What is going to evolve, I do not know. Our present system is done. If it were not for ERP and cumulative purchasing power during the war, we would be in terrible condition.

One of our board of directors, a friend of mine, wrote a book, The Depression Decade, what happened under the New Deal. He did not criticize it, but the title of the concluding chapter of his turns the trick-War to the Rescue.

Mr. Chairman, after 10 years of the New Deal in 1942, returns to ownership and control of property in the United States had increased four times as much as to labor. You cannot touch the tax problem today except as part of the whole economic system. We have got to make changes gradually. My own conviction is now is the time to begin-and Senator George has been listening to me for decades very patiently. The big job is how to combine the unquestioned benefits of private initiative with the need of larger collective ownership. Thank you for your courtesy.

I do not expect you to accept all our suggestions, but the ones you do accept will be gratifying to a very long-suffering people. The CHAIRMAN. I want to thank you, Mr. Marsh.

We will recess until 2 o'clock.

(Whereupon, at 12: 35 p. m., a recess was taken until 2 p. m. of the same day.)

AFTERNOON SESSION

(The committee reconvened at 2 p. m., upon the expiration of the noon recess.)

The CHAIRMAN. The meeting will come to order.

Is Dr. Roos here?

Will you come forward please, Doctor?

We are very glad to have you, and are glad that we were permitted to go ahead this afternoon so that we could hear you.

Will you be seated, Doctor, and give the reporter your full name, your residence, and your occupation?

STATEMENT OF DR. CHARLES F. ROOS, PRESIDENT, THE

ECONOMETRIC INSTITUTE, NEW YORK, N. Y.

Dr. Roos. My name is Charles F. Roos. I live at 817 Fifth Avenue, Manhattan, N. Y.

I am president of the Econometric Institute, Inc., a business research and consulting organization, with offices at 500 Fifth Avenue, New York City.

I am also president of the Index Number Institute, Inc., which publishes the Irving Fisher Price Index and a weekly business service known as Economic Measures.

The CHAIRMAN. Proceed, please.

Dr. Roos. I am an economist, statistician, and mathematician by education and training. I received my B. A. degree in 1921, my M. A. degree in 1924, and my Ph. D. in 1926 from the Rice Institute, Houston, Tex.

After receiving my doctor's degree, I spent 15 months in study at the University of Chicago and 9 months more at Princeton and Chi

cago Universities as a fellow of the National Research Council. I was then appointed assistant professor of mathematics at Cornell University.

In the same year, 1928, I was elected secretary of section K, economics, sociology, and statistics, of the American Association for the Advancement of Science.

In 1931 I left Cornell University to become permanent secretary of this scientific organization.

In 1933 I left this position to do special research on changing economic conditions as a fellow at the Guggenheim Memorial Foundation. While I was engaged in this study in London, England, I was invited to become principal economist and research director of the National Recovery Administration in Washington.

I left this organization in September 1934 to become research director of the Cowles Commission for Research in Economics, which was then in Colorado Springs, Colo., but is now affiliated with the University of Chicago.

In 1937 I left the Cowles Commission to set up my own research and consulting business, which, as I have said, is known as the Econometric Institute, Inc. By way of explanation, I may add that I helped to coin the word "econometric" to mean primarily economic measurement or the development and testing of economic theory. In 1939 I bought the Index Number Institute from Prof. Irving Fisher.

My organizations, composed of about 80 persons, have always specialized in the forecasting of national income, and its components,. incidentally, and production, and in the translation of these forecasts into demand and supply and price levels for industry.

Today there are several hundred major corporations in all fields. of business activity which my organizations serve. I believe that they enjoy the unique position of having always identified correctly the trends of production and income and of having always forecast the turns a few months in advance.

For example, in November 1936, before I organized the Econometric Institute, I questioned the continuance of the upward trend of business. I began to turn bearish on February 9, 1937, and on September 9, 1937, predicted business and financial panic. The Econometric Institute, which began business on April 1, 1938, indicated on May 9, 1938, that the bottom of business had been reached. Financial and commodity markets confirmed this forecast nearly 6 weeks later.

The first new steel capacity to come into production during the war resulted from a forecast of the Econometric Institute in early 1939 that the economy was just entering the capital goods or boom phaseof business for which additional steel capacity would be needed.

Beginning in November 1943 forecasts of income and production were made only by the Index Number Institute. In February 1944 this institute forecast postwar industrial production as measured by the Federal Reserve index at 170 and 185 percent, respectively, of the 1935-39 average level in the first two postwar years, full employment, rising wages, steel and other material shortages, and inflationary price trends.

