Relative Growth of Exports of Manufactures of United States and Other Industrial Countries
Author:
Anne Romanis
Search for other papers by Anne Romanis in
Current site
Google Scholar
Close

THE PURPOSE of this paper is to inquire how far factors other than price competitiveness were responsible for variations in the U.S. share of world exports of manufactures over the period from 1953 to 1958.

Abstract

THE PURPOSE of this paper is to inquire how far factors other than price competitiveness were responsible for variations in the U.S. share of world exports of manufactures over the period from 1953 to 1958.

THE PURPOSE of this paper is to inquire how far factors other than price competitiveness were responsible for variations in the U.S. share of world exports of manufactures over the period from 1953 to 1958.

The U.S. share in exports of manufactures from the 11 leading industrial countries declined sharply from 1953 to 1955, rose slightly from 1955 to 1956, and declined thereafter, particularly from 1957 to 1958 (Chart I). However, the sharp fall in the U.S. share between 1953 and 1955 was due primarily to a reduction in military shipments. In any assessment of U.S. export performance, there is a strong case for excluding changes in military shipments, since they are directly associated with changes in military aid and since the great decline in U.S. military shipments after the early 1950’ s had no counterpart in the exports of the other industrial countries.1

Chart 1.
Chart 1.

U.S. Share of Exports of Manufactures

(In per cent)

Citation: IMF Staff Papers 1961, 001; 10.5089/9781451954135.024.A006

Sources: U.S. exports of manufactures and U.S. exports of manufactures excluding military shipments were derived from U.S. Department of Commerce, Bureau of Foreign Commerce, World Trade Information Service, Part 3, No. 59-13, Total Export and Import Trade of the United States. For sources and description of exports of other countries, and of U.S. “commercial exports,” see Appendix II.1 U.S. share in total exports of manufactures of United States and countries listed in footnote 1, Table 2.2 Share of U.S. “commercial exports” in total of U.S. “commercial exports” and exports of manufactures of countries listed in footnote 1, Table 2.

When military shipments are excluded, the U.S. share in industrial countries’ exports of manufactures declined from 26.5 per cent in 1953 to 25.0 per cent in 1955, rose to 26.2 per cent in 1957, and fell to 23.6 per cent in 1958. However, neither 1953 nor 1957 is an appropriate base year for a comparison of trends in the U.S. share of world exports. Much of the decline in the U.S. share of world trade between 1953 and 1954 was due to the exceptionally rapid increase in German and Japanese exports, which by 1953 had not fully recovered from the effects of the war. In order to lessen the effect of this factor, the average of 1953 and 1954 has been chosen as one benchmark. Again, 1957 was an atypical year, being exceptionally favorable for the United States owing to the effect of the Suez crisis in increasing U.S. exports to Western Europe and limiting Western Europe’ s exports to certain other areas. Therefore, it has seemed preferable to take 1956 as the second benchmark. On this basis, the outstanding feature of the period since 1953, on which it is hoped that the present study will throw some light, is the contrast between the steadiness of the U.S. share of world exports of manufactures2 from 1953–54 to 1956, and the considerable decline from 1956 to 1958.

Summary and Conclusions

The first stage in the inquiry is to see whether changes in the composition and in the direction of world trade were favorable to the United States between 1953–54 and 1956 but unfavorable between 1956 and 1958. The growth, over both periods, of world trade in manufactures and of “commercial exports” of the United States (a concept closely related to nonmilitary exports)3 is analyzed in some detail so as to reveal how far the United States could be expected to benefit (or suffer) more than the average industrial country from the disparate growth of trade in different products and different areas. To this end, trade in manufactures is divided between nine product classes4 and ten5 market areas, chosen in order to distinguish, as far as the source data permit, product classes, or market areas, in which trade was rising more, or less, rapidly than the average, and product classes, or market areas, in which the U.S. share of trade was greater or smaller than the average. An estimate is then made of how far the U.S. share in exports of manufactures would have altered if the United States had maintained its 1953–54 share of exports in each of the nine product classes in each of the nine markets other than the United States between 1953–54 and 1956; and similarly, if it had maintained its 1956 shares between 1956 and 1958.

The conclusion reached is that, while changes in the geographical structure of demand between 1953–54 and 1956 had little effect on the U.S. share of world trade in manufactures, changes in the commodity composition of world demand between 1953–54 and 1956 were slightly favorable to the United States. If the United States had maintained its share in each class in each market, its share of exports of manufactures to the world outside the United States would have risen by nearly 2 per cent. Between 1956 and 1958, on the other hand, changes in the area distribution of demand were very unfavorable to the United States. In particular, exports of manufactures to dollar countries, notably Canada, declined, while exports to other areas continued to rise. Consequently, even had the United States maintained its share in each class in each market, its over-all share of exports to the world outside the United States would have fallen by 3 per cent.

However, the contrast between the steadiness of the U.S. share of world exports of manufactures from 1953–54 to 1956 and its sharp decline from 1956 to 1958 is not fully explained by the change in the trend of the structure of world demand. Of equal importance is the marked deterioration in the performance of the United States in particular classes of manufactures and markets. While the U.S. share of exports in 1956 was only 1.3 per cent less than it would have been if the United States had maintained its 1953–54 share of each product class in each market, the 1958 share of exports was 6 per cent less than it would have been if the 1956 share had been maintained (Table 1).

Table 1.

Percentage Changes in U.S. Share of Exports of Manufactures1

article image
Sources: See Appendix II.

Excluding exports to the United States.

Percentage change in share of U.S. “commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures of industrial countries listed in footnote 1, Table 2. The share was 27.9 per cent in 1953–54 and 28.0 per cent in 1956.

