 41
Box 2.6 Other classifications of products ..................................................................................... 46
Box 2.7 SNA recommendations on partitioning of vertically and horizontally integrated enterprises
....................................................................................................................................................... 53
Box 2.8 Overview of the valuation in SUTs and IOTs ................................................................. 57
Box 2.9 Calculation of output for market and non-market producers .......................................... 60
Box 2.10 Example of derivation of GDP from balanced SUTs .................................................... 64
Box 3.1 Examples of the main recommendations, principles and guidelines provided in this
Handbook ...................................................................................................................................... 85
Box 4.1 Example of in-house custom-built software: Statistics Netherlands ............................. 100
Box 4.2 ERETES ........................................................................................................................ 101
Box 4.3 Data sources generally used .......................................................................................... 121
Box 5.1 Redefinitions ................................................................................................................. 144
Box 5.2 Consistency issues with the CIF/FOB adjustment ........................................................ 150
Box 6.1 Example of a calculation of the values of an input column .......................................... 168
Box 6.2 Classification of individual consumption according to purpose ................................... 174
Box 6.3: Non-durable, semi-durable and durable goods ............................................................ 176
Box 6.4 Classification of the purposes of NPISHs ..................................................................... 179
Box 6.5 Classification of functions of government .................................................................... 181
Box 6.6 Gross fixed capital formation by type of asset .............................................................. 183

ix
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 7.1 Compilation process for trade margins ......................................................................... 221

Box 7.2 Examples of transport costs which do not form transport margins ............................... 231
Box 7.3 Options to consider where no data exists on transport margins .................................... 234
Box 8.1 Standard services components of BPM 6...................................................................... 257
Box 9.1 Treatment of newly introduced and disappearing taxes and subsidies ......................... 297
Box 10.1 Compilation methods used to link SUTs to the institutional sector accounts ............. 315
Box 11.1 Example of discrepancies balanced in current prices and in volume terms ................ 331
Box 11.2 Example of simultaneous balancing comparing volume indices ................................ 331
Box 11.3 Methods used for automated balancing SUTs............................................................. 345
Box 12.1 Clarification of IOTs terminology............................................................................... 373
Box 12.2 Input-output framework for domestic output and imports .......................................... 381
Box 12.3 Basic transformations of SUTs to IOTs ...................................................................... 383
Box 13.1 Selected reference material ......................................................................................... 444
Box 15.1 Fundamental Principles of Official Statistics .............................................................. 488
Box 15.2 Reference metadata in the SDMX metadata structure for SUTs and IOTs ................ 493
Box 17.1 Background papers of each database initiative ........................................................... 541
Box 17.2 Overview of the main features of the various databases ............................................. 543
Box 18.1 Methods for projection of SUTs and IOTs.................................................................. 554
Box 18.2 SUTs and IOTs for Austria, 2005 and 2006................................................................ 568
Box 18.3 Results using the GRAS Method ................................................................................ 570
Box 18.4 Flow diagram of the GRAS method ............................................................................ 571
Box 18.5 Results using the SUT-RAS method ........................................................................... 574
Box 18.6 Flow diagram of the SUT-RAS method...................................................................... 575
Box 18.7 Results using the SUT-Euro method ........................................................................... 578
Box 18.8 Flow diagram of the SUT-EURO method................................................................... 580
Box 19.1 Measurement performance and social progress: overview of Stiglitz-Sen-Fitoussi
Commission 2009 report ............................................................................................................. 588
Box 20.1 Quantities, prices, values and volumes in IOTs .......................................................... 607
Box 20.2 Quantity input-output model ....................................................................................... 620
Box 20.3 Price input-output model ............................................................................................. 621
Box 20.4 Multipliers in the input-output model ......................................................................... 631
Box 21.1 Material product system and Phare projects ............................................................... 660

Figure 1.1 Overview of the links between SUTs and the SNA framework .................................... 9
Figure 2.1: Graphical overview of supply and use tables ............................................................. 27
Figure 2.2 Schematic overview of the compilation of SUTs and IOTs: H-Approach .................. 30
Figure 2.3 System of national accounts in matrix form ................................................................ 41
Figure 2.4 Overview of SUTs and IOTs as part of the SNA compilation .................................... 59
Figure 3.1 Phases of the GSBPM ................................................................................................. 72
Figure 3.2 Simplified business processing model for compiling SUTs, IOTs, and PSUTs ......... 74

x
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3.3 Structure of the SUTs and the links covered in this Handbook .................................. 77
Figure 3.4 Compilation of SUTs and IOTs in current prices and in volume terms ...................... 79
Figure 3.5 Evolution of compiling SUTs and IOTs in the first three years .................................. 80
Figure 3.6 First year of compilation ............................................................................................. 82
Figure 3.7 Second year of compilation ......................................................................................... 84
Figure 4.1 Overview of SUTs and IOTs as part of the SNA compilation .................................. 123
Figure 5.1 Link between valuation matrices in the supply table and the use table..................... 135
Figure 6.1 Three-dimensional view of SUTs .............................................................................. 162
Figure 7.1 Schematic representation of the valuation matrices in the SUTs .............................. 210
Figure 7.2 Alternative distribution channels of goods ................................................................ 227
Figure 9.1 Overview of the compilation schematic layout linking SUTs in current prices and in
volume terms ............................................................................................................................... 274
Figure 9.2 Link between SUTs in current prices and in volume terms ...................................... 278
Figure 10.1 Links between the industry accounts and the institutional sector accounts ............ 304
Figure 10.2 Link between the SUTs and institutional sector accounts ....................................... 309
Figure 11.1 Simplified SUTs system .......................................................................................... 321
Figure 11.2 Six-pack ................................................................................................................... 325
Figure 11.3 Overview of the SUTs balancing framework for simultaneous balancing.............. 333
Figure 11.4 Sources of feedback loops emanating from the balancing process ......................... 342
Figure 12.1 Transformation of SUTs into IOTs ......................................................................... 372
Figure 12.2 Basic transformation models ................................................................................... 378
Figure 13.1 Physical flows of natural inputs, products and residuals ......................................... 431
Figure 13.2 Overview of the compilation system for PSUTs ..................................................... 446
Figure 13.3 Key feedback loops in producing and balancing the PSUTs and environmental
extended IOTs ............................................................................................................................. 449
Figure 13.4 Danish SUTs framework extended with physical flows ......................................... 456
Figure 13.5 From source data to PSUTs ..................................................................................... 458
Figure 14.1 Quarterly GDP production (output) aggregate: data availability and estimation in the
United Kingdom.......................................................................................................................... 470
Figure 14.2 Quarterly GDP expenditure components: data availability and estimation in the United
Kingdom ..................................................................................................................................... 471
Figure 15.1 Release calendar covering SUTs, IOTs and national accounts: Statistics Denmark
..................................................................................................................................................... 490
Figure 15.2 Measuring United Kingdom GDP and SUTs: revision policy ................................ 491
Figure 17.1 Schematic representation of multi-country SUTs (three-country case) .................. 523
Figure 17.2 Schematic representation of multi-country IOTs (three country case) ................... 524
Figure 17.3 System of multi-country SUTs and its conceptual correspondence to a national SUTs
framework ................................................................................................................................... 532
Figure 17.4 Splitting the import matrix by country of origin ..................................................... 535
Figure 17.5 Converting valuation scheme .................................................................................. 536

xi
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 17.6 Formation of the export vector to rest of the world................................................. 538
Figure 17.7 Transformation to multi-country IOTs .................................................................... 540
Figure 21.1: Illustration of a database for the product-flow method used in smaller countries . 649
Figure 21.2 Supply and use table ................................................................................................ 670

Table 2.1: Simplified structure of the supply table ....................................................................... 24

Table 2.2: Simplified structure of the use table ............................................................................ 24
Table 2.3: Supply and use tables framework ................................................................................ 25
Table 2.4 Schematic view of the physical supply table ................................................................ 34
Table 2.5 Schematic view of the physical use table ..................................................................... 34
Table 2.6 Simplified IOT (product by product) ............................................................................ 36
Table 2.7 Links between the use table and functional classifications .......................................... 51
Table 2.8 Simplified table linking the SUTs to the institutional sector accounts ......................... 65
Table 4.1 Examples of the size of published and internal working level SUTs and IOTs ......... 107
Table 5.1 Numerical example of a supply table at basic prices .................................................. 132
Table 5.2 Supply table at basic prices, including a transformation into purchasers’ prices ....... 133
Table 5.3 Data adjustment for external trade of goods and services .......................................... 148
Table 5.4 CIF and FOB adjustment row ..................................................................................... 149
Table 6.1 Use table at purchasers’ prices.................................................................................... 158
Table 6.2 Intermediate consumption of selected inputs into “Manufacture of rubber and plastic
products” ..................................................................................................................................... 169
Table 6.3 Sample product balance for "Gelatine and gelatine derivatives"................................ 170
Table 6.4 Categories of final consumption expenditure ............................................................. 173
Table 6.5 Table linking final expenditures by purpose (COICOP) and product (CPC) ............. 175
Table 6.6 Final consumption expenditure of households (by COICOP headings) ..................... 176
Table 6.7 Table linking final consumption expenditures of NPISHs by purpose (COPNI) and by
product (CPC) ............................................................................................................................. 180
Table 6.8 Table linking final consumption expenditure of general government by COFOG and
CPC ............................................................................................................................................. 182
Table 6.9 Categories of gross capital formation ......................................................................... 183
Table 6.10 Table linking gross fixed capital formation by industries, assets and products ....... 185
Table 6.11 Gross fixed capital formation by investing industry................................................. 187
Table 6.12 Table linking change in inventories industries, assets and products ........................ 192
Table 7.1 Supply table at basic prices, including a transformation into purchasers’ prices ....... 212
Table 7.2 Use table at purchasers’ prices.................................................................................... 213
Table 7.3 Use-side valuation matrices ........................................................................................ 214
Table 7.4 Use table at basic prices .............................................................................................. 216
Table 7.5 Trade turnover and trade margins for wholesale and retail trade margins ................. 224
Table 8.1 Structure of the imports use table ............................................................................... 251
Table 8.2 Numerical example of the imports use table .............................................................. 251

xii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 8.3 Structure of the domestic use table ............................................................................. 252

Table 8.4 Numerical example of a domestic use table ............................................................... 252
Table 8.5 Example of an input table for imports at basic prices................................................. 253
Table 8.6 Input-output table for domestic output at basic prices ................................................ 254
Table 8.7 Input-output table for domestic output at basic prices, net exports with adjustment items
..................................................................................................................................................... 255
Table 8.8 Processing within the country ..................................................................................... 261
Table 8.9 Goods sent abroad for processing ............................................................................... 262
Table 8.10 Industry 1 – Alternative input structures .................................................................. 264
Table 9.1 Supply table in current prices and in volume terms.................................................... 280
Table 9.2 Use table in current prices and in volume terms ......................................................... 281
Table 9.3 Gross domestic product in current prices and in volume terms .................................. 283
Table 10.1 Summary of institutional sectors and subsectors ...................................................... 306
Table 10.2 Main features of the SUT approach and the institutional sector approach ............... 307
Table 10.3 Numerical example showing the table linking the SUTs and institutional sector
accounts....................................................................................................................................... 310
Table 10.4 Goods and services for the whole economy ............................................................. 312
Table 10.5 Production account for the whole economy ............................................................. 313
Table 10.6 Link between GDP and industry GVA ..................................................................... 313
Table 10.7 Generation of income account for the whole economy ............................................ 314
Table 12.1 Product-by-product IOT at basic prices .................................................................... 370
Table 12.2 Numerical example of rectangular SUTs for the transformation ............................. 375
Table 12.3 Numerical example of square SUTs for the transformation ..................................... 376
Table 12.4 Transformation matrix for the product technology assumption ............................... 386
Table 12.5 Product-by-product IOTs based on product technology ........................................... 387
Table 12.6 Transformation matrix for industry technology assumption .................................... 388
Table 12.7 Product-by-product IOTs based on industry technology .......................................... 388
Table 12.8 Matrix for hybrid technology .................................................................................... 389
Table 12.9 Transformation matrix for hybrid technology assumption ....................................... 390
Table 12.10 IOTs based on the hybrid technology assumption .................................................. 390
Table 12.11 Transformation matrix for the fixed industry sales structure assumption ............. 391
Table 12.12 IOTs based on the fixed industry sales structure assumption ................................. 392
Table 12.13 Transformation matrix for the fixed product sales structure assumption for rectangular
SUTs ........................................................................................................................................... 394
Table 12.14 IOTs based on the fixed product sales structure assumption derived from rectangular
SUTs ........................................................................................................................................... 394
Table 12.15 Transformation matrix for the fixed product sales structure assumption for square
SUTs ........................................................................................................................................... 395
Table 12.16 IOTs based on the fixed product sales structure assumption for square SUTs....... 396

xiii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 12.17 Absolute deviation of IOTs based on rectangular SUTs less IOTs based on square
SUTs for Model D ...................................................................................................................... 397
Table 12.18 Alternative presentations of product-by-product IOTs ........................................... 399
Table 12.19 Empirical example of product-by-product IOTs .................................................... 402
Table 12.20 Empirical example of industry-by-industry IOTs ................................................... 404
Table 13.1 General PSUT ........................................................................................................... 432
Table 13.2 Classes of natural input ............................................................................................. 434
Table 13.3 Typical components for groups of residuals ............................................................. 436
Table 13.4 List of individual components of SEEA physical PSUTs ........................................ 443
Table 13.5 Common national data sources and links to SEEA component accounts................. 443
Table 13.6 Single region IOT with environmental data ............................................................. 450
Table 13.7 Single region IOT in hybrid units ............................................................................. 452
Table 13.8 Industry-by -industry IOTs (upper block of Table 13.6) .......................................... 453
Table 13.9 Environmental data by industry (lower block of Table 13.6) ................................... 453
Table 13.10 PSUTs in Denmark ................................................................................................. 460
Table 13.11 SUTs for the Netherlands, 2010 ............................................................................. 462
Table 13.12 PSUTs for the Netherlands, 2010 ........................................................................... 463
Table 14.1 Balancing supply and use of products ...................................................................... 477
Table 16.1 SUTs framework for interregional SUTs.................................................................. 507
Table 16.2 Interregional and international trade flows by province and territory, 2010 ........... 510
Table 17.1 Adjustment targets for national tables of selected countries in the Asian international
input-output table for the year 2000............................................................................................ 530
Table 18.1 Categorization of methods ........................................................................................ 561
Table 19.1 Structure of a social accounting matrix .................................................................... 594
Table 19.2 Numerical example of a social accounting matrix.................................................... 595
Table 19.3 Extended IOT with satellite systems ........................................................................ 599
Table 20.1 IOT at basic prices .................................................................................................... 605
Table 20.2 Input coefficients of IOTs ......................................................................................... 610
Table 20.3 Output coefficients of IOTs ...................................................................................... 612
Table 20.4 Input coefficients for domestic intermediate consumption....................................... 613
Table 20.5 Leontief matrix ......................................................................................................... 614
Table 20.6 Leontief inverse ........................................................................................................ 615
Table 20.7 Quantity input-output model based on monetary data .............................................. 616
Table 20.8 Price input-output model based on monetary data.................................................... 619
Table 20.9 Emission model......................................................................................................... 625
Table 20.10 Input indicators for production activities per unit of output ................................... 628
Table 20.11 Output multipliers (Leontief inverse) ..................................................................... 629
Table 20.12 Multipliers for products .......................................................................................... 632
Table 20.13 Input content of final use by category .................................................................... 635
Table 20.14 Backward linkages .................................................................................................. 637

xiv
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 20.15 Forward linkages..................................................................................................... 637

Table 20.16 Forward and backward linkages ............................................................................. 638
Table 20.17 Normalized forward and backward linkages .......................................................... 638
Table 21.1 Historical benchmark exercises ................................................................................ 664
Table 21.2 Annual business surveys ........................................................................................... 666
Table 21.3 Administrative records .............................................................................................. 667
Table 21.4 SUTs for tobacco products, year 2008, current prices .............................................. 672
Table 21.5 SUT for cleaning and toiletry products, year 2008, current prices ........................... 672
Table 21.6 IOTs for domestic output at basic prices, 2008 ........................................................ 673

xv
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

xvi
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Abbreviations

AIIOT Asian International Input-Output Tables

ANZSIC Australian and New Zealand Standard Industrial Classification

ARIMA autoregressive integrated moving average

BEC classification by broad economic categories

BPM Balance of Payments and International Investment Position Manual

BTDIxE Bilateral Trade Database by Industry and End-Use

CH4 methane

CIF cost, insurance and freight

CO2 carbon dioxide

COFOG Classification of the Functions of Government

COICOP Classification of the Purposes of Non-profit Institutions Serving Households

COPNI Classification of the Purposes of Non-profit Institutions Serving Households

COPP Classification of the Outlays of Producers According to Purpose

CPA Classification of Products by Activity

CPC Central Product Classification

CPI consumer price index

CRAS cell-corrected RAS method

CREEA compiling and refining of economic and environmental accounts

CSPI corporate services price index

EBOPS Extended Balance of Payments Services Classification

ECE Economic Commission for Europe

xvii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

EE-IOT environmentally extended input-output tables

EPI export price index

ERETES equilibre ressources-emplois et tableau entrées-sorties

ESA European system of accounts

EU European Union

FAO Food and Agriculture Organization of the United Nations

FBS finance and business services

FI fixed industry

FIGARO full international and global accounts for research in input-output

analysis

FISIM financial intermediation services indirectly measured

FOB free on board

FP fixed product

GDP gross domestic product

GENESIS online databank of the Federal Statistical Office of Germany

GNI gross national income

GRAS generalized RAS

GSBPM Generic Statistical Business Process Model

GTAP global trade analysis project

GTAP-MRIO multi-region input–output table based on the global trade analysis project
database

GVA gross value added

HS Harmonized Commodity Description and Coding System

ICIO inter-country input-output

ICPIs intermediate consumption price indices

xviii
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

ICT information and communications technology

ILO International Labour Organization

IMF International Monetary Fund

IMTS International Merchandise Trade Statistics: Concepts and Definitions

INSEE Institut national de la statistique et des études économiques

IOTs input-output tables

IPCC Intergovernmental Panel on Climate Change

IPIs import price indices

ISIC International Standard Industrial Classification of All Economic Activities

KLEMS Integrated Industry-Level Production Account

KRAS Konfliktfreies RAS

MPS Material Product System

MRIO multi-region input-output

N2O nitrous oxide

NACE Statistical Classification of Economic Activities in the European Community

NAICS North American Industry Classification System

NMVOC non-methane volatile organic compounds

NOx mono-nitrogen oxides

NPISH non-profit institution serving households

OECD Organization for Economic Cooperation and Development

PIOT physical input-output tables

PPP purchasing power parities

PSUT physical supply and use table

PYPs previous years’ prices

R&D research and development

xix
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

RAS ranking and scaling data reconciliation method

RPI retail prices index

SAM social accounting matrix

SDMX Statistical Data and Metadata Exchange

SEEA System of Environmental and Economic Accounts

SESAME system of economic and social accounting matrices and extensions

SITC Standard International Trade Classification

SNA System of National Accounts

SO2 sulphur dioxide

SUT-Euro Euro method for SUTs

SUTs supply and use tables

TEC trade by enterprise characteristics data

TiVA trade in value added

TLS taxes less subsidies on products

TRAS three-stage RAS

TTM trade and transport margins

UNSD United Nations Statistics Division

UNWTO World Tourism Organization

VAT value added taxes

WCO World Customs Organization

WIOD World Input-Output Database

WPI wholesale price index

WTO World Trade Organization

xx
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Part one

1
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Chapter 1. Introduction

A. Background

1.1. The supply and use tables (SUTs) are an integral part of the System of National Accounts
2008 (2008 SNA) forming the central framework for the compilation of a single and coherent
estimate of gross domestic product (GDP) integrating all the components of production, income
and expenditure approaches, and providing key links to other parts of the SNA framework.

1.2. In their simplest form, the SUTs describe how products (goods and services) are brought
into an economy (either as a result of domestic production or imports from other countries) in the
supply table, and how those same products (intermediate consumption; final consumption by
household, non-profit institutions serving households, and general government; gross capital
formation; and exports) are used in the use table.

1.3. The SUTs also provide the link between components of gross value added (GVA), industry
inputs and outputs. Although typically they show only the industry dimension, SUTs can also be
formulated to show the role of different institutional sectors (for example, non-financial
corporations, government, and others) providing an important linking mechanism to the different
accounts of the SNA framework (the goods and services account, production account, generation
of income account and the capital account).

1.4. Importantly, and by design, these interlinkages facilitate data confrontation and the
examination of the consistency of data on goods and services obtained from different statistical
sources, such as business surveys, household surveys and administrative data within a single
detailed framework. As such, they provide a powerful mechanism for feedback on the quality and
coherency of primary data sources.

1.5. The SUTs do not just provide a framework to ensure the best quality estimates of GDP and
its components: they are also an important analytical resource in their own right, showing the
interaction between producers and consumers. When measured in volume terms, the SUTs provide
the basis for a rich stream of analyses, notably in the field of structural analysis, and in particular
productivity, where in recent years SUTs have been widely accepted as an important tool for

3
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

KLEMS-type 2 productivity measures. Just as important is their growing use as the basis for
deriving the input-output tables (IOTs).

1.6. In many respects, the IOTs, which show the links between final uses and intermediate uses
of goods and services defined according to industry outputs (industry-by-industry tables) or
according to product outputs (product-by-product tables) predate the SUTs. The IOTs also show
separately the consumption of domestically produced and imported goods and services. The
widespread availability of SUTs has meant, however, that the SUTs form the starting point for
constructing IOTs and, in turn, an entire swathe of related analytical products and indicators, such
as the Leontief inverse and other type of analyses, including output multipliers, employment
multipliers, and others.

1.7. The SUTs and IOTs are compiled by many countries in the course of producing their core
national accounts, thereby improving the coherency and consistency of their national account
estimates. The ability to readily create IOTs from SUTs (as shown in chapter 12) has helped to
reinforce the momentum behind the evolution, role and use of SUTs.

1.8. SUTs and IOTs have received much attention in recent years. This is because their
analytical properties allow for a much wider set of analyses, not only of the national economy and
the regions within a nation but also of the interlinkages between economies at the global level and
also of environmental impacts.

1.9. Further momentum has been generated for the role of SUTs and IOTs in step with the
rapidly growing impact of globalization and the international fragmentation of production. For a
full understanding of international interdependencies and their impact on important policy areas,
such as trade, competitiveness and sustainable development, there is increasing need to view
production and consumption through a global value chain lens. In other words, multi-country and
regional SUTs and IOTs have become essential tools to inform policy and policymakers. Over the
past five years, a number of efforts have been made by the international statistics community to
meet these needs, such as the trade in value added database prepared by the Organization for
Economic Cooperation and Development (OECD) and the World Trade Organization (WTO), and
other comparable databases such as the World Input-Output Database (WIOD) and the Handbook
on Accounting for Global Value Chains prepared by the Expert Group on International Trade and
Economic Globalization Statistics.

1.10. Given these developments and, in particular, the heightened importance of SUTs and IOTs,
the timing of the present Handbook is important and highly relevant. The present chapter provides
a general introduction to the various issues considered in greater detail in the various chapters that
follow. Section B of this introductory chapter provides a general overview of the roles and uses of

2
KLEMS is an industry-level growth and productivity research project, based on the analysis of capital (K), labour
(L), energy (E), materials (M) and service (S) inputs.

4
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

SUTs and IOTs. Section C covers the SNA and its links to SUTs and IOTs. Section D covers the
objectives of the Handbook and its new features compared to previous manuals on the subject.
Lastly, section E briefly outlines the structure and content of the Handbook.

B. Uses of SUTs and IOTs

1.11. The uses of SUTs and IOTs are multiple and their statistical and analytical importance has
increased with time and in response to new and emerging issues, such as globalization and
sustainable development, with its three pillars of social, economic and environmental
development. Where possible, the analytical uses of SUTs and IOTs are presented below in
parallel. As SUTs form the basis for the compilation of IOTs, the uses of the two types of tables
are treated in the same way in this section.

1.12. As mentioned above, the SUTs combine in a single framework the three approaches to
measuring GDP, namely, the production approach, the income approach and the expenditure
approach. All three approaches are based on sets of data with various levels of detail and a range
of different sources. Combining the data in a single statistical framework compels compilers to
use harmonized and unique classifications of producers, users and income receivers, together with
harmonized and unique classifications and definitions of products and income categories. Under
these conditions, corresponding data can be related and compared in an organized manner.
Combining the three data sets provides an opportunity to analyse the causes of discrepancies, make
necessary adjustments and fill data gaps when necessary.

1.13. An important objective of national accounts is to estimate year-to-year and quarter-to-

quarter changes in a number of macroeconomic variables. When dealing with production, use and
the generation of value added, it is important to divide the current price changes into volume
changes (representing what is termed “real” growth) and price changes. When SUTs are compiled
simultaneously in current prices and in volume measures (as recommended in this Handbook,
using what is known as the “H-Approach”), there are considerable advantages in the overall quality
and consistency of the information provided. During the entire statistical process – from the
processing and analysis of the source data through to, and including, the balancing of the SUTs –
data in current prices and deflated data are obtained simultaneously and consistently with each
other.

1.14. In addition to annual national accounts, SUTs can be used in the compilation of quarterly
national accounts. This may range from the compilation and balancing of quarterly SUTs to the
mere use of the SUTs framework to highlight possible discrepancies between quarterly product
supply and use. The annual estimates of GVA can, for example, be used as weights in the quarterly
estimate of GDP in volume terms to reflect the most recent period. In addition, SUTs can provide
weighting schemes for price and volume indices.

1.15. The SUTs and IOTs serve also as the basis for compiling a range of accounts –regional,
environmental, labour, tourism, etc. The clear links of these satellite systems with both the SUTs

5
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

and the IOTs ensure the consistency of the satellite systems with the concepts and methods of the
core national accounts and allow for feedback loops with the SUTs during the compilation and
balancing process of the frameworks involved. For instance, the SUTs can support the compilation
of regional accounts by including clear links to variables like regional GVA. When physical
environmental flows are linked to the SUTs and IOTs in the environmental accounts, they provide
feedback loops to the compilation of SUTs by contrasting physical and monetary measures of the
supply and use of products. When SUTs are linked to labour and capital, they can be used for
productivity analyses that link economic growth to the use of intermediate inputs. Lastly, social
accounting matrices elaborate the linkages between SUTs and sector accounts. They capture
transactions and transfers between all economic agents in the accounting system and measures
effects of macroeconomic policies on distribution.

1.16. The SUTs and IOTs also provide the basis for different types of analytical uses at micro
and macro levels (see, for example, United Nations, 2002; Mahajan, 2004a; and Mahajan, 2006).
Various examples are included in the list of additional reading at the end of this Handbook.
Examples include the following:

• Economic analyses: export shares, import penetration, concentration ratios, links between
prices and costs, links between energy production, consumption and emissions, etc.
• Impact and policy analyses: sensitivity analyses, analyses of the impacts of taxation changes,
price changes, introduction of a minimum wage, specific economic crisis, earthquakes, etc.,
analyses of consumption and demand-based accounting and analyses of air emissions,
material flows, energy, water, etc.
• Industrial and sectoral analyses: changes over time to specific sectors, such as information
and communications technology (ICT), oil and gas, food, sport, creative arts, tourism, health,
etc., and, more recently, analyses covering the digital economy, sharing economy,
collaborative economy and also product-specific global value chains.
• Local government type investment planning: construction projects, shopping centres, new
motorways, rural planning, etc.
• Base structures for modelling: computable general equilibrium models, environmental
analyses, supply-side-based models, etc.
1.17. The role of SUTs and IOTs in understanding global value chains is of particular
importance, given the interconnected nature of today’s global economy. SUTs constitute the
centrepiece of the internationally compatible accounting framework for a systematic and detailed
description of the economy, its various components on the supply and use side and its relations to
other economies. The construction of international SUTs and IOTs makes it possible, in
combination with trade statistics, to follow the trade in value added and to understand who
ultimately benefits from the trade of finished goods in terms of value added, employment, and
other factors. The compilation of international or global SUTs and IOTs tables poses a number of

6
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

compilation challenges (including, for example, the recording of goods sent abroad for processing
and the recording of the production abroad and merchanting operations affecting SUTs and IOTs)
and relies on the availability of national SUTs and IOTs on a comparable basis.

1.18. In addition, the inclusion of the environmental dimension in the SUTs and IOTs further
enhances the usefulness of these tables by allowing the integration and consistency of the
economic and environmental information and an understanding of the interlinkages between the
economy and the environment. Incorporating environmental considerations as part of the regular
compilation of SUTs improves the quality, coherence and consistency of the related outputs and
the process provides powerful feedback loops for identifying improvements.

C. System of National Accounts

1.19. The SNA provides an internationally compatible framework for a systematic and detailed
description of a total economy (namely, that of a region, country or group of countries), its
components and its relations with other total economies. The 2008 SNA (United Nations,
European Commission, IMF, OECD and World Bank, 2009) is the latest version of the SNA,
which was adopted by the United Nations Statistical Commission in 2008.

1.20. The SNA describes the basic features of the accounting system in terms of concepts,
principles, statistical units and their groupings, etc. The SNA gives an overview of the sequence
of accounts, the balancing items associated with each account, a brief description of key aggregates
and the role of SUTs and the input-output framework. The key accounting sequence includes the
following stages: production of goods and services, transactions relating to products (goods and
services) and also to non-produced assets, transactions which distribute and redistribute income
and wealth, financial transactions and balance sheets.

1.21. The SNA framework also draws in other aspects, such as price and volume measurement,
population, labour market measures, regional accounts and various specific conceptual issues.
Figure 1.1 provides an overview of how SUTs and IOTs fit within the SNA framework. In
particular, it shows which accounts in the SNA sequence of accounts are more directly linked with
SUTs and IOTs, namely, production accounts, generation of income accounts, use of disposable
income accounts and capital accounts.

1.22. Producing annual SUTs simultaneously both in current prices and in volume terms, not
only ensures consistency for price volume measures, it also allows for the estimation of the volume
of GVA through what is termed “double deflation”, which is recommended in the 2008 SNA.

1.23. As noted above, the SUTs are an integral part of the SNA, determining a single estimate
of GDP both in current prices and in volume terms and linked to the institutional sector accounts.
For example, the goods and services account for the total economy can be directly compiled from
the SUTs through appropriate aggregation. In addition, by using the breakdown of GVA by

7
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

components in the use table, the production account and the generation of income account can
easily be compiled from the SUTs and linked to the institutional sectors.

1.24. Another important aspect linking the SUTs and the institutional sector accounts is the
statistical unit. The SNA uses two types of units and two corresponding ways of subdividing the
economy, which are quite different and serve separate analytical purposes. The units can be
classified to an industry for use in the SUTs and to an institutional sector for use in the institutional
sector accounts.

8
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 1.1 Overview of the links between SUTs and the SNA framework

Compiled by Sanjiv Mahajan, May 2014

1.25. The first purpose of describing production, income, expenditure and financial flows, and
balance sheets, is served by grouping institutional units into institutional sectors on the basis of

9
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

their principal functions, behaviour and objectives. The SNA enables a complete set of flow
accounts and balance sheets to be compiled for each sector and subsector, and also for the total
economy. The five institutional sectors distinguished in the SNA are the following:

• Non-financial corporations
• Financial corporations
• General government
• Households
• Non-profit institutions serving households
1.26. The SNA also describes the transactions between these five institutional sectors and the
rest of the world. These institutional sectors can be further split into subsectors, for example,
general government can be split into central government and local government.

1.27. The second purpose of describing processes of production and for input-output analysis is
served by the grouping of local kind-of-activity units (or establishments) into industries, on the
basis of their type of activity. An activity is characterized by an input of products, a production
process and an output of products.

1.28. In order to ensure consistency between SUTs and the institutional sector accounts, a link
table is compiled as an integrated part of the system. In this link table, a cross-classification of
output, intermediate consumption, components of value added (and other possible variables of
industries) between the industries and the institutional sectors is shown. Thus, this table links the
main macroeconomic variables from the SUTs to the institutional sector accounts, providing a
picture of local kind-of-activity units and one based on institutional units. As both types of units
are classified differently, the link table also provides a picture of the relations of output,
intermediate consumption, value added, and other variables, originating in the different industries
and institutional sectors.

1.29. The SUTs – consistent with the national accounts – are normally produced in connection
with the final or benchmarked versions of the macroeconomic data some two or three years after
first preliminary results of the national accounts are published. The SUTs, however, should play a
more vital role at the heart of national accounts in the production of preliminary annual or even
quarterly accounts. Once the SUTs compilation system is in place on an annual basis, the statistical
benefits are significant.

1.30. SUTs can play various roles in the national accounts. One, for example, is to update SUTs
– often in a more aggregated version – from the previous year with information available for the
preliminary estimates in order to have a complete set of SUTs available that are consistent with
the national accounts. This procedure is a good method for revealing inconsistencies in the
aggregated preliminary national accounts. Another role of SUTs could arise from new information

10
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

in a situation in which new, and more, detailed information on total supply and exports is available
at an early stage, the structure and relationships in the SUTs of the previous year could be used to
project SUTs for domestic output and imports.

1.31. The compilation of SUTs was in the past associated solely with the construction of IOTs.
The SUTs were therefore seen as an intermediate step in the compilation of IOTs. This meant,
effectively, that the SUTs were only compiled out after the compilation of the national accounts
had been completed. This approach, in fact, has significant limitations because the independently
calculated national accounts aggregates had to be kept unchanged despite inconsistencies
identified through the SUTs system.

1.32. SUTs are now seen as more than just as a step in the construction of IOTs: it is the SUTs
that provide the ideal framework guaranteeing the coherency and consistency of supply and use of
products in the system in current prices, and in volume terms, thereby improving the quality of the
national accounts, and in turn the key economic aggregates.

1.33. The compilation of SUTs is thus recommended as part of the regular annual compilation
of national accounts. The annual compilation of SUTs is also one of the recommended data sets
used in assessing the scope of implementation of the 2008 SNA. 3 The compilation of SUTs on a
quarterly basis can also play a role in improving the quality and coherence of quarterly national
accounts (the role played by SUTs in quarterly national accounts is further elaborated in chapter
14).

1.34. The approach to the compilation of SUTs as an integral component of the production of
national accounts may be formulated in general terms as follows:

• SUTs are produced as a central element of the compilation of national accounts with a view
to providing a key link to various parts of the SNA framework.
• SUTs provide a statistical framework representing the most efficient means of incorporating
all basic data – aggregated or detailed – covering the components of the three approaches to
measuring GDP, and linking to the institutional sector accounts in a systematic way.
• SUTs effectively ensure the consistency and reconciliation of results at a detailed level and
thereby improve the overall quality of the national accounts.
• SUTs are compiled and balanced in both current prices and in volume terms.
• SUTs are produced annually or even, if possible, on a quarterly basis, the ideal option.

3
See table 2 of the 2011 report of the Intersecretariat Working Group on National Accounts to the United Nations
Statistical Commission at its forty-second session (E/CN.3/2011/6), available at:
http://unstats.un.org/unsd/statcom/doc11/2011-6-NationalAccounts-E.pdf.

11
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• SUTs can provide a powerful feedback mechanism on the coherency and consistency of
source data, such as business surveys, and on the classification of units on the business
register.
1.35. When balanced, SUTs provide a coherent, consistent and wholly integrated suite of
statistics for a single period (for example, a year), which include:

• A single estimate of GDP in current prices and in volume terms, which is underpinned with
components of the production, income (only in current prices) and expenditure approaches
to measuring GDP
• Detailed goods and services account in current prices and in volume terms (not by
institutional sector)
• Production accounts by industry and by institutional sector in current prices and in volume
terms.
• Generation of income accounts by industry and by institutional sector (both in current prices
only)
• Link to the use of disposable income account through the flows of final consumption
expenditures and capital account through gross capital formation (and its components)
balanced via SUTs.
1.36. These guidelines should form part of the strategic tools used to improve the quality of the
national accounts.

D. Objectives of this Handbook

1.37. The theoretical development of IOTs has a long history. Box 1.1 provides a description of
the evolution of both IOTs and SUTs within the context of national accounts. The United Nations
Statistics Division has followed the theoretical development and the practical work of national
statistics offices on IOTs and SUTs from the outset. Starting in 1996, it has prepared a number of
publications, under the guidance of the United Nations Statistical Commission, such as those listed
among the references under United Nations, 1966, 1973 and 1999, to share practices, update the
methodology in line with the updates of the SNA, and provide guidance on the compilation of
IOTs.

1.38. This Handbook continues those efforts in cooperation with other international
organizations and experts, providing practical and step-by-step guidance for the compilation of
SUTs and IOTs based on the latest international statistical standards set out in the 2008 SNA and
the sixth edition of the International Monetary Fund (IMF) Balance of Payments and International
Investment Position Manual (BPM 6) (IMF, 2009).

12
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 1.1 Evolution of the SUTs and IOTs within the national accounts
The national accounting system is continuously evolving across the various domains to reflect developments in
and improvements to the quality of economic statistics and the evolution of economies, with a view to providing
a relevant measurement of the economy. Over the past four centuries many significant contributions have been
provided by people from various disciplines and countries, resulting in the system as it stands today and how it
relates to SUTs and IOTs. Below is a short description of this evolution.
Wassily Leontief (1905–1999), who is often referred to as the pioneer of input-output-based economics, made
the first of many key contributions with the publication of his article “Quantitative input and output relations in
the economic system of the United States”. This article discussed the construction of an economic transactions
table that Leontief had based on the tableau économique, proposed by François Quesnay in 1758.
The framework was developed and applied as an economic tool with the construction of the first IOTs for the
United States covering the years 1919 and 1929 published in 1936. Later, Leontief developed the first input-
output-based model, which was based on theories developed by Leon Walras published in 1874 and 1877.
Leontief‘s pioneering work was recognized by the award to him of the Nobel Prize in Economics in 1973. As
a result, input-output analysis has become a major tool in developing quantitative economics as a science.
The role of SUTs and IOTs has evolved within national accounts. The 1953 SNA (United Nations, 1953)
included no reference to SUTs or IOTs. The 1968 SNA (United Nations, 1968), however, presented the
integration of an input-output framework into the integrated economic accounts of the SNA. The conceptual
development of the integrated economic accounts of the SNA earned Richard Stone the Nobel Prize in
Economic Science in 1984, for having made fundamental contributions to the development of the SNA and
having thereby greatly improved the basis for empirical economic analysis.
Alongside Leontief and Stone, other Noble laureates include Ragnar Frisch and Jan Tinbergen in 1969, Paul
Samuelson in 1970, Simon Kuznets in 1971, John Hicks in 1972 and James Meade in 1977, who have all
contributed to the foundations of the measurements used in today’s SNA and the interlinkages between various
sectors and activities in an economy.
The latest evolution of SUTs was recognized in the 1993 SNA (United Nations, CEC, IMF, OECD and World
Bank, 1993), chapter XV of which covered both SUTs and IOTs. With the latest version of the SNA, the 2008
SNA, the role and applications of SUTs have been further enhanced, and this in turn will help to meet many
analytical needs, as reflected in chapters 14 and 28 of the 2008 SNA.

