Wassily Wassilyevich Leontief

Who he is and his times

Wassily Wassilyevich Leontief was born on 5 August 1905 in Munich, where his parents were living at the time; his childhood and youth were spent in Petersburg. His father was an economist and a teacher. For a long time 1906 was considered his year of birth, and it is this year that appears in older biographies, including the Nobel one, but modern research and encyclopedias give 1905. In 1921–1924 Leontief studied at the Faculty of Social Sciences of Petrograd University and graduated at the age of 19.

As a student he spoke out for academic autonomy and freedom of speech and, according to his biographers, was detained by the Cheka several times. In 1925 he was allowed to leave the USSR — according to a widespread version, the authorities believed he was terminally ill, and the diagnosis later proved to be mistaken. Leontief went to Berlin, in 1928 defended his dissertation The Economy as a Circular Flow at the University of Berlin under the supervision of Werner Sombart, worked at the Institute for the World Economy in Kiel and in 1928–1929 advised the Ministry of Railways of China.

In 1931, at the invitation of the National Bureau of Economic Research, he moved to the United States and soon began working at Harvard University, where he became a full professor in 1946. In 1948 he founded the Harvard Economic Research Project and directed it until 1973; in 1975 he moved to New York University, where he established the Institute for Economic Analysis. Leontief died on 5 February 1999 in New York.

For the history of Soviet planning, the most important thing is the starting point of his path. In 1926 the Central Statistical Administration of the USSR, headed by its first director P. I. Popov, published The Balance of the National Economy of the USSR for 1923–24 (Balans narodnogo khozyaystva Soyuza SSR 1923–24 goda). Its summary tables contained an asymmetric matrix of up to 79 by 49 items, showing which products of which industries went to productive and non-productive consumption. In B. T. Ryabushkin's assessment, this work had no counterpart in world statistics of the time.

Main ideas

One of Leontief's first scholarly works was a methodological analysis of this balance. The article appeared in 1925, first in German in the Kiel journal of world economics, and then, in an abridged Russian translation, in the journal Planovoe Khozyaystvo under the title The Balance of the Economy of the USSR: A Methodological Critique of the Work of the Central Statistical Administration. According to historians of statistics, the Soviet balance attracted the young economist precisely because it was the first attempt to describe in figures the production and distribution of the social product as a single system of interconnections.

From this idea grew the input-output model, or inter-branch balance (MOB). Imagine a “chessboard” table: the rows are industries as producers, the columns are the same industries as consumers. The cell at the intersection records how much of one industry's output was used by another. By dividing these flows by the output of the consuming industry, one obtains the coefficients of direct inputs: how much steel, coal or electricity is needed per unit of engineering output. By adding final demand — household consumption, investment, exports — one can solve a system of linear equations and find out how much each industry must produce to satisfy that demand.

The main discovery of the method is total inputs. To produce more cars, more steel is needed, but steel requires coal and electricity, coal requires machines, and so on along the chain. The matrix of total input coefficients takes all these direct and indirect links into account at once. This is exactly why the inter-branch balance makes it possible to answer “what if” questions: how output and employment by industry will change if military procurement is cut or exports grow.

Leontief's first work devoted entirely to the method appeared in 1936 in the Harvard journal of economics and statistics and contained a table for the US economy for 1919. In 1941 came the book The Structure of American Economy, 1919–1929, in which the tables covered several dozen industries. The method rested on a strong simplification: inputs were assumed to be proportional to output, that is, the technology of an industry in the model does not change when volumes change. Leontief deliberately set his approach against indirect statistical estimates: he preferred “direct observation” — technological and accounting data collected not by economists but by engineers and accountants.

An important development of the model is associated with the distinction between the closed and open schemes. In the open model, final demand — government purchases, investment, exports, household consumption — is taken outside the system of equations as an external quantity, and the output of industries is calculated as a response to it. It was precisely this formulation that allowed Leontief, in a 1944 article, to ask how ending military purchases of aircraft, guns, tanks and ships would affect the level of employment if it were not offset by growth in other demand. The model became an instrument for scenario calculations, not merely a description of the past year.

What became of the ideas in practice

In the United States the method was put to government use earlier than in the USSR. In 1941 the Bureau of Labor Statistics hired Leontief and began building a 95-industry table for 1939 in order to assess the consequences of demobilization for employment. After 1948, work on the table for 1947 was financed by the US Air Force's interagency project for computing optimal programmes; its initial level of detail reached 450 industrial and 50 autonomous sectors. In 1953, after the Eisenhower administration came to office, the Department of Defense stopped the funding, and the work was resumed only after the recommendations of an expert committee in 1956.

In the USSR the fate of the balance idea was more dramatic. On 27 December 1929, at a conference of Marxist agrarian scholars, Stalin declared that the balance of the national economy published by the TsSU in 1926 was “not a balance but juggling with figures”, and rejected the treatment of the balance problem by Bazarov and Groman. Soviet planning took the path of material balances for individual products, without a model of inter-branch links. Only in the 1960s did the laboratory of Academician V. S. Nemchinov reconstruct the 1923/24 balance in a symmetric format of 79 by 79 industries.