Combined effect of items F and G in Table 5, p. 252.

Effect of item H in Table 5.

In Part II of the paper, an attempt is made, through detailed studies for each of the nine product classes, to assess the importance of factors other than changes in competitiveness in causing the changes observed in the U.S. shares of exports of particular products to individual markets, and so to arrive at an indirect assessment of the trends in U.S. competitiveness in various fields.

Changes in the U.S. share of exports of any one of the nine product classes to any market area may be due to several factors other than genuine changes in competitiveness, namely, (1) reduced (or increased) discrimination against U.S. goods, (2) especially favorable (or unfavorable) composition of U.S. exports within a product class, and (3) other special circumstances tending to increase (or decrease) the U.S. share of exports.

While the impact of changes in the degree of adverse discrimination is easily understood, the other factors require some explanation. The possibility that changes in the U.S. share of exports may arise because the composition of its exports within a particular product class is atypical while the demands for the different products within the class grow at widely divergent rates is, of course, enhanced by the broadness of the commodity grouping employed. “Special circumstances” affecting the U.S. export share refer to changes in competitiveness induced by demand factors that are temporary or otherwise exceptional. Such changes are not indicative of underlying trends in competitiveness in the sense in which we are considering them here.

The very diverse trends in the U.S. shares of exports of the same type in different markets convey an impression of the importance of the three types of factor outlined above, and particularly of changes in the degree of discrimination. While in the dollar area the United States failed to maintain its share of manufactured exports both between 1953–54 and 1956 and between 1956 and 1958, in the non-dollar world6 it succeeded in increasing its share of exports between 1953–54 and 1956, but failed to do so from 1956 to 1958. In view of what happened in the dollar area, it may be doubted whether the United States would have maintained, let alone have increased, its share of trade with the non-dollar world between 1953–54 and 1956 but for demand factors, such as reduced discrimination against U.S. exports, or factors temporarily stimulating U.S. exports of certain types to particular markets in 1956. If it had not been for these factors, the U.S. share of trade in manufactures would have shown a more or less steady decline from 1953–54 (and earlier)—a trend which would seem to require a long-term explanation rather than one based on any change in circumstances from 1956 to 1958.

The general conclusion which emerges from the detailed assessments is that there does not seem to have been a marked acceleration in the deterioration of U.S. competitiveness after 1956, except in two very important groups, transport equipment and nonelectrical machinery. In both these groups, the U.S. competitive position seems to have weakened between 1953–54 and 1956, and even more between 1956 and 1958; in the earlier period, however, this weakening was apparently masked by certain special circumstances tending to favor U.S. exports in both dollar and non-dollar areas and by changes in the degree of discrimination against U.S. exports in non-dollar areas. The U.S. competitive position in metals (i.e., steel) also seems to have weakened from 1953–54 to 1956, although the U.S. share in non-dollar markets increased as a result of specially favorable demand conditions in Western Europe in 1956. If the subsequent reversal of these conditions is taken into account, the U.S. competitive position in metals does not seem to have worsened greatly between 1956 and 1958.

To sum up, the sharp contrast between the maintenance of the U.S. share of world exports of manufactures between 1953–54 and 1956 and the sharp fall between 1956 (or 1957) and 1958 seems to have been due very largely to two factors: (1) the change in the structure of world import demand, which was slightly favorable to the United States in the earlier period but decidedly unfavorable between 1956 and 1958 and (2) the increase in the U.S. share in non-dollar markets from 1953–54 to 1956, brought about by a reduction in the degree of discrimination applied in these markets against dollar imports, and by the development of boom conditions in several Western European countries and in Japan.

But for these factors, the U.S. share of world trade in manufactures would probably have shown a more or less steady decline from 1953–54 or earlier, owing, in the main, to a weakening competitive position in the field of transport equipment and other engineering products. A principal cause of the weakening of the U.S. competitive position over the period as a whole seems to have been the narrowing of the technical lead which the United States possessed in these fields after the war.

I. Influence of Structural Changes on U.S. Share of World Trade in Manufactures

In this section an attempt is made to see how far broad structural changes in the composition or direction of world demand for manufactures (excluding U.S. military shipments) may have been responsible for the 1958 decline in the U.S. share below the level which had been approximately maintained for several years previously.

The concept of U.S. “commercial exports” employed in this paper requires some explanation. Since no data are published concerning the composition of military exports, the changes in U.S. nonmilitary exports of different products can be estimated only by making certain assumptions. It seems that most military shipments are manufactures7 and that a large proportion probably consists of “special category” exports, i.e., of a varying list of strategic items for which data concerning destination (and, in a few cases, composition) are withheld for security reasons. But military exports do not consist exclusively of special category items, although the total value of the latter considerably exceeds that of military shipments. Thus for particular groups of manufactures, such as aircraft, motor vehicles, wireless and electrical apparatus (where both military shipments and special categories are large), the rate of growth of nonmilitary exports can only be guessed, with a fairly wide margin of error, from the figures of exports excluding special category items.

The definition of “commercial exports” used here corresponds, as far as is consistent with the need for adequate data on distribution between markets, to the most reasonable estimate of the value of nonmilitary exports in each class.8 The trend of total “commercial exports” thus defined corresponds quite closely to that of U.S. nonmilitary exports of manufactures (Chart I and Table 2).

Table 2.

Exports of Manufactures of United States and Other Industrial Countries1

(Indices of value; average 1953–54 = 100)

article image
Sources: U.S. exports of manufactures excluding military shipments were derived from data in U.S. Department of Commerce, World Trade Information Service, Bureau of Foreign Commerce, Part 3, No. 59–13, Total Export and Import Trade of the United States. For sources and description of other series, see Appendix II.