1.39. This Handbook may, therefore, be viewed as an update of the United Nations Handbook
of Input-Output Table Compilation and Analysis (United Nations, 1999). In response, however, to
the ever-increasing importance of SUTs in their own right, this Handbook extends the scope of the
previous publication by providing a more detailed description and compilation guidance for SUTs.
As stated in the 2008 SNA, “only supply and use tables provide a sufficiently rigorous framework
to eliminate discrepancies in the measured flows of goods and services throughout the economy
to ensure the alternative measures of GDP converge to the same value” (2008 SNA, para. 14.15).

1.40. The Handbook builds on the experience, practices and guidance available at national and
regional level, such as the Eurostat Manual of Supply, Use and Input-Output Tables (Eurostat,
2008). At the same time, however, it provides an innovative approach to the compilations of SUTs
and IOTs in the following three main areas:

• Underlying use of an integrated approach to statistics

13
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Use of a business model for the compilation of SUTs and IOTs linking the various parts
through an H-Approach compilation scheme
• Mainstreaming of environmental considerations, through the inclusion of the environmental
focus of chapter 13 at the core of part three of the Handbook
1.41. The compilation guidance provided in this Handbook relies on an integrated statistics
approach whereby the production of statistics in the various domains is not seen in isolation but as
part of an integrated process using common concepts, definitions, business registers and frames,
statistical units, estimation methods and data sources to improve the consistency of the statistics
compiled, to reduce the respondent burden, and potentially to reduce the statistical agency costs.
In particular, the consistency of the basic economic information that feeds into the national
accounts and the SUTs with the classifications, concepts and definitions of the 2008 SNA greatly
reduces the discrepancies across data from different sources, thus facilitating their reconciliation
as part of the integration process. The integrated statistics approach is described in the Guidelines
on Integrated Economic Statistics (United Nations, 2013).

1.42. This Handbook follows the Generic Statistical Business Process Model (GSBPM)
(UNECE, 2013) to describe the production of statistics in a general and process-oriented manner.
The underlying concepts and principles of the GSBPM have been followed in describing the
business process and stages of the statistical production processes underpinning the compilation
of SUTs and IOTs. Chapter 3 of this Handbook describes these links in more detail in the context
of SUTs and IOTs. In addition, the chapters in parts two and three of the Handbook are linked to
the different parts of these stages of the statistical production process.

1.43. With the adoption of the System of Environmental-Economic Accounting (SEEA) (United
Nations, European Commission, FAO, IMF, OECD and World Bank, 2014) by the United Nations
Statistical Commission, the extension of SUTs and IOTs to include environmental flows in
monetary and physical terms has become an internationally agreed standard. Including
environmental consideration from the outset in the compilation of SUTs brings a number of
advantages. It facilitates the integration and reconciliation of the information, it enhances the
quality of the information, and it significantly increases the uses to which the tabulations may be
put.

1.44. In line with the United Nations Statistical Commission,4 this Handbook recommends the
annual compilation of SUTs. In addition, the Handbook promotes the compilation of these tables
as an integral part of the compilation of national accounts in order to ensure full consistency of the
basic data and also of the macroeconomic estimates that are derived from the accounts.

4
See the 2011 report of the Intersecretariat Working Group on National Accounts to the United Nations Statistical
Commission at its forty-second session (E/CN.3/2011/6), available online at:
http://unstats.un.org/unsd/statcom/doc11/2011-6-NationalAccounts-E.pdf

14
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

1.45. The Handbook provides a consistent numerical example of SUTs and IOTs that runs
throughout the chapters (as far as practically possible), in order to facilitate understanding of the
various compilation steps. It also provides examples of best practices to illustrate certain aspects
of the compilation of SUTs. It should be noted that, in the numerical examples provided in the
Handbook, the numbers may not add up exactly to the totals because of rounding.

1.46. The target audience for this Handbook mainly includes compilers of SUTs and IOTs with
a basic knowledge and understanding of the SNA. Since, however, the Handbook provides an
overview of the whole statistical production process, managers or staff in charge of the programme
of national accounts and of economic and environmental accounts would also benefit from the
Handbook in gaining an overall understanding of the requirements for the compilation of SUTs
and IOTs. Lastly, analytical users may also benefit from perusing the Handbook, as it would
provide them with a better understanding of the compilation steps, thus increasing the analytical
applications of SUTs and IOTs.

E. Structure of the Handbook

1.47. The Handbook consists of the following four main parts:

• Part one, providing an introduction to the Handbook, set out in chapter 1.

• Part two, describing the overview of SUTs and IOTs, the fundamental building blocks
required, cross-cutting issues and the main stages of the GSBPM – namely, the design,
building and collection phases. Part two comprises chapters 2–4.
• Part three, describing the compilation, balancing and dissemination phases of SUTs and
IOTs. This also includes the physical SUTs (PSUTs), environmentally extended IOTs (EE-
IOTs) and the SUTs links to the quarterly national accounts. Part three comprises chapters
5–15.
• Part four, providing examples of the extensions and applications of SUTs and IOTs. Part
four comprises chapters 16–21.

Part one

1.48. As indicated above, chapter 1 provides an introduction to the Handbook; it describes the
importance of SUTs and IOTs for statistical purposes (for example, compilation of annual and
quarterly national accounts, etc.), for policymaking and for analytical purposes (for example,
economic forecasting, assessing the impact of globalization). It also provides a general description
of the SNA and where the SUTs fit within the SNA framework. This chapter also describes the
overall approach of the Handbook (including in comparison to previous handbooks) to the
compilation of SUTs and IOTs and provides a general outline of its contents.

15
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Part two

1.49. Chapter 2 provides a conceptual overview of SUTs and IOTs and describes the basic
elements determining their structure and compilation. These include the accounting principles of
the SNA, the classifications of economic activities and products, the choice of the statistical units
and how they affect SUTs and IOTs, and the valuation in SUTs and IOTs. The chapter identifies
the advantages of compiling SUTs as an integral part of the national accounts and how the SUTs
are used to obtain consistent estimates of GDP. It also describes in more detail the extended
perspective adopted in this Handbook to SUTs and IOTs, incorporating an environmental
dimension which makes possible an integrated overview of the framework from the very outset.

1.50. Chapter 3 provides an overview of the different phases that constitute the statistical
production process of SUTs and IOTs, based on the stages of the GSBPM and in line with the
United Nations Guidelines on Integrated Economic Statistics (United Nations, 2013). This chapter
also provides an overview of the different institutional set-ups in countries which may have an
impact on the compilation process. The compilation phases specific to SUTs and IOTs are
presented in this chapter together with a link to the relevant chapters of the Handbook.

1.51. Chapter 4 covers specific phases of the GSBPM, namely the specify needs, design, build
and collect phases. It provides a description of the elements that should be considered and carefully
evaluated at the beginning of the compilation process, such as the level of detail of the industry
and products in the tables, the compilation schedule, the revision policy, resources, typical data
sources, and others. These and other issues are covered in this chapter, thus providing a foundation
for the compilation of SUTs and IOTs.

Part three

1.52. Chapter 5 describes the conceptual and practical aspects of the compilation of the supply
table and how the so-called “unbalanced” supply table is put together from the typical data sources
for SUTs, such as business surveys, administrative data and others.

1.53. Chapter 6 describes the conceptual and practical aspects of the compilation of the use table.
As in chapter 5, this chapter shows how an unbalanced use table is constructed on the basis of
typical data sources.

1.54. Chapter 7 describes how to compile the valuation matrices necessary to bridge the different
valuation concepts of the product flows. This chapter covers the main concepts and methodologies
of compiling matrices for trade margins, transport margins, taxes on products and subsidies on
products.

1.55. Chapter 8 describes the structure of the imports use table and the domestic use table and
the steps necessary to disaggregate the use table into an imports use table and a domestic use table.
Historically, the compilation of these tables was largely viewed as an intermediate – though not

16
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

essential – step in the compilation of IOTs. That said, however, the imports use table and the
domestic use table are becoming increasingly important in their own right for analytical purposes.

1.56. Chapter 9 covers the compilation of SUTs in volume terms. It follows the recommendation
that SUTs should simultaneously be compiled in current prices and in volume terms. The
compilation of SUTs in volume terms can start after the SUTs have been compiled in current prices
(although the current price tables do not need to be balanced) but there is need for a simultaneous
presentation of volume and price indices.

1.57. Chapter 10 describes the importance of linking SUTs and the institutional sector accounts,
which involve data by industry that are to be subdivided according to the institutional sectors to
which the units within each industry are assigned. The chapter provides guidance on how to
compile the cross-tabulation between industries and institutional sectors and presents various
approaches that may be followed in establishing the link between the SUTs and the institutional
accounts. It also identified certain issues that may arise in the compilation of the linking table.

1.58. Chapter 11 describes the manual and automated balancing procedures of SUTs in both
current prices and in volume terms. This is important for full consistency of the detailed
information. The various checks related to product, industry and macro identities, benchmarking
with national accounts, and comparison with previous SUTs, if available, are explained. It is
recommended that SUTs should be produced and balanced simultaneously at basic and purchasers’
prices and also for domestic and imported products, all of which should be both in current prices
and in volume terms. A further dimension, and challenge to be surmounted, is the need to cover
both annual SUTs and, if possible, quarterly SUTs.

1.59. The sequence of chapters represents the preferred scenario for the compilation of SUTs.
Different approaches may be followed, however. An increasing number of countries have achieved
the preferred scenario. This scenario for the compilation of SUTs and IOTs may be seen as
ambitious but it can be realized through gradual improvements in source data, production
processes and the information technology environment.

1.60. Chapter 12 provides an overview of the IOTs (product by product and industry by industry)
and describes the methods and the underlying assumptions for transforming SUTs into IOTs. The
compilation of IOTs is quite different in nature from that of SUTs and relies on the availability of
SUTs. The compilation of IOTs is considered more as an analytical step than a compilation process
and, for this purpose, is viewed as a transition from statistics to modelling.

1.61. Chapter 13 describes the structure of the SUTs in physical units where additional rows and
columns are added to show flows from the environment to the economy and vice versa. This
chapter also describes typical data sources for the compilation of these tables and examples of
specific issues in which the SEEA and the SNA differ (for example, the treatment of international
flows and the treatment of goods for processing), and shows how standard economic IOTs in
monetary units may be extended to include information on the environment in physical units in

17
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

the environmentally extended IOTs. Physical IOTs are also an extension of the SUTs framework,
extended to take into account environmental considerations. They consist of a transformation of
the PSUTs. Given certain conceptual and practical issues in the compilation of physical IOTs,
however, the focus of the 2012 SEEA – and thus of this chapter – has shifted towards the
compilation of environmentally extended IOTs rather than physical IOTs. Examples of two
country practices are also presented in this chapter.

1.62. Chapter 14 provides an overview of how SUTs may be used to improve the quarterly
national accounts. Since there are various scenarios that can be used in practice, this chapter
focuses only on three main situations which illustrate the use of SUTs to various degrees in the
compilation of the quarterly national accounts.

1.63. Data dissemination is an important activity for any statistical production process, as it
provides users with a range of statistics produced to internationally agreed guidelines. Presenting
SUTs and IOTs to users in a clear, transparent and user-friendly manner is thus an important task
for the statisticians. Chapter 15 provides an overview of the elements that should be considered
when disseminating SUTs and IOTs, such as the identification of users’ needs in order to tailor
dissemination to the main types of users of SUTs and IOTs, the importance of having a
dissemination strategy and the elements that should be covered in the strategy. Reference to the
Statistical Data and Metadata Exchange (SDMX) for SUTs and IOTs is also provided in this
chapter.

Part four

1.64. Chapter 16 describes the methods for compiling regional (subnational) SUTs and the main
compilation issues, such as the disaggregation of the information at subnational level, among
others. Different issues and challenges are covered through a bottom-up and top-down compilation
approach.

1.65. Although the focus of this Handbook is mainly on the compilation of national SUTs and
national IOTs, there is a growing demand for these instruments to capture the structure and
mechanism of the cross-border fragmentation of production activities. Chapter 17 provides an
overview of multi-country SUTs and IOTs, the main compilation issues, and a simplified
compilation procedure. This chapter also reviews current international initiatives in this area.

1.66. Chapter 18 deals with the projections of SUTs and IOTs. Many users also require
comparable input-output products that are comparable in terms of frequencies and timeliness. For
example, some countries produce quarterly SUTs, some countries produce annual SUTs and some
countries produce SUTs on a less regular basis. Consequently, a variety of methods, techniques
and approaches exist for the projection of SUTs and IOTs and for dealing with data gaps. These
techniques also can help producers, for example, to deal with periods between benchmarked years.
This chapter examines various methods and techniques used, along with a range of literature
available on how to surmount the problem of incomplete data, thus allowing the estimation and

18
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

projection of IOTs. The chapter also presents a numerical example for three methods: the
generalized RAS, the SUT-RAS and the Euro methods.

1.67. Chapter 19 describes the main extensions of supply, use and input-output tables as part of
a satellite system which is routinely used for economic analysis. Several examples of the
disaggregation of the use table and various satellite accounts are reviewed, including such
extensions as social accounting matrices, extended IOTs and other satellite systems.

1.68. Chapter 20 describes the different types of input-output models and provides a broad
overview illustrating the benefits and the approaches used. The traditional quantity model and
price model of input-output analysis are presented for monetary IOTs and physical IOTs. Input
and output coefficients, the Leontief inverse, price and quantity models, indicators, multipliers and
inter-industrial linkages were developed for an empirical extended IOT, with extensions for gross
fixed capital formation, capital stock, employment, energy, air emissions, waste, sewage and
water.

1.69. Chapter 21 provides examples of compilation practices from various countries with
different statistical systems. In general, the compilation practices can vary greatly depending on
the resources available, the statistical infrastructure, registers, surveys, methodologies and other
factors. This chapter provides guidance for countries with limited statistical resources and
illustrates differences and challenges in the compilation of SUTs and IOTs.

19
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Part two

21
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Chapter 2. Overview of the supply and use tables and input-output tables

A. Introduction

2.1. Before providing step-by-step guidance on the compilation of SUTs and IOTs, it is
important to ensure a general understanding of the two sets of tables. The main objective of this
chapter is, therefore, to provide an overview of SUTs and IOTs, which may be found in sections
B and C, respectively. Section D introduces the fundamental elements of the SUTs and IOTs, such
as the underlying classifications, the statistical units and the valuation methods. Some of these
elements are discussed in more detail in subsequent chapters. Lastly, section E elaborates on the
importance of compiling SUTs as an integral part of the national accounts.

2.2. Although this chapter covers a wide range of challenges and issues that must be tackled
when planning and building a new system of SUTs and IOTs, all aspects may not be achievable in
countries with limited resources. It is worth recognizing that a system may be established with a
moderate level of ambition using available data, even if these are incomplete. Nonetheless, it is
preferable to have a SUTs-type environment for reconciliation of the various statistical sources,
rather than only an unbalanced series of national accounts aggregates.

B. Overview of SUTs

2.3. The SUTs describe the whole economy by industry (for example, the motor vehicle
industry) and by product (for example, sports goods). The tables show links between components
of GVA, industry inputs and outputs, and product supply and use. The SUTs link different
institutional sectors of the economy (for example, non-financial corporations) together with details
of imports and exports of goods and services, final consumption expenditure of government,
household and non-profit institutions serving households (referred to as NPISHs), and capital
formation.

2.4. As their name suggests, SUTs consist of two interlinked tables: the supply table and the
use table. The supply table shows the supply of goods and services by type of product and by type
of industry, distinguishing between supply by domestic industries and imports of goods and
services. In other words, the supply table provides information on the output (by product)
generated by economic activities and the imports (by product) from abroad. The totals in the last
column represent the total supply by products and the totals in the bottom row represent the total
output by economic activity and total imports. A simplified supply table is presented in Table 2.1
below.

23
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 2.1: Simplified structure of the supply table

Indus tri es Industries
Agriculture, Mining and Imports Total
… Services
Products forestry, etc. quarrying
Agriculture, forestry, etc.
Ores and minerals, etc. Imports by Total supply
Output by product by industry
… product by product

Services

Total Total output by industry Total imports Total supply

2.5. The second table is the use table, which provides information on the uses of the different
products. The use table shows the use of goods and services by type of product and by type of use,
in other words, as intermediate consumption by industry, final consumption, gross capital
formation or exports. Furthermore, the table shows the components of gross value added by
industry – namely, compensation of employees, other taxes less subsidies on production,
consumption of fixed capital and net operating surplus. While the totals by row represent the total
uses by product, the total by column represent the total output by economic activity, total final
consumption, total gross fixed capital formation and total exports. Table 2.2 below shows the
simplified structure of the use table.

Table 2.2: Simplified structure of the use table

Indus tri es Industries Final uses

Agriculture, Mining and Final Gross capital Total

Products … Services Exports
forestry, etc. quarrying consumption formation

Agriculture, forestry, etc.

Ores and minerals, etc. Total use by
Intermediate consumption by product and by industry Final uses by product and by category
… product

Services

Value added Value added by component and by industry Value added

Total Total output by industry Total final uses by category

Empty cells by definition

2.6. The classification of products, in practice, is often more detailed than the classification of
industries, thus generating rectangular SUTs. For example, the output of the dairy industry is
separately shown in the SUTs for the products of processed milk, butter, yoghurt, cheese and so
forth, and not as only one aggregate product for all dairy products.

2.7. There are three basic identities that hold between the supply table and the use table. The
first identity corresponds to the fundamental identity in national accounts, whereby for each
economic activity the following holds:

Identity (1) Output = Intermediate consumption + GVA

24
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.8. The second identity is that the total supply by product is equal to the total use by product.
This means that the amount of products available for use in an economy must have been supplied
either by domestic production or by imports, and the same amount of products entering an
economy in an accounting period must be used for intermediate consumption, final consumption,
capital formation or exports. This means that, for each product (or group of products):

Identity (2) Output + Imports = Intermediate consumption + Final consumption +

Capital formation + Exports

2.9. Another important identity which is also key when linking the production and income
approaches to calculating GDP and the industry and institutional sector dimension through the
SUTs is the following:

Identity (3) For each industry, the GVA using the production approach equals the
GVA estimate using the income approach.

2.10. These identities are fundamental in the balancing process that is carried out when
compiling SUTs both in current prices and in volume terms, all through a time series dimension.

2.11. Once balanced, the supply table and the use table can be integrated into a single matrix –
often referred to as the SUTs framework, which is shown in Table 2.3 below. This table clearly
shows the two basic identities linking the SUTs. The total supply by product (left part of the bottom
row of Table 2.3) equals the total use by product (top part of the last column of Table 2.3) and the
total outputs by industry are identical in both SUTs (the middle part of the bottom row equals the
middle part of the last column). The schematic view of SUTs in Table 2.3 also serves as the
underlying matrix for projection methods (see chapter 18).

Table 2.3: Supply and use tables framework

Products Industries Final uses
Agriculture, Ores and Agriculture, Mining and Final Gross capital Total
… Services … Services Exports
forestry, etc. minerals, etc. forestry, etc. quarrying consumption formation
Agriculture, forestry, etc.
Products

Ores and minerals, etc. Intermediate consumption by product and by Total use by
Final uses by product and by category
… industry product

Services

Agriculture, forestry, etc.

Industries

Mining and quarrying Total output

Output by product by industry
by industry
…
Services

Value added Value added by component and by industry Value added

Total
Imports Total imports by product
imports
Total Total supply by product Total output by industry Total final uses by category

Empty cells by definition

2.12. SUTs thus bring together the components of each of the three approaches to measuring
GDP– namely, the production, income and expenditure approaches:

(a) Production approach:

25
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

GDP = Output (at basic prices) - Intermediate consumption + Taxes less subsidies
on products

(b) Income approach:

GDP = Compensation of employees + Gross operating surplus + Other taxes less

subsidies on production + Taxes less subsidies on products

(c) Expenditure approach:

GDP = Final consumption + Gross capital formation + Exports - Imports

2.13. When balanced, SUTs show, by definition, a single estimate of GDP both in current prices
and in volume terms. This underlines the importance of the recommendation that SUTs be
compiled as part of the annual regular compilation of the national accounts, as they ensure the
consistency and coherence of the national accounts components, namely, goods and services
account, production account (by industry and by institutional sector) and generation of income
account (by industry and by institutional sector), and make it possible to derive a single estimate
of GDP. The institutional sector links are covered in more detail in chapter 10.

2.14. The SUTs also have links to other accounts, such as disposable income accounts (covering
variables like household final consumption expenditure) and accumulation accounts (covering
variables like gross fixed capital formation as part of the capital account).

2.15. Producing annual SUTs simultaneously both in current prices and in volume terms
(preferably, when two successive years of current price SUTs are available) ensures coherence and
consistency for both price and volume measures. In addition, this approach allows for the
estimation of the volume of GVA through what may be termed “double deflation”, where GVA is
derived by deducting intermediate consumption in volume terms from total output in volume
terms. This can be achieved on the basis of an individual unit, industry, institutional sector, and
for the whole economy.

2.16. SUTs can also be compiled on a quarterly basis to derive official estimates of quarterly
GDP. Developing quarterly SUTs may be highly demanding in terms of resources, time and data
availability but have the advantage of significantly improving the quality of the estimate of
quarterly GDP.

2.17. Figure 2.1 provides a graphical overview of the SUTs, explicitly identifying the main
identities that are ensured in balanced SUTs. Box 2.1 presents a numerical example of balanced
SUTs.

26
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 2.1: Graphical overview of supply and use tables

2.18. The use table records the intermediate consumption and final uses by type of product but
it does not distinguish between the consumption of domestically produced goods and services and
that of imported goods and services. Although such a split is not a necessary condition for the
creation of balanced SUTs in current prices, it is a key step linking SUTs and IOTs. The
disaggregation of the use table into two tables, the domestic use table and the imports use table, is
shown in Box 2.2, with a numerical example.

2.19. The compilation of the imports use table is necessary to ensure good quality volume
estimates (in particular, GVA by industry) and these tables are becoming increasingly important
owing to the growing impact of globalization and the need to measure global value chains and
trade in value added.

27
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.1 Numerical example of the SUTs system

Supply table
Industries
Manufacturing Imports Total supply
Agriculture and Services
construction
Agriculture 270 30 50 20 370
Manufacturing 6 380 87 42 515
Products

Construction 4 50 13 8 75
Trade, transport and communication 10 15 210 7 242
Finance and business services 6 17 240 11 274
Other services 4 8 100 12 124
Total 300 500 700 100 1 600

Use table
Industries Final use
Manufacturing Final Total use
Gross capital
Agriculture and Services consumption Exports
formation
construction expenditure
Agriculture 34 59 93 131 21 32 370
Manufacturing 97 107 57 122 73 59 515
Products

Construction 9 12 4 17 30 3 75
Trade, transport and communication 42 24 11 140 20 5 242
Finance and business services 14 53 42 116 31 18 274
Other services 14 35 22 35 10 8 124
Taxes less subsidies on products 4 5 12 52 6 1 80
GVA 86 205 459 750
Total 300 500 700 613 191 126 2 430

Supply and use tables framework

Products Industries Final use
Trade, transport Finance and Manufacturing Final Gross
Total
Other
Agriculture Manufacturing Construction and business Agriculture and Services consumption capital Exports
services
communication services construction expenditure formation
Agriculture 34 59 93 131 21 32 370
Manufacturing 97 107 57 122 73 59 515
Products

Construction 9 12 4 17 30 3 75
Trade, transport and communication 42 24 11 140 20 5 242
Finance and business services 14 53 42 116 31 18 274
Other services 14 35 22 35 10 8 124
Agriculture 270 6 4 10 6 4 300
Industries

Manufacturing and construction 30 380 50 15 17 8 500

Services 50 87 13 210 240 100 700

Taxes less subsidies on products 4 5 12 52 6 1 80
GVA 86 205 459 750
Imports 20 42 8 7 11 12 100
Total 370 515 75 242 274 124 300 500 700 613 191 126 4 030

= Zero by definition

The supply table shows the supply of goods and services by product and by type of supplier, distinguishing supply
by domestic industries and imports of goods and services. The domestic output of industries is shown by products.
The vector of imports comprises the nation’s total imports of goods and services by product.
The use table shows the use of goods and services by product and by type of use, i.e. as intermediate consumption
by industry, final consumption expenditure, gross capital formation and exports of goods and services. The
intermediate uses and final uses reflect the consumption of domestically produced goods and services and also of
imported goods and services. Furthermore, although the table is shown in summary form, it should be noted that
there are components underlying the headings, for example, GVA can be split between compensation of employees,
other taxes less subsidies on production, consumption of fixed capital and net operating surplus.
Note that, for illustrative purposes, it is assumed that the SUTs presented here are compiled on a consistent valuation
basis.

2.20. Once the imports use table is constructed, the domestic use table can be estimated by
subtracting the imports use table from the use table. The imports use table and the domestic use
table form the basis for the construction of input imports tables and domestic IOTs, respectively.
More detail may be found in chapters 8 and 12.

28
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.2 Numerical example showing a use table split between consumption of domestic
production and imports
Supply table
Industries
Manufact. Total
Agricul- Output Imports
and Services supply
ture
construction
Agriculture 270 30 50 350 20 370
Manufacturing 6 380 87 473 42 515
Products

Construction 4 50 13 67 8 75
Trade, transport and communication 10 15 210 235 7 242
Finance and business services 6 17 240 263 11 274
Other services 4 8 100 112 12 124
Total 300 500 700 1 500 100 1 600

Use table Domestic use table

Industries Final use Industries Final use
Manufact. Final Gross Total Final Gross
Agricul- Agricul- Manufact. and Total use
and Services consumption capital Exports use Services consumption capital Exports
ture ture construction
construction expenditure formation expenditure formation
Agriculture 34 59 93 131 21 32 370 Agriculture 30 50 90 130 20 30 350
Manufacturing 97 107 57 122 73 59 515 Manufacturing 85 90 51 120 70 57 473
Products

Products
Construction 9 12 4 17 30 3 75 Construction 5 10 3 16 30 3 67
Trade, transport and communication 42 24 11 140 20 5 242 Trade, transport and communication 40 20 10 140 20 5 235
Finance and business services 14 53 42 116 31 18 274 Finance and business services 10 50 40 115 30 18 263
Other services 14 35 22 35 10 8 124 Other services 10 30 20 35 10 7 112
Taxes less subsidies on products 4 5 12 52 6 1 80 Imports 30 40 15 5 5 5 100
GVA 86 205 459 750 Taxes less subsidies on products 4 5 12 52 6 1 80
Total 300 500 700 613 191 126 2 430 GVA 86 205 459 750
Output 300 500 700 613 191 126 2 430

Imports use table

Industries Final use
Final Gross
Agricul- Manufact. and Total
Services consumption capital Exports
ture construction
expenditure formation
Agriculture 4 9 3 1 1 2 20
Manufacturing 12 17 6 2 3 2 42
Products

Construction 4 2 1 1 8
Trade, transport and communication 2 4 1 7
Finance and business services 4 3 2 1 1 11
Other services 4 5 2 1 12
Total 30 40 15 5 5 5 100

Empty cells by definition

The domestic use table is derived by subtracting the imports use table from the total use table shown in box 2.1.
The imports of goods and services are then shown separately as a new row denoted as “Imports” in the domestic
use table. The domestic use table shows the input requirements of industries in terms of domestic intermediates,
imported intermediates and primary inputs (GVA). It also shows the use of domestic output of products for
intermediate uses and final uses.
The imports use table includes information on the use of imported products for intermediate consumption and
final uses and the column totals, which match the estimates shown in the “Imports” row.

1. Supply and use tables in current prices and in volume terms: H-Approach

2.21. The SUTs framework not only constrains the current value estimates of supply and use of
products to balance exactly, it also provides a means of ensuring that the corresponding volume
estimates in previous years’ prices are balanced and that the series of prices implied by the
existence of one table in current prices and one in volume terms are strictly consistent. In general,
the best way to ensure mutual consistency is to prepare the SUTs in current values and in volume
terms at the same time (2008 SNA, para. 14.136).

2.22. The compilation and balancing of SUTs in current prices and in volume terms for a
sequence of years also helps to balance the changes in volumes, values and prices in the best
possible way (the key condition for the attainment of this outcome is that SUTs are available in
current prices both for the current year and for the previous year). This approach ensures a high
degree of quality in terms of coherence and consistency over time – and recommended as the best

29
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

approach for the production of SUTs. Producing annual SUTs simultaneously both in current
prices and in volume terms also allows estimation of the volume of GVA through double deflation,
whereby GVA in previous years’ prices is derived by deducting intermediate consumption in
previous years’ prices from total output in previous years’ prices. Then, the change in volume of
GVA between each pair of consecutive years is given by the change of GVA in previous years’
prices compared to GVA of the previous year at current prices of that year.

2.23. The SUTs at purchasers’ prices and at basic prices in current prices and in volume terms
can be compiled and balanced sequentially or simultaneously. In both cases, powerful feedback
loops covering quality in terms of consistency and coherence are available. More details are
provided in chapters 9 and 11.

2.24. Figure 2.2 shows an overview of the H-Approach for an integrated compilation of SUTs
(and IOTs) in current prices and volume terms. The H-Approach is the recommended compilation
approach, which brings together the compilation of SUTs in current prices and volume terms, the
valuation at basic prices, producers’ prices and purchasers’ prices, and the links with the
compilation of IOTs. The matrices covering other taxes on production, other subsidies on
production, trade margins and transport margins are the valuation matrices which link between
basic prices, producers’ prices and purchasers’ prices.

Figure 2.2 Schematic overview of the compilation of SUTs and IOTs: H-Approach

Current prices Volume terms

SUTs SUTs
at purchasers’ prices at purchasers’ prices

- Valuation matrices - Valuation matrices

- Split domestic use table - Split domestic use table
and import use table and import use table
Deflation
processes

SUTs SUTs
at basic prices at basic prices

Assumption applied Assumption applied

IOTs IOTs

30
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.25. The diagram in Figure 2.2 may be visualized as the letter “H” with the left vertical arm
representing SUTs and IOTs in current prices and the right vertical arm representing the SUTs and
IOTs in previous years’ prices. The horizontal transition represents the deflation process using, for
example, a combination of prices, volume indicators and rates of the previous year applied to the
volumes.

2.26. The SUTs in current prices are decomposed into the component parts (imports and the
valuation matrices on the left-hand side of the H-Approach), each of which is deflated separately
as appropriate (the join in the middle), and then added back to get to a purchasers’ prices valuation
in previous years’ prices (the right-hand side). This means that basic prices play the dominant role
in the process, and the initial compilation flow is from top left to middle left and that of deflation
to middle right, and then to top right.

2.27. When final use deflators are deemed to be better for final use components at purchasers’
prices, then the H-Approach allows for the use of higher quality deflators, which are perhaps more
appropriate. In these types of examples, it is possible to work with purchasers’ prices where the
data are believed to be more reliable, making appropriate adjustments, working from top right to
middle right, then onto middle left and to top left, and still ensuring a balance at each stage.
Similarly, if high quality volume indicators are available, then this can better inform, for example,
the step between the middle left and middle right with adjustments as appropriate.

2.28. For balancing purposes, areas such as value added tax (VAT) on products and changes in
inventories may be separated out as balanced matrices (whereby the impact on production, income
and expenditure is equal and the matrices are in balance) to avoid any balancing adjustments, but
this may overcomplicate the system.

2.29. It is important to note that the scheme presented in Figure 2.2 should not be taken as one
to be implemented as a whole. In practice, for example, if a country wants to focus exclusively on
the compilation of annual SUTs, the focus should be on the compilation steps of the SUTs within
the bold line box in Figure 2.2 which are recommended in order to achieve balanced SUTs in
current and in volume terms. If, however, a country wants to compile SUTs and IOTs, all the steps
in Figure 2.2 should be completed in order to guarantee important feedback loops and enhance the
quality of the tabulations.

2.30. When planning for the compilation of SUTs, it is useful to keep in mind the compilation
approach in Figure 2.2, since it is naturally linked to the production of time series of SUTs (and
IOTs) both in current prices and in previous years’ prices using chain-linked volumes. More detail
of this may be found in chapter 9. Although SUTs in volume terms for one period can be compiled
using SUTs in current prices for one period and deflators, the preferred approach contains a time-
series dimension and the following principles:

• To compile SUTs in volume terms, the following are needed:

31
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

 Balanced set of SUTs at purchasers’ prices in current prices for the present year and
the previous year

 Deflators for each flow

• Previous year’s SUTs in current prices of that year are needed to derive volume estimates.
The SUTs in figure 2.2 are balanced for illustrative purposes. In reality, however, they may
be unbalanced and an iterative balancing process may be necessary. This allows for the first
SUTs in previous years’ prices to be compiled.

• Each transition stage is created in a balanced format which provides a much easier systematic
process and build. This means that each of the transition matrices covering taxes, subsidies,
trade and transport margins and import of goods and services will be balanced, in other
words, the supply-side row total will equal the use-side row total.

• For some of the variables, like household final consumption expenditure, there are already
present deflation approaches using consumer price indices to generate the corresponding
estimates in previous years’ prices. These estimates are likely to differ from those generated
with the H-Approach but would also feature in the reconciliation and balancing process of
the estimates, and form an example of working from right to left through balanced
adjustments.

2.31. The H-Approach provides a transparent, coherent and consistent approach for the
compilation and balancing of SUTs. For example, balancing adjustments to one part of the SUTs
can be assessed in terms of their impact on other areas of the SUTs, and also in terms of time
series.

2. PSUTs

2.32. The SUTs described in the previous sections are part of the 2008 SNA framework. As such,
they reflect the production boundary of the SNA and they are compiled in monetary units. The
tables can, however, be extended to include the environment as providers of natural inputs into the
economy and as absorbers of residuals from the economy. The extension of these tables and, more
generally, of the accounting framework of the SNA to include environmental considerations is
carried out in the 2012 SEEA. The SEEA enables the analyses of the interaction between the
environment and the economy, such as an assessment of the use of natural resources, the generation
of waste by the economy and waste flows into the environment.

2.33. The SEEA central framework comprises a sequence of accounts – namely, the SUTs in
monetary and physical units, the asset accounts in physical and monetary units, and environmental
activity accounts and related flows. The present Handbook covers the SUTs of the SEEA and, in
particular, since the monetary SUTs of the SEEA are the same as the SNA, the Handbook focuses
on the PSUTs of the SEEA. For additional information on the complete set of accounts of the

32
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

SEEA, the reader is directed to the publication United Nations, European Commission, FAO, IMF,
OECD and World Bank (2014).

2.34. PSUTs are used to assess how an economy supplies and uses energy, water, materials, and
also their changes in production and consumption patterns over time, and therefore, in combination
with data from monetary SUTs, changes in productivity and intensity in the use of natural inputs
and the release of residuals can be examined. The structure of PSUTs is based on the monetary
SUTs with extensions to incorporate a column for the environment and rows for natural inputs and
residuals.

2.35. Table 2.4 and Table 2.5 provide the simplified structure of the physical supply table and
use table, respectively. In order to address specific environmental domains (for example,
accounting for water, energy, timber and other resources), these tables are compiled for a
disaggregation of products and industries which is relevant for the particular environmental
domain of interest. In the case of energy, for example, the products of interest that can be explicitly
shown in the table include coal, peat and peat products, natural gas and others. The industries of
interest include the main suppliers of energy products (for example, electricity generation,
manufacture of gas and other products) and the main users of energy products (for example,
manufacturing, transportation and others). These tables are compiled in monetary units – within
the SNA context – and in physical units as shown in Table 2.4 and Table 2.5.

2.36. In the PSUTs, the SUTs of the SNA are augmented to include a block of rows for “natural
inputs” and a block of rows for “residuals”. The block for natural inputs is used to describe the
flows from the environment to the economy; in other words, this block describes the extraction of
natural inputs (for example, water, energy resources and others) from their location in the
environment as a part of economic production processes or that are directly used in production.
Natural inputs may be, first, natural resource inputs, such as mineral and energy resources or timber
resources; second, inputs from renewable energy sources, such as solar energy captured by
economic units; or, third, other natural inputs such as inputs from soil (for example, soil nutrients)
and inputs from air (for example, oxygen absorbed in combustion processes) (2012 SEEA, para.
2.89). When an industry, for example, extracts water as part of the economic production process,
this is recorded in the block of natural inputs in the use table in Table 2.5. It is assumed that the
environment provides (that is, supplies) all the natural inputs that are used into the economic
production process.

2.37. The blocks for “residuals” represent the flows of solid, liquid and gaseous materials, and
energy, that are discarded, discharged or emitted to the environment (for example, emission to air)
by establishments and households through processes of production, consumption or accumulation
but that may also flow within the economy, as is the case when, for example, solid waste is
collected as part of a waste collection scheme (2012 SEEA, para. 2.92).

33
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.38. The block for residuals in the supply table (Table 2.4) represents the flows of waste from
the economy to the environment and thus it includes the generation and disposal of waste during
economic production activities (generation of waste by industries) and generated during final
consumption (mainly by households). While the block of residuals in the use table (Table 2.5)
shows, for example, the collection and treatment of waste and other residuals by economic
activities, the accumulation of waste in controlled landfills and the residuals flows direct to the
environment.

Table 2.4 Schematic view of the physical supply table

Indus tri es Industries Gross capital
Final
Agriculture, Mining and Imports formation/ Environment Total
… Services consumtpion
forestry, etc. quarrying Accumulation
Mineral and energy resource Flows from Total supply
Natural
inputs

Water the by natural

environment inputs
…
Agriculture, forestry, etc.
Products

Ores and minerals; etc. Imports by Total supply

Output by product by industry
… product by product

Services
Residuals Residuals from
Residuals

Solid waste
generated scrapping and Total supply
Wastewater Residuals generated by industry
by final demolition of by residuals
… consumtpion produced assets

Empty cells by definition

C ells may contain relevant flows

Table 2.5 Schematic view of the physical use table

Indus tri es Industries Gross capital
Final
Agriculture, Mining and Exports formation/Accu Environment Total
… Services consumtpion
forestry, etc. quarrying mulation
Mineral and energy resource Total use by
Natural
inputs

Water Extraction of Natural inputs natural

inputs
…
Agriculture, forestry, etc.
Products

Ores and minerals; etc. Intermediate consumption by product and by Total use by
Final uses by product and by category
… industry product

Services
Accumulation of Residual
Residuals

Solid waste
C ollection and treatment of waste and other waste in flows direct Total use by
Wastewater
residuals controlled to the residuals
… landfilled enviornment

Empty cells by definition

C ells may contain relevant flows

2.39. The supply and use identity applies to both physical and monetary flows. For each product
measured in physical terms (for example, cubic metres of timber), the quantity of output and
imports (total supply of products) must equal the quantity of intermediate consumption, household
final consumption, gross capital formation and exports (total use of products). The equality
between supply and use also applies to the total supply and use of natural inputs and the total
supply and use of residuals. In addition to the supply and use identity, the PSUTs incorporate the

34
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

input-output identity, implying that the total flows into the economy either are returned to the
environment or accumulate in the economy.

