Harvard's Economic Research Project

Loentief Relates Economic Theory to Fact Professor's Research Project Perfects Input-Output Anaysis By Soma S. Golden December 17, 1959 For the past ten years, Harvard's Economic Research Project has been perfecting a unique analytical tool developed by Wassily W. Leontief, Henry Lee Professor of Economics. Input-output oLoentief Relates Economic Theory to Fact Professor's Research Project Perfects Input-Output Anaysis By Soma S. Golden December 17, 1959 For the past ten years, Harvard's Economic Research Project has been perfecting a unique analytical tool developed by Wassily W. Leontief, Henry Lee Professor of Economics. Input-output or inter-industry analysis, as the tool is known, evolved from a search by Leontief for a way to apply economic theory to a factual analysis of the entire national economy. In Germany during the middle dle twenties, Leontief studied what he considered one of the greatest problems in economics, that of relating abstract theory to actual fact. "Too often," he explained, "one group of people makes the theories while another assembles the facts." His early work was done in "partial theory," by which the market is dissected and sections of it studied. Such problems as the market effect of a change in the price of copper, or an increase in the supply of meat, were tackled by him and his co-workers during these years. Whole Economy of Interest The more Leontief studied, however, the more convinced he became that the most interesting problems were not those related to small, isolated parts of the market, but rather those dealing with the economy as a whole. In early 1932, he came to Harvard, a lecturer in Economics, and secured funds from the College for research. By 1935 he had developed the first input-output table of United States industry. An input-output table is essentially a double-entry chart of all the transactions between different sectors of the economy. It shows for a particular year what each industry supplies to every other industry and to the final consumer. The table is arranged so that each industry, or supplier, appears as a producer down one side of the chart, and again, in the same order, as a consumer along the top. One of the most important aspects of the analysis is that it allows a detailed study of the sectors of the economy and their effects on each other and on the economy as a whole. It can bring to account every transaction in goods and services. Final demand, which appears down the last right-hand column, is the total of intermediate demands of the industrial sectors, private consumers, government, and overseas customers, plus gross investment. The dollar value of final demand for a particular sector is equal to the sum of the inputs that go into that sector. Inputs are totaled in the columns (up and down), outputs are totaled in rows (left to right). By using a double entry system, the chart is self-checking with total output and total input for each sector resulting in equal dollar values. The table can be as detailed as necessary. Leontief's original version had only 15 catalogues; the latest U.S. table, made in 1947, by the government, included 450 industries. Other Products Required Preparation of the table just described, which shows input-output relations in current money values is only the first step of Leonteif's analysis. The second step is to "invert the matrix," producing a table of coefficients that shows the amounts of every other product required to bring into existance a dollar's worth of any given product. From this second table forecasts can be made of the effect throughout all industries of a change in demand for the produce of one industry. The statistical computations for "inverting the matrix" are multitudinous and require the solution of so many simultaneous equations that electronic computers must be used. Presently the Research Project is using a Univac and an IBM 650. Calculations on household consumption are being made with M.I.T.'s 704. ADVERTISEMENT Dynamic Table Needed One of the main limitations of the table is it necessitates (assuming) technological conditions unchanged for the period of the forecast. A large, 450-catagory input-output table takes about two years to complete. This time lag, however, has not proven as serious as critics had predicted. In the United States, calculations based on 1947 figures were found to apply closely to conditions in 1952. Leontief and members of the Project are, however, developing a dynamic model to remove this limitation. International attention has been attracted by the analytical device. Both advanced economies and underdeveloped countries have adopted Leontief's mode to answer specific questions that are pertinent to them. Underdeveloped countries are primarily interested in it for planning purposes; while advanced economies, particularly in Europe, are employing it to help solve trade balance problems. Soon after World War II Italy, Holland, and Norway evolved their own charts. In Italy, under the Marshall Plan and private sponsorship, a table was developed to answer questions about the entire Italian economy and to assist in understanding the problems of relatively underdeveloped Southern Italy. Norway and Holland were largely interested in solving import-export problems, and used the new system along with their national income accounts to see what effect a shift in the trade balance would have on industry. Several South American nations, including Argentina, Peru, Chile, and Equador are also using Leontief's device. Great Britain has employed it for several years. Russians Interested The Russians have also manifested interest in input-output analysis, and are now training people to perform the necessary