SORRY, I AM NOT AWARE OF THAT. 

I think this is the first use of a multinational input-output model:

United Nations, Department of Economic and Social Affairs. 1973.Brief Outline of the United Nations Study on the Impact of Prospective Environmental Issues and Policies on the International Development Strategy.New York: United Nations. 

It is also mentioned in Leontief, W. W. (1986). Input-output economics. Oxford University Press on Demand.

Pedro Gerber Machado

thank you for hopeful information.

I consulted Leontief (1977) p.322. Leontief writes there 

• Major efforts are presently under way to construct a data base for a systematic input-output study not of a single nation economy but of the world economy viewed as a system composed of many interrelated parts. This global study, as described in the official document, is aimed at
•       (quotation from Brief Outline of the United Nations Study ...)

As we see, he is talking in the past progressive form. Was this project was successfully completed? If so, where I can get such IO table? 

Hinted by the Pedro's information, I found a next description of World Input-Output Table in Sam Cole's book. The next line is the read in his book. 

               The United Nations World Input-Output Model (1976)

• The World Input-Output model constructed by Leontief and his team at several universities in the United States forms part of the United Nations study on the Impact of Prospective Environmental Issues and Policies on International Development Strategy initiated in 1973. The objective of this study is to investigate the interrelationships between future economic growth and environmental issues, including questions of natural resource availability, the degree of pollution associated with goods and services, and the economic impact of abatement policies (3). In essence, therefore, the World Input-Output model is again concerned with a reappraisal of the results of the Limits to Growth study. The general aim of the model is: “on a quantitative basis to display various possible interrelationships, as the world economy evolves over future decades, between environmental and other economics policies” (5).
•  An Input-Output model is, however, in several respects quite different from most of the models described here. ( Input-Output modeling goes back 20 // 21 some fifty years to Leontief’s original applications and well before the advent of the electric computer, unlike system analytic and control engineering type models whose origins are essentially post-war. ) In principle the method is far simpler in that it is not dynamic, nor does it take explicit account of the feedback mechanisms considered so important, especially by authors, in contrast to most other models, the United Nations model has “an unusual degree of detail”. It considers fifteen interacting economies each described by forty five sectors of economic activity. This detail is considered to be “advantageous” . . . since it permits use of data specific to individual industries in particular regions and therefore has “the model permits a detailed analysis of prospective changes in technology, cost if production and relative prices” (7)”.   (emphasis by bold letters is mine)

Sam Cole (1977) Global Models and the International Economic Order: a paper for the United Nations. UNITAR. Published by Elsevier Ltd. pp.20-21.

 By this description, World Input-Output Table comprises 15 economies (regions) and 45 sectors of economic activity.  (Thus the endogenous sector comprises 675 ×675 matrix.)

 If this is the first compilation of a world IO table, then we can say that multinational IO table dates 1976.

When I was working with Prof. Bill Gosling at UEA in 1973-75 as an undergraduate, he was chairman of the quite new UK Input-Output Association.  His monographs, in relation to the UK economy, were still very much concerned with the data problems associated with compiling any meaningful level of detail even in such an economy as the UK's with a National Statistical Office of some competence.    A 'worldwide' data set was still a dream, and we had no real hope of computing against such a large set even if one had existed.

It would seem from the Input-Output Model 'Wikipedia' entry that this situation has never been addressed to the point where one can talk of 'An International Input-Output Table'.  Unless one accepts as satisfying the requirement of being 'international' the small domain of those countries able to contribute their own datasets in concert with other capable states and covering the same periods and to do so consistently.   Or tables referring to states within any limited geographic region.

The entry I refer to seems to assert that whereas guidelines and standards were published in the 1970's the actual delivery of tables from the mid 80's has been very patchy, or relatively incomplete.

For your convenience, and further investigation, here are the references and bibliographies of documents drawn upon by that article:

References

Thijs Ten Raa, Input–output economics: theory and applications: featuring Asian economies, World Scientific, 2009

Walras, L. (1874). Éléments d'économie politique pure, ou théorie de la richesse sociale [Elements of Pure Economics, or The Theory of Social Wealth].