Belgium-Luxembourg, Canada, France, Federal Republic of Germany, Italy, Japan, Netherlands, Sweden, Switzerland, and United Kingdom—plus India and Hong Kong for exports of textiles and clothing only.

It is assumed that all U.S. military shipments comprise “manufactures” as defined in Standard International Trade Classification (SITC).

The method of analysis has already been outlined, but it may be helpful to describe briefly the relation of Tables 312 to the analysis.9 Table 3 deals with the commodity composition of exports; it shows the growth of industrial countries’ exports of the nine product classes distinguished, and their relative importance in the total and in U.S. exports of manufactures. Table 4 gives similar information concerning market distribution. Table 5, which is the keystone of the argument, summarizes the estimates of the effect of structural factors and of net losses of shares in exports of each class in each market.

Table 3.

Composition of Exports of Manufactures of Industrial Countries1 and the United States, and Indices of Value Showing Growth of Demand and Growth of U.S. Exports

(All figures relate to exports to world excluding the United States.)

article image
Sources: See Appendix II.

U.S. “commercial exports” and exports of manufactures from countries listed in footnote 1, Table 2.

Average of two years. The percentages were broadly similar in 1956 except that the share of metals increased to 11.4 for the United States and 14.8 for all countries, while the share of textiles declined to 4.6 and 11.0, respectively; also the share of miscellaneous finished goods in U.S. exports declined to 7.9.

Indices of value of industrial countries’ exports to the world excluding the United States.

Table 4.

Distribution of Exports of Manufactures by Nine Markets, and Indices of Value of Exports to Each Market

(All figures relate to exports to world excluding the United States.)

article image
Sources: See Appendix II.

Markets ranked by importance of U.S. share in total exports of manufactures (see Table 8).

For United States, percentages of “commercial exports” of manufactures; for industrial countries, percentages of total U.S. “commercial exports” and exports of manufactures from countries listed in footnote 1, Table 2.

Indices of industrial countries’ exports of manufactures to each area.

Table 5.

Growth of Exports of Manufactures from Industrial Countries1 and from the the United States

article image
Sources: See Appendix II.

U.S. “commercial exports” and exports of manufactures from countries listed in footnote 1, Table 2.

In other words, the average increase in industrial countries’ exports of nine classes of manufactures weighted according to their shares in U.S. “commercial exports” rather than their actual importance in the total exports of manufactures of industrial countries to the world excluding the United States, as in line A.

Table 6.

The Effect, for Nine Product Classes, of Changes in the Market Distribution of Demand, and of Changes in the U.S. Share of Exports to Particular Markets, upon the U.S. Share of Exports of Manufactures to the World Excluding the United States1

article image
Sources: See Appendix II.

Figures show the resulting percentage increases or reductions (—) (from 1953–54 to 1956, or from 1956 to 1958) in the share of U.S. “commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures from other industrial countries to the world excluding the United States (see Tables 1 and 5).

Detail by product class of line G in Table 5.

Detail by product class of line H in Table 5.

Detail by product class of the difference between the actual growth in U.S. exports and the average growth in industrial countries’ exports as it would have been if the nine classes of manufactures had had the same relative weights as in U.S. trade (line D minus line B in Table 5).

Table 7.

Net Effect of Changes in the U.S. Share of Exports of Nine Product Classes in Different Markets upon Its Share of Exports of Manufactures in the Various Markets, and to the World Excluding the United States

article image
Sources: See Appendix II.

Figures show the percentages by which the U.S. share of exports of manufactures (defined below) to each market would have increased or decreased (—) (from 1953–54 to 1956 or from 1956 to 1958) as a result solely of gains or losses of shares within each of the nine product classes, i.e., abstracting from the advantage which the United States derived from the more favorable composition of its exports to each market. The actual U.S. shares in exports of manufactures to these markets in 1953–54, 1956, and 1958 are shown in Table 8.

Figures show the resulting percentage increases or reductions (—) (from 1953–54 to 1956 or from 1956 to 1958) in the share of U.S. commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures from other industrial countries to the world, excluding the United States (see Tables 1 and 5).

Figures for Japan are approximations. Apart from metals, the United States probably more or less maintained its share of trade in both periods.

The sharp decline in the U.S. share of exports to this area was due mainly to very large transfers of new ships from other industrial countries to Liberia in 1956, and to the fact that the impact of the recovery of Japanese textile exports, from an abnormally low level, was felt principally in this area (see Table 9).

Table 8.

U.S. Share of Exports of Manufactures to Eight Markets1

article image
Sources: See Appendix II.

U.S. “commercial exports” of manufactures as per cent of total of U.S. “commercial exports” and exports of manufactures from other countries (listed in footnote 1, Table 2) to each area. See also footnote 1, Table 7.

Changes in the percentage share of exports reflect not only the changes in the U.S. share of exports to markets within each area, but also the effect of changes in the distribution of demand between markets within each area.

Excluding the communist countries.

Table 9.

Net Effect, for Nine Product Classes, of Changes in the U.S. Share of Exports to Particular Markets in the Dollar and Non-Dollar Areas upon the U.S. Share of Exports to the World Excluding the United States

article image
Sources: See Appendix II.

Figures show the resulting percentage increases or reductions (—) (from 1953–54 to 1956, or from 1956 to 1958) in the share of U.S. “commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures from other industrial countries to the world excluding the United States (see Tables 1 and 5). Analysis by areas of figures shown in columns 3 and 4 of Table 6.

Excluding the reduction in the U.S. share resulting from large transfers of ships to Liberia by other countries in 1956. The effect of this is shown in lower line.