operations. Leontief, who visited his native Russia briefly last spring, believes that although they are "having a hard time justifying the use of an analytical tool developed by a capitalist," the Russians will resort to it soon. Certain revisions have to be made in U.S.S.R. statistical methods in order to facilitate use of the analysis, Leontief added. Underdeveloped Countries Underdeveloped countries also have to overcome difficulties in formulating tables to use for planning. Although their economies are relatively uncomplicated, problems do arise in the collection of data, according to Mrs. E. W. Gilboy, assistant director of the Research Project and lecturer in Economics. Another difficulty for them is collecting funds for purchase of the giant computers necessary to construct the chart after the data is secured. Spain, which set up a table several years ago, had to send information to Italy, where computer work was done on Italian machines. The Economic Research Project was set up 10 years ago by a grant from the Rockefeller Foundation to Leontief, "for research in the structure of the United States economy." Beginning in 1950, small Air Force grants were received, but these were discontinued in 1952, when the Administration changed. Ford Foundation then joined in support of the project. Since then a small grant from the National Science Foundation for large scale computation experimentation has also been donated. ADVERTISEMENT Basic Research The Project's work is not to construct actual tables. The cost of collecting the extremely large amount of data is beyond the scope of so small a group. The 25 Faculty members in the project conduct basic research, investigating different facets of Leontief's model. Among the studies now in progress is one on technological change and the methods through which new techniques are diffused through industry--this is coordinated with Leontief's work on a dynamic model. Application of the table to regional problems is being perfected; while various sectors of the economy--particularly household consumption and natural resources--are being given special attention. About every 18 months, the Project prints a report on research to date, which has been sent free to interested people all over the world. The demand for these reports has become so great recently. Mrs. Gilboy remarked, that soon some charge will have to be made. Leontief plans to publish a series of small volumes to present the results of research made since 1953, when his last book (The Structure of the American Economy) on the project, was released. The Research Project building, locat- ed on Massachusetts Ave., near the Law School also serves as a training center for interested visitors from foreign countries. Both Ford and Rockefeller Foundations assign Fellows to study research being conducted by the Project for periods extending from three to six months. Private groups also sponsor visitors. According to Mrs. Gilboy, the growing number of these students is beginning to tax facilities of the Project. Both working space and staff assistance are available to the Foundation-sponsored Fellows. In spite of the adoption of input-output analysis by 35 interested foreign countries, the United States government has completely neglected the system in recent years. The last table devised for the U. S. economy was the 450-category chart made up in 1947. Interest at that time was occasioned by a wish to know the impact throughout the economy of an expansion of government spending on arms--the extent to which other types of production would have to contact, and the "bottle necks" that might arise. Value Demonstrated Wartime experience, when demand is largely in the hands of the Federal Government, demonstrates the value of a realistic guide to such future events: it could assist authorities to control scarce materials, to encourage production or substitution for them well in advance, and to arrange for more imports. Recently, however, the Department of Commerce announced plans to publish a new small table. This move shows that the government is actually interested in the analysis, and Leontief's predictions of wider U. S. government use in the future seem nearer fulfillment. Small - Scale Table The small-scale table that the Department plans, Mrs. Gilboy pointed out, is not really adequate for an economy as large and diverse as this country's. One with at least 450 industrial sectors is needed. Only the government or large scale industries could afford the million dollar cost of such a project. One reason for the government's lack of enthusiasm seems to be a general fear of "centralized planning.' ADVERTISEMENT Interest in European nations has been prompted by import-export considerations. Underdeveloped economies are anxious to pursue a rapid planned growth process, which is greatly facilitated by this type of analysis. Commercial Uses Outside the government American industries and local groups have developed several input-output tables. The Pennsylvania Railroad has based one on industries and commercial establishments that have grown up alongside its tracks. A massive report on the results of its research was published by the railroad. "This is not an academic project," Leontief remarked, "but it is practical." In St. Louis a table based on the metropolitan area has been constructed. A banking house in Berkeley, Calif., has also completed a local chart. Industrialists Benefit The value of input-output analysis to individual industries and local areas is perhaps more limited than its value to a national government, but an industrialist does benefit by knowing the extent to which demand for his product might be affected by economic change. Input-output analysis will not replace, and