Knowledge that Transforms

N September 30, I950, the Canadian authorities suspended the fixed par value (go9. US cents) of the Canadian dollar.1 During the following twelve months, the Canadian dollar fluctuated between 94 and 96 US cents, but in November I95I it started to move upward at a rapid pace. By February I952 it had attained parity with the United States dollar. It continued to rise, reaching a peak monthly average of US $I .04I7 in September. After declining sharply in early November I952 to below US $I.02, the Canadian dollar recovered to an average of US $I.03 for the month of December. During the period from the middle of I950 to November I95I, capital imports from the United States provided the main source of strength for this currency. Since then, however, an improved trade balance has become the important factor. Had the authorities been prepared to buy United States dollars and gold freely, the appreciation of the Canadian currency might have been moderated and official holdings of foreign currencies augmented. During the recent appreciation, however, the Canadian authorities have pursued a policy of nonintervention in the foreign exchange market. Since this policy was maintained even during the third quarter of I952 when the premium exceeded 4 per cent, it would seem that the Canadian authorities have not altered their thinking about exchange rate policy since September I950. When the fixed par value for the Canadian currency was suspended in I950, the authorities announced that they would intervene only to maintain orderly market conditions. This policy was reaffirmed by Finance Minister Abbott in Parliament on February I9, I953, in these words: As I have stated before, our policy has been to allow the exchange rate to be determined by the normal play of economic forces without official intervention, except to ensure orderly conditions in the foreign exchange market. No attempt is made to reverse persistent trends, but only to smooth out excessive short-run fluctuations.2

JETERANS transfer payments have become a prominent segment of the income received by individuals. Although adopted to serve other purposes, the payments have important implications for the per capita income differences among the 48 states. The use of the state distribution of income as a policy determinant and as a tool of economic analysis requires a thorough understanding of the factors with which the differences are associated. The present article isolates a specific type of payment for study, and the results will add to the growing body of knowledge concerning the forces underlying the geographic income aggregates.2 Specifically, it should indicate whether the government programs embodying the veterans payments are offsetting or accentuating the state per capita income differences arising from other sources.

3.5 per cent of national income, while subtractions amount to only 0.4 per cent. (See p. I04.) 4. The so-called technical bias works both ways. It may result in some cases from greater detail of information for I937 than for I899the reason ascribed by Mr. Lebergott-but it results in many more cases simnply from greater diversification of products. High concentration is less indicative of monopoly for a product with many close substitutes than for one with few. In view of the vast proliferation of products and expansion of markets over the last half century, one may doubt the comparability of any specific measure of concentration as applied to both beginning and end of the period. As in the case of growth of output, no index can adequately reflect such qualitative factors. This is not to say that they should be overlooked. 5. The term slight is not used to describe the rise of 6 percentage points in the fraction of private income but the rise of i.9 percentage points in the fraction of national income. The former rise is characterized in my monograph as substantial, a fit counterpart for significant. (See p. 45.) Most of these issues are discussed in my monograph. I cannot agree with Mr. Lebergott that his criticisms fundamentally alter the picture of the growth of monopoly and concentration gathered from recent studies, such as those of Adelman, Stigler, and myself. Nevertheless, such criticisms are of inestimable importance in pointing out errors, no matter how minor they might seem; for the goal is to unearth facts, not to prove cases.

THE analysis of market phenomena by input-output arrays not only provides a practicable means of describing industrial interdependence, but also helps to clarify economic theory. For example, the accepted proposition that a change in the general price level as measured by index numbers is in reality a change in the monetary yardstick, may be easily demonstrated, and index number theory elucidated, by means of theoretical input-output relationships. Table i presents a theoretical input-output array in which the conventional input columns are written as rows comprising a set of simultaneous equations. It is assumed to classify exchanges of goods among certain firms, each marketing a specific commodity. The schedule merely indicates the goods and values brought to the market and the goods and values received, in the completed trading of, let us say, a single day. Monetary and other intermediate exchanges are omitted. With r rows, r i independent equations may be solved for relative, but not absolute, prices. The nature of a perfect, complete market imposes certain limitations on the relationships of the items in Table i. Since the sum of the first column is the sales value of commodity A, it must equal the sum of the first row, which itemizes the value of the goods received in exchange, and so on. to the nth column and row.1 Of course the ultimate units entering into a market are not goods, but the usances of capital and labor, broadly construed; and durable goods embody these usances discounted to their present worth. But such complexities, which have been discussed by Professor Leontief and others, would here only divert the argument. Exchanges may be simply visualized as if they were described in terms of tangible goods. Just as a gold monetary standard means that pricing throughout the economy is relative to the exchange value of a designated quantity of gold, so pricing may also be expressed in terms of physical units of each marketed commodity in turn. That is, prices may be standardized so that a is reduced to unity, then b, and finally n. Equations expressing prices in terms of a changed to unity are obtained by dividing the equations of Table i by a. Analogous equations may be obtained by similarly dividing by b and ... n. The column footings of such equations appear in Table 2. The row totals of Table 2, namely V/a, V/b, ..., V/n, may be taken to represent on a mag-

M ONTHLY statistics of the sterling area's trade and payments position with Western Europe and North America indicate that a substantial volume of short-term speculation in sterling has taken place since the end of I948 and that inward as well as outward movements of capital have occurred. The pressure against sterling resulting from this speculation has been recognized as a significant factor contributing to the I949 devaluation as well as to the loss of reserves incurred in I95I-52. The short-term flows of funds into and out of sterling, which have been induced principally by the expectation of changes in exchange rates, have been of a disequilibrating character. These transactions have taken the form of leads and lags in sterling payments by merchants in connection with their normal trading operations: hence, this speculation has been effected within the framework of rigorous controls over exchange and foreign trade transactions which have blocked short-term money movements of the pre-i939 character.