Excluding effect of reduced U.S. share in the greatly increased exports of metals to Japan in 1956.

OEEC countries and related areas (rest of sterling area and dependent overseas territories) and non-dollar Latin America.

Other Eastern Hemisphere n.e.s. The impact of the recovery of Japanese exports from a low level in 1953–54 was felt principally in this area.

Table 10.

Impact of Rapidly Expanding Exports to Communist Countries,1 1953–54 to 1956, and 1956 to 1958

article image
Sources: See Appendix II.

Albania, Bulgaria, Czechoslovakia, Eastern Germany, Hungary, Poland, Rumania, U.S.S.R., and Mainland China.

Classes ranked in descending order of rate of growth of exports to communist countries.

Percentage of industrial countries’ exports of manufactures, as defined in footnote 1, Table 5, sold to communist countries in 1953–54 or 1956.

Table 11.

The Effect, for Nine Product Classes, of Growth in Other Countries’ Exports to Communist Countries, and of Demand Trends in Other Markets, upon the U.S. Share of Exports of Manufactures to the World Excluding the United States1

article image
Sources: See Appendix II.

Figures show the resulting percentage increases or reductions (—) (from 1953–54 to 1956, or from 1956 to 1958) in the share of U.S. “commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures from other industrial countries to the world excluding the United States (see Tables 1 and 5). Subdivision of effect shown in columns 1 and 2 of Table 6.

The figures show the declines in the U.S. share of exports, as defined in footnote 1, arising from the faster growth of industrial countries’ exports to the world excluding the United States than to the world excluding both the United States and the communist countries, which is effectively the area to which the United States exports.

These estimates are obtained as follows. From Table 3 an estimate is derived for the increase in U.S. exports of manufactures which would have occurred if the United States had maintained its share in exports of each of the nine product classes; this is shown in line B of Table 5. The difference between this increase and the increase in exports of manufactures of all industrial countries in each period measures the U.S. advantage arising from the commodity composition of its exports and the relative trend of demand for the various product classes (line F of Table 5). To take account of the effect of the differing rates of expansion of demand for exports of manufactures in different markets (shown in Table 4), a more refined estimate is then made of the increase in U.S. exports of manufactures which would have resulted if the United States had maintained its share of exports of each of the nine product classes in each of the nine market areas; this increase is shown in line C of Table 5. The difference between this rate of growth and the increase in all industrial countries’ exports of manufactures is taken to measure the U.S. advantage or disadvantage from structural trends in the commodity composition and market distribution of world demand (lines F and G, respectively, in Table 5). The extent to which the actual increase in U.S. exports of manufactures falls short of the increase shown in line C reflects the failure of the United States, on balance, to maintain its share of exports of each product in each market (line H in Table 5); this is referred to in later tables as the “net effect of changes in market shares.”

Tables 612 provide additional detail. Tables 6, 9, and 12 deal with the nine product classes individually. Table 6 shows in which product classes the United States failed to maintain its share of exports and how far this was due to the structural effect of market distribution, on the one hand, or to net losses in market shares, on the other; Table 9 shows net gains and losses of market shares in the dollar and non-dollar areas; and Table 12 gives further detail on gains or losses of shares in seven of the nine markets distinguished in the analysis. (The communist countries are not shown since the U.S. share in exports of manufactures to this market is negligible; nor is Japan, owing to the uncertainty of the estimates for that market.) Table 7 shows the combined effect of gains and losses in the U.S. shares in exports of all nine product classes in each market, and Table 8 shows, for reference, the actual shares of exports of manufactures from the United States to each of the eight markets other than the communist countries. Tables 10 and 11 show the importance of rapidly rising exports to the communist countries as a structural factor contributing to the fall in the U.S. shares in exports of manufactures.

Effect of Changes in Commodity Composition of World Demand

Both from 1953–54 to 1956 and from 1956 to 1958, the United States benefited from the fact that the greatest increase in world demand was for engineering products and chemicals—which comprise a larger share of U.S. exports of manufactures than of those of the other countries combined; textiles, a stagnating group, are much less important among U.S. exports (Table 3). Thus, if U.S. exports had moved in line with world exports of each of the nine commodity groups distinguished in this analysis, total U.S. exports of manufactures would have increased somewhat faster than the average in both periods, with the result that the U.S. share of exports of manufactures to the world excluding the United States would have risen by more than 1½ per cent between 1953–54 and 1956 and by 2 per cent between 1956 and 1958.

Effect of Changes in Market Distribution of World Demand

Since the geographical distribution of U.S. exports differs quite radically from that of the exports of industrial countries in general, the share of the United States in total exports of manufactures may vary substantially as a result of divergent trends in different markets. Dollar markets and markets in the Western Hemisphere are relatively much more important to the United States than to the other countries (when their exports to the United States are excluded), while markets in the Eastern Hemisphere are relatively more important to the other countries than to the United States, and markets in the communist countries10 in particular are of negligible importance to the United States (Table 4).

From 1953–54 to 1956, the relative growth of demand in different markets was not such as either to hamper or to favor the growth of U.S. exports of manufactures versus those of other countries (Table 5). Exports of manufactures to Eastern Europe and Mainland China rose considerably faster than exports to the world excluding the United States, but so did exports to Canada and Japan. From 1956 to 1958, however, demand was tilted very unfavorably for the United States. Exports of manufactures to Canada and dollar Latin America, which account for over half of U.S. exports of manufactures and less than 7 per cent of those of other countries, fell by 2 per cent, while the value of manufactured exports to all areas excluding the United States rose by 8½ per cent.11 Of this increase about one sixth arose from the continued rapid increase of exports to Eastern Europe, the U.S.S.R., and Mainland China. Exports to the world excluding the communist countries and the United States rose by less than 7¼ per cent (see Table 10).