C. Overview of IOTs

2.40. For many analytical purposes, a transformation from a pair of SUTs into a single IOT where
total input (row totals) and total output (column totals) are equal brings considerable advantages.
IOTs have algebraic properties that make them particularly suitable for analyses that enable
estimates of the effects of changing relative prices, labour and capital requirements in the face of
changing output levels, the consequences of changing patterns of demand and so on. They may
also be used as the basis for an expanded version that may be used to estimate the demands made
by the economy on the environment (2008 SNA, para. 28.35).

2.41. An IOT is essentially derived from the use table, where either the columns representing
industries are replaced by products or where the rows representing the products are replaced by
industries through a transformation process which involves a range of assumptions. The resulting
intermediate consumption matrix is then square, showing products in both rows and columns or
industries in both. In both cases, the row totals for the complete matrix match the column totals
for the complete matrix, product-by-product matrix or industry-by-industry matrix as the case may
be (2008 SNA, para. 28.32). Of course, the classifications in the IOTs coincide with those in the
SUTs, as the former is a transformation of the latter.

2.42. It is recommended that the IOTs be derived from SUTs. The IOTs derived from the SUTs
further describe the interrelationships between industries and products, along with the sale and
purchase relationships between producers and consumers within an economy. They can be
produced to illustrate flows between the sales and purchases (final and intermediate) of industry
outputs (referred to as industry-by-industry tables) or to illustrate the sales and purchases (final
and intermediate) of product outputs (referred to as product-by-product tables).

2.43. The derivation of IOTs from the system of SUTs may also reveal inconsistencies and
weaknesses in the SUTs. This is made possible by the powerful quality-related feedback loop from
the IOTs to the SUTs and vice versa.

2.44. Table 2.6 provides a simplified IOT where the columns of the original use table referring
to industry-based structures are transformed into product-based structures. The relations between
output and input are now relations between products and primary inputs necessary to produce
outputs in similar production units. Primary inputs are inputs that are not outputs of other
industries. They include the imports of goods and services and the components of GVA, such as
compensation of employees, and others. They are necessary to the production process but are not
produced anywhere in the domestic economy.

35
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 2.6 Simplified IOT (product by product)

Products Products Final uses

Ores and Total
Agriculture, Final Gross capital
Products minerals; … Services Exports
forestry, etc. consumption formation
etc.
Agriculture, forestry, etc.
Total use
Ores and minerals; etc. Intermediate consumption by product Final uses by product and by category by
… product
Services
Imports Intermediate consumption of imported products Final use of imported products
Value
Value added Value added by component
added
Total Total supply Total final uses by category

Empty cells by definition

2.45. For the transformation of SUTs into IOTs, various assumptions have to be made or
adjustments are required based on industry-by-industry or product-by-product assumptions:

• Product-by-product IOTs: these may be compiled using either the product technology
assumption (whereby each product is produced in its own specific way, irrespective of the
industry where it is produced) or the industry technology assumption (whereby each industry
has its own specific means of production, irrespective of its product mix).

• Industry-by-industry IOTs: these may be compiled using either the fixed industry sales
structure assumption (whereby each industry has its own specific sales structure, irrespective
of its product mix) or the fixed product sales structure assumption (whereby each product
has its own specific sales structure, irrespective of the industry where it is produced).

A mixture of both assumptions can also be applied by implementing a hybrid technology

assumption. The correct use and understanding of the terminology, transformation process and
assumptions applied are covered in more detail in chapter 12 of this Handbook.

2.46. The selection of the appropriate type of IOT – product-by-product or industry-by-industry

– depends on a number of statistical and practical considerations. For example, industry-by-
industry IOTs are closer to statistical sources and actual market transactions. Product-by-product
IOTs are believed to be more similar in terms of cost structure and production activities. This view
has been somewhat weakened, however, by changes introduced in the 2008 SNA, such as the strict
implementation of changes in economic ownership.

2.47. In the IOTs, two identities of the SUTs system are reduced to a single type of identity. It is
typical for IOTs that, for each product or industry, the input equals output and total input equal
total output.

2.48. The figures of total output and total input by product are the same as total supply and total
use by product of the SUTs – this holds for product-by-product IOTs. The industry-based
structures are transformed into product-based structures. In this transformation, the final use data

36
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

are left unchanged. The transformation only rearranges the data on the basis of the production
matrix of the intermediate use table by applying certain analytical assumptions to the relations
between primary and secondary outputs.

2.49. In general, and for analytical purposes, it is recommended to separate the use table into the
use table for domestic output and the imports use table. More detail on the compilation of the
domestic use table and the imports use table may be found in chapter 8.

2.50. Box 2.3 and Box 2.4 show a simplified numerical example of a sequence of tables – based
on the SUTs shown in Box 2.1 and Box 2.2 – necessary for compiling product-by-product IOTs
and industry-by-industry IOTs, respectively.

Box 2.3 SUTs and product-by-product IOTs

Supply table
Industries
Domestic Total
Manufacturing Imports
Agriculture Services supply supply
and construction

Agriculture 270 30 50 350 20 370

Products

Manufacturing and
10 430 100 540 50 590
construction
Services 20 40 550 610 30 640
Total 300 500 700 1500 100 1600

Use table Use table for domestic output

Industries Final use Industries Final use
Final Manufacturing Final Gross
Manufacturing Gross capital Total use Total
Agriculture Services consumption Exports Agriculture and Services consumption capital Exports
and construction formation
expenditure construction expenditure formation
Agriculture 34 59 93 131 21 32 370 Agriculture 30 50 90 130 20 30 350
Products

Products

Manufacturing and Manufacturing and

106 119 61 139 103 62 590 90 100 54 136 100 60 540
construction construction
Services 70 112 75 291 61 31 640 Services 60 100 70 290 60 30 610
Taxes less subsidies
4 5 12 52 6 1 80 Imports 30 40 15 5 5 5 100
on products
Taxes less subsidies on
GVA 86 205 459 750 4 5 12 52 6 1 80
products
Total 300 500 700 613 191 126 GVA 86 205 459 750
Total 300 500 700 613 191 126

Empty cells by definition Imports use table

Industries Final use
Manufacturing Final Gross
Total
Agriculture and Services consumption capital Exports
construction expenditure formation
Agriculture 4 9 3 1 1 2 20
Products

Manufacturing and
16 19 7 3 3 2 50
construction
Services 10 12 5 1 1 1 30
Total 30 40 15 5 5 5 100

Input-output table for domestic output (product by product)

Products Final use
Manufacturing Final Gross Output
Agriculture and Services consumption capital Exports
construction expenditure formation
Agriculture 34.08 52.23 83.69 130.00 20.00 30.00 350.00
Products

Manufacturing and
113.17 111.84 18.99 136.00 100.00 60.00 540.00
construction
Services 73.23 114.19 42.58 290.00 60.00 30.00 610.00
Imports 37.73 46.07 1.20 5.00 5.00 5.00 100.00
Taxes less subsidies on
4.58 4.83 11.59 52.00 6.00 1.00 80.00
products
GVA 87.21 210.84 451.95 750.00
Total 350.00 540.00 610.00 613.00 191.00 126.00

Input table for imports (product by product)

Products Final use
Manufacturing Final Gross Total
Agriculture and Services consumption capital Exports
construction expenditure formation
Agriculture 4.91 10.75 0.34 1.00 1.00 2.00 20.00
Products

Manufacturing and
construction 20.22 21.68 0.11 3.00 3.00 2.00 50.00
Services 12.60 13.65 0.74 1.00 1.00 1.00 30.00
Total 37.73 46.07 1.20 5.00 5.00 5.00 100.00

In product-by-product IOTs, all inputs are allocated to similar production units. They are derived from the SUTs
system on the basis of analytical assumptions (see chapter 12 for further detail on the derivation of IOTs).
Product-by-product IOTs are further away from statistical sources than industry-by-industry IOTs.

37
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.4 SUTs and industry-by-industry IOTs

Supply table
Industries
Manufacturing Domestic Total
Imports
Agriculture and Services supply supply
construction
Agriculture 270 30 50 350 20 370
Products

Manufacturing and
10 430 100 540 50 590
construction
Services 20 40 550 610 30 640
Total 300 500 700 1500 100 1600

Use table Use table for domestic output

Industries Final use Industries Final use
Final
Manufacturing Gross Manufacturing Final
consumptio Total use Gross capital Total
Agriculture and Services capital Exports Agriculture and Services consumption Exports
n formation
construction formation construction expenditure
expenditure
Agriculture 34 59 93 131 21 32 370 Agriculture 30 50 90 130 20 30 350
Products

Products
Manufacturing and Manufacturing and
106 119 61 139 103 62 590 90 100 54 136 100 60 540
construction construction
Services 70 112 75 291 61 31 640 Services 60 100 70 290 60 30 610
Taxes less subsidies
4 5 12 52 6 1 80 Imports 30 40 15 5 5 5 100
on products
Taxes less subsidies on
GVA 86 205 459 750 4 5 12 52 6 1 80
products
Total 300 500 700 613 191 126 GVA 86 205 459 750
Total 300 500 700 613 191 126

= Zero by definition Imports use table

Industries Final use
Manufacturing Final
Gross capital Total
Agriculture and Services consumption Exports
formation
construction expenditure
Agriculture 4 9 3 1 1 2 20
Products
Manufacturing and
16 19 7 3 3 2 50
construction
Services 10 12 5 1 1 1 30
Total 30 40 15 5 5 5 100

Input-output table for domestic output (industry by industry)

Industries Final use
Manufacturing Final
Gross capital Total
Agriculture and Services consumption Exports
formation
construction expenditure
Agriculture 26.78 43.70 72.72 112.31 19.25 25.24 300.00
Industries

Manufacturing and
78.17 90.47 55.30 138.46 85.28 52.32 500.00
construction
Services 75.05 115.83 85.97 305.23 75.47 42.45 700.00
Imports 30.00 40.00 15.00 5.00 5.00 5.00 100.00
Taxes less subsidies on
4.00 5.00 12.00 52.00 6.00 1.00 80.00
products
GVA 86.00 205.00 459.00 750.00
Total 300.00 500.00 700.00 613.00 191.00 126.00

Input table for imports (industry by industry)

Industries Final use
Manufacturing Final
Gross capital Total
Agriculture and Services consumption Exports
formation
construction expenditure
Agriculture 3.71 7.69 2.61 0.86 0.86 1.61 17.34
Industries

Manufacturing and
construction 13.74 16.69 6.16 2.54 2.54 1.83 43.50
Services 12.55 15.62 6.23 1.60 1.60 1.56 39.17
Total 30.00 40.00 15.00 5.00 5.00 5.00 100.00

In industry-by-industry IOTs, inputs are allocated to industries. They are derived from the SUTs system on the
basis of pragmatic assumptions. The intermediate input of industries consists of output of industries rather than
products (of industry adjusted products) (see chapter 12 for details on the derivation of IOTs).
Industry-by-industry IOTs are closer to statistical sources and actual observations than product-by-product
IOTs.

D. Structure of SUTs and IOTs: basic elements

2.51. Defining the structure of SUTs and IOTs is a principal first step and depends on a number
of basic elements which form the backbone of these tabulations. These elements include:

• Principles of the accounting system underlying the SNA applied to SUTs and IOTs
• Classification of economic activities and its level of detail
• Classification of products and its level of detail

38
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Choice of the statistical units

• Valuation
2.52. Considerations on these elements reflect the specific context in which the tables are
compiled, which include the analytical objectives, the data availability, the economic structure of
the country, and others. Each of these elements is described below.

1. Principles of the accounting system underlying the SNA applied to SUTs and IOTs

2.53. The accounting system underlying the SNA is derived from broad bookkeeping principles
and is applied to the structure and links in the SUTs and IOTs. There are three bookkeeping
principles underlying the SNA accounting system:

• Vertical double-entry bookkeeping, also known as double-entry bookkeeping

• Horizontal double-entry bookkeeping
• Quadruple-entry bookkeeping
2.54. The main characteristic of vertical double-entry bookkeeping is that each transaction leads
to at least two entries, traditionally referred to as a credit entry and a debit entry. This principle
ensures that the total of all credit entries and all debit entries for all transactions are equal, thus
permitting a check on consistency of accounts for a single unit. Each transaction requires two
entries.

2.55. The concept of horizontal double-entry bookkeeping is useful for compiling accounts that
reflect the mutual economic relationships between different institutional units in a consistent
manner. It implies that, if unit A provides something to unit B, the accounts of both A and B show
the transaction for the same amount: as a payment in A’s account and as a receipt in B’s account.
Horizontal double-entry bookkeeping ensures the consistency of recording for each transaction
category by counterparties. For example, dividends payable throughout the economy should be
equal to dividends receivable throughout the economy once transactions with the rest of the world
are taken into account.

2.56. The simultaneous application of vertical and horizontal double-entry bookkeeping results
in quadruple-entry bookkeeping – which forms the accounting system underlying the SNA. It deals
in a coherent manner with multiple transactors or groups of transactors, each of which satisfies
vertical double-entry bookkeeping requirements. A single transaction between two counterparties
thus gives rise to four entries. In contrast to business bookkeeping, national accounts deal with
interactions among a multitude of units in parallel, and thus require special care from a consistency
point of view.

2.57. An account records and displays all the flows and stocks for a given aspect of economic
life. In each account, the sum of resources is equal to the sum of uses with a balancing item to

39
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

ensure this equality. Normally the balancing item will be an economic measure which is itself of
interest.

2.58. The accounts can be built up for different areas of the economy by employing a system of
economic accounts which highlight, for example, production, income and financial transactions.
In many cases, these accounts can be elaborated and set out for different institutional units and
groups of units (or institutional sectors). Usually a balancing item has to be introduced between
the total resources and total uses of these units or sectors and, when summed across the whole
economy, these balancing items constitute significant aggregates.

2.59. The accounting structure is uniform throughout the system and applies to all units in the
economy, whether they are institutional units, subsectors, sectors or the whole economy, although
some accounts (or transactions) may not be relevant for some institutional sectors.

2.60. The national accounting system uses two types of units and two corresponding ways of
subdividing the economy, which are quite different and serve separate analytical purposes:

• The first purpose, namely that of describing production, income, expenditure and financial
flows, and balance sheets, is served by grouping institutional units into institutional sectors
on the basis of their principal functions, behaviour and objectives. The national accounting
system enables a complete set of flow accounts and balance sheets to be compiled for each
sector, and subsector, and also for the total economy.
• The second purpose, namely that of describing processes of production and for input-output
analysis, is served by the system grouping establishments into industries on the basis of their
type of activity. An activity is characterized by an input of products, a production process
and an output of products.
2.61. Figure 2.3 shows in matrix form an overview of the structure of the SNA. The degree of
subdivisions of the columns and rows using the relevant classifications determines the degree of
detail of the accounts. The shaded rows and columns for goods and services and production by
industry indicate those parts of the system relevant for the compilation of SUTs and IOTs, and
clearly indicate that SUTs are at the core of the national accounts system.

2.62. The three approaches to measuring GDP (production, income and expenditure) are shown
in Box 2.5 and can be derived from the data in Figure 2.3 generating a single estimate of GDP.

2.63. All the aggregate components and detailed components are included in the SUTs and IOTs-
related part of the system.

40
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 2.3 System of national accounts in matrix form

millions of euros
Expenditures
Assets Opening Goods and Production by Income and Accummu- Rest of the Closing Total excl.
balance services industry consumption lation world balance balance
Receipts

Liabilities 1 2 3 4 5 6 7 8

Assets of
Financial assets
domestic sectors
Opening balance 1 (real and
of the rest of the
w orld
financial)

303 492 226 258 74 612 165 648 770 009

Goods and
2 Intermediate Final Gross capital Exports of goods
services
consumption consumption formation and services

578 360 578 360

Production by
3 Output at basic
industry
prices

33 778 274 868 38 023 346 670

Income and Primary incomes

Taxes less
4 Value added at and current
consumption subsidies on
basic prices transfers from
products
rest of the w orld

79 669 - 5 057 74 612

Deficit on the
Accummulation 5 Liabilities of Liabilities of
Gross savings balance of
domestic sectors domestic sectors
payments

157 871 40 743 198 614

Primary incomes
Financial Financial liabilities
Rest of the world 6 Imports of goods and current
liabilities of the of the rest of the
and services transfers to the
rest of the w orld w orld
rest of the w orld

Assets of
Financial assets
domestic sectors
Closing balance 7 (real and
of the rest of the
w orld
financial )

Total excl.
8 770 009 578 360 346 670 74 612 198 614
balance

Table based on 2011 figures from Austria

Box 2.5 Three approaches to measuring GDP

Production approach Income approach Expenditure approach
millions of euros millions of euros millions of euros
Variable Value Variable Value Variable Value

Output at basic prices 578 360 Compensation of employees 144 343 Final consumptipon 226 258
+ Other taxes less subsidies on production 4 858

- Intermediate consumption - 303 492 + Consumption of fixed capital 53 469 + Gross capital formation 74 612
+ Net operating surplus 72 198
= Gross value added at basic prices 274 868 = Gross value added at basic prices 274 868 + Exports of goods and services 165 648
+ Taxes less subsidies on products 33 778 + Taxes less subsidies on products 33 778 - Imports of goods and services - 157 871
= Gross domestic product 308 647 = Gross domestic product 308 647 = Gross domestic product 308 647

Table based on 2011 figures from Austria

41
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2. Classification of economic activities

2.64. The International Standard Industrial Classification of All Economic Activities (ISIC) is
the international reference classification of economic activities (also referred to as “industries”).
The fourth revision, ISIC Rev. 4, was issued by the United Nations in 2008 (United Nations, 2008).
Its main purpose is to provide a set of activity categories that can be used for collecting and
presenting internationally comparable statistics by economic activity.

2.65. In general, the scope of ISIC covers productive activity, that is, all economic activities
within the production boundary as described in the SNA (with one exception for activities in Class
9820– “Undifferentiated service-producing activities of private households for own use”). The
classification is used to classify statistical units such as establishments or enterprises, according to
the economic activity in which they mainly engage. All categories at each level of the classification
are mutually exclusive. ISIC Rev. 4 is the reference classification of production activities of the
2008 SNA.

2.66. The structure of ISIC consists of 21 sections, 88 divisions, 238 groups and 419 classes. The
principles and criteria used to define and delineate the categories are based on the inputs of goods,
services and factors of production, the process and technology of production, the characteristics of
outputs, and the use to which the outputs are put. At the class level of the classification, preference
has been given to the process and technology of production in defining individual ISIC classes, in
particular in the classes related to services. The list of products that defines a class is called the
principal products of that class. At the division and group levels, characteristics of outputs and the
use to which outputs are put become more important for the creation of analytically useful
aggregation categories.

2.67. At national and regional levels, there may be need for recourse to a level of detail that
reflects specific national and regional circumstances. It is important, however, that these
classifications are compatible with ISIC Rev. 4 at an aggregated level of detail. At its thirty-seventh
session, the United Nations Statistical Commission recommended that countries adapt their
national classifications in a way that allows them to report data at least at the two-digit level of
ISIC Rev. 4 without loss of information. 5 Examples of regional classifications include the
industrial classification used in the European Union, the second revision of the Statistical
Classification of Economic Activities in the European Community (NACE Rev. 2), which is
identical with ISIC Rev. 4 up to the two-digit level (divisions) of the classification. At lower levels,
NACE has provided more detail suitable for the European users of the classification. The
additional detail can always be aggregated to ISIC categories at the three-digit and four-digit
levels, within the same structure. The North American Industry Classification System (NAICS),
although it has a substantially different structure from ISIC, has been designed in a way that

5
See Official Records of the Economic and Social Council, 2002, Supplement No.4 (E/2006/24), chapter I, para. 3,
item 37/105 (b).

42
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

statistical data collected according to NAICS can be re-aggregated into the two-digit divisions of
ISIC Rev. 4. The Australian and New Zealand Standard Industrial Classification (ANZSIC) was
revised in 2006. Its structure broadly follows the ISIC structure, so that categories at the division
and more detailed levels can be aggregated into the two-digit categories of ISIC.

2.68. An economic unit may engage in a variety of production activities. The classification of
the economic unit is made in accordance with the importance of the production activities. In this
regard, the activities of an economic unit are subdivided into principal activity, secondary activity
and ancillary activities. The principal activity of an economic entity is the activity that contributes
most to the value added of the entity, as determined by what is known as the “top-down method”.
This method follows a hierarchical order, starting with the identification of the relevant category
at the highest level (section) and progressing down through the levels of the classification to the
lowest level (classes). The effect of this top-down method is such that the principal activity need
not account for 50 per cent or more of the total value added of an entity or even that its generated
value added exceed that of all other activities carried out by the unit, although, in the majority of
cases, it will do so (United Nations, 2008, para. 57).

2.69. In practice, it is often impossible to obtain information on the GVA of the different
activities performed and the activity classification has to be determined by using substitute criteria,
such as employment and turnover.

2.70. Products resulting from a principal activity are either principal products or by-products.
By-products are products that are necessarily produced together with principal products, for
example, hides produced when producing meat by slaughtering animals. Since normal patterns of
horizontal integration have been taken into account when defining the ISIC classification, such
commonly integrated activities are usually included in the same ISIC class, even though the
outputs of the activities have quite different characteristics. Thus ISIC class 1010 “processing and
preserving of meat” also includes hides, skins, wool and feathers originating from slaughtered
animals (United Nations, 2008, paras. 57 and 120).

2.71. A secondary activity is a separate activity the products of which are ultimately intended for
third parties and that is not the principal activity of the entity in question. The outputs of secondary
activities are called secondary products, including any by-products associated with these outputs.
Most economic entities produce at least some secondary products.

2.72. Traditionally, the existence of by-products has been seen as creating problems in input-
output analysis as they would disturb supply-and-use relationships. Additional demand for the
principal product would therefore also result in more output of the by-product, without there
automatically being any additional demand for that increased output. In the case of more complex
production processes than meat and hides, for example in the chemical and electronic industries,
it will, however, be very difficult or even, in the absence of special technical insight, impossible
for the compilers of SUTs to identify by-products separately.

43
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.73. A distinction is made between principal and secondary activities on the one hand and
ancillary activities on the other. All economic units require some basic, routine services to support
their production activities. When they are provided in-house, they are called ancillary activities. In
general, an ancillary activity is a supporting activity undertaken within an enterprise in order to
create the conditions within which the principal and secondary activities can be carried out (2008
SNA, para. 5.36). Ancillary activities typically produce services that are commonly found as inputs
into almost any kind of economic activity. These outputs are always services and intended for
intermediate consumption within the same entity. They include, for example, the maintenance of
records, files or accounts in written form or on computers; the purchase of materials and
equipment; the provision of electronic and traditional written communication facilities; the hiring,
training, managing and paying of employees; the storing of materials or equipment; warehousing;
the provision of security and surveillance, and others.

2.74. Some of these activities are found in every economic entity. The output of an ancillary
activity is not explicitly recognized and recorded separately in the SNA. It follows that the use of
this output is also not recorded. All the inputs consumed by an ancillary activity – materials, labour,
consumption of fixed capital, and so forth – are treated as inputs into the principal or secondary
activity that it supports.

2.75. The following activities are not to be considered ancillary: producing goods or services as
part of gross fixed capital formation and research and development activities, which are considered
to be part of gross fixed capital formation in the 2008 SNA. These items will therefore appear as
either principal or secondary output. Goods that become embodied in the output of the principal
or secondary activities are not outputs of ancillary activities.

2.76. More details on principal products, secondary products and ancillary products specific to
the construction of the supply table may be found in chapter 5.

(a) Considerations for the compilation of SUTs and IOTs

2.77. In SUTs and IOTs, industries should be classified according to ISIC Rev. 4. The major
advantage of using established international industrial classifications is that comparability with
other types of economic statistics and the national accounts is not compromised. The choice is
therefore not which industrial classifications should be used in the SUTs and IOTs but rather the
level of detail.

2.78. At the working level, it is recommended to use the most detailed level of classification of
industry, taking into consideration user needs, the availability of data and the level of detail used
in the national accounts. Furthermore, certain compilation aspects also influence the choice of
working level, such as the distinction between industries which are allowed to deduct VAT and
those that are not, the distinction between market and non-market producers, and the explicit
identification of certain industry subdivisions that are relevant for the compilation of the trade and
transport margin matrices. In addition, the link between SUTs and the institutional sector accounts

44
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

should be reflected. These considerations are further elaborated in chapters 4–7. In general, the
level of detail in the published and disseminated SUTs differs from that at the working level: SUTs
tend to be published at a more aggregated level of detail which takes into account users’ needs and
confidentiality.

3. Classification of products

2.79. The international reference classification of products is the Central Product Classification
(CPC). The latest revision, CPC Version 2.1, was issued by the United Nations in 2015 (United
Nations, 2015). The primary purpose of CPC is to classify all goods and services that are the result
of production in the economy. CPC presents categories for all products that can be the object of
domestic or international transactions or that can be entered into stocks. It includes products that
are the output of economic activity, including transportable goods, non-transportable goods and
services. CPC in general follows the definition of products within the SNA.

2.80. The importance of the industrial origin of goods and services was underscored by the
attempt to group into one CPC subclass mainly the products that are the output of a single ISIC
class. Through their linkage to the criterion of industrial origin, the input structure, technology and
organization of production characteristics of products are also reflected in the structure of CPC.
The criterion of industrial origin of products is one of the classification principles applied by ISIC.

2.81. CPC was developed primarily to enhance harmonization among various fields of economic
and related statistics and to strengthen the role of national accounts as an instrument for the
coordination of economic statistics. It provides a basis for transforming basic statistics from their
original classifications into a standard classification for analytical use. As a general purpose
classification, CPC provides less detail than other specific classification systems in areas or
applications for which such systems are available, for example the Harmonized Commodity
Description and Coding System (Harmonized System). The Harmonized System 6 codes provide
building blocks for the part dealing with transportable goods and take into account the basic
categories of economic supply and use of products as specified in the SNA such as intermediate
consumption, final consumption, capital formation, and imports and exports.

2.82. CPC is a system of categories that are both exhaustive and mutually exclusive. This means
that if a product does not fit into one CPC category, it must automatically fit into another. In CPC
Version 2.1, in total there are 10 sections, 71 divisions, 329 groups, 1,299 classes and 2,887
subclasses. Each subclass in sections 0–4 of CPC is defined as the equivalent of one heading or
subheading or the aggregation of several headings or subheadings of the Harmonized System,
since the Harmonized System is a detailed classification of transportable goods that is widely
accepted for use in international trade statistics by virtually all countries. Other classifications of

6
The Harmonized Commodity Description and Coding System is the classification used for international trade
statistics.

45
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

products may be used at country and regional level, however, these classifications are in general
broadly consistent with CPC Version 2.1. The Classification of Products by Activity (CPA) is
based on NACE – and therefore follows a different aggregation structure than CPC – and detailed
categories that are mostly aligned with CPC. Exceptions exist for areas where CPC deviates from
the Harmonized System, since CPA maintains a closer link with the Combined Nomenclature,
which is the European version of that classification.

2.83. CPC and ISIC are both general purpose classifications, with ISIC representing the activity
side. Each subclass of CPC consists of goods or services that are predominantly produced in a
specific class of ISIC. The relationship between industries and their products is complex, however,
and changes over time, and it should be noted that there has been no intention of establishing a
one-to-one correspondence between CPC and ISIC. Such an effort is considered neither practical
nor desirable as it might lead to an inadequate description of CPC categories, in particular at the
higher levels.

2.84. The classification of a product in the service part of CPC does not automatically imply that
the product cannot be a principal output of a goods producing industry. Thus the two CPC
divisions: (87) Maintenance, repair and installation (except construction) services; and (88)
Manufacturing services on physical inputs owned by others, both appear in the business and
production services section of CPC but the units carrying out these activities on a fee or contract
basis are classified in the same ISIC category as units producing the same goods or services for
their own account. The correspondence table between CPC Version 2.1 and ISIC Rev. 4 (see the
United Nations Statistics Division classification website at:
http://unstats.un.org/unsd/class/default.asp) shows that 125 subclasses of CPC division 88 are
defined to correspond to 125 manufacturing industry classes of ISIC. This implies that these
manufacturing services are the principal output (and not as might have been expected, the
secondary output) of the corresponding manufacturing activities. In other words, there are no
service industries producing these services. This example shows the need to ensure that services
of these kinds are correctly entered into the domestic output matrix, requiring a considerable
number of products.

2.85. Box 2.6 shows other classification of products, such as the Harmonized System, the
Standard International Trade Classification (SITC), the Classification by Broad Economic
Categories (BEC) and the Extended Balance of Payments Services Classification (EBOPS) and
how they relate to CPC. The basis for grouping products in the SUTs (and IOTs) is thus most
commonly an aggregation of CPC sections, divisions or groups (2008 SNA, para. 14.22).

Box 2.6 Other classifications of products

The Harmonized Commodity Description and Coding System, generally referred to as the Harmonized
System or simply HS, is a multipurpose international product nomenclature developed by the World Customs
Organization (WCO). The system is used by more than 200 countries and economies as a basis for their
collection of external trade statistics for goods. It is also extensively used by governments, international

46
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

organizations and the private sector for many other purposes, such as internal taxes, trade policies, monitoring
of controlled goods, rules of origin, freight tariffs, transport statistics, price monitoring, quota controls,
compilation of national accounts, and economic research and analysis.
As a result of the intensive world-wide use and degree of detail that the Harmonized System provides, it is a
fundamental classification system and provides a key link for the definitions of all other classifications of
goods (including the goods part of CPC), as well as for the definition of the classes of ISIC. The latest version
now available is HS 2012.
The Harmonized System explanatory notes are a part of a commodity database giving the Harmonized System
classification of more than 200,000 products actually traded internationally. This high number of background
products also makes it evident that, at the level of external trade statistics (usually 5,000–10,000 products, as
most countries apply further subdivisions of the 5,000 Harmonized System codes), there will be no
homogeneous products, and of course, even less so at the much higher level of aggregation applied in a SUTs
system.
SITC and BEC are both classifications of goods defined in terms of the Harmonized System, and also
primarily used in relation to external trade data. SITC distinguishes around 3,000 products at its most detailed
level. It is primarily used as an alternative to the Harmonized System publication level of external trade
statistics, and there will usually be no advantage in applying it in SUTs, rather than using the Harmonized
System directly. Following the breakdown of products according to BEC (food, materials, fuel, capital goods,
transport equipment, consumer goods), these groupings may be used as a reference when deciding on uses of
some products but this breakdown is not applicable as the main product classification in the SUTs system.
Furthermore, BEC is not an international standard classification in the same sense as the Harmonized System
or SITC.
The relationship between CPC and SITC is similar to that between CPC and the Harmonized System, since
SITC also uses the subheadings of the Harmonized System as building blocks to create product groupings that
are more suitable for the economic analysis of trade. BEC is related to CPC through its close correlation with
SITC and is designed to serve as a means for converting external trade data compiled by use of the SITC into
end-use categories that are meaningful within the SNA framework. It is generally possible to rearrange entire
CPC subclasses into BEC categories through the correspondences between CPC and SITC, and between SITC
and BEC.
EBOPS 2010 (United Nations, European Commission, IMF, OECD and WTO, 2011) is a classification of
trade in services that was developed to provide further breakdowns of the BPM 6 classification so as to meet
a number of user requirements, including the provision of information required under the General Agreement
on Trade in Services. It builds upon the BPM 6 classification of services. In BPM 6, 12 main service categories
are identified and broken down into a list of standard and supplementary components. EBOPS 2010 consists
of a further breakdown of these components into more detailed sub-items. EBOPS 2010 also includes several
supplementary items for the recording of useful additional information regarding services transactions in
various sectors such as, travel and tourism or insurance services. Like the BPM 6 services classification,
EBOPS 2010 is primarily a product-based classification. Items of these classifications may be described in
terms of CPC. Correspondences cannot, however, be established in the areas of travel, construction, and
government goods and services, n.i.e., which focus on the mode of consumption of goods and services or the
status of the transactor, rather than on the type of product consumed. A detailed correspondence between
EBOPS 2010 and CPC Version 2, may be found online at
http://unstats.un.org/unsd/tradeserv/TFSITS/msits2010/annexes.htm.

47
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(a) Considerations for the compilation of products in the SUTs and IOTs

2.86. Based on the description of the product classifications and their level of detail, it is obvious
that products in the SUTs, even when a high level of detail is applied, as is the case with 2,000
products, will nonetheless represent aggregated groups of products when compared with the detail
applied in basic statistics, and even more so when compared with the real-world variety of
products. Accordingly, analogies to the notion of homogeneous products, which are often assumed
in standard economic theory, will in general be inappropriate, as there can be no homogeneous
products or production processes at the SUTs or IOTs level of aggregation. Many economies
usually consist of hundreds of thousands of producing units, of which virtually no two are
completely identical, and there are millions of different products and even more production
processes. It is therefore important to realize that national accounts and SUTs record economic
transactions, and not physically identifiable products or related technical production processes,
which will in general be outside the sphere of official statistics.

2.87. Even very detailed basic statistics already represent highly aggregated data when compared
to the number of real-world products. As previously mentioned, the HS contains descriptions of
200,000 products. Statistics on products are collected at a maximum detail of, say, 10,000
products, and only in selected areas such as external trade statistics and output from manufacturing
industries. Furthermore, products that are identical in a physical sense may be different in an
economic sense when they are sold at different prices to different purchasers. This may, for
example, happen because of the way transportation costs are invoiced. The concept of basic prices
is defined specifically to include this possibility. Statistics on the breakdown of products for
intermediate consumption will often give less detail than production statistics and may sometimes
be collected from enterprise units rather than establishment units, and in most cases the statistical
coverage of purchases is irregular or limited to certain industries, for example, mainly
manufacturing industries, but even in this case the compilers of SUTs may be confronted with the
task of further aggregation.

2.88. To gain a better understanding of the level of aggregation, it is useful to consider the
product definitions required when selecting items and collecting prices for compiling price indices
such as consumer price indices and producer price indices. Each item must be defined more
precisely than by just referring to even the most detailed product classification. The same applies
when collecting prices for use in the International Comparison Programme. Official statistics have
in these cases to make selections from a product universe at a much more detailed level than 10,000
product groups, in order to compile a sound price index. This places the notion of “homogeneity”,
as often applied in connection with SUTs and IOTs, in perspective (see 2008 SNA, para. 14.144).

2.89. As a result, the term “homogenous” in the context of the SUTs system usually means
“mutually exclusive”. As outlined above, international activity and product classifications aim at
mutually exclusive classification criteria. Yet within any group of products fulfilling this criterion,
there may be considerable “non-homogeneity” depending on the level of aggregation. The

48
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

classification of products in this (mutually exclusive) sense is statistically possible at any level of
aggregation but a product in the SUTs will usually represent a basket of products, and the contents
of the basket will, furthermore, vary from one cell to another along the rows of both the supply
table and the use table. For the classification of producer units into industries the same “mutually
exclusive” conditions basically hold, although the situation is somewhat different as the
statistically observed input structures will usually represent a mixture of intermediate consumption
structures for many individual products, some of which will also be produced in other industries.
As a consequence, industries producing mutually exclusive products can only be derived on certain
assumptions which do not in general form part of the compilation of the SUTs. Redefinitions (see
chapter 5) may be seen as an exception.

4. Other classifications relevant for SUTs and IOTs

2.90. The SUTs system distinguishes a large number of products and industries. Final uses,
however, often distinguish only final consumption, gross capital formation and exports at a very
aggregate level. The functional classifications help to support the compilation of SUTs and allow
for a wide range of other analyses. It is mainly the disaggregated SUTs which allow us to identify
the different purposes of expenditure on a product basis.

2.91. The SNA uses special classifications to analyse consumption and other outlays according
to the purpose for which the expenditure is undertaken. Such functional classifications and
associated detail – the Classifications of Expenditure According to Purpose (United Nations,
2000a) – can be found in chapter 29 of 2008 SNA, on satellite accounts and other extensions.
These classifications include, in particular, the classification of the functions of government
(COFOG); the classification of individual consumption according to purpose (COICOP), the
classification of the purposes of non-profit institutions serving households (COPNI) and the
classification of the outlays of producers according to purpose (COPP). The main purpose of these
classifications is to provide more detailed statistics for a wide variety of analytical uses. In 2018,
the United Nations issued the first revision of COICOP, COICOP 2018 (United Nations, 2018), to
reflect users’ need for more detail and several other issues that required a revision of the
classification.

2.92. For the SUTs, it is recommended that the lower-level detail be produced in the form of
disaggregated matrices as subsystems feeding into the central compilation of SUTs in current
prices and in volume terms. As a result, the correspondence between categories of these functional
classifications and CPC makes it possible to bring the basic data into the use tables.

2.93. The correspondence table between categories of CPC and COICOP has been established
and available on the United Nations Statistics Division classification website at
http://unstats.un.org/unsd/class/default.asp. When making decisions on the details of the product
classification to be applied in SUTs, the possibility of establishing transformation tables to
COICOP at group levels or class levels and to make use of the reverse transformation, from the

49
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

products surveyed in the household budget surveys to the products of the SUTs, should be taken
into consideration. These transformation matrices are keys to the use of the consumer price index
(CPI) in the volume estimates, as sub-indices of the CPI will usually be based on the COICOP
classification. In addition, for the purposes of household budget surveys and purchasing power
parities (PPPs), COICOP is applied at a more detailed level, including as many as 300 or more
subclasses.

2.94. Table 2.7 shows the types of links and extensions. Some of the key areas are covered in
this section but more detail in terms of compilation is provided in chapter 6.

2.95. COPP provides detailed information on outlays of producers for current production,
infrastructure research and development, environmental protection, marketing and human resource
development. It should be noted that COPP is included here more for completeness of presentation
of the functional classifications. COPP is not widely used and does not fit well in the SUTs
framework, as its outlays include wages and other types of costs in addition to intermediate
consumption. In principle, COPP applies to all producers, whether market or non-market or for
own final use.