The inter-branch balance returned with the Thaw and the development of the economic-mathematical school. On the instructions of the government, the Central Statistical Administration of the USSR, under the direction of V. A. Sobol and M. R. Eidelman, compiled a reporting inter-branch balance of the production and distribution of output for 1959: a value balance for 83 industries, as well as a balance in physical terms. For it, a special sample survey of inputs was carried out in all branches of material production. After that, “large” reporting balances were compiled for 1959, 1966, 1972, 1977, 1982 and 1987, and “small” ones annually from 1975 to 1990, and, starting with the report for 1966, for the union republics as well.

Planning balances were built on the basis of the reporting balances: according to researchers, the first planning inter-branch balances in value and physical terms were built in 1962. The work was carried out by the TsSU, the Research Economic Institute attached to the USSR Gosplan and academic institutes; in 1968 a group of developers including Eidelman, Yershov, Klotsvog and Shatalin received the USSR State Prize. The balance was used most extensively in planning with the creation of the automated system of plan calculations of the USSR Gosplan. For 1987 a physical-and-value balance was compiled with a breakdown by 26 ministries and agencies.

What worked and what did not

The function of checking consistency worked. An inter-branch table is built on the principle of double entry: the deliveries of each industry must match the inputs of all the others. That is why it mercilessly exposes gaps in statistics. American historians of statistics wrote that the main lesson of the work on the 1947 table was the realization of how incomplete, unsystematic and contradictory the basic economic data of the United States were. Reconciliation with the national accounts forced, for example, a revision of the estimate of construction output. Today input-output tables are built in more than 90 countries; in Russia, within the framework of the system of national accounts, starting with data for 1995.

Practice also changed the method itself. Staff of the Bureau of Labor Statistics, in particular the national accounts specialist Marvin Hoffenberg, separated the current and capital expenditures of industries in order to reconcile the table with the national accounts; the 1939 table acquired an investment column and a depreciation row. Thanks to this, the inter-branch balance became compatible with the system of national income accounts. This is a good example of how a methodology improves not only at a university but also when it collides with real data and the requirements of government accounting.

What did not work in the USSR was the link between calculation and decision. V. V. Kossov, himself a participant in developing the balances, writes that the CPSU Central Committee did not want to take into account calculations made on the basis of models. He gives an example: model calculations justified a share of capital investment in agriculture of about 18 %, but a share of 27.2 % was adopted. There were also technical obstacles: multiple prices for the same products and the incompatibility of the reporting classification with the planning one. The instrument existed, prizes were awarded for it, but it was never given a decisive voice.

There are also limitations of the method itself. Constant input coefficients describe the short term well but describe technical progress and the substitution of some resources for others poorly; economists criticized this from the very beginning. In addition, the inter-branch balance works at the level of industries: 83 or even 118 items are aggregates, not specific products. It shows how much metal is needed in general, but not which section, for which plant and by what date. The gap between an item of the plan and a specific product, which in Soviet practice was bridged by tolkachi (“pushers”) and informal exchange, was not eliminated by the inter-branch balance.

Debate and criticism

In the USSR the first criticism was ideological. The TsSU balance and the school of statisticians associated with it were rejected not for methodological errors but as a non-Marxist approach, and dropped out of practice for a long time. Leontief himself, an émigré and an American professor, could not serve as a positive reference in Soviet science until the end of the 1950s. The return of the method became possible together with the rehabilitation of mathematical methods in economics, associated with the names of Nemchinov, Kantorovich and Novozhilov.

In Western scholarship the debate was methodological. Economists, including the circle of the Cowles Commission, preferred models that allowed for the substitution of resources and the estimation of parameters by regression methods. Leontief replied that reliable measurement rests not only on sophisticated statistical technique but also on systematic direct observation. This debate over what matters more — the theoretical flexibility of a model or the empirical transparency of its data — has not been settled even today.

There was a political side as well. A method that makes it possible to calculate how much each industry must produce looked like an instrument of centralized planning, and in the United States the ending of military funding in 1953 coincided with a change of administration. At the same time, history shows that the method is ideologically neutral: the same tables were used in the planned economy of the USSR and in market economies, and eventually became part of the international system of national accounts.

Throughout his career Leontief himself remained a critic of his own profession. He opposed economists' fascination with theoretical assumptions not grounded in observable facts and insisted on the systematic collection of primary data. This position makes him an inconvenient ally both for supporters of a centralized plan and for supporters of purely market theory: he demanded the same thing from both sides — show the data and the calculation.

Legacy today

In 1973 Leontief received the Prize in Economic Sciences in Memory of Alfred Nobel “for the development of the input-output method and for its application to important economic problems”. Four of his doctoral students later also became Nobel laureates: Paul Samuelson, Robert Solow, Vernon Smith and Thomas Schelling. In 1977 The Future of the World Economy, prepared at the request of the UN, was published; it used inter-branch models. In 1988 Leontief was elected a foreign member of the USSR Academy of Sciences.