With the exception of transport equipment, the unfavorable changes in the pattern of demand constituted a major factor for each of the seven product classes for which the U.S. share of world exports declined between 1956 and 1958. (There was no fall for textiles or metal manufactures.) As Table 6 shows, the unfavorable effects of market distribution accounted for all of the decline in the U.S. share of chemicals and miscellaneous finished goods, most of the decline in electrical machinery, more than half the decline in base metals, and about 40 per cent of the decline in nonelectrical machinery. The very rapid growth of other countries’ exports to the communist countries was an important element in the decline of the U.S. share of world exports of chemicals, nonelectrical machinery, and base metals.12

The contrast between the development of demand in the period from 1953–54 to 1956 and the development from 1956 to 1958 may be summed up as follows: If the United States had maintained its share of exports in each of the nine classes in each of the nine markets distinguished, its share of exports of manufactures would have increased by 1.7 per cent from 1953–54 to 1956, owing mainly to the favorable composition of its exports. Between 1956 and 1958, however, its share of exports would have fallen by 3.0 per cent, an unfavorable market distribution outweighing a favorable commodity composition (see Tables 1 and 5).

II. Export Performance of the United States in Different Product Classes and Markets

Superimposed on this adverse development in the broad structure of world demand from 1956 to 1958 was the marked deterioration in the U.S. performance in particular classes of manufactures and markets, mentioned above. This marked worsening was by no means general, but rather surprisingly concentrated. In four13 of the eight market areas the net fall in the U.S. shares in the nine product classes was smaller than the similar fall between 1953–54 and 1956 (Table 7). In five of the nine product classes, there was little or no decline in the U.S. share of trade between 1956 and 1958 over and above the decline which might have been expected as a result of changes in the distribution of demand between areas (Table 6). This pattern is hardly what would be expected if the decline in the U.S. share of world trade in manufactures had been due primarily to reduced competitiveness of U.S. manufactures in general. It suggests that a major part of the deterioration in the U.S. performance between 1956 and 1958 may have reflected demand factors specific to certain market regions or classes of exports.

In view of the prevalence of discrimination against U.S. exports in the early 1950’ s, it seems possible that changes in the degree of discrimination against U.S. exports may have had some bearing on the changes in U.S. shares of exports to particular markets. As shown in Table 7, the United States failed to maintain its shares of different classes of manufactured exports in the markets of the dollar area from 1953–54 to 1956, and did only slightly worse from 1956 to 1958; in the non-dollar markets, however, there was no corresponding decline of shares from 1953–54 to 1956, but a substantial fall occurred between 1956 and 1958.

In three of the non-dollar markets (OEEC countries, Rest of Sterling Area, and non-dollar Latin America), the U.S. share of manufactured exports rose considerably from 1953–54 to 1956; and in the two areas where the U.S. share fell heavily (Japan and “Other Eastern Hemisphere”), the fall was due largely to circumstances affecting one or two classes of manufactures.14 Thus the fall in the U.S. share of Japanese imports of manufactures arose predominantly from the fall in its share of greatly expanded Japanese steel imports. But where exports of a certain product to a particular market are usually very small, and drawn predominantly from one country, it would seem natural for that country’ s share to fall considerably if an enormous expansion in demand makes it worthwhile for other suppliers to enter the market. This may not be due to reduced competitiveness on the part of the usual supplier. The marked fall in the U.S. share of exports to the “Other Eastern Hemisphere” market is explained principally by very large transfers of ships from other industrial countries to Liberia in 1956. The U.S. share of textile exports to this area also declined considerably as Japan regained its prewar share of world trade in textiles.

The striking contrast in the development of U.S. trade with dollar and non-dollar areas between 1953–54 and 1956 is shown in Table 9. In the dollar area, the United States lost ground in every product class except nonelectrical machinery; the great increase in U.S. direct investment in dollar countries was a favorable factor for U.S. exports of this class.15 In the non-dollar area, the United States showed marked gains in virtually every class except the two machinery classes; the degree of discrimination against U.S. imports was probably initially lower for machinery than for other manufactures.

The earlier gains in the U.S. share of exports of nonelectrical machinery to dollar areas were reversed from 1956 to 1958, and there was a much greater fall in the U.S. share of exports of transport equipment, electrical machinery, and chemicals to these markets than had occurred from 1953–54 to 1956. On the other hand, the U.S. share of exports of base metals, textiles, and miscellaneous finished goods to dollar markets increased from 1956 to 1958, and its share of metal manufactures and “other” manufactures fell less than during the earlier period.

In non-dollar areas, the worsening of the U.S. performance from 1956 to 1958, when compared with the previous period, was at least as pronounced as in dollar areas in regard to nonelectrical machinery, transport equipment, and chemicals. In addition, there was a sharp reversal of former gains in base metals and “other” manufactures, and signs of weakening in metal manufactures and textiles.16 It would seem that reductions in discrimination against U.S. goods and special circumstances in particular non-dollar markets favored the maintenance of the U.S. share of certain exports to non-dollar markets from 1953–54 to 1956, and caused a sharper contrast to appear when such factors were operating less strongly, or were even reversed, from 1956 to 1958.