Nikaido, H. (1970). Introduction to Sets and Mappings in Modern Economics. New York: Elsevier. pp. 13–19. ISBN 0-444-10038-5.

Zera. "Quantity-Directed Socialism". EconomicTheories.Org. Retrieved 21 December 2013.

Towards A New Socialism, 1993, by Paul cockshott and Allin Cottrell. Coronet Books Inc. 978-0851245454. "Planning in the USSR", (P.79)

About SNA, UN

Bibliography

Dietzenbacher, Erik and Michael L. Lahr, eds. Wassilly Leontief and Input-Output Economics. Cambridge University Press, 2004.

Isard, Walter et al. Methods of Regional Analysis: An Introduction to Regional Science. MIT Press 1960.

Isard, Walter and Thomas W. Langford. Regional Input-Output Study: Recollections, Reflections, and Diverse Notes on the Philadelphia Experience. The MIT Press. 1971.

Lahr, Michael L. and Erik Dietzenbacher, eds. Input-Output Analysis: Frontiers and Extensions. Palgrave, 2001.

Leontief, Wassily W. Input-Output Economics. 2nd ed., New York: Oxford University Press, 1986.

Miller, Ronald E. and Peter D. Blair. Input-Output Analysis: Foundations and Extensions. Prentice Hall, 1985.

Miller, Ronald E. and Peter D. Blair. Input-Output Analysis: Foundations and Extensions, 2nd edition. Cambridge University Press, 2009.

Miller, Ronald E., Karen R. Polenske, and Adam Z. Rose, eds. Frontiers of Input-Output Analysis. N.Y.: Oxford UP, 1989.[HB142 F76 1989/ Suzz]

Miernyk, William H. The Elements of Input-Output Anaysis, 1965.[2].

Polenske, Karen. Advances in Input-Output Analysis. 1976.

Pokrovskii, Vladimir N. Econodynamics. The Theory of Social Production, Springer, Dordrecht, Heidelberg et cetera, 2011.

ten Raa, Thijs. The Economics of Input-Output Analysis. Cambridge University Press, 2005.

US Department of Commerce, Bureau of Economic Analysis . Regional multipliers: A user handbook for regional input-output modeling system (RIMS II). Third edition. Washington, D.C.: U.S. Government Printing Office. 1997.

Robin

From Wikipedia (in spanish):

A pesar de que se acredite la teoría a Leontief el procedimiento fue descrito en 1758 por el economista francés Francois Quesnay que desarrolló una versión más rudimentaria denominada Tableau économique. El economista Karl Marx fue el primero en traducir la obra de Quesnay a un sistema matricial de ecuaciones, en los llamados modelos de reproducción simple y reproducción ampliada que aparecen en el volumen II de El Capital. Posteriormente otro economista francés Léon Walras aplicó nociones de la mecánica newtoniana a la economía dando lugar a su teoría del equilibrio generalizado en economía. En su trabajo de 1874 empleó un sistema de coeficientes. Finalmente Leontief presentó un modelo de tablas para la economía de EEUU en 1919 y 1929 fijando la estructura metodológica empleada. En los años cincuenta Leontief revisó su trabajo ofreciendo una metodología de trabajo que posteriormente divulgó en conferencias y publicaciones de carácter económico.1 La aparición de computadoras electrónicas popularizó el método y ya en los años sesenta comenzó a ser estudiado de forma global por diversos autores.

Robin,

it is surprising that you worked on IO tables in 1970's. In 1960's and 70's, there were a kind of fever in IO analysis. But for these 30 years or so, IO analysis became unpopular. Wikipedia (eg) explains it by the decline of interest in microeconomics. It may be a factor, but it is also noteworthy to point that fundamental vision on how economy  works differs between Leontief and other general equilibrium theorists.

However, it seems the trend is changing. By the increasing importance of global value chain, we see a resurgence of IO analysis. OECD started to launch Trade in Valu-Added  (TiVA) data based on International Input-Output Tables. Analysts are beginning to be interested in IIOT in order to see what is happening in the world trade.