50
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 2.7 Links between the use table and functional classifications
Use table
INDUSTRIES FINAL USE
Final consumption expenditure Gross Total
Agricul- Other
.. Total General capital Exports
ture services Households NPISH
government formation
Agriculture
Products

:
Other services
Value added
Total

Final consumption expenditure by households (COICOP 2018)

COICOP
Alcoholic beverages, tobacco

protection and miscellaneous

Food and non-alcoholic

Clothing and footwear

Personal care, social

goods and services

and narcotics
beverages

Total
…

Agriculture
Products

:
Other services
Total

Final consumption expenditure by NPISH (COPNI)

COPNI
and culture
Recreation

Services
Housing

Health

n.e.c.

Total
…

Agriculture
Products

:
Other services
Total

Final consumption expenditure by government (COFOG)

COFOG
General public

Public order
and safety

protection
services

Defence

Social

Total
…

Agriculture
Products

:
Other services
Total

51
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5. Statistical units

2.96. In general, the same statistical unit is the basis for compiling the use table and for compiling
the supply table. Different choices of statistical units are available for the compiler and it is
important to have a clear understanding of the impact of the choice of different statistical units has
on the SUTs and on the IOTs.

2.97. Different types of statistical units may be defined (for example, enterprise group, local unit,
kind-of-activity unit, and others). For SUTs, however, the focus is on two specific statistical units:
enterprises and establishments (local kind-of-activity units).

2.98. An enterprise is defined as the view of an institutional unit as a producer of goods and
services (where institutional units are economic entities that have autonomy of decision making
and have clear links with the legal units). An establishment is an enterprise or part of an enterprise
that is situated in a single location and in which only a single productive activity is carried out or
in which the principal productive activity accounts for most of the value added.

2.99. The impact of globalization and the way in which multi-national businesses control and
operate their activities pose a number of challenges, including the basis of the statistical unit for
measurement of national activity versus global activity. Following the recommendations of the
2008 SNA, however, the establishment is the unit that is more suitable for the analysis of
production in which the technology of production plays an important role. The establishment is
therefore the recommended unit for the compilation of the production part of the national accounts
and therefore the compilation of SUTs. This means, as a rule, that multi-product enterprises must
be partitioned into smaller and more uniform units with regard to their kind of production, if
possible.

2.100. Trying to collect data on sub-establishment production processes as part of the input-output
compilation is an approach that has no natural limitation and that will, apart from the costs, almost
invariably become skewed by the specific knowledge and insight that the compilers happen to
possess and lead to non-transparent and uneven compilation processes.

2.101. In practice, the extent of partitioning enterprises into establishments varies across
countries, depending on whether the creation of establishments is based on a relatively modest
breakdown of institutional units or whether, alternatively, the starting point is a register of all local
producer units. The latter case follows the formal definitions set out in the 2008 SNA and would
lead to a purer activity classification than the former. Recommendations for partitioning vertical
and horizontal integrated enterprises are briefly outlined in Box 2.7.

52
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.7 SNA recommendations on partitioning of vertically and horizontally integrated

enterprises
A horizontally integrated enterprise is one in which several different kinds of activities that produce different
kinds of goods or services for sale on the market are carried out simultaneously, using the same factors of
production (2008 SNA, para. 5.21).
Horizontal integration occurs when an activity results in end products with different characteristics. This could
theoretically be interpreted as activities carried out simultaneously using the same factors of production. In this
case, it will not be possible to separate them statistically into different processes, assign them to different units or
generally provide separate data for these activities. Another example would be the production of electricity
through a waste-incineration process. The activity of waste disposal and the activity of electricity production
cannot be separated in this case.
Within the SNA, a separate establishment should be identified for each different kind of activity wherever
possible (2008 SNA, para. 5.22).
A vertically integrated enterprise is one in which different stages of production, which are usually carried out
by different enterprises, are carried out in succession by different parts of the same enterprise (2008 SNA, para.
5.23). Vertical integration of activities occurs wherever the different stages of production are carried out in
succession by the same unit and the output of one process serves as the input to the next process. Examples of
common vertical integration include tree-felling and subsequent on-site sawmilling, mining of metal ores and
manufacture of basic iron and steel, operation of a clay pit combined with a brickworks or production of synthetic
fibres in a textile mill.
While the procedure for the treatment of vertically integrated activities could be applied to any unit, it should be
noted that the SNA recommends that, when a vertically integrated enterprise spans two or more sections of ISIC,
at least one establishment must be distinguished within each section. With such a treatment, activities of units
engaged in vertically integrated activities will not cross section boundaries of ISIC (2008 SNA, para. 5.26). If
this approach has not already been followed in basic statistics, the compilers of SUTs will exceptionally have to
deal with individual producer units.

6. Valuation in the SUTs

2.102. More than one set of prices may be used to value outputs and inputs depending on how
taxes and subsidies on products, and also transport charges and trade margins, are recorded. The
2008 SNA distinguishes three main valuation concepts of the flows of goods and services: basic
prices, producers’ prices and purchasers’ prices.

2.103. The valuation of the data for the use table (for example, intermediate and final
consumption) is different from the valuation of the data for the production side of the supply table.
In fact, the valuation of use table is based on the actual price paid by the users for the goods and
services (i.e., purchasers’ price) while the valuation of the production data in the supply table is
based on output at basic prices – this in line with the 2008 SNA.

2.104. In order to balance the SUTs, the same valuation should be used. For this purpose, specific
matrices have to be compiled for trade and transport margins and taxes and subsidies on products.
The compilation of these valuation matrices is an important component of the compilation of SUTs

53
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

and IOTs. Chapter 7 provides a detailed description of the compilation steps for the valuation
matrices and the compilation issues.

2.105. An overview of the three different valuations – basic prices, producers’ prices and
purchasers’ prices – is provided below. They differ as a result of the treatment of taxes on products
less subsidies on products, and trade and transport margins.

(a) Basic prices

2.106. Basic prices are the preferred method in the 2008 SNA for valuing output in the accounts.
This price basis reflects the amount receivable by the producer from the purchaser for a unit of
goods or services, minus any taxes payable, and plus any subsidy receivable on that unit as a
consequence of production or sale (for instance, the cost of production).

2.107. The value of output at basic prices reflects the sum of intermediate consumption of goods
and services at purchasers’ prices, compensation of employees, return to capital for market
producers’ own capital formation, and other taxes less subsidies on production. Other taxes on
production include items such as property taxes and business rates, business licences, motor
vehicle licenses, mission permits issued by governments under cap-and-trade schemes, and others.
Basic prices exclude any transport charges invoiced separately by the producer. When a valuation
at basic prices is definitely not feasible, then a proxy as close as possible to basic prices should be
used.

2.108. The basic price valuation is the preferred valuation for the construction of IOTs which in
turn are used in constructing structural models of the economy or modelling particular features of
economic behaviour. When compiling the IOTs, it is therefore necessary also to value the
purchases by products at basic prices, a process which is further explained in chapter 7.

(b) Producers’ prices

2.109. Producers’ prices may be thought of as the prices of goods and services “at the factory
gate”, so to speak. This valuation includes all taxes on production and taxes on products, for
example excise duties. Producers’ prices relate to basic prices as follow:

Producers’ prices equals basic prices

plus taxes on products (excluding invoiced VAT)
less subsidies on products.

2.110. Although the producers’ price valuation is valid and noted in the 2008 SNA, it not
recommended for use in the 2008 SNA. At the same time, this valuation still forms the basis for
some business survey data. Accordingly, if relevant, specific steps are needed to change data based
on business survey to basic prices, as appropriate, for use in national accounts and SUTs.

54
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(c) Purchasers’ prices

2.111. Purchasers’ prices are those prices payable by the purchaser and include transport costs,
trade margins and taxes (unless the taxes are deductible by the purchaser). Purchasers’ prices are
defined as follows:

Purchasers’ prices equals producers’ prices

plus any non-deductible VAT or similar tax payable by the
purchaser
plus transport costs paid separately by the purchaser and not
included in the producers’ price.
plus trade margins.

2.112. Where taxes and subsidies on products and other taxes and subsidies on production are
concerned, some short definitions are provided below:

• Taxes on products include, in particular, value added taxes, taxes and duties on imports, and
taxes on products such as stamp taxes on the sale of petrol, diesel, alcoholic beverages and
tobacco.

• Subsidies on products include import subsidies and other subsidies on products.

• Other taxes on production consist of all taxes that enterprises incur as a result of engaging in
production, independently of the quantity or value of the goods and services produced or
sold. These may be payable on the land, fixed assets, business and property rates or labour
employed in the production process or on certain activities or transactions.

• Other subsidies on production consist of subsidies which resident producer units may receive
as a consequence of engaging in production, including in particular subsidies on payroll or
work force, subsidies to reduce pollution and grants for interest relief.

2.113. In the use table, transactions are recorded at purchasers’ prices. In the supply table,
domestic production is recorded at basic prices and imports by type of product at cost, insurance
and freight (CIF) prices. In the SNA and the balance of payments, total imports of goods are valued
at free on board (FOB) prices. Further details on these connections and the adjustments required
may be found in chapter 5, section D. Accordingly, additional columns are included in the supply
table in order to complete the valuation gap between total use and total supply of products. These
include information on trade and transport margins, taxes on products and subsidies on products.

(d) Value added tax

2.114. VAT is a wide-ranging tax usually designed to cover most or all goods and services. In
some countries, VAT may replace most other forms of taxes on products but it may also be levied
in addition to certain other taxes on products, such as excise duties on tobacco, alcoholic beverages

55
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

or fuel oils. VAT is a tax on products collected in stages by enterprises. Producers are required to
charge certain percentage rates of VAT on the goods or services that they sell. VAT is shown
separately on the sellers’ invoices so that purchasers know the amounts that they have paid.
Producers are generally not required, however, to remit to the government the full amounts of the
VAT invoiced to their customers because they are permitted to deduct the VAT that they
themselves have paid on goods and services purchased for their own intermediate consumption,
resale or gross fixed capital formation.

2.115. Deductible VAT is the VAT payable on purchases of goods or services intended for
intermediate consumption, gross fixed capital formation or for resale that producers are permitted
to deduct from their own VAT liability to the government in respect of VAT invoiced to their
customers. Non-deductible VAT is VAT payable by purchasers that is not deductible from their
own VAT liability, if any.

2.116. The SNA requires that the net system of recording VAT should be followed. In the net
system, outputs of goods and services are valued excluding invoiced VAT; imports are similarly
valued excluding invoiced VAT; and purchases of goods and services are recorded including non-
deductible VAT.

(e) Valuation in SUTs and IOTs

2.117. Box 2.8 presents an overview of the valuation in the compilation of SUTs and IOTs in a
simplified numerical example. this overview underlines the different valuation layers: the supply
table at basic prices including the transformation into purchasers’ prices is considered with the use
table at purchasers’ prices (total supply equals total use). In a second step, valuation matrices are
compiled – one for the trade and transport margins and the other for the taxes less subsidies on
products – in order to transform the use table from purchasers’ prices to basic prices. In this way,
the supply table at basic price can be considered in relation to the use table at basic prices (total
supply at basic prices equals total use at basic prices). The use table at basic prices is further split
between the domestic use table and imports use table at basic prices. The SUTs at basic prices are
the starting point for the compilation of IOTs, which are compiled at basic prices.

56
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.8 Overview of the valuation in SUTs and IOTs

BALANCED SUPPLY AND USE SYSTEM
Supply table at basic prices including a transformation into purchasers' prices Use table at purchasers' prices
Industries Taxes Total supply Industries Final use
Output Supply at Trade and Total use at
less at Final Gross
Agricul- Manuf. and at basic Imports basic transport Agricul- Manuf. and purchasers’
Services subsidies purchasers’ Services consumption capital Exports
ture constr. prices prices margins ture constr. prices
on prices expenditure formation
Agriculture 270 30 50 350 20 370 28 20 418 Agriculture 38 65 103 155 24 33 418
Manufacturing 6 380 87 473 42 515 80 21 616 Manufacturing 115 123 64 167 85 62 616
Construction 4 50 13 67 8 75 22 4 101 Construction 12 16 6 24 39 4 101
Products

Products
Trade, transport and Trade, transport and
communication 10 15 210 235 7 242 - 130 13 125 communication 21 2 2 98 1 1 125
Finance and business Finance and business
services 6 17 240 263 11 274 15 289 services 14 54 43 128 32 18 289
Other services 4 8 100 112 12 124 7 131 Other services 14 35 23 41 10 8 131
Total 300 500 700 1 500 100 1 600 80 1 680 GVA 86 205 459 750
Output at basic prices 300 500 700 613 191 126
COMPILATION OF VALUATION TABLES
Trade and transport margins
Industries Final use
Final Gross
Agricul- Manuf. and Total
Services consumption capital Exports
ture constr.
expenditure formation
Agriculture 3 5 4 13 2 1 28
Manufacturing 16 14 5 33 10 2 80
Construction 3 4 1 5 8 1 22

Products
Trade, transport and
communication - 22 - 23 - 10 - 51 - 20 - 4 - 130
Finance and business
services
Other services
Total
Taxes less subsidies on products
Industries Final use
Final Gross
Agricul- Manuf. and Total
Services consumption capital Exports
ture constr.
expenditure formation
Agriculture 1 1 6 11 1 20
Manufacturing 2 2 2 12 2 1 21
Construction 1 2 1 4

Products
Trade, transport and
communication 1 1 1 9 1 13
Finance and business
services 1 1 12 1 15
Other services 1 6 7
Total 4 5 12 52 6 1 80

TRANSFORMATION OF SUPPLY AND USE TABLES TO BASIC PRICES

Supply table at basic prices Use table at basic prices
Industries Industries Final use
Output Supply at
Final Gross
Agricul- Manuf. and at basic Imports basic Agricul- Manuf. and Total
Services Services consumption capital Exports
ture constr. prices prices ture constr.
expenditure formation
Agriculture 270 30 50 350 20 370 Agriculture 34 59 93 131 21 32 370
Manufacturing 6 380 87 473 42 515 Manufacturing 97 107 57 122 73 59 515
Construction 4 50 13 67 8 75 Construction 9 12 4 17 30 3 75
Products

Products

Trade, transport and Trade, transport and

communication 10 15 210 235 7 242 communication 42 24 11 140 20 5 242
Finance and business Finance and business
services 6 17 240 263 11 274 services 14 53 42 116 31 18 274
Other services 4 8 100 112 12 124 Other services 14 35 22 35 10 8 124
Taxes less subsidies on
Total 300 500 700 1 500 100 1 600 products 4 5 12 52 6 1 80
GVA 86 205 459 750
Total 300 500 700 613 191 126

COMPILATION OF SEPARATE USE TABLES FOR DOMESTIC OUTPUT AND IMPORTS

Imports use table at basic prices Domestic use table for domestic output at basic prices
Industries Final use Industries Final use
Final Gross Final Gross Total use at
Agricul- Manuf. and Total Agricul- Manuf. and
Services cons. capital Exports Services consumption capital Exports basic prices
ture constr. ture constr.
exp. form. expenditure formation
Agriculture 4 9 3 1 1 2 20 Agriculture 30 50 90 130 20 30 350
Manufacturing 12 17 6 2 3 2 42 Manufacturing 85 90 51 120 70 57 473
Construction 4 2 1 1 8 Construction 5 10 3 16 30 3 67
Products

Products

Trade, transport and Trade, transport and

communication 2 4 1 7 communication 40 20 10 140 20 5 235
Finance and business Finance and business
services 4 3 2 1 1 11 services 10 50 40 115 30 18 263
Other services 4 5 2 1 12 Other services 10 30 20 35 10 7 112
Total 30 40 15 5 5 5 100 Imports 30 40 15 5 5 5 100
Taxes less subsidies on
products 4 5 12 52 6 1 80
Empty cells by definition GVA 86 205 459 750
Total 300 500 700 613 191 126

E. Compiling SUTs as an integral part of the national accounts

2.118. As mentioned before, the compilation of SUTs should be seen as an integral part of the
compilation of the national accounts. Figure 2.4 provides a general overview of how the
compilation of SUTs and IOTs fits within the compilation of national accounts conforming to the
same statistical standards (for example, 2008 SNA, BPM 6, 2012 SEEA, IMF Government Finance
Statistics, and others), and using the same basic sources generally used for the compilation of
national accounts.

57
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.119. One important feature of the approach outlined in Figure 2.4 is the level at which the
traditional annual and quarterly balancing process of the national accounts and balance of
payments system takes place. Balanced macroeconomic data can be derived at a more aggregated
level by applying the production, income and expenditure approaches. A recommended, better
quality option is that the system be balanced at the same time for the institutional sector accounts
and the SUTs at a lower-level disaggregation of products and industries. In many countries, the
annual and quarterly estimates of GDP are obtained from the production, income and expenditure
approaches and reconciled using SUTs. Some countries have a long tradition and much experience
in using detailed production data based on establishments (local kind-of-activity units) as the
statistical unit for compiling GDP estimates following the production approach.

58
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 2.4 Overview of SUTs and IOTs as part of the SNA compilation

System of National Accounts

Supply and use tables Accounts

Supply and
Production Income
use tables at Valuation Expenditure
approach approach
purchasers' matrices approach GDP
GDP GDP
prices (unbalanced) (unbalanced)
(unbalanced) (unbalanced)
(unbalanced)

Balancing

Supply and
use tables at Valuation
Single estimate of GDP
purchasers' matrices
prices (balanced)
(balanced) (balanced)

Institutional sector accounts

1. non-financial corporations
Supply and use tables at 2. financial corporations Total Rest of
basic prices 3. general government economy the world
(balanced) 4. households
5. NPISH

Goods
Product-by- Industry-by- Distribution Accumu-
and Production
product industry and use of lation
services account
IOTs IOTs income accounts
account
accounts

1. Different approaches to measuring GDP

2.120. The three approaches to measuring GDP form the basis of estimating GDP both quarterly
and annually. The use of three different methods which, as far as possible, use independent sources
of information avoids sole reliance on one source and is conducive to greater confidence in the
overall estimation process. This in turn also underpins not only the quality of the key aggregates
but also of the underlying details. The SUTs combine the three approaches in a consistent manner.

59
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(a) Production approach

2.121. The production approach looks at the contribution of each economic unit by estimating the
value of their output less the value of goods and services used up in the production process to
produce their output, this is also known as GVA. Using the production approach:

GVA at basic prices equals output at basic prices

less intermediate consumption at purchasers’ prices
and then,

GDP equals GVA at basic prices

plus taxes on products
less subsidies on products

GDP is also the balancing item of the production account for the whole economy.

2.122. The distinction between market and non-market producers (see 2008 SNA, para. 6.133, for
the definitions) is important for the determinants of both total output and gross value added which
is covered in this section. While the output of market producers is determined from the revenue
side, the output of non-market producers is calculated as the costs of all inputs including labour
cost and consumption of fixed capital. Box 2.9 provides an overview of the calculation of output
for market and non-market producers.

2.123. The estimate of output for producing units in the non-market sector is derived by summing
their costs, for example, intermediate consumption, compensation of employees, other taxes (less
subsidies) on production and consumption of fixed capital. GVA is the sum of compensation of
employees, other taxes (less subsidies) on production and consumption of fixed capital.

2.124. The production approach to measuring GDP, and the estimates of GVA, can be
implemented by using an industry dimension or by an institutional sector dimension. GVA is the
variable used when producing labour productivity estimates and also output per worker uses GVA
as the output measure.

Box 2.9 Calculation of output for market and non-market producers

Market producers and producers for own final use
Total output equals total sales of goods and services (as invoiced, excluding VAT)
(at basic prices) plus changes in inventories of work-in-progress and finished goods
plus output produced for own use, for example research and development (R&D), computer software
and construction (also known as own account capital formation) and household production of
agriculture products for own use
less purchases of goods or services for resale without further processing (thereby only including the
gross margin within output)
plus income earned in kind
less any taxes on products
plus any subsidies on products

60
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Total intermediate equals total purchases of goods and services for use as inputs to the production process (excluding
consumption (at employment costs and fixed capital formation)
purchasers’ prices) less changes in inventories of materials and fuels
less any purchased or bought-in R&D, computer software (treated as capital expenditure, assuming
this is included in the purchases in the first place)
plus Financial intermediation services indirectly measured (FISIM)
plus any imputed insurance premium supplements
less any payments to employees such as income earned-in-kind
Gross value added equals total output (at basic prices)
(at basic prices) less total intermediate consumption (at purchasers’ prices)
Non-market producers
Total output equals total intermediate consumption (at purchasers’ prices)
(at basic prices) plus compensation of employees (labour costs)
plus imputed charge for consumption of fixed capital (sometimes called depreciation)
plus other taxes on production and imports
less other subsidies on production
Gross value added equals compensation of employees (labour costs)
(at basic prices) plus imputed charge for consumption of fixed capital (depreciation)
plus other taxes on production and imports
less other subsidies on production
Final consumption expenditure equals total intermediate consumption at purchasers’ prices
(at purchasers’ prices) plus gross value added at basic prices
equals total output at basic prices
less market output
less payment for non-market output
less output produced for own final use
equals non-market output

(b) Income approach

2.125. Using the income approach, GDP is obtained by adding together the income components
that make up value added. GDP by income approach covers only the income generated within the
domestic economy:

GDP equals compensation of employees

plus gross operating surplus and gross mixed income
plus other taxes less subsidies on production
plus taxes on products and imports.
less subsidies on products

The above income approach provides estimates of GDP market prices.

2.126. As its name suggests, the income approach adds up all income earned by resident
individuals or corporations in the production of goods and services and is therefore the sum of uses
in the generation of income account for the total economy (or alternatively the sum of primary
incomes distributed by resident producer units).

61
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2.127. The income approach to measuring GDP can be analysed either by industry, by institutional
sector or by type of factor income. The type of factor income approach is often linked to the source
data and allows for the incorporation of various administrative data sources. These include, for
example, generating direct estimates of mixed income (using labour force data and administrative
data) and gross trading profit and loss (using company accounts data) as complementary estimates
and not as residuals.

2.128. Based on factor incomes, to estimate gross operating surplus the following categories are
added:

• Self-employment income (mixed income and quasi-corporations)

• Gross trading profits of private financial corporations
• Gross trading profits of private non-financial corporations
• Gross trading surplus of public corporations (financial and non-financial)
• Rental income
• Non-market consumption of fixed capital
and the following categories are deducted:
• Holding gains and losses on inventories
• Intermediate consumption of financial intermediation services indirectly measured (referred
to as FISIM)
2.129. Producing all three dimensions in a single, integrated SUTs framework provides a natural
link between the production account and generation of income account, both by industry and by
institutional sector. This approach also ensures a high degree of consistency and coherency across
the accounts.

2.130. It should be noted that the income approach to measuring GDP cannot be used to calculate
chained linked volume measures directly because it is not possible to separate income components
into prices and quantities in the same way as for goods and services. However, a chained linked
volume measure of the income based total can be obtained indirectly. The expenditure based GDP
deflator at market prices (also known as the index of total home costs) can be used to deflate the
current market price income based total estimate to provide a chained linked volume measure of
the total income component of GDP for balancing purposes.

(c) Expenditure approach

2.131. In the expenditure approach, GDP is obtained by adding the final expenditures or uses by
consumers and producers of goods and services produced within the domestic economy. The total
is obtained from the sum of final consumption expenditure on goods and services by households,

62
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

NPISHs and government, gross capital formation (gross fixed capital formation on tangible and
intangible fixed assets, changes in inventories and acquisitions less disposals of valuables) and net
exports of goods and services.

2.132. Using the expenditure approach:

GDP equals Final consumption expenditure (households, NPISHs and government)

plus gross fixed capital formation
plus change in inventories
plus acquisitions less disposals of valuables
plus exports
less imports

2.133. The data for these categories are estimated from a wide variety of sources, including
business surveys, expenditure surveys, the government’s internal accounting system, surveys of
traders and the administrative documents used in importing and exporting goods.

2.134. To avoid double counting in this approach, it is important to classify consumption

expenditures as either final or intermediate. Final consumption expenditure involves the
consumption of goods purchased by or for the ultimate consumer or user. These expenditures are
final because the goods are no longer part of the economic flow or being traded in the market place.
Intermediate consumption, on the other hand, is consumption of goods and services that are used
or consumed in the production process. Gross capital formation is treated separately from
intermediate consumption as the goods (or services) involved are not used up within the production
process in an accounting period, except for depreciating over time.

2.135. Exports include all sales to non-residents, and exports of both goods and services have to
be regarded as final consumption expenditure, since they are final as far as the domestic economy
is concerned. Imports of goods and services are deducted because they are not part of the
production of the domestic economy but produced in another economy.

2.136. The expenditure approach to measuring GDP is also used to estimate chain-linked volume
measures of GDP. The chained-linked volume measure shows the change in GDP after the effects
of inflation have been removed.

2.137. Box 2.10 shows a numerical example of how a single estimate of GDP can be derived from
a balanced SUTs system by extracting the components of the production, income and expenditure
approaches to measuring GDP from the supply table and use table.

63
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Box 2.10 Example of derivation of GDP from balanced SUTs

Supply table at basic prices, including a transformation into purchasers' prices Millions of euros
INDUSTRIES Total VALUATION Total
Trade, Finance and supply at supply at
Other Imports Wholesale Retail trade Transport Taxes on Subsidies
Agriculture Manufacturing Construction transport and business Total basic Total purchasers’
services trade margins margins margins products on products
communication services prices prices
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (15)
Agriculture (1) 8 782 0 0 0 0 0 8 782 3 271 12 052 1 052 873 274 386 - 107 2 479 14 532
Manufacturing (2) 796 182 982 643 1 808 133 44 186 405 124 590 310 995 29 777 19 061 2 540 21 041 - 49 72 370 383 364
Construction (3) 83 961 43 060 734 255 179 45 272 563 45 835 0 0 0 1 542 0 1 542 47 377
Products

Trade (4) 1 4 773 311 54 204 640 257 60 187 600 60 787 - 31 301 - 21 040 0 586 0 - 51 755 9 032
Transport (5) 13 465 66 25 538 128 125 26 335 8 150 34 485 0 0 - 2 800 628 - 448 - 2 620 31 865
Communication (6) 160 1 781 139 43 912 1 253 982 48 228 6 234 54 463 472 1 021 9 3 592 - 34 5 059 59 522
Finance and business services (7) 29 8 902 698 7 588 106 909 3 381 127 508 7 061 134 569 0 0 - 22 4 865 0 4 842 139 411
Other services (8) 3 85 13 1 053 143 74 346 75 643 824 76 467 0 85 0 1 777 0 1 861 78 329
Total (9) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 151 293 729 653 0 0 0 34 416 - 638 33 778 763 431
Adjustments

CIF/FOB adjustments on imports (10) 0 0 0 0 0 0 0 - 97 - 97 0 0 0 0 0 0 - 97

Direct purchases abroad by

(11) 0 0 0 0 0 0 0 6 675 6 675 0 0 0 0 0 0 6 675
residents
Total (12) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 157 871 736 230 0 0 0 34 416 - 638 33 778 770 009
Total of w hich:
Market output (13) 9 763 195 916 41 462 127 401 88 330 18 116 480 989 0 480 989 0 0 0 0 0 0 480 989
Output for ow n final use (14) 104 4 029 3 468 2 134 19 890 2 670 32 295 0 32 295 0 0 0 0 0 0 32 295
Non-market output (15) 0 4 0 5 302 1 241 58 528 65 075 0 65 075 0 0 0 0 0 0 65 075

Use table at purchasers' prices Millions of euros

INDUSTRIES FINAL USE
Total use at
Trade, Finance and Final consumption expenditure Gross fixed Changes Changes
Other purchasers’
Agriculture Manufacturing Construction transport and business Total General capital in in Exports Total
communication services
services Households NPISH prices
government formation valuables inventorie
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
Agriculture (1) 2 583 6 570 16 371 34 49 9 623 3 595 0 0 180 0 - 27 1 161 4 909 14 532
Manufacturing (2) 2 205 107 190 12 441 16 874 6 015 8 797 153 522 71 438 0 3 180 26 756 2 183 3 034 123 252 229 842 383 364
Construction (3) 105 2 440 9 528 2 446 3 907 1 604 20 029 1 667 0 0 25 155 0 - 38 563 27 348 47 377
Products

Trade (4) 33 1 883 119 2 240 259 308 4 842 3 325 0 0 67 45 0 753 4 189 9 032
Transport (5) 14 4 386 267 8 399 822 321 14 208 5 833 0 3 370 0 0 0 8 453 17 656 31 865
Communication (6) 34 2 563 299 9 359 5 919 1 833 20 008 26 444 0 121 5 976 0 67 6 905 39 514 59 522
Finance and business services (7) 457 13 578 4 736 20 359 29 166 9 134 77 430 38 838 0 1 006 11 170 0 - 178 11 145 61 981 139 411
Other services (8) 8 382 59 1 171 415 1 794 3 829 14 923 5 416 53 373 113 107 1 567 74 500 78 329
Total at purchasers’ prices (9) 5 440 138 991 27 466 61 219 46 538 23 839 303 492 166 063 5 416 61 050 69 418 2 335 2 859 152 800 459 939 763 431
CIF/FOB adjustments on exports (10) 0 0 0 0 0 0 0 0 0 0 0 0 0 - 97 - 97 - 97
Adjustments

Direct purchases abroad by

(11) 0 0 0 0 0 0 0 6 675 0 0 0 0 0 0 6 675 6 675
residents
Purchases on the domestic
(12) 0 0 0 0 0 0 0 - 12 945 0 0 0 0 0 12 945 0 0
territory by non-residents
Total at purchasers’ prices (13) 5 440 138 991 27 466 61 219 46 538 23 839 303 492 159 792 5 416 61 050 69 418 2 335 2 859 165 648 466 517 770 009
Compensation of employees (14) 551 30 679 10 239 37 906 22 997 41 971 144 343 0 0 0 0 0 0 0 0 0
VALUE ADDED

Other taxes less subsidies on

(15) - 1 627 1 077 546 1 755 2 004 1 103 4 858 0 0 0 0 0 0 0 0 0
production
Consumption of fixed capital (16) 1 845 12 750 1 542 10 917 18 934 7 480 53 469 0 0 0 0 0 0 0 0 0
Net operating surplus (17) 3 658 16 453 5 138 23 040 18 989 4 921 72 198 0 0 0 0 0 0 0 0 0
Gross operating surplus (18) 5 503 29 203 6 680 33 957 37 923 12 401 125 667 0 0 0 0 0 0 0 0
GVA (19) 4 427 60 959 17 465 73 618 62 923 55 475 274 868 0 0 0 0 0 0 0 0 0
Total input at basic prices (20) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 0 0 0 0 0 0 0 0 0

Table based on 2011 figures from Austria

The box below shows how a single estimate of GDP at market prices can be derived from the above balanced SUTs
system by extracting the components of the production, income and expenditure approaches to measuring GDP
from either the supply table or the use table.
Calculation of gross domestic product
Millions of euros
Production approach Income approach Expenditure approach
Total output at basic prices 578 360 Compensation of employees 144 343 Final consumption expenditure by Househods 159 792
- Intermedate consumption at purchasers’ prices - 303 492 + Other taxes less subsidies on productio 4 858 + Final consumption expenditure by NPISH 5 416
+ Consumption of fixed capital 53 469 + Final consumption expenditure by General government 61 050
+ Net operating surplus 72 198 + Gross fixed capital formation 69 418
= GVA 274 868 = GVA 274 868 + Changes in valuables 2 335
+ Changes in inventories 2 859
+ Taxes less subsidies on products 33 778 + Taxes less subsidies on products 33 778 + Exports 165 648
- Imports - 157 871
= GDP 308 647 = GDP 308 647 = GDP 308 647

Table based on 2011 figures from Austria

2. Linking SUTs to the institutional sector accounts

2.138. It is important to link the SUTs to the institutional sector accounts in order to have a
complete, consistent and integrated set of accounts, as highlighted in Figure 1.1. The SNA uses
two types of units and two ways to subdivide the economy. Both are quite different and serve
different analytical purposes. In order to describe production, income, expenditure and financial

64
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

flows, and balance sheets, the SNA uses institutional units which, on the basis of their principal
functions, behaviour and objectives, are grouped into institutional sectors like non-financial
corporations and financial corporations. For the institutional units, the full set of accounts is
covered in the system.

2.139. A simplified version of a table covering the main institutional sectors is shown in Table
2.8. Further details may be found in chapter 10, on linking the institutional sector accounts to the
SUTs.

Table 2.8 Simplified table linking the SUTs to the institutional sector accounts

INDUSTRIES
INSTITUTIONAL SECTORS Total
1 2 … n
1. Non-financial corporations
Total output
Market output
Output for own final use
Non-market output
Intermediate consumption
GVA at basic prices
Compensation of employees
Other net taxes on production and imports
Consumption of fixed capital
Operating surplus, net
Gross fixed capital formation
2. Financial corporations
Total output
:
Gross fixed capital formation
3. General government
Total output
:
Gross fixed capital formation
4. Households
Total output
:
Gross fixed capital formation
5. Non-profit institutions serving households
Total output
:
Gross fixed capital formation
6. Total
Total output
:
Gross fixed capital formation

2.140. When describing the processes of production (and input-output analyses), the system uses
the establishment as the statistical unit and groups it into industries on the basis of its principal
activity. For the establishment, only a limited set of accounts is feasible, namely those accounts of
the SUTs framework.

2.141. In order to show the relationships between the accounts of the production processes and
the accounts of the institutional units, a link table can be compiled as an integrated part of the
system. In this link table, a cross-classification of output, intermediate consumption, components

65
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

of GVA (and possible other variables of industries) between the industries and the institutional
sectors is shown. This link table should help to ensure consistency of data compiled on the basis
of establishment and on the basis of institutional units. As both units are classified differently, the
link table also provides a picture of the relationships between output, intermediate consumption,
GVA, and other variables, originating in the different industries and institutional sectors.

3. Benefits of compiling SUTs as an integral part of the national accounts

2.142. There are a number of advantages of producing SUTs as an integral part of the national
accounts and this approach is therefore recommended in this Handbook.

2.143. From a methodological point of view, the following may be identified:

• SUTs provide the ideal framework for integrating the components of the three approaches
to measuring GDP both in current prices and in volume terms.
• When statistical discrepancies exist amongst the macroeconomic aggregates, it is less clear
where adjustments could be applied. Through their detailed examination of the supply and
use of products, however, the SUTs provide a powerful approach to identifying which areas
could be adjusted.
• SUTs allow for the data confrontation of different primary sources by bringing them together
into a single framework, and facilitate efforts to prioritize how resources could be allocated
to seek quality improvements.
• Where statistical information is incomplete or contradictory, as may happen with gross fixed
capital formation or household final consumption expenditure, alternative estimates can be
made in a transparent way using the SUTs framework, ensuring consistency and coherence.
• SUTs provide a full framework for establishing the connection between the various valuation
concepts in national accounts, from basic prices through to purchasers’ prices.
• SUTs form the ideal framework for estimating GVA through double deflation and GDP in
volume terms, while also ensuring coherence of deflation across the different areas.
2.144. In terms of practical benefits:

• SUTs offer new options to incorporate all existing information, including from primary
sources, on a consistent basis. This is also true for information that is only periodically
available, as well as a framework for making reliable estimates, including plausible
restrictions and identities.
• When SUTs are produced as an integral part of the national accounts, it is relatively easy to
compile IOTs. These IOTs derived from SUTs will be fully compatible and consistent with
all figures from the national accounts, adding to the credibility and analytical usefulness of
both products.

66
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• SUTs that are consistent with the national accounts are normally produced in connection
with benchmarked macroeconomic data some two or three years after the initial preliminary
results of the national accounts are published. SUTs can also play a vital role in the
production of preliminary annual or even quarterly accounts.
2.145. Once the SUTs system is in place on an annual basis, the benefits are significant and can
take various forms:

• SUTs from the previous year can be updated with information available for the preliminary
year in order to produce a complete set of SUTs (albeit at a more aggregated level) that are
consistent with the preliminary figures. This procedure is a good method for revealing
inconsistencies in the aggregated preliminary figures at an early stage.
• SUTs can be used to incorporate new information; for example, when new detailed
information on total supply and exports is available earlier, then the structure of SUTs of the
previous year could be used to project SUTs for domestic output and imports.

67
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Chapter 3. Business processes and production stages

A. Introduction

3.1. The compilation of monetary and physical SUTs and thus IOTs is viewed as part of a
statistical production process which starts from the identification of the objectives and users’ needs
to the dissemination of the tabulations and the evaluation of the production process. The various
stages of compilation of SUTs (and IOTs) presented in this Handbook follow those of the GSBPM
(United Nations Economic Commission for Europe, 2013). The GSBPM explicitly identifies and
organizes the compilation steps and the interdependencies between them in a generic statistical
business process. Thus it provides a useful flexible framework by which to describe the
compilation process for SUTs and IOTs.

3.2. It should be mentioned that country practices in the compilation of SUTs and IOTs vary
considerably, since they are specific to the particular context in which they take place. For
example, they depend on the specific institutional arrangements of the statistical system, the
statistical legal framework, the legal, political, regional and taxation arrangements, the statistical
units, the business registers, the range of processes, publication schedules, revision policies,
resources, data availability, confidentiality, and also the final outputs. Despite the great variability
in country practices, there are common steps in the compilation of SUTs and IOTs. In the
description of the compilation stages of the GSBPM in this Handbook, these common steps that
are flexible and applicable to all countries are identified.

3.3. There is an overarching framework within which the statistical production process takes
place and this should be taken into consideration in the design of the compilation process and also
in the actual compilation of SUTs and IOTs. This includes the statistical institutional arrangement
in the country and the data and metadata quality framework.

3.4. The objective of this chapter is to provide an overview of the compilation steps for SUTs
and IOTs. Section B presents an overview of the different institutional set-ups in various countries.
Section C outlines the GSBPM compilation stages that relate to SUTs and IOTs and, lastly, section
D provides a schematic summary of the compilation steps and their links with the relevant chapters
of the Handbook and a summary of the main recommendations, principles and guidelines for the
compilation of SUTs, IOTs, PSUTs and EE-IOTs that are covered by this Handbook. Annex A to
chapter 3 provides examples of institutional arrangements for the compilation of economic
statistics in countries.