The peculiar development of U.S. trade with non-dollar Latin America should, however, be noted at this point. As Table 8 shows, this is the one area where the United States increased its share of manufactured exports not only from 1953–54 to 195617 but also from 1956 to 1958. It is also the only area in which the United States gained a substantially larger share of the market for machinery and transport equipment between 1956 and 1958.18 This development would seem to reflect an increase in pro-dollar discrimination resulting from the sharp increase in credits from the Export-Import Bank of Washington to Argentina, Brazil, and Chile from 1956 to 1958.19

In Appendix I, an attempt is made to estimate indirectly how far the changes in the U.S. shares of exports of each of the nine product classes in different markets reflect genuinely increased (or reduced) competitiveness. This is done by assessing the possible importance of other factors, mentioned earlier, namely, changes in the degree of discrimination against U.S. goods and special demand factors tending to favor (or hamper) the growth of U.S. exports, either within a particular market or within a particular class of manufactures. The reader is referred to Tables 6, 9, and 12 and the earlier description of their content.

The conclusions of these detailed studies may be summarized as follows:

The U.S. share of world trade in manufactures was much more affected by decreases in the U.S. shares of exports of each product to each market from 1956 to 1958 than from 1953–54 to 1956. However, the marked acceleration in the deterioration of the U.S. competitive position was confined to the two groups, transport machinery and nonelectrical machinery. In both these groups, the U.S. competitive position seems to have been weakening between 1953–54 and 1956, and still more between 1956 and 1958; in the earlier period this weakening was masked by especially favorable demand developments in both dollar and nondollar areas, and by changes in the degree of discrimination against U.S. exports.20 In the other engineering group (electrical machinery and apparatus), the U.S. share of exports dropped substantially between 1953–54 and 1956 in the face of an increasing German export drive. The slight gain in this group in the non-dollar area from 1956 to 1958 arose principally from the increase in the U.S. share of exports to non-dollar Latin America.

The U.S. competitive position in base metals also seems to have weakened between 1953–54 and 1956, although its share in non-dollar markets increased as a result of especially favorable demand conditions in 1956. If the reversal of these conditions is taken into account, the U.S. competitive position in the group of metals as a whole does not seem to have changed very much between 1956 and 1958.

The United States appears to have maintained its share of chemical exports to particular markets both from 1953–54 to 1956 and from 1956 to 1958. The group data must be regarded with caution, however, since they may conceal an exceptionally favorable composition of U.S. exports, particularly in areas outside the Western Hemisphere. The marked increase in the U.S. share of chemicals to most non-dollar markets between 1953–54 and 1956 probably reflected reduced discrimination, and especially favorable demand conditions in Europe in 1956, while the continued rise in the U.S. share of chemical exports to OEEC countries may have been due to a favorable commodity composition and further liberalization of imports from the United States. Even when account is taken of these factors, however, it does not appear that the United States was losing ground in chemicals.

In consumer goods—textiles, metal manufactures, and miscellaneous finished goods—reduced discrimination against dollar supplies in many countries enabled the United States to increase its share of exports between 1953–54 and 1956. However, the U.S. share of trade was sharply reduced in a few markets where the impact of the rapid increase in Japan’ s exports from their low level in 1953–54 was concentrated. Especially after 1956, the United States tended to benefit from the composition of its exports in these groups.

A principal cause of the weakening in the U.S. competitive position in transport equipment and other engineering groups over the whole period between 1953–54 and 1958 seems to have been the narrowing of the technical lead which the United States possessed in these fields after the war. Thus a significant part of the decline in the U.S. share of exports of transport equipment, for example, was due to the leadership which British aircraft producers captured, at least temporarily, in 1958, in supplying turbo-jet planes for commercial service. The sharp declines in the U.S. share of exports of motor vehicles reflect the fact that heavy investments, often by U.S. subsidiaries, coupled with greater availability of steel supplies and skilled manpower to the export trades, have in recent years enabled European producers to put ever-increasing quantities of the most up-to-date models on the world market, and so reap the full advantage of their lower wage levels for the first time since the war—perhaps for the first time since the United States went ahead in the mass production of automobiles during World War I. Similar considerations apply throughout the field of engineering products. The trend toward a gradual lowering of European steel prices—if the effect on these prices of the boom and the subsequent easing of demand are discounted—may also reflect relatively greater technical progress from a lower level of efficiency and smaller scale operation.

On the other hand, it is perhaps not surprising that the U.S. competitive position seems to have been maintained in chemicals, since this, together with electronics, is probably the field in which the most substantial advances in technique have been introduced in the United States over the last few years.

It is also perhaps not surprising that there does not seem to have been any very marked change in the U.S. competitive position in textiles, metal manufactures, or the diverse group of miscellaneous finished goods. For whereas the desire to acquire the latest type of machinery or transport equipment is likely to be the decisive factor when one country is ahead in developing a new line, considerations of relative cost are likely to be more important for nondurable goods or household furnishings. Not only is innovation generally less significant (excluding fashion changes which consist essentially of ringing the changes), but in these fields there had been few developments comparable to the speeding up of technical advance in the engineering industries of the United States under the impact of war production. The element of a narrowing technical leadership was therefore, in general, less important for consumer nondurable goods. Moreover, the U.S. competitive position was initially less favorable for many consumer goods than for engineering products, metals, or chemicals, since wages constitute a larger element of the cost of most consumer goods, so that the higher absolute level of U.S. wages is likely to be a more serious obstacle to competitiveness in these fields. Even in 1953–54, the relatively small U.S. exports of these groups went mainly to nearby markets, such as Canada and dollar Latin America, where lower transport costs and easier access to U.S. suppliers gave certain U.S. exports a considerable natural protection against increasing competition from other industrial countries.