It is possible that Word Input Output Model was compiled once or twice, but they could not continue because of difficulty in getting necessary information for the table.

Yoshinori

 Dear Luis Enrique Quintanilla Chacón,

The Spanish article gives the prehistory of IO tables. It is now common that (national) IO Tables are compiled by some government office in most of countries. International (or global or world) IO tables are different. They came to be officially complied in this century. 

The article emphasize that Leontief received influence from Walras. This is true and Leontief endorsed it in various occasion. However, it is important to distinguish general equilibrium à la Arrow and Debreu and general interdependence. Please read Leontief's paper: 

Leontief, Wassily W. (1937) Interrelation of Prices, Output, Savings, and Investment. Review of Economics and Statistics 19(3): 109-132.

At the first sentence of this paper, Leontief wrote as follows:

• This article describes an attempt to apply the economic theory of general equilibrium -or better, general interdependence-to an empirical study of interrelations among the different parts of a national economy as revealed through covariations of prices, outputs, investments, and incomes. (p.109)

Subtitle of this paper reads:

• A STUDY IN EMPIRICAL APPLICATION OF THE ECONOMIC THEORY OF GENERAL INTERDEPENDENCE

Leontief uses expression "general equilibrium" four times, but he uses this in the meaning of "general interdependence." He did not believed in the input substitution. This is the essential and crucial difference between Leontief's IO table vision and Arrow-Debreu type General Equilibrium Theory vision.

Dear  Rossitsa Rangelova,

thank you for interesting story. You have engendered me a series of questions.

Do you still think that IO tables helped much in managing the centrally planned economy? If the most refined version of IO table counts 100 sectors, it does not seem sufficient to dress balances of all goods and services.

Do you know any information on an initiative or plan to compile international IO tables among COMECON countries?  

As I have written to an response to Robin Jarvis above, popularity of IO tables went down during (circa) 1980 to 2010. Can you explain why this phenomenon happened?

Were you aware of differences concepts general equilibrium à la Arrow and Debreu versus general interdependence that W. Leontief preferred? Were there any arguments with this regards in Bulgaria or in Eastern countries before 1989?

I find the subject of input/output tables as applied to a closed national economy to be hard enough. In fact, except for my own work ("Consequential Economics" which is here on LinkedIn) I have not see such a table applied logically to the full activities of a single country. Can anybody correct me?

Dear Yoshinori,

At UEA we were aware of course of the use of IO in the Comecon countries, but we had only a small ICL computer which took 20 minutes, of undergradute computing budget, to determine the general interdependence in an 'unsectored' macroeconomic model of about 10 functions and forecasting (via regression) across a 7-year time series.  Much more data was not available.

I proofed Gosling's monograph on I-O Coefficients, but it was the months after the 1973 oil crisis and was offered to pursue this modelling at post-graduate level with the aim of providing data on the likely effects of the oil price rise throughout the UK economy. But I had to return to the RSA where I substituted by studying for a Cost Accountancy qualification and worked in a large factory. I re-entered the ICT world before completing these.

In these days of 'big data' and government 'intrusion' it should be practical to consider all production companies to submit, with their monthly VAT returns, their fully detailed, "anonymized" cost accounts, for direct and indirect inputs to the national statistical offices for local use and onward aggregation by a UN affiliated office (not the WTO, ILO, IMF or World Bank or any  associates please!!  Some semblance of trust is preferable to none!!).  This cost accounting data would provide the base data set, of a reasonable level of completeness and consistency that might be a good beginning for developing the data sets that will be increasingly important to this work.

In any alternative theoretical narrative which provides a viable alternative to neoclassical equilibrium models, its models, planning and policy developments will depend crucially upon data sets of this granularity.

Robin

I would suspect that the long lag between the periods covered, by quality data, and their availability were too great to allow for timeous analysis and utilisation in policy formulation as the pace of technological change and product proliferation accelerated.  It is an effect similar to the extinction of species which are unable to breed fast enough to evolve adaptations in the time period during which a critical change occurs within their environment from which they cannot migrate to escape.