Leontief also applied input-output analysis to testing economic theories. In 1953 he examined how much labour and capital was embodied in US exports and imports and obtained a result that entered textbooks as the “Leontief paradox”: American exports turned out to be relatively more labour-intensive than imports, contrary to the expectations of standard international trade theory. This example shows that the input-output table does not belong to one economic system but is a way of checking theory against data.

Today the inter-branch balance is a standard part of official statistics. In Russia, benchmark input-output tables are compiled within the framework of the system of national accounts; a government directive of 2009 provided for the compilation of benchmark tables for 2011. The method is used to assess the multiplier effects of investment, the consequences of sanctions and price shocks, and for regional calculations. It is one of the few instruments of the Soviet statistical tradition that has outlived both the Soviet system and its critics.

For Open Planning, Leontief matters twice over. First, he provided a way to calculate the consequences of decisions along the entire chain of links, not only in the industry that a decision directly concerns. Second, his story is a warning: a model that exists alongside a decision but is not required to explain that decision turns into decoration. The calculation must be open and reproducible, and any divergence between the decision taken and the calculation must be public.

Links between ideas

Whom the author argued with, whose ideas they developed, whom they answered and whose fate they shared. All links are on the shared graph. Graph of ideas

• Vladimir Bazarov and Vladimir Groman

  Develops ideas ←

  In 1925 Leontief analysed the TsSU balance for 1923/24 and later developed it into the input-output method.

• Vasily Nemchinov

  Develops ideas →

  The inter-branch method returned to the USSR together with the economic-mathematical school headed by Nemchinov.

• Viktor Novozhilov

  Parallel path →

  In Nemchinov’s assessment, the pre-war works of Novozhilov and Kantorovich in a certain sense anticipated inter-branch analysis.

• Nikolai Fedorenko

  Practice and institutions →

  The inter-branch balance reached its widest scope in Gosplan’s ASPR, the project linking Gosplan and TsEMI.

• Paul Cockshott and Allin Cottrell

  Develops ideas →

  Cockshott and Cottrell solve the inter-branch balance system by successive approximations to calculate labour values.

Timeline

1. 1905

   Born in Munich (some older sources give 1906)

2. 1925

   Critique of the balance of the national economy of the USSR in Planovoe Khozyaystvo; leaves the USSR

3. 1928

   Dissertation The Economy as a Circular Flow at the University of Berlin

4. 1931

   Moves to the United States, works at the National Bureau of Economic Research and Harvard

5. 1936

   First article on the input-output method, with a table for the US economy

6. 1941

   Book The Structure of American Economy, 1919–1929; start of work with the Bureau of Labor Statistics

7. 1959

   The Central Statistical Administration of the USSR compiles the first reporting inter-branch balance, for 1959

8. 1973

   Prize in Economic Sciences in Memory of Alfred Nobel

9. 1988

   Elected a foreign member of the USSR Academy of Sciences

10. 1999

    Dies in New York on 5 February

Key works

• 1925The Balance of the Economy of the USSR: A Methodological Critique of the Work of the Central Statistical Administration
• 1928The Economy as a Circular Flow (Die Wirtschaft als Kreislauf; dissertation, University of Berlin)
• 1936Quantitative Input and Output Relations in the Economic System of the United States
• 1941The Structure of American Economy, 1919–1929
• 1953Studies in the Structure of the American Economy (with co-authors)
• 1966Input-Output Economics
• 1977The Future of the World Economy (report for the UN)
• 1990Economic Essays: Theories, Research, Facts and Policy (Ekonomicheskie esse: teorii, issledovaniya, fakty i politika), Russian edition

Sources

1. Нобелевский комитет (NobelPrize.org). The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel 1973: Wassily Leontief — Facts. 1973
2. Большая российская энциклопедия. Леонтьев Василий Васильевич
3. А. В. Воронцовский. В. В. Леонтьев — выдающийся экономист ХХ столетия. 2007
4. Ю. Н. Иванов, Б. Т. Рябушкин. К 90-летию первого баланса народного хозяйства: истоки и эволюция макроэкономической статистики России. 2016
5. Б. Т. Рябушкин. Межотраслевой баланс — информационно-методологическая основа структурного макроэкономического анализа: этапы развития
6. В. В. Коссов. Возрождение межотраслевого баланса в СССР. 2014
7. Martin C. Kohli. Wassily Leontief and the Bureau of Labor Statistics, 1941–54: Developing Interindustry Relations. 2002
8. Л. В. Шелехова, З. У. Блягоз, А. В. Нагоев, В. А. Тешев. Межотраслевой баланс и модель «Затраты — выпуск»: история создания и перспективы развития. 2015
9. И. В. Сталин. К вопросам аграрной политики в СССР. Речь на конференции аграрников-марксистов 27 декабря 1929 г.. 1929
10. Wassily W. Leontief. Quantitative Input and Output Relations in the Economic Systems of the United States. 1936
11. Википедия (англоязычная). Wassily Leontief

Prepared from open sources. Dates, titles and quotations were checked against the source list at the bottom of the page; where sources disagree, the text says so.