APPENDICES

I. Notes on Changes in U.S. Shares of Exports of Nine Product Classes in Various Markets.

Machinery, Nonelectrical (Group 71)

Between 1953–54 and 1956, the United States increased its share of exports of this class in dollar markets, and almost maintained its share in non-dollar markets (Tables 9 and 12). The increase in the dollar markets was probably connected with the exceptionally great expansion in U.S. investments in that area, especially in oil exploitation and pipelines. The subsequent sharp decline in U.S. direct investment probably contributed to the decrease between 1956 and 1958 in the U.S. share of exports to that area. The decline in the U.S. share of exports to non-dollar markets from 1953–54 to 1956 would have been greater but for an exceptional expansion in the U.S. share of exports to OEEC countries. The reversal of this trend may in turn explain a part of the very marked decline in the U.S. share of machinery exports to those countries from 1956 to 1958. U.S. exports to non-dollar Latin America seem to have been favored by special circumstances. To sum up, the gain in the U.S. share of exports from 1953–54 to 1956 appears to have been due to exceptionally favorable circumstances. It seems probable that there would otherwise have been a significant decline in the U.S. share of trade in this period and consequently a less marked decline than actually occurred from 1956 to 1958.

Table 12.

The Effect, for Nine Product Classes, of Changes in the U.S. Share of Exports to Particular Markets Upon the U.S. Share of Exports of Manufactures to the World Excluding the United States1

article image
Sources: See Appendix II.

Figures show the resulting percentage increases or reductions (—) (from 1953–54 to 1956, or from 1956 to 1958) in the share of U.S. “commercial exports” of manufactures within the total of U.S. “commercial exports” and exports of manufactures from other industrial countries to the world excluding the United States (see Tables 1 and 5). Further detail of figures given in Tables 7 and 9.

Arrows indicate cases where there was apparently some reversal of abnormally favorable (or unfavorable) demand conditions after 1956.

Excluding reduction in U.S. share that was due to exceptionally large transfers of ships to Liberia by other countries in 1956 (see Table 9).

Transport Equipment (Group 73)

In both periods, the United States was clearly losing ground in this field. Its performance after 1956 does seem to have been considerably worse than during the earlier period, even when allowance is made for the fact that the previous increase in the U.S. share of exports to certain non-dollar markets reflected reduced discrimination against U.S. vehicles. As Table 12 shows, between 1953–54 and 1956 the United States substantially increased its share of exports to the Rest of the Sterling Area, the Dependent Overseas Territories, and non-dollar Latin America; and comparison with the dollar area suggests that the small loss in the OEEC markets would have been greater but for the same factor. Owing to special circumstances, the changes in the degree of discrimination were likely to be less favorable in the “Other Eastern Hemisphere” market,21 where the U.S. share fell between 1953–54 and 1956.

A Department of Commerce study22 suggests that well over three fourths of the fall in the U.S. share of world exports of transport equipment between 1954–56 and 1958 was accounted for by the reduced U.S. share in exports of motor vehicles to all markets. Much of the remainder arose from the fall in its share of aircraft exports already alluded to,23 which occurred in three markets, Canada, Western Europe, and the Far East. The U.S. share in world exports of railway vehicles actually increased, owing to a marked rise in its share of exports to Latin America and a significant rise in its share of exports to Africa.

Machinery, Electrical (Group 72)

As Table 9 shows, apart from miscellaneous finished goods (Section 8), this is the only class in which the U.S. shares of exports to non-dollar markets were better maintained from 1956 to 1958 than from 1953–54 to 1956. One reason for this is that, in the earlier period, exports from the Federal Republic of Germany—which still had not regained its “natural” share of world exports of this particular class by 1953–54—were expanding very rapidly; after 1956 their growth slowed down.24 This factor was of comparatively slight importance in dollar markets, because of the much smaller share of exports accounted for by Germany, particularly in the Canadian market.

The fact that the U.S. share of exports of this class to non-dollar markets actually increased from 1956 to 1958 seems to have been due to special circumstances. The marked gain in the U.S. share of exports to non-dollar Latin America already referred to25 was one very important element; another was the slight increase in the U.S. share of exports to OEEC countries. It is possible that this may reflect a reduction in the coverage of special category items between 1956 and 1958, and the consequent inclusion in “commercial exports” of items formerly excluded.26

It may therefore be suggested that the U.S. competitive position weakened between 1953–54 and 1956, and might have further deteriorated after 1956, not only in dollar markets but also in non-dollar markets, but for special circumstances affecting trade with non-dollar Latin America and the OEEC countries.

Chemicals (Section 5)

Of the nine classes covered, the competitive position of the United States appears to have been relatively strongest in chemicals. While conclusions drawn from the over-all figures for this diverse grouping must be regarded as tentative, the data suggest that, even when the effect of reduced discrimination and of the relatively favorable composition of its exports to certain markets is discounted, the U.S. share in world exports of the principal types of chemical and pharmaceutical products did not decline seriously in either period.

A few points in Table 12 are revealing. Between 1953–54 and 1956 the United States significantly increased its share of chemicals to every non-dollar market except the Dependent Overseas Territories. Reduced discrimination against U.S. exports was almost certainly a major factor, coupled, in the case of the OEEC countries (where the increase in the U.S. share was particularly marked), with especially heavy demands for U.S. basic industrial chemicals (arising from boom conditions in Europe) and for explosives. Only in dollar Latin America did the U.S. share decline significantly; this was the one area in which U.S. exports had to meet the expanding German export drive without benefit of offsetting factors. Because German exports to Canada are relatively small, the United States was able to maintain, and even increase, its share.