The production index averaged 170 in the first postwar year and 187 in the second. The accuracy of the other forecasts is well known.

Late in 1946, when bearish forecasts were sweeping the world, the institute's studies indicated rising production, rising income, and rising prices and its forecasts to that effect are on record.

On November 17, 1947, the institute indicated that the high of the price level for several years would be made in the period of February to May of 1948.

We have no crystal balls and no telephone connections with God. All we have to offer is tested economic measures of business and financial pressures which are likely to maintain or change a business trend. The expected rates of taxation are of fundamental im portance, particularly when business is operating at capacity.

We have found that in making scientific forecasts of business a good starting point is the detail of the Federal Reserve Index of Industrial Production.

As you know, that is an index of physical volume; not of dollars, but of units.

Industrial production is related to a man-hours worked and the man-hours times the average wage per hour yields pay rolls. Moreover, man-hours required in the distribution and service industries are closely related to this industrial production and can be forecast from it. Man-hours in all industry, together with hourly earnings, determine the demand for agricultural products, and the relation of this demand and export demand to the supply determines the price level of agricultural products and thus farm income.

In 1947 wages and salaries of workers engaged in the production and distribution of industrial and farm products and in government constituted about 71 percent of the personal income of the entire population. Entrepreneurial profits of shopkeepers, professional persons, and others working for themselves are closely related to this income and in 1947 accounted for about 12 percent more of personal income.

Interest payments, rents, royalties, pensions, unemployment insurance, military and other bonuses, and dividends constituted the remaining 17 percent. We have charts which enable us to translate industrial production into the various components of income and so from forecasts of production to arrive at forecasts of income.

The industry break-down as presented by the Federal Reserve index of industrial production can be subdivided and rearranged as shown in chart I.

Now, I have here a copy of that chart.

The CHAIRMAN. May we have it for the record?

Dr. Roos. You may have it for the record.

(The chart referred to will be found on p. 139.)

Dr. Roos. In this form the highly variable segments are pointed up, and it is thus shown wherein lies the core of the forecasting problem. Even a casual glance at the chart shows that the highly variable components of production are capital goods, construction materials, and consumers' durable goods. Indeed, production of consumers' perishable goods and consumers' semidurable goods can be forecast even by a novice with relatively good accuracy from population data and growth in consumption relative to this population. On the other hand, the forecasting of production of the durable-goods components is much more difficult. Yet, if one is to make a useful forecast of total production and, as I shall show, of national income, he must make a reliable forecast of production of these durable goods.

The capital-goods component, which is the most variable segment of economic activity, is a unit-volume index. It may be forecast directly from the level of demand for consumers' goods and construction materials, the ratio of this expected demand to capacity and the rate of interest on long-term bonds.

A better forecast can be obtained by working with the dollar figures representing producers' durable-equipment expenditures and then converting these forecasts of dollar expenditures to unit volume by means of a price index of machinery and machinery products.

Chart II taken from my paper, The Demand for Investment Goods, which I presented in December 1947 to a joint meeting of the American Economic Association and the Econometric Society, shows the relationship that exists between expenditures for producers' durable equipment and a composite of several variables 6 months previous. The chart shows that producers' durable equipment can be forecast 6 months ahead from present values of (1) corporate profits, (2) interest rates, and (3) the ratio of the Bureau of Labor Statistics index of prices of commodities other than farm products to the same Bureau's index of prices of metals and metal products. The agreement between the expenditures and the forecasting variable is as close as one finds in the physical sciences, except in the war years when we know that the War Production Board purposefully held down equipment activity by licensing.

(The chart referred to will be found facing this page.)

Dr. Roos. Despite the sharp increase in bond yields brought on by the tightening money policy of the Federal Reserve Board and the excessive-tax rates of the present period, corporate profits and relative prices are sufficiently strong to indicate expenditures for producers' durable equipment at an annual rate of about $17,000,000,000 during the first half of 1948. You will notice from the chart that this would be below the figure for 1947. However, this amount would be augmented by deferred demand of about one more billion dollars.

In the second quarter of 1948, however, if present tax rates continue, corporate profits are likely to decline as a result of increasing competition and the excessive drain on consumer purchasing power by Treasury receipts, coming at a time when wages and other costs are increasing. This same competition will prevent nonfarm prices from continuing outrun prices of metals and metal products.

I understand that a steel price hearing is in progress now to ascertain the reason for the recent rise in prices of steel. My earlier testimony indicated that such a rise would occur since I predicted that prices of