I can say that in the 70's at UEA we had already observed that in 'classical' regression modelling of the macro-economy the major coefficients determined for use in the models seemed to undergo paradigmatic change at frequencies which were less than is required to accumulate sufficiently long data series with which to establish a statistically significant population size (which was necessarily also the entire sample).  This also had the implication that you already knew when building any 'valid' model that one could not apply it with any confidence into any future period... except by further reducing the sample size from which you were extrapolating.  And that you could only do for a few years, in between coefficient shifts.

It am not sure that the availability of monthly or quarterly date series validly solved the problem of statistical significance... it seems to raise so many other issues.  Although history would seem to suggest they were swept under the carpet in the name of pragmatism and expediency.

It is not very surprising that, in the absence of good data and computing tools, and an increasingly dynamic economic environment that IO tables failed to gain wide use and drive the improvement of data availability... the necessary 'virtuous circle'.    The mainstream clearly had enough work to do finding equilibria and optimisation solutions using time series of high-level data series!

Robin

I have read with much interest the concurrent discussions of Leontief’s original IO analysis, its similarities and dissimilarities with Walras’ general equilibrium theory, although started with a specific question about the first publication of the ‘international’ IO table. I fully agree with Yoshinori Shiozawa that Leontief’s IO models are distinguished from Walrasian models in theoretical scope, despite repeated references by Leontief himself to ‘general equilibrium’. If the former is predicated upon variable factor proportions, the latter, by comparison, upon the fixed proportions. However, in the literature, both have been referred to as general equilibrium models, the latter interpreted as a simplified (simpler) version of the original Wlaras’ theory. In early 1930s Takata Yasuma, Kyoto University economics professor, introduced to his young associate, Kei Shibata, a simplified version of Walras’ GE theory in terms of Gustav Cassel’s fixed proportions system.

The ambitious Shibata then quickly discovered something common in Walras-Cassel’s general equilibrium systems and Karl Marx’s so-called reproduction schema (of capital and labor power). Shibata is also well known to have demonstrated Cassel’s system as mathematically inadequate. (Cassel was known as a full professor of mathematics!)

Anyhow in early 30s Shibata was ‘cordially invited’ by Joseph Schumpeter to visit Harvard where he met Leontief among many other rarely gifted young scholars and discussed matters of mutual interest. Of particular interest to them were Shibata’s theory of ‘inter-related industries’ (生産連携論) and/or Leontief’s IO tables. Both were in terms of, or better yet, predicated upon fixed input coefficients a la Cassel. It may thus be Cassel’s simplified version of Walras’ general equilibrium that inseminated Leontief’s IO foundation, which owes not a little to his discussion with Kei Shibata. (Sadly, this latter economist has long been almost forlorn in the literature. Related obviously to this was the subsequent WWII. The war caused not only their academic correspondences to halt, but also all the more on the part of Shibata as the great economist was purged right after the war from all the public activities, including academic research and teaching in higher education until justice done quite belatedly!)

Anyhow if Leontief is indebted to Shibata on the novel ideas of horizontally related industries represented by input-output tables of coefficients, so are both to Cassel’s simplified GE, and even to Marx’s reproduction schema. In any event they are predicated upon fixed input coefficients. Simple tables of fixed coefficients proved to go a long way. Thus, as we have observed, Leontief’s IO has kept expanded from national IO, regional IO, international IO, and so on along with the number of researchers involved over time to this date. Good luck to them all. Hiroshi Ohta, retired professor, emeritus, or demeritus?   

Dear Hiroshi,  A most interesting contribution!  I have little to add.  Only that it is because we know that IO_coefficients are not fixed that we run into problems of data availability cycles when trying to use IO-models pro-actively in monitoring and policy cycles. However the underlying tables do lend themselves, if made to be accurate in near real-time through statutory requirements (as we go forward with establishing their importance). to accommodating their variability within simulation computations within the overall approach of an evolutionary economics.  Robin

Dear Professor Ohta,

thank you for giving us such interesting information. I have read two of his own books and several papers on him but I did not know this story. 