was not intended to replace the entrepreneur's vital role of seeking profits by anticipating changes in taste and technology, but it does provide throughout the economic system many useful indicators of the results of a change--large or small--in one of the sectors Want to keep up with breaking news? Subscribe to our email newsletter.r inter-industry analysis, as the tool is known, evolved from a search by Leontief for a way to apply economic theory to a factual analysis of the entire national economy. In Germany during the middle dle twenties, Leontief studied what he considered one of the greatest problems in economics, that of relating abstract theory to actual fact. "Too often," he explained, "one group of people makes the theories while another assembles the facts." His early work was done in "partial theory," by which the market is dissected and sections of it studied. Such problems as the market effect of a change in the price of copper, or an increase in the supply of meat, were tackled by him and his co-workers during these years. Whole Economy of Interest The more Leontief studied, however, the more convinced he became that the most interesting problems were not those related to small, isolated parts of the market, but rather those dealing with the economy as a whole. In early 1932, he came to Harvard, a lecturer in Economics, and secured funds from the College for research. By 1935 he had developed the first input-output table of United States industry. An input-output table is essentially a double-entry chart of all the transactions between different sectors of the economy. It shows for a particular year what each industry supplies to every other industry and to the final consumer. The table is arranged so that each industry, or supplier, appears as a producer down one side of the chart, and again, in the same order, as a consumer along the top. One of the most important aspects of the analysis is that it allows a detailed study of the sectors of the economy and their effects on each other and on the economy as a whole. It can bring to account every transaction in goods and services. Final demand, which appears down the last right-hand column, is the total of intermediate demands of the industrial sectors, private consumers, government, and overseas customers, plus gross investment. The dollar value of final demand for a particular sector is equal to the sum of the inputs that go into that sector. Inputs are totaled in the columns (up and down), outputs are totaled in rows (left to right). By using a double entry system, the chart is self-checking with total output and total input for each sector resulting in equal dollar values. The table can be as detailed as necessary. Leontief's original version had only 15 catalogues; the latest U.S. table, made in 1947, by the government, included 450 industries. Other Products Required Preparation of the table just described, which shows input-output relations in current money values is only the first step of Leonteif's analysis. The second step is to "invert the matrix," producing a table of coefficients that shows the amounts of every other product required to bring into existance a dollar's worth of any given product. From this second table forecasts can be made of the effect throughout all industries of a change in demand for the produce of one industry. The statistical computations for "inverting the matrix" are multitudinous and require the solution of so many simultaneous equations that electronic computers must be used. Presently the Research Project is using a Univac and an IBM 650. Calculations on household consumption are being made with M.I.T.'s 704. ADVERTISEMENT Dynamic Table Needed One of the main limitations of the table is it necessitates (assuming) technological conditions unchanged for the period of the forecast. A large, 450-catagory input-output table takes about two years to complete. This time lag, however, has not proven as serious as critics had predicted. In the United States, calculations based on 1947 figures were found to apply closely to conditions in 1952. Leontief and members of the Project are, however, developing a dynamic model to remove this limitation. International attention has been attracted by the analytical device. Both advanced economies and underdeveloped countries have adopted Leontief's mode to answer specific questions that are pertinent to them. Underdeveloped countries are primarily interested in it for planning purposes; while advanced economies, particularly in Europe, are employing it to help solve trade balance problems. Soon after World War II Italy, Holland, and Norway evolved their own charts. In Italy, under the Marshall Plan and private sponsorship, a table was developed to answer questions about the entire Italian economy and to assist in understanding the problems of relatively underdeveloped Southern Italy. Norway and Holland were largely interested in solving import-export problems, and used the new system along with their national income accounts to see what effect a shift in the trade balance would have on industry. Several South American nations, including Argentina, Peru, Chile, and Equador are also using Leontief's device. Great Britain has employed it for several years. Russians Interested The Russians have also manifested interest in input-output analysis, and are now training people to perform the necessary operations. Leontief, who visited his native Russia briefly last spring, believes that although they are "having a hard time justifying the use of an analytical tool developed by a capitalist," the Russians will resort to it soon. Certain revisions have to be made in U.S.S.R. statistical methods in order to facilitate use of the analysis, Leontief