69
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

B. Institutional arrangements

3.5. The institutional arrangements are generally understood as a set of agreements between the
agencies involved regarding the division of responsibilities in the collection, processing,
compilation and dissemination of data. They are fundamental to an effective statistical system and
essential for the management of an integrated economic statistics programme. The functions and
responsibilities of the lead statistical agency in the country can be carried out more efficiently if it
is supported in its role by institutional arrangements such as advisory committees, relationship
meetings, memorandums of understanding, service-level agreements, technical cooperation and a
legal framework that protects the confidentiality and integrity of the data while allowing for the
sharing of data between partner statistical agencies (United Nations, 2013, para. 3.23).

3.6. Apart from the legal framework and other factors, institutional arrangements depend on the
kind of national statistical system that exists in a given country, namely, whether it is centralized
or decentralized. A national statistical service is considered centralized if the management and
operations of the statistical programmes are predominantly the responsibility of a single
autonomous government agency, and decentralized if the statistical programmes are managed and
operated under the authority of several government departments. Under this arrangement, a
particular agency is usually entrusted with the responsibility of coordinating the statistical
activities of the various departments.
3.7. In economic statistics, countries have different institutional arrangements under which
such bodies as, for example, the national statistical office and the central bank have different roles
and responsibilities. Countries often follow a decentralized approach, under which the collection
of economic statistics is split across different institutions within the country, so that, for example,
the national accounts (non-financial accounts) are compiled by the national statistics office, the
balance of payments and financial accounts are compiled by the central bank, and the government
finance statistics covering the public sector are compiled by the finance ministry.

3.8. When countries are considering either building or redesigning their systems or changing
the roles and responsibilities of the various institutions involved, the undertaking should be
approached with the aim of producing integrated economic accounts throughout the entire
statistical production process. The motivation for integrated economic statistics comes from the
benefits that such data sets provide for coordinated national and global policy initiatives in an
increasingly interconnected world. The integration of economic statistics involves the use of
common concepts, definitions, estimation methods and data sources for statistical reconciliation,
helping to improve the coherence and consistency of a wide range of economic statistics and to
reduce the respondent burden and overall costs. Integration therefore is not specific either to the
type of statistical system (centralized versus decentralized) or to the level of development of the
statistical system. This approach has the following prerequisites:

70
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Adoption of the conceptual framework of the SNA and the SEEA as the umbrella framework
for organizing economic statistics

• Alignment of the interdependencies of the components of the statistical production process

(statistical units, classifications and others)

• Establishment of enabling institutional arrangements for statistical integration (United

Nations, 2013, para. 2.5)

3.9. Examples of institutional arrangements in different countries may be found in the annex to
this chapter. In general, it could be said that, beyond being conducive to the coherence and
consistency of official economic statistics, a centralized arrangement may provide more
comparability and harmonization both within the statistical system and with the statistical system
of other countries. Although the transition to an integrated system may incur large investment
costs, it would generate great benefits in terms of improved quality and reduced costs in both the
short term and the long term.

3.10. The roles and responsibilities of the various institutions in countries evolve over time and
aspects of the historical evolution of these arrangements are reflected in the country examples
covered in this chapter. One such example is that of Finland, where, in 2014, the compilation of
the balance of payments was transferred from the country’s central bank to its statistics office.
Finland now follows the practice of other countries such as Denmark, Ireland, Luxembourg, Malta,
Norway and the United Kingdom, where the balance of payments is compiled alongside the
national accounts within the statistical office and not in the central bank.

C. Overview of the Generic Statistical Business Process Model

3.11. The GSBPM describes, and defines, the set of business processes needed to produce
official statistics. It provides a standard framework and harmonized terminology to help statistical
organizations to modernize their statistical production processes, and to share methods and
components. The GSBPM can also be used for integrating data and metadata standards, as a
template for process documentation, for harmonizing statistical computing infrastructures, and to
provide a framework for process quality assessment and improvement. The GSBPM is a reference
model that can be used in a flexible manner to describe, document, organize and communicate the
statistical production process in question.

3.12. The GSBPM consists of a sequence of eight phases: (1) Specify needs; (2) Design; (3)
Build; (4) Collect; (5) Process; (6) Analyse; (7) Disseminate; and (8) Evaluate. An overview of
the phases, together with the sub-elements of each phase, may be seen in Figure 3.1.

71
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3.1 Phases of the GSBPM

3.13. The GSBPM is not a rigid framework in which all steps must be followed in a strict order;
rather, it helps to identify the possible steps in the statistical business process, and the
interdependencies between them. Although presentation of the GSBPM follows the logical
sequence of steps in most statistical business processes (for example, business surveys), in
different circumstances the elements of the model may occur in a different order. In addition, in
compiling SUTs and IOTs, some sub processes will be revisited a number of times, forming
iterative loops, in particular within the “Process” and “Analyse” phases.

3.14. This section focuses on the business processes in national accounts, in particular, the
compilation of SUTs, PSUTs and IOTs. The business process and stages of production covered in
this chapter therefore reflect the application of the underlying GSBPM. As a result, Figure 3.2
provides an overview of a simplified business processing model specific for the compilation of
SUTs, PSUTs and IOTs.

3.15. In the compilation of SUTs and IOTs, the sequential stages in the compilation of the
GSBPM may be summarized as follows, and as presented in Figure 3.2:

• Phases 1–3: Specify needs, design and build. This stage includes tasks related to the phases:
1 – “Specify needs”, 2 – “Design”, and 3 – “Build” of the GSBPM set out in Figure 3.1. It

72
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

covers all the pre-collection activities of setting up the system, including identifying users’
needs, designing the system, determining the size of the SUTs and IOTs, and other tasks.
• Phase 4: Collect. This relates to the activities of data gathering from various sources. In
general, the compilers of SUTs (and IOTs) rely on data already collected for the purposes of
national accounts which have often already been adjusted to fit into the national accounting
framework.
• Phase 5: Process. This stage corresponds to a number of activities related to the data
cleaning, adjustments and transformation that are needed in order to start putting the data
into an unbalanced SUT. This stage is very important in the compilation of SUTs and IOTs
and is therefore separated into two steps in Figure 3.2. The first corresponds to all the
activities necessary to put the data in the initial unbalanced SUTs. This involves data
cleaning, pre-processing, aggregation and disaggregation of the basic data and any other
adjustment to the basic data to fit into the national accounts concepts of the SUTs. The
second step in this phase corresponds to all the activities of setting up an initial (unbalanced)
set of SUTs at purchasers’ and basic prices and in current prices and volume terms.
• Phase 6: Analyse. This stage corresponds mainly to the activities of balancing (manual and
automated) SUTs and IOTs and the feedback loop to the source data to resolve
inconsistencies. As a result, there is a continuous loop between this and the previous phase,
making it possible to achieve balanced SUTs and IOTs. In this stage, the final output of the
compilation process is prepared, validated and finalized.
• Phases 7 and 8: Disseminate and evaluate. This stage refers to the activities related to the
dissemination of the output tables, which include the preparation of printed publications,
press releases and websites, the promotion of dissemination products and other tasks, as well
as the activities related to the evaluation of the production process and also of the output in
the light of internal or external feedback.
3.16. The grey boxes for each stage in Figure 3.2 include examples of the types of functions
undertaken. They are listed in no particular order of importance and are linked with one another.

3.17. The broad approach is to move and process data from left to right, with minimal backward
loops, even though effective feedback loops are critical at each phase, and the incorporation of
new, or improved, data deliveries are unavoidable. Good data version control at each stage is
needed, enabling the generation of a wide-range of outputs, articles and analyses such as revision
analysis.

73
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3.2 Simplified business processing model for compiling SUTs, IOTs, and PSUTs

3.18. Each phase should be viewed as cumulative, even when allowing for the iterative nature of
the balancing process. The incorporation of balancing adjustments should be viewed as a
cumulative step and not as creating a loop.

3.19. It is important to prepare proper documentation throughout the various compilation stages
and in particular during the stage of balancing and adjustment. The steps and links between the
source data through to the balanced data should be recorded and documented separately and
reviewed in subsequent balancing exercises to investigate source data incoherence, bias and other
factors. For example, moving from the original source data (such as business survey data,
administrative-based data, company accounts-based data, and other types of data) to the validated
2008 SNA data, a number of adjustments may need to be made in such areas as the following:

• Coverage (including exhaustiveness) adjustments

• Conceptual adjustments
• Quality adjustments

74
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Balancing and coherence adjustments

3.20. In general, for all the stages of the compilation process, it is important also to have in place
the following:

• Data version control – for example, data storage, conventions allowing easy access and
revision analyses, and also pre and post automated balancing analyses
• Clear controls and disciplines – for example, read and write access for each stage, setting
out operational standards, change controls and testing, and other measures
• Appropriate staffing – for example, ensuring that all staff are trained and skilled to undertake
the different functions and ensuring that sufficient staff are in place for each phase
• Clear organizational structure of the staff involved – for example, clear roles and
responsibilities of staff, as staff members can have more than one role in more than one phase

D. Overall strategy for the compilation of SUTs and IOTs

3.21. Within the stages of the overall production process presented in the previous section, the
“Process” and “Analyse” phases (5 and 6) have a particular structure in the compilation of SUTs
and IOTs. This section provides an overview of the steps that are generally undertaken to construct
SUTs and IOTs after the data have been gathered. In addition, since the compilation of SUTs and
IOTs is not seen as a one-time exercise but as part of a continuous programme, this section also
provides the strategy for compiling SUTs and IOTs in current prices and in previous years’ prices
for the first year of compilation and the subsequent years.

3.22. The first step in compiling SUTs and IOTs is to populate the various separate parts of the
supply table and use table (as shown in Figure 3.3) with the available data. This leads to the
construction of unbalanced SUTs which are then subjected to a balancing process to reconcile all
the entries.

3.23. The steps that are generally used by countries to construct an unbalanced version of the
SUTs are presented below:

• Step 1 – construction of the supply table: This consists of filling the available data into an
initial unbalanced supply table, which covers domestic output by product (part 1 in Figure
3.3) and the imports of goods and services and the valuation matrices comprising
information on taxes less subsidies on products, trade margins and transport margins (part
2 in Figure 3.3). These valuation matrices allow the transformation of total supply of
products at basic prices (formed by summing the domestic output and the imports) to total
supply of products at purchasers’ prices. The construction of this initial unbalanced supply
table is presented in chapter 5 of this Handbook.

75
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Step 2 – construction of the use table: In a manner similar to step 1, this step consists of
filling the use table with the available data, which cover: the intermediate consumption at
purchasers’ prices (showing the input requirements of goods and services for the
production of the domestic output of each industry shown in the supply table – part 3 in
Figure 3.3); final uses at purchasers’ prices and for each category, such as final
consumption and gross fixed capital formation, for which separate compilation steps will
be needed (part 4 in Figure 3.3); and production-based GVA at basic prices shown by
industry (part 5 in Figure 3.3). This compilation step is covered in chapter 6 of this
Handbook.
• Step 3 – compilation of the valuation matrices: These matrices are essential to the
preparation of SUTs at basic prices. They expand the valuation columns in part 2 of the
supply table in Figure 3.3 into corresponding matrices for intermediate consumption and
final consumption of the use table. This compilation step is described in chapter 7 of this
Handbook.
• Step 4 – compilation of the imports use table and domestic use table at basic prices: This
step is essential to increasing the analytical uses of SUTs by distinguishing the use of
imported and domestic products. This compilation step is presented in chapter 8 of this
Handbook.
• Step 5 – compilation of the SUTs in volume terms (previous years’ prices): When balanced
both in current prices and in volume terms, the SUTs ensure coherent and consistent
deflation of the components of the production and expenditure approaches to measuring
GDP as well as coherent and consistent estimates of price and volume indices. This requires
that SUTs are compiled in volume term at this stage of the compilation process. The
compilation of SUTs in volume terms is described in chapter 9 of this Handbook.
• Step 6 – Linking SUTs with the institutional sector accounts: Linking SUTs and the
institutional sector accounts is an important step in the compilation of SUTs, ensuring the
full integration and consistency of SUTs with the national accounts. This link is provided
by compiling a linking table between the sectors and industries (part 6 in Figure 3.3). The
compilation of the linking tables is presented in chapter 10 of this Handbook.
3.24. These six steps above are generally followed in that order; however, there is a significant
level of interdependency in the compilation process. For example, trade and transport margins and
taxes less subsidies on products are necessary for the transformation of the use table from
purchasers’ prices to basic prices and also for conversion of the supply of products at basic prices
to purchasers’ prices in the supply table, to enable the balancing of products at purchasers’ prices.
This information may partly be derived from estimates based on the use tables and linked to
estimates from the supply table at basic prices.

76
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

3.25. On the other hand, estimates of certain final uses may require basic supply side information
if, for example, the product flow 7 method is being applied. Nevertheless, allowing for
interdependencies in the compilation of these tables, it is vital that the tables are viewed and
accepted as primary estimates.

Figure 3.3 Structure of the SUTs and the links covered in this Handbook

3.26. Once these six steps are completed, the result is unbalanced SUTs at purchasers’ prices and
basic prices. This represents the start of a balancing procedure which is an iterative procedure
integrating the following aspects:

• Balancing of SUTs at purchasers’ prices

• Compilation of valuation matrices
• Transformation of SUTs into basic prices

7
Following the terminology used in 2008 SNA (para. 14.2), in this Handbook the expressions “product balance” and
“product flow” methods are used in preference to “commodity balance” and “commodity flow method”, as
reflecting the more recent usage of the word “product” in place of “commodity”. It is noted, however, that the
change in terminology does not indicate a change in methodology.

77
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Compilation of separate use tables for use of domestic output and use of imports of goods
and services
• Balancing of SUTs at purchasers’ prices and at basic prices
• Balancing the production-based GVA and income-based GVA, providing the link to the
institutional sector accounts
All of the above should be balanced with time series in mind to ensure consistent movements of
levels and growth rates.

3.27. One of the key reasons why the SUTs are balanced first at purchasers’ prices is to reflect
as closely as possible the basis of the survey data that is being fed into the use table. The
intermediate uses and final uses, for example, are collected close to the economic reality of the
prices paid by purchasers, in other words, the purchasers’ prices. In addition, no valuation issues
exist with such variables as compensation of employees and other taxes less subsidies on
production.

3.28. These aspects should, however, be viewed alongside the domestic output part of the supply
table reflecting data collected from producers whereby the output is valued at basic prices. Thus a
balance between the two is needed.

3.29. For the SUTs balanced at purchasers’ prices, the two key identities are:

• Total supply of products at purchasers’ prices equals total uses of products at purchasers’
prices.
• Total output of industries at basic prices equals total input of industries at basic prices.
3.30. Balancing is not just necessary in order to achieve the above identities but also makes it
possible to trace inconsistencies of basic data and estimation methods. Ideally, the balancing of
the SUTs system should be done both in current prices and in volume terms simultaneously. In
fact, balancing in this manner means that the process is not complete until the transformation into
basic prices and the separation of the use of domestically produced products from the use of
imported goods and services have been achieved, as these are key steps in producing the SUTs in
volume terms. These steps are in practice interrelated and provide a powerful feedback loop in
terms of quality and validity of the various component estimates.

1. Compilation of SUTs in current prices and volume terms

3.31. The SUTs framework in Figure 3.3, when treated in summary form, can be combined with
the H-Approach to show a simplified version of the compilation schematic configuration when the
SUTs are compiled in current prices and in volume terms. Figure 3.4 illustrates the sequence of
steps involved in the compilation of SUTs, PSUTs, and IOTs. The inner box outlined in red focuses
on the compilation of SUTs. Thus, countries that intend to compile only monetary SUTs can focus
on the steps within the red box and follow the compilation sequence indicated by the arrows in the

78
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

figure. The order of compilation of PSUTs and IOTs in the figure does not imply a compilation
sequence for these tabulations. The compilation of these tables reflects the country’s priority.

3.32. The simplified illustration provided in Figure 3.4 of the compilation of SUTs, PSUTs and
IOTs can be seen in relation, on the one hand, to phase 4, “Process”, when it comes to compiling
unbalanced SUTs and PSUTs and, on the another hand, to phase 5, “Analyse”, when it comes to
compiling balanced SUTs, PSUTs, and IOTs.

Figure 3.4 Compilation of SUTs and IOTs in current prices and in volume terms

Current prices Volume terms

SUTs SUTs
BOX BOX
A
at purchasers’ prices at purchasers’ prices E
PxI PxI

Separation of valuation Addition of valuation

and imports matrices and imports matrices
Deflation
processes
SUTs SUTs
BOX BOX
B
at basic prices at basic prices D
PxI PxI

BOX
G PSUTs Assumption Assumption
applied applied

BOX IOTs IOTs BOX

C PxP and IxI PxP and IxI F

PxP: product-by-product
PxI: product-by-industry
IxI: industry-by-industry

3.33. In compiling the seven boxes, in the sequence from box A to box G, a further dimension
of their evolution needs to be reflected. In year 1 of the compilation process, boxes A, B and C
representing current prices are produced in that sequence covering the economy for year (T) for
SUTs and IOTs, and box G covering PSUTs which are linked to the outputs of boxes B and C.

3.34. As mentioned in chapter 2, SUTs in volume terms for one period can be compiled using
SUTs in current prices for one period and deflators. The preferred approach, however, includes a
time-series dimension and boxes D, E and F representing the previous years’ prices should not be
compiled in year 1 as there are no SUTs in current prices for the previous year (T-1).

79
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

3.35. It is therefore essential to have two consecutive years of SUTs in current prices to enable
the production of the first year of SUTs in previous years’ prices. If the SUTs are produced less
frequently, say, every five years, it is much more difficult to produce SUTs in volume terms.

3.36. In year 2, boxes A, B and C are produced covering the economy for year (T+1) together
with any revisions to the data for boxes A, B and C for the year (T). In addition, the first set of
SUTs in previous years’ prices can be compiled for year (T+1). In each year thereafter, the process
will extend the availability of SUTs by an extra year in current prices and previous years’ prices,
while also incorporating any revisions to SUTs for earlier periods to ensure consistent time series.

3.37. Figure 3.5 provides a summary of the evolution dimension for the first three years. As time
passes, different challenges will evolve, such as the need to retain an ever-increasing number of
years of SUTs on a consistent basis, the need for a revisions policy, data version and vintage
control, managing the production of consistent levels and growth rates, the organizational
arrangement of resources which may not increase each year, among others. It is thus important to
ensure that this process is properly planned and managed from the start.

Figure 3.5 Evolution of compiling SUTs and IOTs in the first three years

Note: CP: current prices; PYP: previous years’ prices; (r) revised tables.

3.38. Based on the overall strategy for the compilation of SUTs and IOTs, it is possible to provide
step-by-step guidance. This is provided for the first year of compilation, and then subsequent years

80
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

of compilation, as there are some additional steps that need to be considered in order to ensure a
fully consistent time series of SUTs in current prices and in previous years’ prices.

(a) Step-by-step summary for the first year of compilation

3.39. Figure 3.6 provides a description of the various phases in the compilation of SUTs and
IOTs for year 1 of the compilation, together with the reference to the relevant chapters of the
Handbook. The figure also contains references to boxes A–G of Figure 3.4 and the compilation
stages of Figure 3.2. Although boxes A, B and C form key outputs, there are various intermediate
stages and intermediate outputs as denoted in the separate stages in Figure 3.6.

3.40. Before starting the first year of compilation, as illustrated in Figure 3.6, compilers should
ensure that the overall framework is well in place, comprising: standards, definitions,
classifications and methods, statistical units, business registers and sample frames, census, survey
and administrative data collection.

3.41. The following key features of Figure 3.6 should be noted:

• It is consistent with Figure 3.5 and covers the simplified business process model for
compiling SUTs and IOTs. Each block of work in Figure 3.6 is split according to the type of
work as indicated in the six different stages in Figure 3.2.
• It follows the underlying principles and features of the GSBPM.
• The flow of work is kept as logical and sequential as possible and follows the H-Approach
as covered earlier. As mentioned earlier in this chapter, however, the compilation of SUTs
and IOTs includes several interrelated processes and dependencies which must be reflected
and retained. Furthermore, in some cases, there is more than one approach available, for
compiling trade margins using a supply-side approach or use-side approach or both.
• The flows in Figure 3.6 do not present backward loops, although effective feedback loops
are critical at each phase and improve the process. For example, the compilation and
balancing of PSUTs provides an important feedback loop to the compilation of monetary
SUTs, thus enhancing the quality of physical and monetary SUTs.
• Integrated links bring the PSUTs together with such input data as the prices and quantities
(levels) and material flow accounts alongside the SUTs and IOTs.
• Each phase of work is also linked to the main chapters in parts A and B of this Handbook,
providing much more detail on the compilation – these links provide the key chapters but
not all references.
• The same approach has not been applied to part C (Extensions and applications) of the
Handbook as there are many variations and options.

81
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3.6 First year of compilation

82
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(b) Step-by-step summary for the subsequent years of compilation

3.42. Having completed the year 1 results, Figure 3.7 provides a detailed stage of production for
year 2 with the focus on the SUTs in previous years’ prices, which, as mentioned, can only be
compiled when SUTs have been compiled for the current year and the previous year.

3.43. In compiling SUTs in previous years’ prices for the first time, there may be several years
of SUTs in current prices already available. If so, then the compilation is just an extended version
of the process indicated in Figure 3.7, and it is better in terms of quality and consistency as there
is a time series dimension in place immediately.

3.44. Other features to note in Figure 3.7 are the following:

• The focus of the steps in Figure 3.7 is the right-hand side of the H-Approach of Figure 3.4
and builds on the detail available from the left-hand side – assuming the left-hand side
products are available.
• The deflation approach follows the underlying H-Approach and is covered in chapter 9, on
compiling supply and use tables in volume terms.
• Other approaches are available but this is the recommended approach.
• The compilation of IOTs in volume terms is not essential but is not resource-intensive if all
the other parts are available.
3.45. An additional feature is also achieved whereby GVA in volume terms is arrived at using
the SNA recommended approach, namely, double deflation. However, the results from this
approach need additional quality assurance against other indicators. This is to ensure that the
quality of the GVA estimate in volume terms (and, in turn, GDP) is not reduced as a consequence
of the errors in either the current price estimates of output and intermediate consumption or
inappropriate deflation of these two variables.

83
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3.7 Second year of compilation

84
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2. Summary of the main recommendations, principles and guidelines of the Handbook

3.46. Box 3.1 provides a list of the main recommendations, principles and guidelines relevant
for the compilation of SUTs, IOTs, PSUTs (and EE-IOTs) and related products presented in this
Handbook, covering various aspects such as the organizational or institutional environment,
compilation, data strategy and requirements, and balancing.
3.47. The recommendations and guidelines presented in box 3.1 may be viewed as aspirational
as they provide the model scenario for the compilation and dissemination of SUTs, IOTs and
related products. Countries can gradually implement the recommendations and guidelines in
accordance with their specific situation in terms of such factors as data availability, resource
constraints, legal framework and others, and in line with their priorities.

Box 3.1 Examples of the main recommendations, principles and guidelines

provided in this Handbook

A. Organizational or institutional environment

1. The organization of the economic statistics system should follow an integrated economic statistics approach. The use
of the GSBPM to organize the statistical production process would facilitate the compilation of SUTs, IOTs and related
products.

2. National accounts should have very close links with all its suppliers, in particular, the business register, business
surveys and administrative sources.

3. The compilation of the various components of the SNA framework should be coordinated and integrated in terms of
production processes, such as production schedules, feedback loops, coherence, and other features:

• National accounts (including balance of payments and monetary financial statistics, government finance
statistics)
• SUTs and IOTs together with PSUTs and EE-IOTs
• Environmental-economic accounts to be closely linked with the compilation of SUTs
• Regional accounts
• Prices
• Labour market statistics
4. The compilation of SUTs and IOTs should be performed as part of the regular compilation of the national accounts
and within the core national accounts. This would have the following effects:

• Leads to better quality, coherence and consistency of national accounts, balance of payments and related
statistics
• Creates effective and powerful data quality and coherence feedback loops, which in turn help to address
structural issues and biases and to prioritize resources to targeted improvements
5. The final estimates of the national accounts aggregates should be derived from the balanced SUTs framework and not
the other way around. For example, the SUTs based estimates should be not confined to predetermined estimates or
already published estimates.

85
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

6. The compilation of SUTs and IOTs should reflect stakeholder interests. This can be achieved by organizing regular
meetings with data suppliers and users together with other regular stakeholders.

7. Appropriate internal governance should be exercised to ensure accountability and guidance, supported by programme,
project and process management, including risk management, under a framework reflecting:

• Schedules, timetables and customer-supplier service-level agreements set in place to ensure a regular supply
of source data, briefings and evaluation reviews
• Various standards and policies, such as revision policy, confidentiality and disclosure controls, and others
• Staff recruitment, retention and skill development
8. Skill development needs to take into consideration the following types of training requirements:

• National accounts – Technical skill focus covering national accounts concepts, methods, processes and
guidance, and also such functions as developments, compilation, coordination, balancing, analyses and
dissemination
• Systems – IT systems, programming, data management (standards and principles), data dissemination, also
covering website management, and other systems, including the role of dedicated IT professionals supporting
economic statistics
• Management – staff management, effective leadership, communication, and other aspects
9. For effective and sustainable production of SUTs and IOTs, it is important to have sufficient computing capacity in
place that includes:

• Robust, reliant, structured, quick and well-documented systems

• Database software and hardware, speed, structure, flexibility, statistical functionality, data management and
links to web dissemination
10. It is important that the statistical production process is well documented and kept up to date, reflecting:

• Operational, methodological, system, metadata and recording-specific issues, adjustments, etc. for each
quarterly or annual exercise
11. The compilation of SUTs and IOTs is to be performed with due consideration for the costs and resources available
and also other criteria such as data availability, data quality and time.

B. Compilation

1. SUTs (and IOTs) should be compiled annually and, if possible, on a quarterly basis, following the H-Approach for the
production of SUTs and IOTs in current prices and in previous years’ prices (including valuation and imports matrices).
The application of the H-Approach makes it possible for the volume of GVA to be estimated using a double deflation
method and also ensures greater coherence, linking SUTs to various other parts of the SNA framework.
2. SUTs should be produced first, then IOTs derived from the SUTs, using additional information and assumptions.
3. Rectangular SUTs should be compiled with more products than those provided for by industries:

• The greater the detail, better the quality – while more detail will increase the burden on business, systems and
resources, it can improve the quality of balancing
• Improved matching between prices and values, thereby ensuring better quality of the data in volume terms
• Compilation (and balancing) should be undertaken at the greatest level of detail available – time, quality and
resources permitting. Due to confidentiality-type criteria, however, the level of publication may – or will – be
aggregated to a higher level

86
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4. Standard international statistical classifications (such as ISIC, CPC, COICOP, and others) should be used at
appropriate detailed levels to ensure international comparability. Within these classifications, greater granularity may be
desired for specific economies.
5. Consistent statistical units should be used throughout the process, from the business register and business surveys
through to the SUTs.
6. Ideally, SUTs and PSUTs (and EE-IOTs, as appropriate) should be based on sound and complete data sources,
reflecting:

• Common concepts, definitions and classifications

• Comprehensive and up-to-date statistical business register
• Wide range of (preferably annual) regular business surveys (including structural detail), household surveys,
administrative data, prices, and other sources
• Benchmarking and reconciliation – preferably conducted annually, reflecting rapidly changing economies
(minimizing the use of fixed factor or stability assumptions)
• Incorporation of labour and capital information ensuring improved coherence for productivity estimates
• Appropriate choice of index number formulae and base year
7. All the data building blocks should be recorded separately, namely, source data, coverage adjustments (including
exhaustiveness), conceptual adjustments, quality adjustments, balancing adjustments, and others
8. A table should be compiled linking the SUTs and the institutional sector accounts, including:

• Goods and services

• Production accounts by industry and by institutional sector
• Generation of income accounts by industry and by institutional sector
• Parts of the use of disposable income account (such as household final consumption expenditure) and parts of
the capital account – by industry and by institutional sector (such as gross capital formation and its
components)
9. Use of the bottom-up approach should be preferred in the compilation of regional SUTs, which should be reconciled
with national SUTs.
10. For the derivation of IOTs, the following should be the methods most frequently used:

• Model A (product-by-product) IOTs using the product technology assumption

• Model D (industry-by-industry) IOTs using the fixed product sales structure assumption
• Hybrid – mix of technologies usually chosen to avoid having any negatives
11. Comprehensive documentation should be prepared on operational methods and methodology, including appropriate
metadata and revision analysis.
12. Efforts should be made to keep up to date with, and contribute to, internationally evolving and agreed changes to
concepts, methods and systems developments.

C. Data strategy and requirements

1. SUTs are data-hungry and a range of timely, comprehensive, consistent and coherent data sources are needed. The
data strategy should reflect a range of aspects.
2. These should include data-handling aspects such as:

• Data collection (for example, questionnaire design, electronic data capture, receipt of all the data that a
company can provide, etc.)
• Data processing, data editing, metadata and data warehousing

87
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Data quality frameworks

• Data dissemination and use of SDMX standards
3. The following structural and data-collection issues should be reflected:

• Comprehensive and up-to-date statistical business register used as the sampling frame for all business surveys
• Use of as many data sources as possible, censuses, business and household surveys, administrative data,
company accounts, regulatory accounts, company websites, and other sources
• An international business unit handling all aspects of multinational enterprise groups, from profiling the
business structures to data collection and data reconciliation and feeding coherent data through to the various
statistical domains. In addition, the need to develop links and share data with other national statistics offices
and national central banks for statistical purposes only.
• Frequency of information – monthly, quarterly, annually or five-yearly. The more regular, the better the
information reflects rapidly changing industry structures of sales and inputs, changing patterns of household
consumption, impact of globalization on trade flows, and other factors
• Sufficient, appropriate and relevant, price indices matching the current price values for deflation or use of
suitable volume only indicators where price information may be unavailable
• Strategy for handling, and reviewing, areas where data may be missing
4. The following more general needs should be included:

• Need to minimize the burden on business

• Need to have confidentiality and disclosure testing processes

D. Balancing

1. SUTs should be balanced in current prices and in volume terms, thus leading to:

• A single estimate of GDP incorporating the components of production, income and expenditure approaches
to measuring GDP
• Volume estimates of GVA through double deflation
• Balance between supply of products and use of products and between industry inputs and industry outputs
2. The balancing process should simultaneously encompass:

• SUTs at basic prices and at purchasers’ prices

• SUTs in current prices and in volume terms (preferably, previous years’ prices)
• SUTs links to IOTs, PSUTs and EE-IOTs (as appropriate)
• Link with the institutional sector accounts
3. Balancing should strongly promote integration of the following:

• Goods and services, production account, generation of income account, parts of the capital account and use of
disposable income account
• Incorporation of PSUTs and EE-IOTs (as appropriate)
• Productivity estimates (labour, capital and multifactor)
4. Simultaneous balancing should be preferred to sequential balancing. If this is not possible, sequential balancing (first
in current prices, then in volume terms), with quick and effective feedback loops, should be considered as an alternative.
5. The organization of the balancing function can be set up in different ways across teams. A centralized balancing
approach should, however, be preferred to the decentralized balancing arrangement whereby the balancing of the various
elements related to SUTs and IOTs (such as current and constant prices for a single year and for a time series, links with

88
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

productivity, regional accounts, among others) is carried out at the same time and within the same unit in order to ensure
the full consistency of all SUTs-related products.

6. The production and balancing of SUTs should enable the identification of source data incoherence. Mechanisms should
be developed to provide feedback to data suppliers and help prioritize areas for improvement and allocation of resources.

7. An annual review and evaluation of the balancing adjustments should be carried out, to identify and address any
evolving biases.

89
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Annex A to chapter 3: Examples of institutional arrangements in countries

A3.1 This annex presents examples of institutional arrangements in selected countries. The
examples cover the cases of centralized and decentralized statistical systems.

A. Centralized production of economic statistics: Canada

A3.2 As a centralized national statistics office, Statistics Canada is accorded the legal mandate
to collect and disseminate a broad range of statistics by a federal act (the Statistics Act). Provisions
in the Act also serve to protect data confidentiality and assure political neutrality and an arms-
length relationship with policymakers.

A3.3 Users are regularly consulted, and the office, through various channels, ensures that priority
requirements are established and met. These channels include national advisory committees,
federal-provincial consultations and regular bilateral meetings with key policy partners such as the
federal finance department and the Central Bank.

A3.4 Statistics Canada produces a full suite of macroeconomic accounts, including:

• National accounts (including financial and wealth accounts)

• Balance of payments
• Government finance statistics
• Productivity measures
• Environmental accounts (natural resource stocks, along with physical flows of energy use,
greenhouse gas emissions and water use)
• Selected satellite accounts covering tourism, culture and pensions
A3.5 The compilation processes are integrated to assure data coherence across components of
the Canadian macroeconomic accounts, and regional SUTs serve as the integrating mechanism for
the production dimensions. The integration is achieved through annual benchmarking and
reconciliation processes with current price measures of GDP income and expenditure, real GDP
by industry and labour and multifactor productivity. Data coherence is a requirement for key policy
applications, such as the input to fiscal formulas for the sharing of sales tax revenues among the
Federal Government and provincial jurisdictions or to formulas for equalizing fiscal capacity
among Canadian provinces.

90
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

A3.6 Economic surveys, along with labour market data, price statistics and international trade
statistics, are produced within Statistics Canada, ensuring the alignment of priority-setting for
feeder programmes to the macroeconomic accounts.

A3.7 All business surveys are linked to a central business register maintained through regular
updates from administrative files. The survey content is harmonized, as are approaches to
collection, processing and estimation within an integrated business statistics programme
framework. The use of administrative data is optimized throughout all stages of the process, and
continuous access to required files is assured through formalized arrangements with data providers,
such as the Canada Revenue Agency and national and provincial regulatory authorities.

A3.8 In recent years, Statistics Canada has made significant progress towards implementing
consistent classification standards across all feeder programmes, thereby facilitating the
compilation of SUTs. The North American Industry Classification System (NAICS) serves as the
basis of industry statistics and North American Product Classification (NAPCS) the basis of
statistics on products. Continuing efforts are being made to ensure compliance and to coordinate
input from the macroeconomic accounts and feeder programme areas into the development of
updated standards.

B. Centralized production of economic statistics: Norway

A3.9 Statistics Norway has overall responsibility for official statistics in Norway, and also for
the conduct of extensive research and analysis activities. Statistics Norway reports to the Ministry
of Finance, with the support of the Statistics Act of 1989. Statistics Norway is a professional,
autonomous organization with the mandate to determine what it publishes, and when and how the
publishing shall take place.

A3.10 Statistics Norway is responsible for the production and maintenance of the business
register, along with the business surveys using samples drawn from this register.

A3.11 The Department of National Accounts and Industry Statistics comprises nine divisions,
with the following responsibilities:

• National accounts
• Primary industry statistics
• Manufacturing and research and development statistics
• Construction and service statistics
• Transport, tourism and ICT statistics
• Energy statistics
• Natural resources and environmental statistics

91
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Accounting statistics
• Business register
A3.12 The division for national accounts is responsible for the following:

• Quarterly and annual national accounts (including SUTs, IOTs and regional accounts)
• Quarterly and annual non-financial accounts
• Quarterly balance of payments
The balance of payments has been an integral part of the national accounts since the 1950s.
Satellite accounts are also prepared by the division for national accounts but not on a regular basis.

A3.13 The Department of Prices, Financial and External Trade Statistics comprises six divisions,
with the following responsibilities:

• Financial market statistics (including financial accounts)

• Public finance statistics
• Financial corporations
• External trade statistics (in liaison with the Customs Department)
• Price statistics
• Development cooperation
Banking statistics were originally under the responsibility of Statistics Norway and were then
moved to Bank of Norway before being moved back to Statistics Norway.

C. Centralized production of economic statistics: United Kingdom

A3.14 The structure of the United Kingdom statistical system has evolved over many decades,
helped by several reorganizations of statistical departments and changes in legislation,
consolidating the responsibility for almost all economic statistics under the Office for National
Statistics and the Government Statistical Service. The United Kingdom system continues to
evolve, for example, by developing better links and access to administrative data.

A3.15 Currently, the United Kingdom has in place resources, systems and processes for producing
detailed, integrated and timely quarterly and annual economic accounts. The Office for National
Statistics, as an independent statistical body with a central role, is wholly responsible for the
compilation of the national accounts, balance of payments, public sector finance statistics, labour
market statistics and price statistics. The compilation of SUTs is central to the annual national
accounts system. The Office also produces regional accounts, environmental accounts and IOTs
(Mahajan, 2016).

92
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

A3.16 The Office for National Statistics is one of the few national statistics offices with such
centralized responsibility and coverage of economic statistics – this has only been the case since
the late 1980s. Furthermore, since 2011, all the above economic statistics are being produced at
the same location.

A3.17 The independent status of the Office for National Statistics is supported by national
legislation, pursuant to which it reports to the United Kingdom Statistics Authority. The Statistics
Authority, which was established on 1 April 2008 under the Statistics and Registration Service Act
2007, is a non-ministerial department overseen by Parliament and not by a government minister.

Figure 3A.1: Integrated process of compiling national accounts and balance of payments
United Kingdom

Compiled by Sanjiv Mahajan March 2014

D. Decentralized production of economic statistics: Chile

A3.18 Economic statistics in Chile are mainly produced by three institutions:

• Central Bank of Chile, which is responsible for the compilation of most of the
macroeconomics statistics, namely, national accounts (non-financial and financial accounts),
balance of payments and international investment position
• Finance Ministry, which produces the government finance statistics
• National Statistics Office, which undertakes the data collection covering economic and
business surveys and the compilation of price indicators, labour market indicators and socio-
demographic data

93
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

A3.19 The organizational structure described above allows the Bank to achieve a high level of
consistency between external statistics and national accounts in terms of both the statistics
themselves and the methodology.

A3.20 In the production of macroeconomics statistics, the Bank uses a significant amount of data,
provided mainly by the Tax Revenue Service, the National Customs Service, the General
Comptroller’s Office and the National Statistics Office, the latter being the main provider of
statistics for national accounts compilation. Dependency on the statistics from the National
Statistics Office entails a high degree of coordination between both institutions. To this end, a
framework agreement is in place to ensure that the requirements and conditions for the provision
of statistical products are met. In addition, a committee with members from both institutions
regularly meets to coordinate issues related to data collection and the specific needs of national
accounts.

A3.21 There are strong links between the Central Bank of Chile and the National Statistics Office,
buttressed by a continuous programme to improve the cooperation and the quality of the links and
the data flows between the customer and supplier.

A3.22 Other salient features of the Chilean system include the following:

• The SUTs and IOTs are compiled within the national accounts in the Central Bank of Chile.
• Where the balance of payments is concerned, the Central Bank of Chile collects the data on
international trade in services to supplement the data on the international trade in goods
collected and provided by the National Customs Service.
• The Central Bank of Chile also produces regional GDP figures on an annual basis.
• The national statistics office produces a business register which, in turn, is employed by the
Central Bank of Chile after making improvements and modifications.
• Although the environmental accounts are not produced for Chile, various efforts have been
undertaken by the Minister of the Environment to produce a range of environmental
indicators.
A3.23 The diagram below shows the components of the statistical system in Chile.