Dear Robin, Thanks for correcting my slip of tongue as if IO tables were made of fixed coefficients. but as you say, "IO-coefficients are not fixed." And thanks also go to Yoshinori Shiozawa for a quick response with his learned scholarship. My hat off, if the 'two of his own books' you have read are the two thick volumes of his dissertation submitted to Kyoto University. I suspect the copies are now hardly available, but perhaps your library carries them. Anyhow the info I provided is mostly from 'learning by ear,' so would be prepared to contrary evidence. Hiroshi Ohta

My memory is not very clear. One of two books was a small book in a pocket book style (In Search for Economic Laws). It was a kind of autobiography. Another book must be the first volume of the two volume book (Theoretical Economics). I only looked inside to see what kind of topics were treated there. I could not understand what was his own contribution. So, I do not deserve your hat-off.

What I retained in my memory is Kaika no Hōsoku (壊禍の法則).  I remind of the word but I cannot even explain what it really meant. 

I have hardly finished my draft that is planed to be a chapter of a book. The next year 2017 is a bicentennial of the publication of Ricardo's The Principles of Political Economy. We are planning to publish on this occasion a book concerning Ricardo's trade theory and its modern development.

The point I asked in this question page  will appears in  §15. International Input-Output Table. Comments are welcome.

Working Paper The New Theory of International Values: An Overview

Thanks for your rejoinder. K. Shibata was my mentor in theoretical economics training, which was unique and different from the standard graduate education that I belatedly encountered when the US Fulbright Grant enticed me to study in Texas under Mel Greenhut. It is with Mel that I have had the pleasure of writing many papers jointly since. Those outcomes, I must admit, owe much to the past training under my Japanese mentor, K. Shibata.

It is ironic, however, that Shibata had started (after the last war) to dismiss the standard Anglo-Saxon economics entirely! According to him not only Marx is 'upside down', but also Walras' system of 'general equilibrium' and Keynes' 'general theory' are both 'indeterminate.' Anyhow I miss both of my two mentors I loved so much. 

Dear Ohta sensei,

your story on Kei Shibata is interesting. He is an "unknown" economist but deserves more attention. I made a question page on Shibata. Please post all you know about Shibata.  Your story about Shibata's IO table connection is not known widely but deserves to be known much wider.

https://www.researchgate.net/post/What_do_you_know_about_Kei_Shibata_chaitianjing_What_are_the_most_important_contributions_of_Shibata_to_economics

 Dear Shiozawa and Ohta sensei,

Thank for your information, I love all researchers on input-output compilation and analysis, I like Miyazawa model on interregional I/O and demographic - economics more than, I also love the way for expansion them of my teacher Takeo Ihara

 Dear Bui Trihnh,

you are discussing trade in value-added. Then, please read my recent paper (See #20 post in this question page [the last post in p.2]).

My theory of international values has a close relationship with International IO Tables. Although there are one difference on the concept of "country-industry", I believe we have much to do with international IO Tables. 

Dear Bui Trihnh,

I uploaded a wrong version of my paper. Some of figures were broken. I have replaced it with the new PDF and renewed the link. I think now you can see figures in a good state. 

Dear Yoshinori Shiozawa sensei,

Thanks for your reply. I can not get your paper. Could you send me at [email protected]

I knew your book "A New Construction of Ricardian Theory of International Values", Can you give me full text of thí book? Or how can I get it? I also believe international IO to be good for researching on  trade in value added but it dose not exist for Vietnam. Trade relations between Vietnam and China is huge, so if we have the bilateral io between VN-CN will be interesting for us. Thank you sir!

Bui Trinh,

it is possible that you tried to download my paper, while I was replacing the old one with the new PDF version. I have send the same PDF by e-mail.

Yoshinori

Dear Sajad Rajabi

I am not sure when the first IO table was published. In 1976, Leontief and his collaborators published United Nations World Input-Output Model. This may be the first one.

Please see my paper: The New Theory of International Values: An Overview. In Section 15, I have argued how the new theory gives theoretical basis for the International Input-Output Tables. The question is in what sense we can assume that Leontief coefficients are stable. This is important, because we lose most of significance of using Leontief inverse matrix without this stability of coefficients when the world final demand is changed.