added. Underdeveloped Countries Underdeveloped countries also have to overcome difficulties in formulating tables to use for planning. Although their economies are relatively uncomplicated, problems do arise in the collection of data, according to Mrs. E. W. Gilboy, assistant director of the Research Project and lecturer in Economics. Another difficulty for them is collecting funds for purchase of the giant computers necessary to construct the chart after the data is secured. Spain, which set up a table several years ago, had to send information to Italy, where computer work was done on Italian machines. The Economic Research Project was set up 10 years ago by a grant from the Rockefeller Foundation to Leontief, "for research in the structure of the United States economy." Beginning in 1950, small Air Force grants were received, but these were discontinued in 1952, when the Administration changed. Ford Foundation then joined in support of the project. Since then a small grant from the National Science Foundation for large scale computation experimentation has also been donated. ADVERTISEMENT Basic Research The Project's work is not to construct actual tables. The cost of collecting the extremely large amount of data is beyond the scope of so small a group. The 25 Faculty members in the project conduct basic research, investigating different facets of Leontief's model. Among the studies now in progress is one on technological change and the methods through which new techniques are diffused through industry--this is coordinated with Leontief's work on a dynamic model. Application of the table to regional problems is being perfected; while various sectors of the economy--particularly household consumption and natural resources--are being given special attention. About every 18 months, the Project prints a report on research to date, which has been sent free to interested people all over the world. The demand for these reports has become so great recently. Mrs. Gilboy remarked, that soon some charge will have to be made. Leontief plans to publish a series of small volumes to present the results of research made since 1953, when his last book (The Structure of the American Economy) on the project, was released. The Research Project building, locat- ed on Massachusetts Ave., near the Law School also serves as a training center for interested visitors from foreign countries. Both Ford and Rockefeller Foundations assign Fellows to study research being conducted by the Project for periods extending from three to six months. Private groups also sponsor visitors. According to Mrs. Gilboy, the growing number of these students is beginning to tax facilities of the Project. Both working space and staff assistance are available to the Foundation-sponsored Fellows. In spite of the adoption of input-output analysis by 35 interested foreign countries, the United States government has completely neglected the system in recent years. The last table devised for the U. S. economy was the 450-category chart made up in 1947. Interest at that time was occasioned by a wish to know the impact throughout the economy of an expansion of government spending on arms--the extent to which other types of production would have to contact, and the "bottle necks" that might arise. Value Demonstrated Wartime experience, when demand is largely in the hands of the Federal Government, demonstrates the value of a realistic guide to such future events: it could assist authorities to control scarce materials, to encourage production or substitution for them well in advance, and to arrange for more imports. Recently, however, the Department of Commerce announced plans to publish a new small table. This move shows that the government is actually interested in the analysis, and Leontief's predictions of wider U. S. government use in the future seem nearer fulfillment. Small - Scale Table The small-scale table that the Department plans, Mrs. Gilboy pointed out, is not really adequate for an economy as large and diverse as this country's. One with at least 450 industrial sectors is needed. Only the government or large scale industries could afford the million dollar cost of such a project. One reason for the government's lack of enthusiasm seems to be a general fear of "centralized planning.' ADVERTISEMENT Interest in European nations has been prompted by import-export considerations. Underdeveloped economies are anxious to pursue a rapid planned growth process, which is greatly facilitated by this type of analysis. Commercial Uses Outside the government American industries and local groups have developed several input-output tables. The Pennsylvania Railroad has based one on industries and commercial establishments that have grown up alongside its tracks. A massive report on the results of its research was published by the railroad. "This is not an academic project," Leontief remarked, "but it is practical." In St. Louis a table based on the metropolitan area has been constructed. A banking house in Berkeley, Calif., has also completed a local chart. Industrialists Benefit The value of input-output analysis to individual industries and local areas is perhaps more limited than its value to a national government, but an industrialist does benefit by knowing the extent to which demand for his product might be affected by economic change. Input-output analysis will not replace, and was not intended to replace the entrepreneur's vital role of seeking profits by anticipating changes in taste and technology, but it does provide throughout the economic system many useful indicators of the results of a change--large or small--in one of the sectors Want to keep up with breaking news? Subscribe to our email newsletter.