94
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3A.2 Statistical system in Chile

E. Decentralized production of economic statistics: United States of America

A3.24 The United States of America has a highly decentralized statistical system, under which
responsibility for producing a substantial portion of official federal economic statistics is divided
among 13 agencies that have statistical work as their principal mission.

A3.25 There are also numerous other entities that are considered part of the statistical system in
the United States but statistical work is not their principal mission. Most of the country’s primary
economic indicators are produced by one of three main federal statistical agencies, while the
United States Census Bureau conducts economic censuses and surveys. The three main agencies
and their responsibilities are the following:

• Bureau of Economic Analysis: this body relies primarily on data generated by other agencies
to compile the national accounts (non-financial accounts) and the balance of payments.
• Federal Reserve Board (the United States central bank): this compiles the financial accounts
and government finance statistics.
• Bureau of Labor Statistics: this body prepares the labour market statistics and price statistics.
A3.26 The Bureau of Economic Analysis also undertakes a number of business surveys. At the
same time, however, most of the statistics used by the Bureau in preparing GDP and input-output

95
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

accounts come from non-Bureau sources, including other statistical offices. The Census Bureau
provides most of the other expenditure components of GDP and output and intermediate purchases
within the input-output framework. For the period 1997–2017, the Bureau of Economic Analysis
produced SUTs at basic prices with a transformation to purchasers’ prices, make/use tables at
purchasers’ prices for benchmark years and at producers’ prices annually, together with IOTs.
Annual make/use tables at producers’ prices at a more aggregated level of detail are also available
for the period 1947–1996.

A3.27 The agencies each produce and maintain their own business register, often created using
different sources, as detailed below:

• The Census Bureau’s business establishment list is compiled mainly from federal tax forms
and used as the primary sampling frame for the five-year economic censuses and many of
the economic surveys.
• The Bureau of Labor Statistics business establishment list is based on information collected
in connection with the joint federal and state unemployment insurance programme and used
by Bureau establishment surveys, including the producer price index (PPI) survey.
Figure 3A.3 Overview of the products produced by the main agencies

96
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 3A.4 Overview of the statistical system structure

A3.28 Data sharing between agencies, as practised in the United States, may significantly improve
data harmonization and eventually lead to savings. In recent years, the Bureau of Economic
Analysis has signed memorandums of understanding with other agencies and ministries to
facilitate the exchange of data, including confidential data.

A3.29 Although there may be differences in the concepts of statistics employed by different
agencies and in their statistical coverage (for example, productivity statistics are published in the
United States by the Bureau of Labor Statistics and then used by the Bureau of Economic Analysis
in its measurement of national accounts), the confrontation of the data themselves or of the data
processing steps raises the data validation to another level and enhances the quality of statistics.

97
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Chapter 4. Specify needs, design, build and collect stage

A. Introduction

4.1. The objective of this chapter is to describe the activities specific to the compilation of SUTs
and IOTs which take place during the phases of the GSBPM relating to the specification of needs,
designing, building and collecting, referred to together as the “specify needs, design, build and
collect stage”. The chapter will also explore specific elements that need to be considered when
compiling SUTs and IOTs. These phases of the GSBPM is of particular importance in the
compilation process because of its direct implications for all the remaining phases and their
elements need to be continuously reviewed and corresponding adjustments made to the process.

4.2. The following section of the chapter focuses on the “specify needs”, “design” and “build”
phases of this stage. The level of detail of the industry and products in the tables must be carefully
evaluated at the beginning of the compilation process, together with other elements such as the
compilation schedule, the revision policy and others. Section C moves on to the collect phase and
describes the main data sources used for SUTs and IOTs.

B. Specify needs, design and build phases

1. Specify needs

4.3. The identification of user’s needs is a fundamental step in the compilation of any statistics,
as it aims to identify what statistics need to be compiled, in which format, when and for what
purpose. All these elements affect the planning of the compilation process of SUTs and IOTs since
they have implications, for example, for the choice of the level of industry and product detail of
the SUTs. Thus an assessment of the objectives of these tabulations has to take place during the
specify needs phase of the statistical production process and this assessment must be regularly
reviewed in the light of feedback from users, to ensure the relevance of the compiled SUTs and
IOTs. During this phase, consultation with relevant stakeholders, through meetings, workshops
and surveys, is of key importance.

4.4. Other elements of this phase include the identification of the statistical outputs that are
required to meet the user needs and checking the data availability to see if existing data sources
can meet the user requirements, if there are alternative data sources that would be more suitable
for the specific statistics, and if there are data gaps to fill.

99
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.5. There are many different users – policymakers, analysts, researchers and others – and uses
– planning, modelling, monitoring and so forth – of SUTs and IOTs. It is therefore important to
maintain links with the users to ensure that their needs are met through a single efficient
compilation process. For example, in the context of growing concerns about the environment, if a
specific environmental topic, such as water, energy, fish, or forests, is considered as a key user
need to be addressed, it is important to develop and design a compilation process which includes
these elements from the outset, rather than adjusting ex post facto the statistical output through
modelling based on various assumptions.

4.6. The compilation of any statistical output depends to a large extent on the availability of
appropriate infrastructure for information technology and human resources. The technology has
changed enormously over the last fifty years. Statistics which were once compiled with calculators
are now processed in seconds by modern computers, laptops or even smartphones. This rapid
development has facilitated the work of statisticians and improved the timeliness of their statistical
outputs. When compiling SUTs and IOTs, a variety of software, databases and custom-designed
platforms is available and can be adapted to the specific compilation process in any given country.

4.7. It is therefore important to have a clear understanding of the information technology

requirements necessary for all the phases of the compilation process. In practice, more than one
software package is often required and the ones selected are those which best meet the diverse
functional requirements of the specific phase and which are able to communicate with one another
in an easy, effective and efficient manner. If the links between the software packages are
cumbersome or time-consuming, alternatives should be sought. Many national statistical offices
have developed in-house tailored software to meet their needs – this has advantages but may entail
greater overhead maintenance and training requirements.

4.8. Box 4.1 and Box 4.2 provide examples of custom-made software produced and maintained
by, in the first case, Statistics Netherlands and, in the second, the Institut national de la statistique
et des études économiques (INSEE) – the French national statistical office – and Eurostat – the
statistical office of the European Union.

Box 4.1 Example of in-house custom-built software: Statistics Netherlands

Statistics Netherlands has a long-standing tradition of compiling SUTs both in current prices and in volume
terms, and also IOTs, and therefore has extensive experience in tackling the various challenges associated
with the use of computer systems to produce and maintain long-run series of SUTs and IOTs.
Statistics Netherlands has been publishing SUTs since 1990 (relating to the benchmark revision for the year
1987) and IOTs since the 1950s (for years starting from 1948).
Statistics Netherlands has always used in-house custom-built software for the compilation of the balanced
SUTs and IOTs. There are two separate software tools which are both updated continuously on an SQL
database. These are combined with a graphical user interface in Visual Basic for Applications (VBA) code,
enabling data to be accessed and adjusted by the national accounts staff. Both systems can handle quarterly
and annual data.

100
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

The first application includes tools for the transformation of source data to national accounts definitions and
standards, both in current prices and in volume terms.
The second application includes tools for simultaneous balancing of the SUTs in current prices and in volume
terms, for compiling the valuation matrices and for transforming the SUTs to industry-by-industry IOTs.
The present systems are based on a major overhaul carried out in 1995 and have been continuously updated.
A key rebuild took place in 2004/2005, reflecting a new programming language and also including new
features.

Box 4.2 ERETES

ERETES – the acronym for “équilibre ressources-emplois et tableau entrées-sorties” (supply-use
balances and input-output tables) – is a computer system designed to assist national accountants to
compile the SUTs and integrated economic accounts (including sector accounts), which complies with
the principles and guidelines set out by the SNA. ERETES was developed by INSEE and Eurostat. In
2014, it was being used by several countries in Africa and Latin America and the Caribbean. Further
countries are expected to adopt the system. The ERETES system is available in French, Spanish and
English.
Although the objective of ERETES is to generate SUTs and the integrated economic accounts, it can also
be used by countries that have limited data resources. The minimum data requirements are an enterprise
survey and a household budget survey, foreign trade statistics, government accounts, balance of
payments and banking statistics. With these data, ERETES can help countries to generate estimates of
GDP in current prices. If sufficient price and volume indices are available, then estimates of GDP in
volume terms can also be generated. The compilation of SUTs would also require information on
intermediate consumption and trade and transport margins.
One key advantage of ERETES over other computer systems is that it is supported by a permanent
secretariat that can call on a group of multilingual national accountants and information technology
experts with extensive experience of applying the system in a number of developing countries. ERETES
is regularly updated and improved. ERETES is available at: http://www.eretes.net/EN/index.htm.

4.9. Another successful example of custom-built software produced by one national statistical
office, and then made available for use by other countries under specific terms and conditions, may
be seen in the Norwegian software SNA-NT. Here, Statistics Norway provided both the software
and the associated human resources for training in the use of the applications by, for example,
Malawi, the Czech Republic and Slovakia.

4.10. When choosing the software and hardware to support the compilation of SUTs and IOTs
as part of the national accounts, consideration should be given to various criteria such as the
database environment, in particular its flexibility and structure, its statistical functionality and
diagnostic tools required, the necessary availability of mathematical functions such as matrix
calculations, the resources and costs, the training programme, compatibility with data suppliers,
data management, and the data dissemination platform envisaged.

101
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.11. Across the world a range of different products are used to this end, such as Oracle, JAVA
programming, MATLAB, SPSS, SQL, SAS, Excel and custom-made software built to meet
specific requirements. For example, the use of Excel or input systems such as an Oracle database
could provide an effective solution in the “collect” and “process” phases, for preparing and
validating the data, while SAS could be used for the further processing of SUTs and IOTs. Tools
such as Excel, an output system and web-tools for dissemination may offer the best means of
validating and balancing, and also of analysing and disseminating the data.

4.12. Skilled and trained human resources are also a fundamental pillar for the compilation of
SUTs and IOTs. It is thus important to recruit and retain skilled and effective staff and develop
and use internal and external training opportunities on the theoretical and practical aspects of the
compilation of economic statistics.

4.13. An important step in this phase is the preparation of a document summarizing the findings
of all the activities mentioned above – namely: needs identification; establishment of output
objectives; checking of data availability and information technology requirements, and others – in
the form of a business case, with a view to securing approval to implement the new or modified
statistical business process. Such a business case would need to conform to the requirements of
the approval body but would typically include elements such as a description of the “as-is”
business process (if it already exists), with information on how the current statistics are produced,
highlighting any inefficiencies and issues to be addressed; the proposed “to-be” solution (with
clear improvements and benefits); detailing how the statistical business process will be developed
to produce the new or revised statistics; an assessment of costs and benefits, and any external
constraints.

2. Design and build phases

4.14. The design phase comprises all the activities undertaken to define the statistical output and
the concepts, methods, collection instruments and operational processes necessary. Accordingly,
this phase includes all the design elements needed to define or refine the statistical output identified
in the previous phase, all relevant metadata, ready for use later in the statistical business process,
and quality assurance procedures.

4.15. These activities make substantial use of international and national standards, in order to
reduce the length and cost of the design process and enhance comparability and usability of
outputs. Organizations are also encouraged to reuse or adapt design elements from existing
processes. In addition, outputs of design processes may form the basis for future standards at the
national and international levels.

4.16. The design of the statistical output for SUTs and IOTs consists of the size and layout of
the tables; the breakdown of industries and products; disclosure control methods; processes
governing access to any confidential information; and the identification of the statistical variables
needed, which is then linked to the data collection phase.

102
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.17. In the build phase, the production solution is put together and tested to the point where it
is ready for use in a live environment. This phase is broken down into several activities, which
include a review of data sources; the configuration of the workflow from data collection through
to dissemination; and testing of the statistical business process. For statistical outputs produced on
a regular basis, the design and build phases usually occur for the first iteration, and following a
review or a change in methodology or technology, rather than for every iteration.

4.18. During the design and build phases a number of specific issues in the compilation of SUTs
and IOTs should be considered. These include, for example, the choice of the level of detail of
industries and products of SUTs and IOTs at working level and in the dissemination phase; how
to handle confidentiality throughout the compilation process; the schedule for the compilation and
dissemination of SUTs and IOTs; the revision policy and analysis; the resources required to sustain
the compilation; the benchmarking; and the choice of the index formula and the base year. In
addition, it is important to create and maintain documentation for all phases of the compilation
process to serve as a quality control measure of the process. All these activities in the design and
build phases are further elaborated in the following sections.

(a) Level of industry, product detail and size of SUTs and IOTs

4.19. The level of industry and product detail of the published and disseminated SUTs and IOTs
greatly depends on the objectives of the tabulations and their uses. The industries and products
explicitly identified in the disseminated tables reflect to a great extent the users’ needs and the
specific policy concern of interest. For example, if a specific environmental domain is of interest,
such as energy, specific industries and products are likely to be explicitly identified in the
disseminated tables in order to address the specific domain. The ultimate level of aggregation of
the disseminated SUTs and IOTs has an impact on – and at the same time is affected by – the data
availability and the data collection, compilation and balancing procedures.

4.20. The number of products in the SUTs is usually higher than the number of industries, thus
showing more than one primary product for each industry, casing the SUTs to be rectangular. Their
size and shape will have appropriate implications for IOTs, the physical tables and other related
products analyses (for example, productivity).

4.21. The level of detail of industries and products at the working level is generally very
disaggregated and the recommendation is to work with the most detailed level of aggregation
taking into consideration the constraints posed by the available data, resources, and burden on
business. Various aspects need to be considered, including the user needs, the availability of data,
and the level of detail used in national accounts. For example, compilation aspects that influence
the level of detail (since they facilitate the compilation and validation of the data at the working
level) include:

• Distinction between industries which are allowed to deduct VAT and those that are not
allowed to deduct VAT, and different VAT rates per product and categories exempt from

103
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

VAT to facilitate the compilation of the valuation matrices in particular for taxes and
subsidies.

• Distinction between market and non-market activities to facilitate the understanding of the
input structure of GVA.

• Subdivisions of industries according to institutional sectors which have an impact on the

linking table between SUTs and the institutional sector accounts.

• Separate identification of certain industry and product subdivisions to facilitate the

compilation of the trade and transport margin matrices, such as section G and its divisions
of ISIC Rev. 4.

• Links to international industry, product and functional classifications.

• Links between output structures, input structures, price indices and values for deflation,
along with the availability of price statistics to generate estimates in volume terms.

• Links between domestic supply and exports, and also intermediate consumption and
imports, to study the input structure of the industries.

• Links to the environmental accounts.

• Enterprises that trade internationally and have links to global value chains (see extended
SUTs and the OECD trade by enterprise characteristics (TEC) database).

• Availability and quality of source data.

• Size and value of output and value added of the industry: if the industry is too large and
heterogeneous then it should be further broken down. The same can apply to products.

• Benchmarking (annual as opposed to five-yearly) using comprehensive sources and

censuses, since the level of detail in benchmarking years is much larger than that during
regular annual compilation of SUTs and IOTs.

• Annual chain-linking the volume estimates.

• Staff resources, time schedules for production and publication, confidentiality and system
infrastructure.

4.22. An appropriate choice of the level of industry and product detail in the SUTs at working
level will facilitate the compilation and the search for causes of inconsistencies. For many
products, it is possible, by their nature, to identify the industry in which they are used. For example,
fertilizers are mainly used in agriculture, crude oil in oil refineries, concrete in construction, and
so forth. For some products, it may also be possible to identify whether they are used as

104
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

intermediate consumption or final consumption. Haircut services, for example, can be assumed to
be mainly consumed by households and thus recorded in household final consumption expenditure.
The more detailed the classification of products used in SUTs, the easier it is to use expert
knowledge to supplement surveys in allocating products to different uses.

4.23. Linking and matching products to valuation-related elements (taxes, subsidies, trade and
transport margins) make the composition of transactions more transparent and clear-cut and
significantly facilitates their analysis.

4.24. If the products are more detailed, it will also mean that the number of users of a certain
product is greatly reduced. Where there is only one producer and one user of a product, the search
for the cause of inconsistencies would only require investigating two source statistics. When a
product has 20 users, for example, the search becomes more complicated.

4.25. Questions relating to the layout of SUTs tables also arise for the final use part of the SUTs.
It may be useful to integrate the functional classifications in the final consumption data, showing
final consumption by products and by consumption purpose. On the other hand, it may be better
to keep such detail and transformation separate, but this integration will still be needed in some
form or other in the compilation of SUTs and the other accounts of the system.

4.26. In the dissemination phase, the size and breakdown between industries and products shown
in the SUTs (and thus in the IOTs) mainly reflect the user’s needs and the objective of the
tabulations, taking into account confidentiality considerations. Other presentational considerations
include, for example, the size and relative value of output and value added for the industries and
the size and value of supply for the products. Industries or products that are not economically
significant or relevant for a particular economy may be aggregated together, while a more detail
breakdown may be shown for economic activities that contribute substantially to GDP, in order
more effectively to analyse the cost structures and the interdependencies with other economic
activities.

4.27. It should be mentioned that, when compiling consistent annual, or quarterly, SUTs, the
stability of the level of detail of the applied classifications is also important, as many ratios and
proportions will usually be taken as a starting point in the estimates for the following year.

4.28. A higher degree of product detail also supports the use of certain estimation methods, for
example the product flow method of compiling national accounts (namely, balancing the supply
and use of products) by taking into account the relevant differentiation concerning product tax
rates, margin rates and homogeneity in prices. Moreover, it is much easier to distribute
disaggregated products and services across industries and final use categories with the product
flow method than at a higher aggregate level. Detailed product accounts also help in the balancing
procedure, as it is easier to explore and detect the causes of imbalances if the basis is determined
by homogeneous single products rather than aggregate groups of products. The work on a detailed
product level certainly increases the data quality but has resource and systems implications. At

105
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

higher levels of aggregation, problems of imbalance might not even be detected at all and therefore
not remedied.

4.29. Table 4.1 provides an example of the size of SUTs and IOTs compiled by selected
countries. It is worth noting that the internal working level of the industry and product detail used
for compilation and balancing is much higher than which is actually published. For example, in
the United States, the working level in producing the SUTs is over 800 industries, whereas in
Denmark, there are around 2,350 products at the working level but the SUTs are published only at
a level of 64 products by 64 industries.

4.30. It is important to distinguish between the detail required for the compilation and balancing
work at the working level as opposed to the information required for the publication. The in-house
operating detail should be the same or, as in most cases, in greater detail in terms of number of
industries and products than that allowed for disclosure by the publication. For example, many
countries distinguish hundreds or even thousands of products but do not publish at these levels, as
a great deal of confidential information would thereby be released. It should be noted, however,
that countries often allow people outside the national statistical office to have access to more
detailed data, albeit confidential and under signed agreements, for analytical purposes.

106
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 4.1 Examples of the size of published and internal working level SUTs and IOTs

National Supply and Use Tables National Input-Output Tables

Volume terms
Current prices Current prices
or PYPs
Internal working / Do you Internal working /
Country(1),(2) Published levels produce such compilation levels Published levels
compilation levels
tables:
Number of Number of Number of Number of Y (yes) Number of Number of
products industries products industries N (no) industries / products industries / products
or PxP IxI PxP IxI
P (plan to) Tables Tables Tables Tables
Argentina (3) 271 162 271 162 N n/a 124 n/a 124
Australia (4) 301 67 n/a n/a Y n/a 114 n/a 114
Austria 573 135 74 74 P 74 n/a 74 n/a
Belgium (5) 355 135 64 64 P 135 n/a 64 n/a
Brunei Darussalam (6) 324 74 74 74 N 74 74 74 74
Canada 490 230 490 230 P n/a 230 n/a 230
Chile 275 160 180 111 Y n/a 111 n/a 111
Columbia 369 61 61 61 Y 61 61 61 61
Costa Rica 183 146 183 138 Y 183 136 183 128
Czech Republic 252 120 88 88 Y 184 184 82 82
Denmark(7) 2 350 117 117 117 Y n/a 117 n/a 117
Estonia 247 98 64 64 Y 64 n/a 64 n/a
Finland 776 179 64 64 Y n/a 179 n/a 64
Germany 86 63 85 63 P 73 n/a 72 n/a
Hungary(8) 820 242 64 64 Y 88 88 64 64
Iceland(9) 561 142 n/a n/a P n/a n/a n/a n/a
India (10) 142 126 140 66 N 130 n/a 130 n/a
Indonesia (11) 244 81 n/a n/a P 251 n/a 185 n/a
Ireland(12) 82 82 58 58 Y 82 n/a 58 n/a
Kuwait(13) 43 43 n/a n/a N 43 43 n/a n/a
Mexico(14) 819 814 262 262 P 814 262 814 262
Netherlands 614 128 85 76 Y n/a 128 n/a 76
New Zealand(15) 299 118 201 106 P n/a 106 n/a 106
Norway 860 156 64 64 P n/a 156 n/a 64
Republic of Korea 1 851 328 384 328 N 1 851 n/a 384 n/a
Saudi Arabia (16) 59 59 18 18 N 59 59 n/a n/a
Serbia 216 88 n/a n/a N n/a n/a n/a n/a
Singapore 71 71 71 71 N n/a 71 n/a 71
Slovakia 290 88 64 64 Y 88 n/a 64 n/a
Slovenia (17) 350 230 64 64 Y 64 n/a 64 n/a
South Africa 104 293 104 62 N n/a 50 n/a 50
Sweden(18) 403 97 62 64 P 62 n/a 62 n/a
United Kingdom (19) 112 112 112 112 P 112 n/a 112 n/a
United Republic of Tanzania (20) 250 59 250 59 P n/a n/a n/a n/a
United States of America (21) 4 988 819 73 71 P 73 71 n/a n/a
Submissions to European
Commission reflect EU Member 64 64 64 64 P 64 64 64 64
States(22)

Key: PYP – previous year’s prices

P x P – product-by-product
I x I – industry-by-industry

(1) The above table was compiled by Sanjiv Mahajan (Office for National Statistics, United Kingdom) at the time of preparation
of this Handbook. The numbers are indicative for the reference year 2014 (as at December 2016) unless denoted otherwise.
(2) Other differences will exist in terms of comparability, given that the tables are compiled on different bases, in terms of the
frequency of the tables, classifications used, the SNA version, latest reference year, latest benchmark year, valuation of SUTs
(basic prices or producers' prices or purchasers' prices), assumptions underpinning the IOTs, etc.
(3) For Argentina the SUTs reference year is 2004 and the IOTs reference year is 1997.
(4) For Australia the IOTs working level operates at 1268 products and 114 industries.

107
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(5) For Belgium the IOTs reference year is 2010.

(6) For Brunei Darussalam the reference year for SUTs and IOTs is 2005.
(7) For Denmark the number of products can vary from year to year. Detailed data can be made available outside Statistics Denmark
for analytical purposes.
(8) For Hungary the IOTs reference year is 2010 (compiled five-yearly).
(9) For Iceland the SUTs are produced but not yet published. The plans are to publish 64 products x 64 industries.
(10) For India the SUTs reference year is 2012/13 and the IOTs reference year is 2007/08. Industry outputs are published in full
whereas industry uses are aggregated to higher levels.
(11) For Indonesia the SUTs and IOTs reference year is 2010.
(12) For Ireland the SUTs reference year is 2013 and the IOTs reference year is 2011.
(13) For Kuwait the SUTs and IOTs reference year is 2010.
(14) For Mexico the reference is year 2008, the PxP IOTs reference year is 2008 (814 products), updated version for 2012 (259
products), and the IxI IOTs reference year is 2012.
(15) For New Zealand the production of SUTs in volume terms is being planned but there are no plans for publication.
(16) For Saudi Arabia the SUTs and IOTs internal working levels reference year is 2011 and the published SUTs reference year is
2013.
(17) For Slovenia the SUTs reference year is 2013 and the IOTs reference year is 2010. SUTs and IOTs will be published in 2017,
both with a reference year of 2014.
(18) For Sweden the SUTs in PYPs are produced but not yet published. A version is, however, submitted to the Commission
(Eurostat) and the plan is to publish at some stage.
(19) In the United Kingdom the IOTs PxP tables use the hybrid assumption approach
(20) For the United Republic of Tanzania the SUTs reference year is 2007. IOTs for the reference year 2007 are being finalized.
(21) For the United States, in benchmark years, the SUTs are published at the 389 product and 389 industry level of detail. While
IOTs themselves are not currently published, the IxI Leontief Inverse tables calculated from IOTs are published at the 71
industry level of detail, and PxP tables at the 73 product level of detail.
(22) All EU Member States are expected to provide:
Annual SUTs both in current prices and PYPs as well as five-yearly IOTs under the ESA 2010 Transmission
Programme after the expiry of any National Derogations.
The SUTs and IOTs are to be supplied using 64 products and 64 industries.

4.31. Confidentiality is a fundamental principle of official statistics (see United Nations, 2013b).
It ensures that individual data collected by statistical agencies for statistical compilation, whether
they refer to natural or legal persons, are to be strictly confidential and used exclusively for
statistical purposes. It is important therefore that procedures are put in place to ensure the
confidentiality of the information disseminated.

4.32. Countries may apply different criteria to decide whether specific data may be disclosed or
not. This decision is likely to be driven by the legislation in place underpinning the collection of
data from businesses. The decision is normally influenced by the number of enterprises observed
in an industry or by whether the data can be disclosed through a process of deduction. One solution
would be to choose a higher aggregation level with a sufficient number of enterprises in an industry
to overcome any disclosure problems. There may not always be an easy solution for some
industries or the products to which they should be allocated. The price is a loss of information due
to aggregation, resulting in the increased heterogeneity of the SUTs system. Other methods might
therefore be explored or combined, such as creating or redefining products.

4.33. Cases where there are one or two dominant producers in an industry, such as mining,
extraction of crude oil, sugar, pharmaceuticals and others, pose a different challenge. In these
cases, it is recommended, when necessary, that specific permission is sought from the business
when their data are publicly available from other public sources, for example published company

108
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

annual reports and accounts. If permission is not granted, then suppression of the relevant cells in
the SUTs should be considered. 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.

(b) Schedule of the work programme

4.34. Aligning the monthly, quarterly and annual timetables covering data collection processes,
compilation processes, data supply, validation, balancing and publication for all the accounts and
outputs is a key step in ensuring coherence and consistency. This should include the compilation
of SUTs and other input-output-related products, as appropriate.

4.35. The overall process needs therefore to be split into well-defined blocks of work with clearly
defined processes and linkages between the processes, in such a way that they all fit within realistic
schedules (including contingency planning and risk management) with clear roles and
responsibilities for the staff and management involved. The governance of the programme should
be clear, with regular monitoring and meetings scheduled at key junctures, for example, linked to
key milestones in the process. The project management of the process should ensure that dedicated
resources are attributable to this support function.

4.36. The schedule needs to incorporate deadlines of both data providers and data users, together
with various internal intermediate deadlines. For annual business surveys, the time lag from
changing the questionnaire to incorporating the new results and publication in SUTs and national
accounts could be around three years; thus it is important to retain schedules, which are regularly
reviewed and reflect the incorporation of continual data improvements.

4.37. In general, it is useful to put in place service-level agreements with data providers and also
with data users. Agreements with data providers would cover the types of data to be provided, the
quality criteria, briefings, schedules and the format in which the data will be delivered. Important
elements to consider in such service-level agreements include:

• Clear ownership – senior representatives from both the supplier side and the customer side
• Reasons for data requirement
• Publication of results including disclosure requirements
• Process of data delivery by data provider (for example, format)
• What data are required (need to specify the variables needed)
• Timing of data deliveries
• Required briefing to accompany the data
• Handling of customer queries

109
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Quality (covering such criteria as consistency, credibility, revisions, precision and

communication)
• Methodological notes supporting the data (for example, sources, methods, coverage, and
others)
• Development, improvements and consultation
• Arrangements for review of the process, among others
4.38. Service-level agreements with data users, on the other side, would cover, for example,
users’ deadlines (such as those linked to policy agendas, research schedules, and other factors) to
be reflected in the statistical production schedule. For example, the finance ministry or the central
bank or both may require structural updates on the economy to fit with their policy review. Such
user requirements may form part of regular annual schedules for the producers of SUTs.

4.39. In general, there should be a regular review, possibly on an annual basis, of all aspects of
the process (including timetables, data quality, implementation of future changes, and other
considerations) with both data providers and data users, to ensure continuous improvement and to
guide changes as necessary.

4.40. The schedule for the compilation of SUTs naturally depends on the periodicity and
frequency of SUTs and IOTs. In general, it is recommended that SUTs be compiled annually, in
line with the United Nations Statistical Commission recommendations on the scope of the
implementation of the 2008 System of National Accounts. 8 While it is recommended that a
benchmark system of SUTs based on specific survey results be compiled every five years, rapid
changes in the economy, the external impacts of globalization, the increasing rate of change of
technology and its impact, new products, new industries, the impact of digitization and other
factors may affect the production process of SUTs and the periodicity of benchmarked tables, with
the result that an annual benchmark process may be preferred.

4.41. Schedules setting the frequency of SUTs compilation may be assessed against the revision
policy and the uses of intra-annual sources (such as quarterly and monthly short-term indicators),
with a view to incorporating if necessary the revision guidelines and indicators policy.

(c) Revision policy and analysis

4.42. Revisions to time series data are an important part of the production process. Changes to
published data can occur for many different reasons. For example, forecast data may be replaced
by survey data, the reclassification of industries, methodological changes to the way in which data
are estimated or just through the correction of errors. Changes due to the correction of errors should
be identified as corrections and distinguished from revisions that are more commonly associated

8
See the recommended data set in the report of the Intersecretariat Working Group on National Accounts
(E/CN.3/2011/6), available at http://unstats.un.org/unsd/statcom/doc11/2011-6-NationalAccounts-E.pdf.

110
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

with improving estimates as more information is gathered over time (Mahajan, 2015). There is a
conflict between the release of timely estimates and that of accurate estimates. If statistical offices
and central banks waited to publish the most accurate data possible, given the nature of data
collection, there would be a large time lag between the date to which the data refer and the date of
publication.

4.43. Bringing together data for the purposes of compiling SUTs and other input-output-related
products integrated within the national accounts involves aligning production timetables (quarterly
or annual) and schedules, and also revision policies. Ideally, a highly effective revision policy to
ensure that revisions are implemented in a coordinated and coherent fashion across the accounts
should cover the national accounts, balance of payments and government finance statistics, should
include the SUTs and IOTs, and should also extend to primary source data and other domains,
such as regional accounts and environmental accounts.

4.44. The revision policy should reflect appropriate criteria to assess each revision, including
when best to implement the changes, for example, in a quarterly exercise (for example, for short-
term revisions) or in an annual exercise (for example, for long-period revisions). This will have an
impact on how to compile SUTs and IOTs and the revision guidelines should be operated flexibly,
reflecting such issues as economic significance and practical aspects such as their impact on
resources and systems.

4.45. The usual guidelines applied to all the accounts cover:

• Revisions to the latest quarters for an incomplete year (these can affect annual SUTs)
• Revisions to past recent quarters since the last full benchmark year
• Revisions made through the annual process to recent years, say, between three and five
completed years
• Revisions to a longer period, sometimes viewed as a major revision. Many earlier years may
be revised depending on whether they meet certain criteria such as those relating to
methodological improvements, correction of errors and economic significance
4.46. Any changes to back data will also have an impact on the monthly and quarterly seasonally
adjusted estimates.

4.47. In some countries, occasional or major revisions of national accounts are usually carried
out every five years and require more resources if the revision is implemented on the basis of a
large SUTs system. A revision at a more aggregated level is always easier and less demanding.

4.48. The various revision practices have different pros and cons:

111
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Regular revisions to SUTs and national accounts help to preserve good quality levels and
growth rates. Some users do not welcome regular revisions, however, whereas others do
not welcome the “big bang” approach to revisions.
• Five-yearly revision exercises may mean more significant changes for a number of years
until the next revision window.
• Revisions only at the aggregated level may provide problems with the detail and provide
discontinuities in the long time series of annual SUTs
• Revisions also at the level of IOTs ensures that SUTs and IOTs remain consistent.
4.49. Revisions applied only to part of the accounts and not all the relevant outputs would
generate incoherence across different outputs and not help users, for example, in some countries,
SUTs may be revised but the IOTs are not revised, thus the links between the two products are out
of line. This situation should be avoided.

4.50. Analysis of revisions can provide information about the reliability of estimates and how
they change between the first estimate and final estimate as well as a source for the identification
of any biases, (Mahajan, 2004b). Note that revision analysis does not give information on the
accuracy of an estimate, for example, the final estimate may not be accurate in terms of sampling
error or non-sampling error.

4.51. The knowledge of the source and the reasons for the revisions is key and helps producers,
and users, to better understand the data. Sometimes with major revisions going back in time,
understanding the changes can be quite complicated, such as changes in definitions, classifications
and data. For analytical users of SUTs and IOTs, it is important to understand the reasons and
impact, especially when there is a mix of revisions, such as, for example, with the introduction of
2008 SNA, necessitating the revision of SUTs and IOTs going back 30–40 years or more. The
forthcoming handbook on backcasting methodology (United Nations, forthcoming) will provide
more information on this topic.

(d) Resources

4.52. When considering the resources and human requirements for the full integration of SUTs
and IOTs (and also physical SUTs) in the compilation of national accounts and balance of
payments, it is essential to distinguish between the first compilation of the SUTs, the recurrent
production of SUTs and the process surrounding major backward revisions.

4.53. A substantial amount of resources is required to build up an integrated SUTs framework

for the first time. This work involves establishing all the planning, conceptual and methodological
work, data collection needs, requirements of all the individual industry and product balances, the
development of appropriate techniques for incorporating the primary sources and new software for
handling the SUTs system and the necessary training and investment in staff and systems. The

112
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

investment may lead to considerable changes in working procedures towards a better integration
of activities from data collection through to publication, following an integrated statistics
approach.

4.54. The resources needed to establish an integrated SUTs framework should, however, be
viewed against the way in which the development will evolve, for example, the level of integration
and the organization of roles and responsibilities across the statistical domains. The
implementation of the recommendations and guidelines provided in this Handbook can be carried
out in a gradual manner taking into account countries’ specific situation in terms of the resources
available and their national priorities. Countries’ practices may vary considerably, among such
scenarios as the following:

• Only annual SUTs in current prices

• Annual SUTs in both current prices and in volume terms

• Annual SUTs complemented with a quarterly SUTs system

• Annual and quarterly SUTs in current prices and in volume terms

• SUTs at basic prices or at purchasers’ prices

• Benchmarking annually or, say, at least every five years

• Links between SUTs and IOTs:

o SUTs only

o SUTs and IOTs, both in current prices (possibly also in volume terms)

o Only IOTs and no SUTs – this is clearly the least favoured scenario but it has a
historical legacy in some countries

4.55. Developing a new SUTs and national accounts system poses different challenges involving
the change of existing production systems while maintaining business-as-usual activities.

4.56. Where periodicity is concerned, at least every five years a benchmark system of SUTs
should be compiled which is based on more exhaustive specific survey and administrative data
results. As described earlier in this chapter, however, when considering schedules for SUTs (and
IOTs), with rapidly changing and developing economies, the impact of globalization and
digitization, and other factors, it is recommended that the development of new SUTs systems
should reflect an annual benchmarking and reconciliation process. This will also help to avoid
significant revisions and distortions to the levels of data. Together with annual chain-linking, this
will mean better quality measurement of the growth rates of GDP in volume terms, in particular
for the more recent periods. Appropriate techniques have been developed and the trends of

113
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

structural change (for example, composition of outputs and intermediate consumption) during the
previous years can be used to take forward the structures in SUTs, if no new structural information
is available, for example, beyond benchmark years.

4.57. The experience from countries which have integrated the SUTs framework into the
compilation of their national accounts suggests that the resources needed are similar to the
resources for those countries following a more traditional approach with a separate SUTs
compilation. It is therefore recommended that SUTs are compiled as an integrated process and a
regular part of the compilation of the national accounts. The fact that IOTs are produced with
relatively few additional resources if a SUTs system is in place militates in favour of an integrated
and regular approach for the compilation of IOTs.

(e) Benchmarking, extrapolation using indicators, price and volume information

4.58. Planning a new system of SUTs is generally linked to a benchmark year for which the most
important areas of the economy are covered by censuses and surveys. This is especially important
when policy decisions are based on the levels of the figures, for example, the level of gross national
income (GNI). Some detailed data sources are collected at more or less regular intervals and will
not all be available for the benchmark year. Hence, figures that do not relate to the actual year will
need to be corrected for the changes that have taken place between the reference period of the data
and the period for which they are being used. This can be done using indicators for value or volume
and price indicators.

4.59. When a balanced benchmark SUT exists, the compilation of SUTs for following (or
previous) years will usually be considerably easier. It is possible to use information on the
structures of the benchmark table to fill the gaps between those cells for which no new source data
are available or to extrapolate as appropriate. For example, input structures will change over time
and information on new input structures can then be used as it becomes available, replacing the
extrapolated information. Taking into account that the other estimated structures are subject to
uncertainty, it may be sufficient to review them at intervals of a few years, one at a time. Even
when such structures are reused from the previous year, they will change over time as a result of
balancing the SUTs. This could happen even more rapidly with the impact of globalization and the
development of new products.

(f) Features of GDP in volume terms

4.60. When looking at the change in the economy over time, the main concern is often whether
more goods and services are actually being produced now than at some time in the past. With
productivity, however, the point of interest is whether this output is increasing relative to the
inputs.

4.61. Over time, changes in current price GDP show changes in the monetary value of the
components of GDP. These changes in value can reflect changes in both price and volume, making

114
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

it difficult to establish how much of an increase in the series is due either to increased activity in
the economy (volume change) or to an increase in the price level. For productivity measures, only
volume changes are used. It is therefore useful to measure GDP in volume terms (preferably in
previous years’ prices), meaning that price effects are excluded, including in current prices. In
most cases, the revaluation of current price data to remove price effects (known as deflation) is
carried out by using price indices such as component series of the retail prices index or producer
price index, to deflate current price series at a detailed level of disaggregation.

4.62. At the international level, constant price estimates and chain-linked volume measures are
two common measures for volume change of GDP. Under the constant price method, a certain
year is selected as the base year; constant price volume estimates of GDP for subsequent years are
the aggregation of its components computed by multiplying the price of each component in the
base year by its volume in the current year, and the real growth is derived from the comparison of
constant price volume estimates at different years.

4.63. Under the chain-linked volume measures method, the annually reweighted chain-linking
approach is adopted to compile the volume measures of GDP and its components. First, the volume
estimates of major components of GDP in the current year are revalued at preceding year prices,
which in practice are calculated by deflating the current price values of subcomponents by the
relevant price indices. Second, the short-term volume indices for different years, calculated by
dividing the volume estimate of GDP from the initial step by the current price GDP in the previous
year, are chain-linked to a selected reference year in order to obtain a continuous time series of the
chain volume indices of GDP and its components. The chain-linked volume index series can be
converted into the chained monetary series by multiplying the chain-linked volume index by the
current price value in the reference year.

4.64. For some series, price indices for particular goods and services are used to deflate the
current price series, such as components of the following:

• Consumer price index (CPI)

• Retail price index (RPI)

• Producer price index (PPI)

• Corporate services price index (CSPI)

• Import prices

• Export prices

4.65. The process known as “double deflation” is the preferred method for the estimation of
GVA in volume terms. This is achieved by deflating the value of output and the value of
intermediate inputs separately to get corresponding volume measures, and then subtracting the

115
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

latter from the former. This double deflation approach means that an industry’s total output is
deflated by the price of its primary and secondary output, while each intermediate input is deflated
by its own price index.

4.66. This is in contrast to the single deflation method, whereby GVA in current prices is deflated
directly using an output-based deflator to arrive at GVA estimates in volume terms. The single
indicator volume estimates can also be derived in other ways, for example, by deflating output
with output price indices, assuming a constant GVA to total output ratios from the base year, or
using volume indicators directly. This direct price deflation of GVA is not recommended by the
SNA when using a single indicator method.

4.67. The SUTs offer a major advantage and a natural framework that enables double deflation
to be applied in a coherent and consistent manner across the national accounts.

4.68. Chain-linked volume measure series are expressed as index numbers in which the series
are simply scaled proportionately to a value of 100 in the reference year. These index numbers are
volume indices of the so-called “base weighted” or “Laspeyres” form.

4.69. Aggregate price indices are of the “Paasche” or “current-weighted” form. They are
generally calculated indirectly by dividing the current price value by the corresponding chained
volume measure and multiplying by 100. Examples are the GDP deflator and the households’
consumption deflator.

4.70. Value indices are calculated by scaling current price values proportionately to a value of
100 in the reference year. By definition, such a value index, if divided by the corresponding volume
index and multiplied by 100, will give the corresponding price index.

4.71. From the point of view of production, GDP at market prices is at best estimated with
reference to annually compiled SUTs both in current prices and in previous years’ prices. The
SUTs are compiled in previous years’ prices in order to achieve an accurate breakdown of value
changes in subsequent years according to volume and price changes.

4.72. The base-year table provides the specific weights for each industry and product, used in
the index formulae by which the price data are aggregated.

4.73. The great statistical benefit of a system based on previous years’ prices is the fact that the
weights in the index formulae are always up-to-date, thus reflecting the structure of the recent past,
and in turn optimizing the quality of the GDP growth rates in volume terms for more recent periods.

(g) Choice of index number formulae

4.74. In order to calculate price and volume measures, a number of methodological choices have
to be made, for example:

116
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Which index number formulae will be applied

• Whether a fixed base year or an annually changing base year will be applied

4.75. Different index formulae may be applied using different weighting schemes. It is beyond
the scope of this manual to discuss in depth the theoretical and practical considerations relating to
this choice. Chapter 15 of the 2008 SNA and Eurostat (2016) provide much more detail on the
choice of index formulae.

4.76. Economic theory suggests that an index formula that assigns equal weight to the current
year and the base year is to be preferred. This is one of the reasons why the SNA has a preference,
albeit not a strong one, for the so-called “superlative” indices, like Tornqvist and Fisher.

4.77. Although the superlative indices have a number of attractions, it should be noted they also
have notable disadvantages:

• Superlative indices are demanding in their data requirements and will increase the work
burden significantly.

• Superlative indices are less intuitive than Laspeyres and Paasche indices.

• Superlative indices are not additively consistent, which is a serious constraint when applied
in an accounting framework.

• Values change does not always equal volume change times price change.

4.78. From a practical point of view, a number of requirements can be imposed on the index
numbers:

• The applied index formulae should be a good approximation of the changes obtained by
the superlative indices.

• A change in value must be divided into a price change and a volume change without a
residual.

• Values in volume terms for aggregates should equal the sum of values in volume terms of
constituent parts, applying the same index formulae.

• In addition, it is sensible for the index formulae to be relatively straightforward and easy
to interpret for users.

4.79. The last three requirements can only be met with the application of the Laspeyres volume
index formula and the Paasche price index formula. The formulae underpinning these indices are
shown below

117
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Laspeyres volume index Paasche price index

𝑞𝑞𝑡𝑡 ∑ 𝑝𝑝𝑡𝑡 𝑞𝑞𝑡𝑡 ∑ 𝑝𝑝𝑡𝑡 𝑞𝑞𝑡𝑡
∑ 𝑝𝑝0 𝑞𝑞𝑡𝑡 ∑ 𝑝𝑝0 𝑞𝑞0 𝑞𝑞0 𝑃𝑃𝑝𝑝 = =
𝐿𝐿𝑞𝑞 = = ∑ 𝑝𝑝0 𝑞𝑞𝑡𝑡 ∑ 𝑝𝑝 𝑞𝑞 / 𝑝𝑝𝑡𝑡
∑ 𝑝𝑝0 𝑞𝑞0 ∑ 𝑝𝑝0 𝑞𝑞0 𝑡𝑡 𝑡𝑡 𝑝𝑝
0

where 𝑝𝑝 is the price and 𝑞𝑞 is the quantity.

4.80. The characteristic for the Laspeyres volume index is that the volume changes of individual
goods are weighted with the value of the transaction concerned in the base year..

4.81. The characteristic for the Paasche price index is that the price changes of individual goods
are weighted with their value of the transaction concerned in the current year. The deflated values
derived with this index formula combination can easily be explained as values in prices of the base
year.

4.82. It can be easily shown that the decomposition of value changes, in terms of volume and
price changes, do not have a residual.

∑ 𝑝𝑝𝑡𝑡 𝑞𝑞𝑡𝑡
𝑃𝑃𝑣𝑣 = = 𝐿𝐿𝑞𝑞 × 𝑃𝑃𝑝𝑝
∑ 𝑝𝑝0 𝑞𝑞0

4.83. The deflation of values in current prices by using a Paasche price index gives:

∑ 𝑝𝑝𝑡𝑡 𝑞𝑞𝑡𝑡
𝑃𝑃𝑣𝑣 /𝑃𝑃𝑝𝑝 = � 𝑝𝑝𝑡𝑡 𝑞𝑞𝑡𝑡 / = � 𝑝𝑝0 𝑞𝑞𝑡𝑡
∑ 𝑝𝑝0 𝑞𝑞𝑡𝑡

4.84. This illustrates that the deflated aggregate equals the sum of deflated components, which
means that additivity in the SUTs for volume estimates is assured. The use of Laspeyres volume
indices and Paasche price indices ensures that the current price identities also hold in volume terms.
This also means that, after balancing, for every product, the total supply equals total use, and for
every industry, the total output equals total intermediate consumption plus gross value added and,
in volume terms, this is:

Total supply in volume terms equals Total use in volume terms

Total output in volume terms equals Total intermediate consumption in volume terms

plus Total value added in volume terms

(h) Choice of the base year

4.85. By applying the Laspeyres volume index number formula, the volume changes are
weighted with the values of the concerning transaction in a base year. The question arises which
year should be chosen as the base year. Generally speaking, there is a choice between a fixed base
year and a changing base year.

118
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.86. With the method of fixed weights for a series of years, the weights are derived from a single
year in the past. An advantage of this method is that, with long time series of values at prices of
the base year, the deflated components of aggregates add up exactly to the deflated aggregate. A
serious disadvantage, however, is that volume changes of aggregates are calculated with outdated
weights. This disadvantage is especially severe when relative prices change rapidly and, as a result,
economic growth can be significantly misrepresented. In addition, the disappearance of products
(such as vinyl records, cathode ray tube televisions, and so forth) or the appearance of new products
(such as mobile telephony, pharmaceutical tablets, iPads, and the like) can lead to notable
distortion in the estimates of economic growth. Even a fixed base year has to be changed every
five years, and then all previously published real growth rates will be revised – this change is often
not welcomed by users.

4.87. Applying a changing base year means that the weights are updated every year and are
usually derived from the previous year. Since those weights are more up-to-date, a better
approximation of the volume changes is obtained than with the method of using fixed weights.
The time series can be obtained by multiplying separately estimated year-to-year volume indices
– this is called “chaining”.

4.88. An important advantage of the chaining method is that the above-mentioned

misrepresentation of growth rates is avoided. In fact, chain-linked volume measures provide a
more reliable measure of volume growth, provided individual prices and quantities tend to increase
or decrease steadily over time. They also have a key disadvantage, however, in that when the time
series are in prices of the previous year, the deflated components of an aggregate then no longer
add up exactly to the deflated aggregate. As a result, mathematical discrepancies will appear that
cannot be removed without distorting the underlying actual volume and price movements.

(i) Documentation

4.89. As the compilation of SUTs is a complex process, a thorough documentation of the basic
data and the methods used, the problems encountered, the solutions applied, and the results
achieved is highly recommended. Such an annual (or quarterly, if appropriate) inventory is not
only worthwhile for purposes of publication but also for internal use in the compilation process
itself and future exercises. When SUTs have to be balanced, information on, in particular, the
sources and methods of estimation for each single supply and use element is needed to evaluate
and analyse industry and product imbalances. The documentation helps to evaluate the data quality
and outline the strategy for balancing. The balancing steps should also be documented, of course,
in order to avoid repeating changes and the destruction of already balanced data.

4.90. Documentation of the various compilation steps can also point to missing data issues and
problems of basic data quality. It is important that such findings are used as feedback to primary
statistics and to identify priorities in improving the compilation methodology. A documentation

119
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

system for the SUTs should be seen in the frame of the overall documentation for the system of
national accounts data.

4.91. Producing a balanced set of SUTs for several years is like solving a puzzle. At first, all
macroeconomic data, survey results, census results and other valid economic data on supply and
use of products in the economy have to be collected. In a second step, missing data have to be
estimated on the basis of harmonized methodologies and documented procedures. In the third and
final stage, the balancing of the SUTs system generates a consistent set of macroeconomic
variables in current prices and in previous years’ prices. Thus the documentation of all stages of
the compilation of hard data in terms of sources and soft data in terms of estimates and adjustments,
ideally for each cell of the SUTs system, is key. In turn, a quality assessment using clear criteria
for each cell can also be developed over time.

4.92. Links between survey data and final national accounts data should be maintained in the
system, in particular separately recording documentation on the survey data, coverage adjustments,
conceptual and valuation adjustments, quality adjustments and balancing and coherence
adjustments. Analyses of these types of adjustments over time and over successive exercises can
also help to highlight any biases, incoherence in data sources, and other shortcomings and, in turn,
help in developing priorities and strategies for further improvements and investment.

C. Collect phase

4.93. The collect phase of the GSBPM (see figure 3.2) consists of all the activities concerning
the collection and gathering of all necessary information (data and metadata), using different
collection modes (including extractions from statistical, administrative and other non-statistical
registers and databases), and their storage in an appropriate structured environment for further
processing. While it may include the validation of dataset formats, it does not include any
transformations of the data themselves, as these are all effected in the “process” phase. For
statistical outputs produced regularly, this phase occurs in each iteration.

4.94. Generally, the compilation of SUTs and IOTs relies on the data sources used for calculating
GDP according to the production, income and expenditure approaches. An overview of the range
of data sources generally used is provided in Box 4.3.

4.95. It should be noted that this situation is reversed in some countries, such as Chile and Japan,
where the requirements of SUTs (and IOTs) are first clearly defined and then the surveys (regular
and ad hoc) are undertaken to collect data to meet these requirements. This may be ideal for SUTs
(and IOTs), but it may also be more costly.

4.96. The compilation of SUTs is data demanding and in principle could be based on many
sources that could help to populate the SUTs. It is strongly recommended, however, for purposes
of timeliness, that more data sources should be regularly available, and that these are also based
on official statistics, preferably using the same business register sampling frame. It is also worth

120
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

noting that many sources can provide data feeding into both the supply table and the use table for
the respective industries and products, as appropriate. For example, annual structural surveys
provide details on both sales and purchases.

Box 4.3 Data sources generally used

Monthly, quarterly, annual, regular and ad hoc business surveys based on the business register sampling frame:
• Monthly business surveys covering production, distribution, construction and service industries and
financial information such as employment and turnover
• Monthly surveys collecting price-related details such as producer (input and output) prices covering
all industries (for example, manufacturing and services), retail and consumer price indices, import
prices, export prices, and earnings and wage prices
• Quarterly business surveys covering areas like capital expenditure, inventories and profits
• Annual business surveys covering structural business statistics for each industry:
o Detailed information covering such variables as turnover, employment, purchases, capital
expenditure, inventories, holding gains, taxes and subsidies, and others
o Detailed sales by type of product
o Detailed purchase data by type of product, covering intermediate-type products and capital-
type product separately
• Range of quarterly and annual surveys covering financial services, in particular financial assets and
liabilities, international trade in services and foreign and direct investment
• Economic censuses every three to five years.
Household-based surveys:
• Living costs and food surveys – collecting details on expenditure by households
• International passenger surveys – expenditure by residents abroad and non-residents’ expenditure in
the domestic economy
• Labour force surveys – collecting details on employment including hours worked
Administrative data:
• Census – population estimates for grossing purposes for use in household-based surveys
• Pay and profits data relating to tax and employment records from collecting government departments,
along with data covering self-employed incomes
• VAT payments data and turnover subject to VAT by industry (and by product where differential rates
exist) from the tax collecting departments
• Imports and exports of goods data collected by customs
Other government departments:
• Agriculture industry data from the agricultural ministry
• Banking industry data from the central bank
• Data on government incomes and expenditures from the finance ministry – expenditure will need to
be split between the individual consumption and collective consumption categories. This source can
also provide the full range of taxes receivable and subsidies payable
Other sources:

121
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Company annual report and accounts (in general, the direct use is limited, however these are often
used for public utility companies like electricity, gas, water, postal services, telecommunications and
others, in many countries)
• Regulatory bodies’ accounts
• Insurance data from the insurance industry regulators
• Airline data from the airline industry regulators
• Financial details from company websites which supplement the company annual report and accounts

4.97. The compilation of SUTs, whether carried out as an integral part of the compilation of
national accounts (as recommended in this Handbook) or as a separate compilation from that of
the national accounts, is based to a large extent on the same data sources as those used for the
national accounts. Figure 4.1 expands on figure 2.3 in chapter 2 to show the typical data sources
used for the compilation of national accounts, which also feed into the compilation of SUTs and
IOTs. Often the same source can provide data feeding into more than one of the approaches to
measuring GDP. For example, the agricultural data from agricultural departments often feeds into
all three approaches to measuring GDP, thereby ensuring natural consistency and coherence of the
data used in SUTs.

4.98. In general, there is a strong correlation with the level of detail used within the SUTs and
the quality of the product balances and the aggregates. The more disaggregated the level of
industries and products, the higher the degree of matching of individual products in terms of
allocation of uses, prices and other characteristics and hence the better the quality. If the product
level is too aggregated, the individual products may be too broad and heterogeneous, and therefore
of lesser quality.

122
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure 4.1 Overview of SUTs and IOTs as part of the SNA compilation

Business register

System of national accounts

Supply and use tables Accounts

Supply and
Business Production Income Expenditure Consumer
use tables at Valuation
census approach approach approach price
purchasers' matrices statistics
GDP GDP GDP
prices (unbalanced)
Annual (unbalanced) (unbalanced) (unbalanced)
business
(unbalanced) Producer
price
survey
statistics

Construc- Balancing Capital

tion survey expenditure
survey

Income Supply and Material

survey use tables at Valuation input
Single estimate of GDP statistics
purchasers' matrices
prices (balanced)
Household (balanced)
expendi-
(balanced) Balance of
ture survey payments

Govern- Sector accounts

Foreign
ment 1. non-financial corporations trade
budget Supply and use tables at 2. financial corporations Total Rest of statistics
basic prices 3. general government economy the world
4. households
Financial
5. NPISH Production
statistics
statistics

Monetary
Agriculture
statistics Goods
Product-by- Industry-by- Distribution Accumu- statistics
and Production
product industry and use of lation
services account
Employ- IOTs IOTs income accounts
ment account Population
statistics
accounts census

1. Typical data sources

(a) Structural surveys

4.99. Many countries rely on annual structural type surveys which fit in naturally to the needs of
the SUTs framework. One of the advantages of such a set-up is that it allows the collection of a
range of variables from a single source – the statistical unit – thereby ensuring the consistency and
coherence of each variable and across variables. Thus, the employment data will be consistent with
the turnover data and, in turn, with the derived GVA data.

123
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.100. Where annual structural surveys exist, data for a number of key variables are collected with
more detail supplemented from other surveys (for example, detailed purchases data). These
variables can include:

• Sales of goods and services (and an appropriate product breakdown)

• Purchases of goods and services (and an appropriate product breakdown)
• Purchases and sales of goods for resale without any further processing (this helps to
produce the trade margins by type of product)
• Changes in inventories (a split between materials and fuels, work-in-progress and finished
goods not sold is required, as the first category affects intermediate consumption and the
latter two categories affect output)
• Capital expenditure (and an appropriate product breakdown)
• Employment costs (and an appropriate breakdown)
• Taxes on products and production (covering business rates, excise duties and others)
• Subsidies on products and production (covering agriculture, transport and others)
• Areas such as research and development, and international trade in services, which link to
other related surveys
4.101. Often data compiled to respond to statistical regulations are used in the compilation of
SUTs. For example, in the European Union, statistical regulations for member States cover various
business statistics derived from monthly, quarterly and annual surveys providing short-term
indictors and structural business statistics. Data from many of these sources also feed into the
compilation of SUTs. The Handbook on the Design and Implementation of Business Surveys
(Eurostat, 1998), provides extensive detail on the conduct of such surveys.

(b) Administrative data

4.102. Administrative data constitute a key data source for the compilation of both quarterly and
annual national accounts and, in some countries, administrative data may be the main data source.
The use of administrative data is growing under the influence of a number of factors, such as good
coverage, declining resource impact on national statistical offices and lessening of the response
burden.

4.103. Administrative data have statistical strengths and weaknesses vis-à-vis sample surveys.
Apart from the low cost of obtaining administrative data, their major strength is that they
commonly have complete – or nearly complete – coverage of the fields to which they relate. As a
result, there are no sampling errors and any non-sampling errors, such as those arising from an
out-of-date business register and inadequate new business provisions, are scarce and minor.

124
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.104. The weaknesses of administrative data arise from the fact that they are by-products of
administrative systems which are not generally designed to meet the needs of the national accounts.
Examples of these weaknesses include the following:

• Available data do not meet national accounting definitions (for example, wages rather than
compensation of employees, or a measure of depreciation that differs from the national
accounting concept of consumption of fixed capital).
• Purchases registered in the VAT system will usually include both purchases for
intermediate consumption and for gross capital formation.
• Data are not recorded on an accrual basis (for example, exports and imports from customs
are recorded as they cross the customs frontier and not when they change ownership).
• Data are incomplete (for example, movement of oil rigs in and out of territorial waters are
excluded from customs data).
• Data may not be disaggregated in a desirable way (for example, government expenditures
may not distinguish between wages and intermediate consumption or new motor vehicle
registrations may not distinguish between household and business use).
• Administrative data may be untimely (for example, company tax data).
• Administrative data may undergo change as a result of a change in policy.

(c) Business and company accounts-based statistics

4.105. The values of outputs, inputs, gross capital formation and other elements have their
counterparts in business accounts or government accounts, but the concepts used in business
accounting often do not consistently follow national accounts definitions. Further details on this
issue may be found in United Nations (2000b) and Mahajan (2013). A few examples are listed
below:

• Differences between concepts: financial intermediation services indirectly measured

(referred to as FISIM), insurance services, and others.
• Change in inventories: different valuations, correction for holding gains and losses, and
others.
• Distinction between intermediate consumption and capital formation: acquisitions of
machinery and equipment included in current expenses, and others.
• Distinction between intermediate consumption and compensation of employees: fringe
benefits, links to own account production, and others.
• Use in business accounts of a time frame that does not follow the calendar year: often
accounts that are closest to the calendar year can be used as an approximation for the annual

125
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

values but it may be appropriate to correct figures for some big enterprises that have
accounts that cover other periods or may have large seasonal patterns, such as the gas
supply industry.
• Use in government accounts of a fiscal year that differs from the calendar year. It can be
misleading to use the information for the nearest fiscal year as if it were identical to
calendar year data. Annual data can be calculated by weighting together data for the two
fiscal years that overlap the actual year. A better method is, however, to use quarterly or
monthly information to split the data from each fiscal year into the shares belonging to
different calendar years where such information exists.

(d) VAT-based statistics

4.106. The VAT system usually provides statistics on those units that are covered by the business
register, which will also use VAT-registered businesses as a source. The business register generally
includes the units that collect VAT on all or a part of their turnover and those that can deduct VAT
on all or part of their purchases (capital or current).

4.107. The VAT-based statistics may exist in published form but, even when not published, they
can usually be obtained from the authorities that are responsible for collection of VAT, with
appropriate service level and confidentiality agreements in place. Typically, this source will
contain information on VAT-liable, zero-rated and VAT-exempt turnover and also on deductible
and non-deductible purchases with a classification by industries.

4.108. The VAT-based statistics tend to be available on both a quarterly and an annual basis, and
are usually available shortly after the reference period. A correction based on the final dates of
payment may be necessary if the statistics show payments instead of accruals.

4.109. There are pitfalls to bear in mind when using VAT-based statistics for national accounts
purposes:

• The concepts used in VAT-based statistics are different from those used in national
accounts. As VAT-based turnover covers sales of products from own production, sales of
traded goods and sales of used capital equipment, the VAT purchases cover purchases of
goods and services for use as inputs, goods intended for resale and also purchases of capital
equipment. Before these figures can be used in national accounts, the VAT purchases must
be split into the different shares based on details payments.
• VAT-based statistics do not show figures for units with activities that are not VAT-liable
and they may not include units with turnover below certain thresholds. Such thresholds
differ from country to country but will often exclude a significant share of the smaller
enterprises. The informal economy can – by its very nature – be assumed not to be included
in the VAT-based register sources.

126
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• The industry classification of the VAT-based units may not be the same as the classification
used in national accounts.
• For practical reasons, the units accepted in VAT-based statistics may be enterprises, kind-
of-activity units, establishments, or even conglomerates of enterprises which are allowed a
joint registration for payment of VAT.
4.110. Despite such caveats, the VAT-based statistics may still be the most reliable data source
for the size of some industries that are poorly covered by other sources. The figures from VAT-
based statistics should, however, be seen as the minimal size of the industries in question. It can
be necessary to add values for units below the threshold values and VAT-exempt units, including
those operating in the informal and hidden economy.

(e) Missing data

4.111. If certain data are not available in the official statistical system, the first option would be
to check whether such data are available outside official statistics. One such example is when
intermediate data on advertising costs are not available as a separate item in the business surveys.
In this case, one possible source could be data from relevant trade associations observing the
advertising market. Despite the fact that the data are often not sufficiently comprehensive, or the
classifications differ from the official classifications, so that the data do not fully conform to the
required concepts, these data certainly give a good indication of the advertising market across the
various industries.

4.112. There may be a full set of data observations for a period, and then again for a later period.
Various modelling techniques exist for generating estimates for the intervening periods to populate
the SUTs, such as the basic Holt-Winters approach (Holt, 1957; Winters, 1960). When balancing
SUTs, however, these estimates should be treated as being of much lower quality than the more
reliable and up-to-date estimates.

4.113. Furthermore, if no specific data are available, the expert advice of chambers of commerce,
trade associations, research institutes or other similar organizations could be useful.

4.114. In certain industries, a single company or a few companies are the big players in that
market. Those could also be specifically contacted for expert input or to request some of their
internal data on a confidential basis. For example, telecommunication companies may provide
their revenue data by type of customer; supermarket chains may be asked to provide data on their
sales by products; major railway companies may provide data on the goods transported, and so on.

4.115. Annual company reports and accounts, publications of regulatory bodies and trade
associations, and internet company websites are very useful sources of financial data for businesses
and households.

127
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

4.116. Certain estimates can be based on the identities and coherence of the SUTs framework.
This holds true for the application of the product flow method, where detailed supply data are used
to estimate certain use data. The product flow method basically applies fixed allocations which
will need to be reviewed each year. The method should be applied with great caution in populating
SUTs and will depend on the level of product detail. The collection of primary data from various
sources with data confrontation provides the best approach to the populating of SUTs and the
attainment of good quality results.

(f) Exhaustiveness: methods of grossing up

4.117. Statistical sources usually exclude units with employment or turnover below certain
thresholds, while national accounts data should include estimates for these missing units. The
methods used for grossing up will typically be based on an estimate of employment in the excluded
units, and assumptions on output, input, capital formation and other factors by employee. These
assumptions should as far as possible reflect the conditions in comparable units but when the small
units are not covered by source data, the grossing up procedure will necessarily add to the
uncertainty of the estimated totals. It can also be expected that the structures of outputs and inputs
of small units are somewhat different from those found in the units covered by collected data.

4.118. There is considerable interest in the phenomenon of the non-observed economy. This term
is used to describe activities that, for one reason or another, are not captured in regular statistical
enquiries, because they are underground, illegal or informal, consist in household production for
own final use, or simply because of deficiencies in the basic data collection system. Guidance in
this regard is provided in the handbook on measuring the non-observed economy (OECD, ILO and
Interstate Statistical Commission, 2002) and in the manual on the informal sector (ILO, 2013).

4.119. In countries where a significant share of total output and input is found in the informal
economy, it can be appropriate to conduct specific surveys of this activity. With a view to
confronting data on the supply and use of labour, useful information on this subject may actually
be found in population censuses, household budget surveys or labour force surveys. In this respect,
the SUTs framework, in which available source data are combined and balanced, offers the most
promising means of arriving at exhaustive estimates of economic activity.

128
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Part three

129
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Chapter 5. Compiling the supply table

A. Introduction

5.1. The first step in the compilation of SUTs and IOTs is the construction of an initial and
unbalanced version of the supply table. The values entered into the tables should reflect, as far as
possible, all available knowledge and data on the product structure of each column, although many
values may need to be changed when the SUTs system is balanced. This applies to estimated totals
and also to the values of supply of specific products.

5.2. Before balancing takes place, the estimates for domestic supply, imports of goods and
services should be checked for credibility, and if necessary adjusted as appropriate. These will
then form the starting point for the balancing process.

5.3. This chapter focuses on the steps and data sources needed to compile this initial,
unbalanced version of the supply table. Section B provides a more detailed overview of the
structure of the supply table. Section C focuses on the compilation of the domestic output table
and the necessary compilation steps and section D focuses on the compilation of the imports of
goods and services. Annex A to chapter 5 provides an example of a questionnaire on the collection
of sales of goods and services, inventories of goods and trade-related data.

B. Structure of the supply table

5.4. The supply table shows the supply of goods and services for a given period of time by type
of product of an economy and distinguishes between the output of domestic industries and imports
by type of product. The supply table is generally compiled first at basic prices, reflecting the
valuation of the data sources. As illustrated in Table 5.1, the supply table at basic prices consists
of two main parts: domestic output and imports of goods and services.

5.5. The domestic output matrix contains information on the supply of products by the different
industries. The column for the imports of goods and services contains information on the total
imports by product. The matrices for domestic output and imports of goods and services have the
same row structure, defined by product categories. This structure allows for the horizontal
aggregation of all the elements and the transition from total supply of products at basic prices to
total supply at purchasers’ prices.

131
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Table 5.1 Numerical example of a supply table at basic prices

Millions of euros
INDUSTRIES
Total supply
Trade, Finance and Output Imports at basic
Agricul- Manufac- Construc- Other
transport and business at basic prices
ture turing tion services
communication services prices
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Agriculture (1) 8 782 0 0 0 0 0 8 782 3 271 12 052
Manufacturing (2) 796 182 982 643 1 808 133 44 186 405 124 590 310 995
Construction (3) 83 961 43 060 734 255 179 45 272 563 45 835
PRODUCTS

Trade (4) 1 4 773 311 54 204 640 257 60 187 600 60 787
Transport (5) 13 465 66 25 538 128 125 26 335 8 150 34 485
Communication (6) 160 1 781 139 43 912 1 253 982 48 228 6 234 54 463
Finance and business services (7) 29 8 902 698 7 588 106 909 3 381 127 508 7 061 134 569
Other services (8) 3 85 13 1 053 143 74 346 75 643 824 76 467
Total (9) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 151 293 729 653
Adjustments

CIF/FOB adjustments on imports (10) 0 0 0 0 0 0 0 - 97 - 97

Direct purchases abroad by residents (11) 0 0 0 0 0 0 0 6 675 6 675

Total (12) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 157 871 736 230
Total of w hich:
Market output (13) 9 763 195 916 41 462 127 401 88 330 18 116 480 989 0
Output for ow n final use (14) 104 4 029 3 468 2 134 19 890 2 670 32 295 0
Non-market output (15) 0 4 0 5 302 1 241 58 528 65 075 0

Table based on 2011 figures from Austria

5.6. The supply table at basic prices is then transformed into the supply table at purchasers’
prices, through the addition of valuation adjustments represented by valuation matrices covering
trade margins, transport margins, taxes on products and subsidies on products. Table 5.2 shows
the valuation adjustments which are added to the columns of the supply table at basic prices to
arrive at the total supply of each product at purchasers’ prices.

5.7. The first step in the compilation of an initial version of the supply table therefore involves
the compilation of data for total domestic output at basic prices and imports valued at CIF prices
aggregated to total supply at basic prices. The second step involves the compilation of trade and
transport margins, taxes on products less subsidies on products which are used to convert the total
supply of products at basic prices to the total supply of products at purchasers’ prices.

5.8. The data in the domestic output matrix are valued at basic prices, which is the amount
receivable by the producer from the purchaser for a unit of a good or service produced as output
minus any tax payable, plus any subsidy receivable by the producer as a consequence of its
production or sale. The value of output of goods excludes any transport charges invoiced separately
by the producer.

5.9. Data on imports by product from foreign trade statistics are usually valued at CIF prices.
In the 2008 SNA and BPM 6, however, in which total imports of goods are valued FOB, an extra
row has to be added for the CIF/FOB adjustments on imports, in order to reconcile the different
valuations. These adjustments are shown in row (10) of Table 5.1 and explained in detail in section

132
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

D. In addition, a further adjustment is added in the supply table to account for direct purchases
made abroad by residents: this is shown in row (11) in Table 5.1.

5.10. These adjustments in the supply table (rows (10) and (11) in Table 5.1 and Table 5.2) have
corresponding entries in the use table (rows (10) and (11) in table 6.1) under the columns for
exports and final consumption expenditures by households. It should be noted that some countries
do not show these estimates in separate rows but consolidate the values across the product groups
in the respective columns, thereby providing a different product balance.

Table 5.2 Supply table at basic prices, including a transformation into purchasers’ prices
Millions of euros
INDUSTRIES
Output Total supply
Trade, Finance and at basic Imports at basic
Agricul- Manufac- Construc- Other
transport and business prices prices
ture turing tion services
communication services

(1) (2) (3) (4) (5) (6) (7) (8) (9)

Agriculture (1) 8 782 0 0 0 0 0 8 782 3 271 12 052
Manufacturing (2) 796 182 982 643 1 808 133 44 186 405 124 590 310 995
Construction (3) 83 961 43 060 734 255 179 45 272 563 45 835
PRODUCTS

Trade (4) 1 4 773 311 54 204 640 257 60 187 600 60 787
Transport (5) 13 465 66 25 538 128 125 26 335 8 150 34 485
Communication (6) 160 1 781 139 43 912 1 253 982 48 228 6 234 54 463
Finance and business services (7) 29 8 902 698 7 588 106 909 3 381 127 508 7 061 134 569
Other services (8) 3 85 13 1 053 143 74 346 75 643 824 76 467
Total (9) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 151 293 729 653
Adjustments

CIF/FOB adjustments on imports (10) 0 0 0 0 0 0 0 - 97 - 97

Direct purchases abroad by

(11) 0 0 0 0 0 0 0 6 675 6 675
residents
Total (12) 9 867 199 950 44 931 134 837 109 461 79 314 578 360 157 871 736 230
Total of w hich:
Market output (13) 9 763 195 916 41 462 127 401 88 330 18 116 480 989 0
Output for ow n final use (14) 104 4 029 3 468 2 134 19 890 2 670 32 295 0
Non-market output (15) 0 4 0 5 302 1 241 58 528 65 075 0

Total VALUATION MATRICES Total supply

supply at at
basic Trade Transport Taxes on Subsidies purchasers'
VAT Total
prices margins margins products on products prices
(9) (10) (11) (12) (13) (14) (15) (16)
Agriculture (1) 12 052 1 926 274 329 57 - 107 2 479 14 532
Manufacturing (2) 310 995 48 838 2 540 13 175 7 866 - 49 72 370 383 364
Construction (3) 45 835 0 0 1 529 13 0 1 542 47 377
PRODUCTS

Trade (4) 60 787 - 52 341 0 575 11 0 - 51 755 9 032

Transport (5) 34 485 0 - 2 800 558 71 - 448 - 2 620 31 865
Communication (6) 54 463 1 493 9 3 375 217 - 34 5 059 59 522
Finance and business services (7) 134 569 0 - 22 2 706 2 159 0 4 842 139 411
Other services (8) 76 467 85 0 1 201 576 0 1 861 78 329
Total (9) 729 653 0 0 23 447 10 969 - 638 33 778 763 431
Adjustments

CIF/FOB adjustments on imports (10) - 97 0 0 0 0 0 - 97 - 97

Direct purchases abroad by

(11) 6 675 0 0 0 0 0 6 675 6 675
residents
Total (12) 736 230 0 0 23 447 10 969 - 638 40 356 770 009
Total of w hich:
Market output (13) 0 0 0 0 0 0 0 0
Output for ow n final use (14) 0 0 0 0 0 0 0 0
Non-market output (15) 0 0 0 0 0 0 0 0

Table based on 2011 figures from Austria

133
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.11. A distinction may be made in the SUTs between the three types of production: market
output; output for own final use; and non-market output. In the domestic output matrix, however,
these three categories of production are usually grouped together in the relevant industries and
shown in three supplementary rows for each industry. Thus government services are distributed in
the system according to the various activities in which the government is engaged: for example,
public administration services, education services, health services, recreation services, social
welfare services, and others, but are shown together with the corresponding market producers. For
example, health services provided by market and non-market producers (within the same industry)
are shown as a total. Furthermore, for some industries, the supplementary rows are useful for the
link with the institutional sector accounts.

5.12. Although the supplementary rows make it possible to split output by industry into the three
categories of output, there is no product dimension. Ideally, each industry should be shown
separately (also reflecting different structures and links between the output and the inputs) or
additional analyses produced for the user.

5.13. Imports of goods and services are classified by type of product. Since this table is designed
to show the total supply by type of products, the valuation of imports of goods should be
compatible with the valuation of the domestic production of goods. Imports by type of product are
therefore valued at CIF prices which are comparable with the domestic output at basic prices.
Although total imports of goods are valued at FOB prices, it is not easy to move the imports of
goods by type of product to CIF prices, as this depends upon the source data and the person
providing the transportation (see 2008 SNA, para. 14.77).

5.14. The addition of the two components – production and imports – gives the total supply of
products at basic prices.

5.15. The supply table at purchasers’ prices is obtained by adding various valuation matrices
(earned on both domestic output and imports) to the total supply at basic prices, thus enabling
movement from one valuation to another. The valuation matrices include:

• Trade margins
• Transport margins
• Taxes on products (with non-deductible VAT treated separately from other taxes on
products)
• Subsidies on products (which are deducted)
5.16. It should be noted that, when the supply table is shown with the final column summing to
purchasers’ prices, it is referred to as the supply table at purchasers’ prices. This is actually just
the supply table at basic prices with the addition of the valuation columns. The production and
import sections of the supply table have not been changed and remain valued at basic prices.

134
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.17. The task of compiling SUTs is a highly integrated process. This is particularly true for the
estimation of the valuation vectors and matrices, where it is often necessary to rely also on
estimates from the use table side in order to obtain the valuation vectors entered into the supply
table. Figure 5.1 provides an overview of how the valuation matrix in the supply table is linked to
a sequence of valuation matrices in the use table. This figure also demonstrates the
interconnections between the valuation matrices linking the supply table and use table. The
estimation of the valuation matrices, considering both the supply table and the use table, is dealt
with in chapter 7.

5.18. The rest of this chapter focuses on the compilation of the supply table at basic prices.

Figure 5.1 Link between valuation matrices in the supply table and the use table
Supply table at purchasers' prices Use table at purchasers' prices
INDUSTRIES Total VALUATION Total supply INDUSTRIES FINAL USE Total use
Output supply at Taxes Subsidies at Final Gross at
Agricul- Other Imports Trade Transport Other
.. at basic basic VAT on on Total purchasers' Agriculture .. Total consumption capital Exports Total purchasers
ture services prices margins margins prices services ’ prices
prices product products expenditure formation
(1) .. (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (1) .. (6) (7) (8) (9) (10) (11) (12)
Agriculture (1) Agriculture (1)
Manufacturing (2) Manufacturing (2)
Construction (3) Construction (3)
PRODUCTS

PRODUCTS
Trade (4) Trade (4)
Transport (5) Transport (5)
Communication (6) Communication (6)
Finance and Finance and business
(7) (7)
business services services
Other services (8) Other services (8)
TOT (9) Total (9)
Compensation of
(10)
employees
Other taxes on production (11)
Consumption of fixed
GVA

(12)
capital
Net operating surplus/Net
(13)
mised income
Total (14)
Total input at basic prices (15)

Subsidies on products
Agriculture (1)
PRODUCTS

: :
: :
Other services (8)
Total (9)

Taxes on products
Agriculture (1)
PRODUCTS

: :
: :
Other services (8)
Total (9)

Value added tax (VAT)

Agriculture (1)
PRODUCTS

: :
: :
Other services (8)
Total (9)

Transport margins
Agriculture (1)
PRODUCTS

: :
: :
Other services (8)
Total (9)

Trade margins
Agriculture (1)
PRODUCTS

: :
: :
Other services (8)
Total (9)

C. Domestic output

1. Structure of the domestic output table

5.19. The first and most elaborated part of the supply table is the domestic output matrix. This
records data on the production of the economy classified along two dimensions: the rows represent
the type of products (based on CPC Version 2.1) and the columns represent the different industry
groupings (based on ISIC Rev. 4). Thus the rows show the domestic output matrix, a single product
by producing industry, and the columns show all the products produced by a single industry.

135
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Although this is consistent with CPC and ISIC, countries may however use different and more
detailed classifications, for example, those which reflect country-specific activities.

5.20. The domestic output matrix reflects the principal and secondary products of industries
including by-products. It is the principal activity of the statistical unit that determines its
classification to a specific industry. In special cases where the domestic output matrix is square
(the number of products being equal to the number of industries), and the sequence of products
arranged to reflect the sequence of the industries (based on their principal activities), the principal
activity of an industry is reported on the diagonal of the matrix, while the secondary activities of
an industry are listed as off-diagonal entries.

5.21. In practice, however, it is common to have more products than industries. For this reason,
the production part of the supply table is usually a rectangular matrix with more rows than
columns, as shown in Table 5.1. This reflects the fact that it may be of more interest to specify, for
example, different kinds of agricultural crops, in the case of agriculture, and of less interest or
practical use to distinguish farms specializing in each possible type of crop. In this case, all the
crops would still form the principal output of agriculture, whereas, for example, the production of
wine or the construction of buildings for own use would be treated as secondary outputs of the
industry. The greater the level of product detail, the more scattered the entries will be around the
principal products. In these cases, it is not possible to observe directly the distinction of principal
products versus secondary products or production in the rectangular domestic output matrix.

5.22. Annex A to this chapter provides an extract of a survey questionnaire collecting data on
sales of goods and services by type of product, in addition to other variables by product, such as
opening stocks (inventories), closing stocks (inventories) and trade margins.

5.23. Even though the industry concept is already being applied in the national accounts, the
existing level of detail or precise delimitation should not be viewed as a constraint when compiling
SUTs and, in particular, when compiling benchmark tables. On the other hand, the way in which
statistical units are defined and classified in the business register and covered in basic statistics
represents a real constraint on the possible choices concerning industries in the SUTs. Even though
industries may, in the process of compiling SUTs, be redefined to some extent or otherwise
modified in terms of their basic statistics, the options are much more limited than the range of
choices available when it comes to deciding what product classification should be applied.

5.24. The choice of the level of detail for industries and products to be used in the SUTs must be
based on a thorough examination of the available statistics and considerations concerning the
advantages of using product details in balancing, in estimating margins and taxes on products by
uses, final uses by purpose, in volume estimates and in other applications. The general
recommendation, however, is to work with as much detail as possible, as any aggregation of basic
statistics will also entail a loss of information that could at some stage have contributed to the
overall quality of the balanced SUTs (on this issue, see also chapter 4).

136
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.25. It is also necessary to clarify any user requirements about the format of the final table,
including international reporting. In general, it would be an advantage to work at a more detailed
level than that warranted by current uses, in order to extract the maximum information from
available data sources and to be prepared for the emergence of new uses and for transformations
necessary to comply with future changes to economic activity and product classifications.

2. Primary statistics and data sources

5.26. The structure of economic entities varies from small enterprises engaged in one or a few
activities that are undertaken either at, or from, a single geographical location to large and complex
enterprises engaged in many different activities. These enterprises may be horizontally or
vertically integrated, and their activities may be undertaken either at, or from, many geographical
locations. The way in which producer units are defined, measured statistically, broken down or
aggregated is of fundamental importance when compiling SUTs.

5.27. In practice, compilers of SUTs will not deal with individual economic units but only with
the aggregates of units in the form of industries, usually based on current business statistics by
economic activity. To arrive at a full understanding of the role of these statistics in the compilation
of SUTs, it is necessary to assess the delimitation of units that influence the properties of the
industries.

5.28. The most important prerequisite for the collection of basic statistics is the business register
and the types of economic units that it holds. Ideally, business registers will contain two types of
unit: enterprise units and establishments.

5.29. Usually the enterprises form the core units of the business register, as they are easier to
identify and track on a current basis because of their legal status. The number of establishments
created depends on the register policy adopted (in other words, how many enterprises are
partitioned into establishments). Different geographical locations of the production units will be
one of the main criteria for subdividing an enterprise into several establishments.

5.30. In the collection of basic statistics, the enterprise will usually serve as the collection entity
and, to the extent that the enterprise is made up of several establishments, the enterprise will be
requested to report a separate range of statistics for each of those establishments. This has
implications for both the supply table and the use table, as some types of costs can only be reported
at the enterprise level, whereas it is possible to report all regular production costs for the individual
establishments.

5.31. As the large majority of enterprises are small or medium in size, and tend to engage in one
kind of activity only, the enterprise and the establishment units may be identical in these cases.
Large enterprises, however, which often contribute the bulk of the production of an economy, will
often cover different kinds of economic activity and therefore, in formal terms, be made up of
several establishment units.

137
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.32. It is important to note that many primary sources, such as enterprise surveys and production
surveys, which are used to collect data feeding into the domestic output matrix, often also collect
data at the same time through the same survey questionnaire but feed these into the use table (for
example, data on the industries’ input structures and gross fixed capital formation). This approach
provides the data feeding into the SUTs with a high degree of coherence and consistency.

5.33. The estimates of the domestic output matrix are usually based on two main types of
information sources: enterprise surveys and production surveys. Additional information, such as
administrative sources, company accounts and others, will also be used. Figure 5.2 provides a
simplified view of the different types of information used in compiling the domestic output matrix.

Figure 5.2 Different types of information used in compiling the production matrix
INDUSTRIES
Total
Trade, transport Finance and domestic
Agricul- Manufac- Construc- Other
and business output
ture turing tion services
communication services
Production
(1) (2) (3) (4) (5) (6) (7) type surveys
Agriculture (1)
Manufacturing (2)
PRODUCTS

Construction (3)
Data on total
Trade, Transp. and comm. services (4) sales or
Finance and business services (5) output by type
of product
Other services (6)
Total (7)

= Principal activities

= Secondary activities

Surveys of enterprises or establishments

Collect data on total sales by enterprises or establishments plus a range of other

information, preferably with a product breakdown, such as:
(a) Sales by type of products, which are then allocated by CPC products
(b) Changes in inventories (split by asset type, for each an opening and closing
level)
(c) Own account production (by type of product)
(d) Other taxes and subsidies on products
(e) Trade activity (by type, for example, wholesale or retail)
(f) Total sales, sales for export (by type of product)

Data collected Principal activity

under (a), (b), (c) will link to Secondary activity
and (e) Ancillary activity

5.34. Starting with the enterprise survey, the principal objective is to supply information on the
main structural characteristics of the different economic activities. The basic unit of this type of
survey tends to be the enterprise. From this source, it is possible to estimate the total production
by activity, starting with its private accounting business systems. On the other hand, the production
surveys allow the estimation of the total production by type of product.

5.35. Combining both sources of information, enterprise statistics and production statistics, it is
possible to combine the data and obtain the production by type of product, by principal activities
of the enterprise and by principal activities of the establishment that belongs to this enterprise.
Consequently, the principal production and the secondary production of a product can be
identified, primarily for industrial products. In many cases, lack of information makes it necessary

138
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

to use reasonable assumptions about what products are constitute the industries’ secondary
production.

5.36. Many enterprises may perform some construction work, for example, own-account gross
capital formation and minor maintenance and repair work. Enterprises in the manufacturing and
service industries are often involved in either wholesale trade or retail trade or even both. Many
service industry enterprises may provide retail trade services as a secondary activity. Lastly,
activities involving the rental of real estate and leasing of equipment are often secondary activities.

5.37. In basic statistics, output by products will usually be available for goods-producing
industries such as agriculture, mining and manufacturing industries – at least for enterprises and
establishments above a certain threshold – and similarly imports and exports of goods will be
covered in great detail by external trade statistics. For service industries, a breakdown of output
by individual kinds of services (as defined in the CPC classification) is less common, although in
recent years many countries have developed such statistics. If there is a lack of product statistics
for services, the output by the most detailed service activities of ISIC may be used as proxy,
assuming that all output consists of the services characteristic for that particular industry.
Concerning product breakdowns, construction will be placed between these two extremes.

5.38. For manufacturing units below a certain threshold, output statistics by products will usually
be missing, whereas total output will be estimated based on either business surveys or
administrative records. Working at the most detailed activity level available, output from these
small units can be broken down into the products of the system, for example, by assuming that the
composition by product is identical to that which has been observed for the smallest category of
those units for which output statistics by product exist. During the balancing process, this
assumption may be modified and the output redistributed by product.

5.39. The products recorded in the domestic output part of the supply table should be output
valued at basic prices at the time it is completed. For manufacturing industries, it is usually only
sales by product that are given in the surveys and adjustments would need to be made for change
in inventories of finished goods and work-in-progress, in order to move from sales to output.

5.40. When information exists about opening and closing inventories by industry, it can be
assumed that the composition by product is identical to the sales by product. The change in
inventories of finished products and work-in-progress by product can be derived by applying
relevant price indices and assumptions about inventory valuation principles used by enterprises.
As the reliability of these data by product is limited, however, and as, by definition, these data
should always be identical on the supply side and the use side, there is no need to adjust the sales
figures or enter them into the final use category of changes of inventories at this stage. In the
system, it is the actual sales figures that are relevant for the distribution by users, and the estimated
data for change in inventories can therefore be simply imposed on the system after the balancing
has been completed. This is, however, not the case for change in inventories of materials and fuels

139
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

(recorded in the intermediate use part of the use table) and in trade. Changes in inventories of
agricultural products and mining products will usually have to be included in the system from the
outset, as the output data will often refer not to sales but to actual output.

3. Principal and secondary products

5.41. The distinction between principal and secondary production has traditionally played a
prominent role in I-O literature, as the existence of secondary production requires certain
assumptions for IOTs to be compiled. It should be noted, however, that a match between products
and industries (determining in which industry a product is the principal output) is really only
necessary in those cases where the chosen techniques for deriving IOTs as a starting point requires
the SUTs to be aggregated to square tables, cases where the sequence of the aggregated products
is made comparable to the sequence of industries. With regard to other techniques for the
compilation of IOTs and for the purpose of the SUTs, there is no need to match products and
industries. It should thus be noted that, when industry-by-industry IOTs are derived on the
assumption of fixed product sales structures, there is no need first to aggregate the rectangular
SUTs.

5.42. When necessary, the match between product and principal producer can be derived either
theoretically (by identifying for each product the principal producer according to the ISIC
definitions of the principal products of each industry – the correspondence keys are available on
the United Nations Statistics Division classification website at
http://unstats.un.org/unsd/class/default.asp) or empirically (established by observation of the
actual domestic output matrix, the industry being the main producer of each product).

5.43. In principle, the empirical match will be the most precise in the sense that it depicts the
production relationships as they actually exist in this particular economy. The theoretical match
may be the preferred approach when considering time series and international comparisons. These
matching methods also demonstrate that the product classification applied when compiling SUTs
can be chosen completely independently of whether or not it may subsequently be necessary to
derive square SUTs.

5.44. When the domestic output matrix is aggregated to a square matrix and arranged so that the
entries for the primary products fall on the diagonal, the off-diagonal elements show the extent of
secondary production. This refers to that part of a product which is produced by industries other
than the one where it principally belongs either formally according to the industrial classification
(theoretical aggregation key), or according to the industry which is actually the main producer
(empirical aggregation key).

5.45. As the secondary production observed in the domestic output matrix depends on the level
of aggregation both of products and of industries, secondary production does not possess any
observable characteristics of its own. The elusive character of the concept of secondary production
makes it difficult to justify that a product should be of particular interest statistically just because

140
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

it is produced in two or more industries at a certain level of industry or product aggregation. When
the industry and product classifications to be used in SUTs have been decided (inclusive of
possible redefinitions), the principal versus secondary distinction plays no role in the subsequent
elaboration and balancing within the SUTs framework.

5.46. For most countries, the domestic output matrix is characterized by showing secondary
production almost exclusively for manufacturing industries, whereas for most other industries,
practically all the production is found on the diagonal elements (or in the rectangular table – in
what is known as the “diagonal field”). There are three main reasons for this:

• Basic statistics for manufacturing industries have traditionally included detailed product
statistics and thus make identification of secondary production possible.

• For service industries, the diagonal structure is simply due to the fact that, more often than
not, very limited detail has been collected on the type of product breakdown of these service
activities. Thus, the total output from establishments (or even enterprises) must be assumed
to be the primary output of the industries to which the units are classified in the business
register.

• The activities of industries such as agriculture, construction and trade are often defined in a
purer form (the industries covering all their principal products, and only those) in the national
accounts and SUTs than in the business register. In this case, all secondary agricultural,
construction and trade activities in other industries would have been transferred to their main
activities. Alternatively, data for some activities may have been constructed in such a way
that from the outset no secondary production exists, for example, agricultural activity is
measured as the sum of all agricultural products, construction activity as the value of new
construction and repairs, and so forth.

5.47. When the rectangular SUTs have been balanced, there may be a need to aggregate them
into a square system either for dissemination purposes or for compiling IOTs by methods that
require square SUTs. In a square system, the number of product groups must be identical to the
number of industries, and furthermore, the products aggregated in such a way that the resulting
product groups can be given corresponding industry names, indicating the industry of which they
are principal products. If aggregation is made solely for dissemination purposes, the product
aggregation could also be carried out to, for example, higher levels of CPC – which, as mentioned
above, have no direct correspondence with ISIC defined industries.

4. Ancillary activities

5.48. When the production of an enterprise takes place in two or more different establishments,
certain ancillary activities may be carried out centrally for the collective benefit of all the
establishments. If, in such a case, a producer unit is undertaking purely ancillary activities and is
statistically observable, in that separable accounts for the production that it undertakes are readily

141
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

available, or if it is in a geographically different location from the establishments that it serves, it

may be desirable and useful to consider it as a separate unit and to allocate it to the industrial
classification corresponding to its principal activity. Another exception is the case when some
products are used both for own ancillary use and also supplied to another unit (see 2008 SNA,
para. 14.33). In the 2008 SNA, however, it is recommended that statisticians do not make
extraordinary efforts to create separate establishments for these activities artificially in cases where
suitable basic data are not available.

5.49. The fact that establishments, even at a detailed level, are classified to the same activity,
does not mean that they are in all respects identical. Each establishment has its own unique
institutional and organizational characteristics, which may influence the composition of its
purchases as much as the underlying technical production processes. Two establishments
producing identical products may have quite different input structures, depending on the degree of
reliance on purchased semi-fabricated products, outsourcing of certain activities (see also chapter
8, section D, on goods sent abroad for processing), whether it owns the capital equipment and
buildings it uses rather than leasing or renting them, and so forth, and in general, on the degree of
vertical integration of the various stages of the production processes. There is no way that these
institutional characteristics inherent in the original source data should be eliminated from the SUTs
(or subsequently from the IOTs) nor does the SNA expect the compilers to try to do so.

5.50. Institutional arrangements of production not only differ between establishments classified
in the same industry but also across countries and over time. It is obvious that there are serious
limitations to the view that the SUTs (and the IOTs) portray the technical characteristics of a
production system. From a statistical point of view, the achievable elimination of institutional
arrangements is obtained by using establishments as production units (with the possible additional
partitioning of vertically integrated enterprises as discussed above), given that the establishments
are designed with this purpose in mind and there are no official statistics providing production
structures below this level.

5.51. In some countries, the recommended establishment unit approach may not be achievable
in practice, since, for legal, practical or historic reasons, statistics are only collected for enterprise
units. Even though compilers of the SUTs may in this situation try to break down the most
important multi-activity enterprises into their constituent establishment units, there is in general
no feasible alternative to working with the existing data. In this case it is still possible to compile
SUTs, although the overall picture of the productive system will become less precise and to some
extent blurred, which will also have an adverse effect on the resulting SUTs and IOTs.

5.52. It should, however, be recognized that the important objectives of compiling the SUTs may
also be achieved when the data are based on enterprise units (see SNA 2008, para. 14.21), although
some product-flow and common-sense procedures may be more difficult to apply because of the
less stringent definition of industries, as the composition of output from the enterprise units will
also be crossing the borderlines between sections of ISIC. For these reasons, an enterprise-based

142
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

approach will in general require more thorough coverage of statistical source data. It should also
be borne in mind that there are no automatic methods available that can disentangle this dataset
and transform it into SUTs or IOTs with analytical properties comparable with those resulting
from SUTs based on establishment-type units. Depending on the specific circumstances, it may in
such cases be decided to compile SUTs alone and not IOTs.

5. Redefinitions

5.53. Redefinitions refer to adjustments made to the source data relative to the way in which they
are obtained from the primary statistics, in order to obtain “purer” industries, so to speak, for use
in the SUTs. This is an exception to the previously mentioned rule that SUTs compilers should not
attempt to create their own versions of basic statistics. That would not be cost-effective nor would
it be conducive to the comparability of SUTs with other economic statistics or on an international
scale. In practice, deviations from the way in which enterprises or establishments are defined in
the business register and reflected in primary statistics should be limited in scope.

5.54. Such redefinitions may be seen as implementing the SNA recommendation to partition
vertically or horizontally integrated enterprises or establishments that have production in two or
more sections of ISIC Rev. 4 (2008 SNA, paras. 5.52–5.54). Redefinitions are generally carried
out manually, using product-specific input structures based on specific insight into the activities,
leading to results that will come as no surprise nor give rise to negative elements, as might have
been the case had more automatic methods been applied. By reducing secondary production,
redefinitions facilitate the subsequent compilation of IOTs, and compilers of SUTs should be
aware of how the choice of compilation techniques will affect the subsequent calculation of the
IOTs.

5.55. Redefinitions (more background is provided in Box 5.1) are usually concentrated on a few
major activities, such as agriculture, energy, construction and trade, or a few major enterprises,
such as mining operations. Redefinitions affect all those activities from which secondary output is
being removed. For some activities, redefinition-type adjustments may have been carried out
already in the source data, as in the following cases:

• The European Union System of Agricultural Accounts requires that all agricultural activity
is covered by these accounts and there are very limited possibilities for the retention of non-
agricultural secondary production within the system definition of agriculture.
• All rented dwellings are usually grouped together in one single industry (together with
owner-occupied dwellings) independently of the activity of the actual owner.
• Trade activities outside the trade industries (trade as secondary activity), by definition, have
already been separately identified when compiling the national accounts, as only the trade
margins, and not the gross turnover of the traded products, should be counted as output, and
may have been grouped together with trade as primary activity.

143
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

• Construction activities are also frequently redefined to form one single, “pure” construction
activity, often because total output has been defined by adding up the values of specific types
of construction output rather than output from building establishments, or alternatively
inputs have been determined from the supply of construction materials. Any of these
approaches will also facilitate the distribution of building materials for intermediate
consumption.
Box 5.1 Redefinitions
In Miller and Blair (2009), on page 141, redefinitions are defined as: “Factoring out the amount of
secondary products produced as well as the inputs used in that production and reassigning both to the
industry for which the product is classified as primary”.
In addition, a distinction is made between specific redefinition and mechanical redefinition (page 215),
where the former is the “by hand” procedure and the latter refers to the various mathematical procedures
that can be applied to eliminate secondary production when producing IOTs from SUTs (covered in
chapter 12 of Miller and Blair).
The specific redefinition or two-step process emerges from the practice in several countries. It is explained
in detail for the United States in Guo J. and others (2002). The Bureau of Economic Analysis paper was
presented at the fourteenth International Conference on Input-Output Techniques, held in Montreal,
Canada, in 2002. The article also analyses the differences between the resulting tables when redefinitions
are not applied (case 1), and when they are applied (case 2).
The redefinition method is also used in Canada and Denmark, whereas the industry-by-industry IOTs in
Norway are more of the case 1 type, in that they retain the micro-macro link to a maximum degree.
The industry-by-industry IOTs of the Netherlands seems to fit somewhere between case 1 and case 2.
In France, the first step (redefinition) is based on enterprise units and is carried out to an extent that the
supply table becomes diagonal. The use tables thereby also form the IOTs, and the second step (compiling
the IOTs) becomes superfluous.

5.56. Although the redefinitions serve the purpose of creating purer activities and thus facilitate
I-O analysis, their main purpose is to arrive at an activity classification that is applicable for use
in the national accounts, and thus conducive to the integrated compilation of SUTs and national
accounts. Three different situations can be distinguished:

• Case 1: no redefinitions take place in the national accounts, the SUTs and the industry-by-
industry IOTs.

• Case 2: redefinitions have been carried out for all national accounts data and in the SUTs
prior to the calculation of the industry-by-industry IOTs.

• Case 3: redefinitions are not carried out when the current national accounts are compiled but
applied when the SUTs and the industry-by-industry IOTs are compiled.

5.57. In the first two cases, the consistency and comparability of the current national accounts
(tables by industry), and of the SUTs and IOTs classifications are upheld, but that is not so in the
third case. Ideally, the choice of redefinitions should already be introduced in the general

144
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

classification of industries used in national accounts. Not only will the manually prepared
redefinitions be more precise at these earlier stages, but they will also facilitate the balancing of
the system as the need to make a large number of small input entries to many cells of the use table
will be obviated.

D. Imports of goods and services

1. General description and definition

5.58. The second part of the supply table covers the total imports of goods and services. In
national accounts, imports refer to transactions that occur when there are changes of economic
ownership of goods between residents and non-residents, whether or not there are corresponding
physical movements of goods across frontiers.

5.59. International merchandise trade statistics represent the main source of data for imports of
goods. International standards are specified in International Merchandise Trade Statistics:
Concepts and Definitions (IMTS 2010) (United Nations, 2011). For imports of services, the main
source of data is either the details available in the balance of payments statistics or specialized
statistics on international trade in services (for example, business surveys), according to the
international standards given in the 2010 Manual on Statistics of International Trade in Services
(United Nations, European Commission, IMF, OECD and WTO, 2011) in connection with product
classifications.

5.60. Some differences exist, however, between the concepts used in international trade statistics
and the 2008 SNA and BPM 6, and adjustments therefore need to be made to the basic statistics in
order that they can be used in the SUTs. The BPM 6 identifies sources of difference between the
IMTS and the 2008 SNA and BPM 6 concepts of imports that may occur in countries. In this
regard, it recommended that a standard reconciliation table be compiled to assist users in
understanding these differences.

5.61. One major difference is the valuation used to record imports of goods. While IMTS 2010
uses a CIF valuation for imports, the 2008 SNA and BPM 6 use a uniform FOB valuation for both
exports and imports of goods. The 2008 SNA states, in paragraph 3.149:

Imports and exports of goods are recorded in the SNA at border values. Total imports and
exports of goods are valued free-on-board (FOB, that is, at the exporter’s customs frontier).
As it may not be possible to obtain FOB values for detailed product breakdowns, the tables
containing details on foreign trade show imports of goods valued at the importer’s customs
frontier (CIF, that is, cost, insurance and freight), supplemented with global adjustments to
FOB values. CIF values include the insurance and freight charges incurred between the
exporter’s frontier and that of the importer. The value on the commercial invoice may of
course differ from both of these.

145
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.62. The adjustments for the FOB and CIF valuation of imports are described in more detail in
the next section.

5.63. Another difference is the time of recording. In the 2008 SNA and BPM 6, the time of
recording of imports and exports corresponds to the time that ownership of the goods is transferred.
By contrast, IMTS follows the timing of customs processing. While this timing is often an
acceptable approximation to the change of economic ownership principle, adjustments may be
needed in some cases, such as for items with large values or goods sent on consignment (that is,
dispatched before they are sold). It should be noted that, in the case of goods sent abroad for
processing with no change of economic ownership, the values of goods movements are included
in the IMTS-based recording but are to be excluded from the ownership-based recording in the
national accounts and balance of payments. It is recommended, however, that the values of goods
movements be entered as supplementary items in the balance of payments, to indicate the nature
of these arrangements.

5.64. Other adjustments to IMTS may be needed to bring estimates into line with the change of
economic ownership of goods, either generally or because of the particular coverage of each
country. Possible examples include:

• Merchanting
• Non-monetary gold
• Goods entering or leaving the territory illegally
• Goods procured in ports by carriers
• Goods moving physically but where no change of economic ownership has taken place such
as in cases of operating leases
5.65. To maintain consistency with BPM 6, the 2008 SNA introduced new treatment relating to
merchanting and goods sent abroad for processing. Merchanting is a process whereby a unit in
economy X purchases goods from economy Y for sale in economy Z (sometimes within economy
Y itself). The goods legally change ownership but do not physically enter the economy where the
owner is resident. By convention, the purchases of the goods intended for resale is shown as
negative exports. When the goods are sold, they are shown as positive exports. When the purchase
and sale take place in the same period, the difference is shown as an addition to exports. If the
purchase takes place in an accounting period, the negative export is offset by an increase in
inventories of goods for resale, even though those goods are held abroad.

5.66. The surplus on this item in the foreign trade statistics is by its nature a trade margin and
should be included in the output of the industry. In the main, this activity takes place in the trade
industry. In exceptional cases, this may lead to an overall deficit on the item in the foreign trade
statistics but the trade margin would usually still remain positive (the deficit added to changes in
inventories). As indicated, trade margins from merchanting activity primarily occur in the trade

146
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

industry but can occur in many other industries, unless all trade is redefined to the trade industry.
Given that business statistics provide source data as a starting point for the compilation of SUTs,
merchanting activity can then appear in various industries, for example, oil companies and
pharmaceutical companies.

5.67. The new treatment of goods sent abroad for processing is dealt with in more detail in
chapter 8 of this Handbook.

5.68. A special category within imports is the direct purchases abroad by residents. This item
covers all purchases of goods and services made by residents while travelling abroad for business
or pleasure. Two categories must be distinguished because they require different treatments:

• Expenditure by resident business travellers: this item refers to intermediate consumption of

several industries to which the travellers belong (in the use table) and imports of services (in
the supply table).
• Expenditure by other resident travellers on personal trips: this expenditure is recorded in
final consumption expenditures by households (in the use table) and imports of services (in
the supply table).
5.69. Imports broken down by products in the SUTs do not include direct purchases abroad by
residents. Consequently, these must be included in an adjustment row to obtain the overall value
of imports (row (11) in Table 5.1).

5.70. In tables 5.1 and 5.2, the estimates for CIF/FOB adjustments on imports and the direct
purchases abroad by residents are shown separately in the rows. It should be noted, however, that
some countries do not show these estimates in separate rows but consolidate the values in the
product groups in the respective columns. This situation in turn leads to different product balances
but does not change the imports aggregate total. This is often due to the coverage of the data
sources and, in these cases, appropriate adjustments should be applied to extract the corresponding
entries to generate the separate entries.

5.71. Goods procured in ports by carriers may be included in a similar adjustment row. It should
also be noted that imports and exports of ships and aircraft may have to be given special attention,
as these transactions may follow special recording procedures in the external trade statistics that
are not consistent with the way in which output or gross fixed capital formation should be recorded
in the national accounts.

5.72. Imports of goods and services in SUTs are dealt with in more detail in chapter 8.

2. Valuation for imports: CIF and FOB valuation

5.73. In the 2008 SNA and BPM 6, the total imports of goods are valued FOB. The data on
imports by detailed products from the foreign trade statistics used in the SUTs are usually available
at CIF prices, however, following the International Merchandise Trade Statistics (United Nations,

147
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

2011). To reconcile the different valuations used for total imports of goods and for the product
components of imports, two types of adjustments are needed. These adjustments are presented
below.

(a) Data adjustment

5.74. The first type of adjustment must be made to the data of the balance of payments prior to
entering data from this data source into the SUTs system. This adjustment is necessary in order to
ensure as a starting point a consistent set of data for imports and exports of goods and services that
can be balanced across the SUTs. This adjustment is illustrated in Table 5.3.

5.75. The starting point is the account for the rest of the world, as shown in columns (1) and (2)
of Table 5.3 (which mirrors the balance of payments according to BPM 6), where only the entries
for goods and services are shown and where imports of goods are valued FOB (372 in column (2)
of Table 5.3). This is the value of the goods at the point of exit from the exporter’s economy,
including transport charges and trade margins up to the border point. The CIF value of imports
(382 in column (6) of Table 5.3) of goods measures the value of a good delivered at the point of
entry into the importing economy. The difference between the two values (10 in column (4) of
Table 5.3) is made up of the costs of transportation, insurance and other expenditures between the
point of exit of the exporter’s country and the point of entry into the importer’s country.

5.76. The services linked to the difference between the FOB and CIF values can be delivered by
either resident producers or non-resident producers. To the extent that non-resident producers are
involved, the BPM 6 imports of services must be reduced with their services (7 in column (4) of
Table 5.3) to avoid double counting, as these services are now included in the CIF value of the
imported goods. Adjustment for the services delivered by resident producers (3 in column (3) of
Table 5.3) is a bit trickier, as a service that, according to the BPM 6 definition, is a purely domestic
transaction will now appear as an import of services included in the CIF value of imported goods.
As this import originates from resident producers, it is necessary to introduce a balancing service
export of the same value.

Table 5.3 Data adjustment for external trade of goods and services
SNA/BPM balance of payments Introducing imports CIF SUTs balance of payments
Uses (FOB) Resources (FOB) Uses Resources Uses (CIF) Resources (CIF)
(1) (2) (3) (4) (5) (6)
Imports of goods 372 10 382
Exports of goods 462 462
Imports of services 84 -7 77
Exports of services 78 3 81
Total 540 456 3 3 543 459
Balance 84 0 84

Note: In practice there will be a further breakdown of both goods and, in particular, services in the
balance of payments, and therefore also for the adjustments in columns (3) and (4).

148
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.77. In Table 5.3, all data adjustments are shown in columns (3) and (4), and the resulting “SUTs
Balance of Payments” in columns (5) and (6). It is noted that the balance of the adjustment items
is zero, and consequently, the surplus on the transactions in goods and services – 84 in column (2)
– is identical in the two alternative ways of presenting the external transactions.

5.78. The “SUTs balance of payments” represents the framework of source data for external
trade for SUTs with the appropriate product breakdowns. The composition by specific services
making up the CIF and FOB difference will usually be available from the working tables of the
balance of payments compilers, as their starting point for the FOB recording of imports will usually
have been imports of goods from the external trade statistics valued at CIF Regular surveys may
also have been carried out to illuminate the CIF and FOB difference and the related service
structure.

5.79. It is important to underline that the above data adjustment is not the CIF and FOB
adjustment often seen as a separate row in SUTs or IOTs. The data adjustment must be made
before starting compiling SUTs. At the detailed product level, the supply and use of the individual
services are adjusted so that they can meaningfully be balanced under the CIF valuation of goods,
and these data adjustments will not be separately identifiable in the completed SUTs.

(b) CIF and FOB adjustment row

5.80. The CIF and FOB adjustment is an ex post facto adjustment made at the macro level to the
totals for exports and imports of goods and services to derive the corresponding totals found in the
SNA (the goods and services account and the rest of the world account).

5.81. In principle, the purpose of this adjustment is to demonstrate that the data in SUTs are
consistent with the rest of the national accounts and to avoid the double-counting of CIF-type
services provided by residents. The CIF and FOB adjustment row has no balancing or other
methodological functions in the SUTs, and it may be omitted from the SUTs and also the IOTs if
there is no special need to maintain the exact conceptual relationship to the national accounts.

5.82. Table 5.4 illustrates the place and content of the CIF and FOB adjustment row in SUTs,
albeit limited here to external trade data.

Table 5.4 CIF and FOB adjustment row

Supply Table Use Table
Im ports Exports
Goods Services Goods Services
SUT total 382 77 462 81
CIF/FOB adjustment -10 7 -3
BOP total 372 84 462 78

5.83. The “SUT total” row includes the totals for imports and exports of goods and services in
the balanced SUTs system, consistent with the “SUTs Balance of Payments” in Table 5.3.

149
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

5.84. In order to obtain totals for the external transactions identical to those found in the rest of
the national accounts (and the balance of payments), the adjustments shown in the “CIF/FOB
adjustment” row of Table 5.4 are introduced. These adjustments mirror those that were made as
data adjustments in Table 5.3. The two types of adjustments have quite different roles, however:

• Those in Table 5.3 relate in principle to columns of the SUTs and must necessary be carried
out prior to the balancing, and there is no way to avoid this adjustment.
• On the other hand, the CIF/FOB adjustment in Table 5.4 is a kind of “memo” row of the
SUTs that can be added ex post facto, or even omitted if there is no need to demonstrate
consistency with the national accounts.
5.85. It should be noted that, if goods and services are lumped together in SUTs, the CIF/FOB
adjustment row will only include the adjustment item, -3, for both imports and exports.

5.86. From a bookkeeping perspective, the data adjustment for exports of services (3 in Table
5.3) could alternatively be recorded as a negative import, even though this action entails a less
logical explanation of how the domestic output of services are disposed of and also requires the
existing imports of those services to be sufficient to prevent a negative net result.

5.87. With this approach, there would be adjustments in Table 5.3 for imports only, showing
identical numerical changes for goods and services, respectively. The CIF/FOB adjustment row in
Table 5.4 would in this case have entries only for imports (-10 for goods and +10 for services),
and if imports were not shown separately for goods and for services, the CIF/FOB adjustment row
would be empty. Further details covering issues of consistency in the SNA are provided in Box
5.2.

Box 5.2 Consistency issues with the CIF/FOB adjustment

The CIF/FOB adjustment is dealt with in both the Eurostat Manual of Supply, Use and Input-Output
Tables (Eurostat, 2008) and in the supply and use table chapters 14 and 28 of the 2008 SNA.
In the numerical example in Eurostat (2008) (pages 60, 70 and 122), external trade in goods and
services are lumped together and, as explained above, for this case, the CIF/FOB adjustment row
therefore contains identical negative adjustments for imports and exports. In the more extensively
elaborated numerical example (pages 113–115), where comparisons are also made to the treatment
in the 1993 SNA, ex ante data adjustments are mixed up with the ex post CIF/FOB adjustment in a
complicated manner.
The CIF/FOB adjustment table has both a column and a row dimension, and the final outcome is
incorrect because the ordinary exports of services linked to exports of goods are included in the
adjustment.
The exposition of the CIF/FOB adjustment in the 2008 SNA is unclear because it starts out from the
assumption that the SUTs have been balanced using inconsistent data, namely imports of goods
valued CIF, and services as defined as in the BPM 6 based on imports being valued FOB. This
shortcoming can obviously not be remedied by ex post adjustments to columns and rows, as any new
column data would require a new balancing of the SUTs.

150
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

The final outcome in the 2008 SNA are SUTs with a CIF/FOB column in the supply table (table
14.12 of the 2008 SNA), in addition to the CIF/FOB adjustment row with adjustments for imports
only (the resident producers’ delivery of services linked to imports CIF being treated as negative
imports).
If the CIF/FOB adjustment column in the supply table (table 14.12 of the 2008 SNA) is added to the
column for imports of services, services as defined in the SUTs balance of payments are obtained,
so that in principle this could be taken to indicate the ex ante data adjustment. This is not easily
understood from the exposition, however, and, to compound the lack of clarity, the CIF/FOB
adjustment is being distributed by user (in table 14.15 of the 2008 SNA), a step for which there is no
explanation.

151
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

152
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Annex A to chapter 5: Sample questionnaire collecting sales of goods and

services, inventories of goods and trade-related data

A5.1 The extract shown in figure A5.1 is from a business survey questionnaire from the
Statistical Office of Serbia. Data are collected for each industry and by product covering the
following areas:

• Sales of goods produced by the enterprise

• Closing stocks of products and work-in-progress

• Sales of merchandise

• Trade margin

• Closing stocks of goods for resale

A5.2 Additional tables collecting data on the sales of industrial and non-industrial services
provide the full coverage of goods and services needed to calculate the industry totals. An extract
of these tables may be seen in figure A5.2. These data make it possible to calculate the industry
output by product and trade margins required to populate the domestic output part of the supply
table and the trade margins column, as shown in Table 5.2. In some countries, opening stock values
are also collected.

153
Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications

Figure A5.1 Extract of questionnaire covering sales of goods, inventories of goods and
trade activity
Closing
Sales of
stocks of Closing
goods
products and Sales of Trade stocks of
produced by
work in merchandise margins goods for
No. Code Product description the
progress (group of amount resale
enterprise
(group of accounts 60) rate % (group of
(group of
accounts 10 accounts 13)
accounts 61)
and 11)
(1) (2) (3) (4) (5) (6) (7) (8)
1000 TOTAL
AGRICULTURAL PRODUCTS, RAW AND UNPROCESSED PRODUCTS OF PLANT
AND ANIMAL ORIGIN
1001 01.11.1–01.11.4 Cereals, all kinds (except rice), cereal seeds
1002 01.11.6, 01.11.7 Green leguminous vegetables (beans, peas, lentils and other)
1003 01.11.8 Soya beans, groundnuts (row) and cotton seed
1004 01.11.9 Other oil seeds–sunflower, sesame, flax, etc.
1005 01.11.12 Rice, not husked
01.13 except
1006 Vegetables, raw and seeds
01.13.7
1007 01.13.7 Sugar beet and sugar beet seed
1008 01.13.8 Mushrooms and truffles
1009 01.15 Unprocessed, raw tobacco
1010 01.16 Fibre crops (flax, cotton, hemp and other, used in textile industry)
1011 01.19.1 Forage crops and vegetative matter for livestock feeding unprocessed form
1012 01.19.2 Flower and flower seeds
1013 01.21 Grapes
1014 01.22, 01.23 Tropical and subtropical fruits, all kinds (including citrus, figs, etc.)
01.24, 01.25
1015 Other fruits, tree and bush fruits, except nuts (apples, pears, cherries, berries, etc.)
except 01.25.3
1016 01.25.3 Nuts (almonds, hazelnut, walnuts, etc.)
1017 01.26 Olives, coconuts (raw, unprocessed)
1018 01.27 Coffee beans, tea leaves, cocoa beans, not roasted
1019 01.28 Spices, aromatic, drug and pharmaceutical crops
01.11.5, 01.14, Vegetables and fruit seeds, other seeds; grass, unprocessed straw and other
1020 01.19.3, 01.29, residues of cereals; seeds for trees and seedings; planting materials, sugar cane and
01.3 other raw, unprocessed and untreated products of plant origin not elsewhere
01.4. except Live animals and animal products (unprocessed milk, eggs, natural honey; seeds and
1021
01.45.3 & 01.49.3 embryos of animals, except raw skins, shorn wool and skins, etc.)
Raw fur skins, shorn wool, skins (excluding products of slaughterhouses and
1022 01.45.3, 0