Between 1956 and 1958 the effect of differences in composition was probably much stronger than during the earlier period, as exports of basic industrial chemicals, pharmaceuticals, and explosives continued to increase in value, while exports of other products, such as dyestuffs and fertilizers, declined. The two areas where the U.S. share of chemical exports fell sharply from 1956 to 1958 were dollar and non-dollar Latin America. The reason for this may be that dyestuffs and fertilizers constitute a larger share of United States’ than of other countries’ chemical exports to these areas. Analogously, one reason why the United States continued to do relatively well in Eastern Hemisphere markets may be that in those markets the composition of the other countries’ exports was similarly unfavorable to them. The fact that practically no fertilizers are exported by the United States to OEEC countries was one reason why the U.S. share of total chemical exports to that area increased considerably between 1956 and 1958. There was also a sharp increase in U.S. exports of medicinal products; whether this was because the United States gained an increased share in the pharmaceutical market, or because it merely benefited from a rapid increase in demand for pharmaceuticals (one field in which it is a relatively large supplier to OEEC countries) is not clear without further analysis.

Base Metals (Group 68)

The apparent worsening of the U.S. performance in exports of base metals from 1956 to 1958, compared with the period from 1953–54 to 1956, was due largely to the reversal of special circumstances which had tended to raise its share of these exports to three markets in 1956. Thus the considerable gain in the OEEC market from 1953–54 to 1956 was primarily a reflection of the strong boom conditions, and consequent excess demand for U.S. metals, in these countries; the increase in the U.S. share of exports to non-dollar Latin America and the “Other Eastern Hemisphere” market was similarly due to the existence of boom conditions in Europe and Japan which tended to limit metal exports, particularly from France, Belgium-Luxembourg, and Japan.27 A decline in the U.S. share of these markets was therefore to be expected after 1956.

Textiles (Group 65)

The U.S. share of textile exports is likely to be influenced by atypical composition, since the United States exports scarcely any wool textiles and the newer synthetics comprise a much larger proportion of its textile exports than of those of other countries.

Probably as a result of reduced discrimination against U.S. goods, the U.S. share of the total value of textile exports to most non-dollar markets rose quite substantially from 1953–54 to 1956. But there was a marked fall in the U.S. share in the “Other Eastern Hemisphere” market, where the main impact of the recovery in Japanese exports was felt and where the initial degree of discrimination against the United States was probably smaller. The fall in the U.S. share of textile exports to the dollar area was concentrated in Canada. A principal reason why the United States did relatively worse in Canada than in dollar Latin America was that the U.K. share of Canadian trade increased, while its share of dollar Latin American trade declined even more sharply than that of the United States. The contrast in British exports to the two areas reflects the fact that in the Canadian market the value of British wool textiles increased somewhat, while in the dollar Latin American market it fell by one third. Probably only part of this difference is to be explained by the protection enjoyed by British exports in the Canadian market, so that some of the contrast lies in the fact that demand for wool textiles in Canada tended to rise, or was at least maintained, between 1953–54 and 1956, while it seems to have dropped quite substantially in dollar Latin America. If so, changes in the composition of textile demand were less favorable to the United States in Canada than in dollar Latin America between 1953–54 and 1956.

From 1956 to 1958, the United States broadly maintained its share of textile exports to non-dollar markets, except “Other Eastern Hemisphere,” and its share in that market fell much less than during the preceding period. In the dollar area its share increased. Thus, the United States did less well than from 1953–54 and 1956 in most nondollar markets, because reduced discrimination was a less important factor, but it did considerably better in the “Other Eastern Hemisphere” and dollar areas. At least two factors favored the United States during this period. The slowing down of Japanese exports to certain areas, partly as a result of voluntary restriction schemes, actually gave a measure of discrimination in favor of the United States in the dollar area; and the much greater fall in wool prices than in those of other textile fibers also favored the United States.

Miscellaneous Finished Goods (Section 8)

The general pattern of development in this class is similar to that for textiles, as Table 12 shows. One noticeable difference is that the United States apparently succeeded in increasing its share of exports to the “Other Eastern Hemisphere” market between 1956 and 1958. A factor which may have helped the United States, particularly in this area, is the small proportion of its exports in this class which consisted of cheap clothing, footwear, etc., whose prices fell significantly as raw material prices declined.

Metal Manufactures (Group 69)

As for textiles and miscellaneous finished goods, the U.S. share of exports of this class to the dollar area fell less from 1956 to 1958 than from 1953–54 to 1956. This was due to a slowing down of the earlier, very rapid, expansion of Japanese and German exports, from an exceptionally low level in 1953–54. Composition probably does not influence the U.S. showing in this small group. The figures for the non-dollar areas seem to suggest that reduced discrimination against U.S. exports was of importance in both periods.

“Other” Manufactures (Section 6)

The composition of this class varies substantially from country to country, and changes in the shares of different countries are strongly influenced by divergent demand trends for the different types of product included.28 The figures in Table 12 suggest that reduced discrimination against the United States may have favored the United States in nondollar markets during the earlier period. The sharp decline between 1956 and 1958 in the U.S. share of trade with OEEC countries—which is the most marked development of that period—was, however, due largely to a fall in shipments of certain types of tire.29

II. Notes on Statistical Sources and Methods

Figures for exports of Belgium-Luxembourg, Canada, France, Federal Republic of Germany, Italy, Netherlands, Sweden, Switzerland, United Kingdom, and United States have been taken from Organization for European Economic Cooperation (OEEC), Statistical Bulletins, Series IV, with certain adjustments described below.

Figures for exports of Japan have been estimated from the area distribution for three main types of manufactured goods—chemicals, machinery and transport equipment, and other manufactures—from the following sources: 1953–57 data from United Nations, Monthly Bulletin of Statistics, February 1959; 1958 data from United Nations, Commodity Trade Statistics, January-December 1958; and the area distribution of Japanese textile exports published in Contracting Parties to the General Agreement on Tariffs and Trade (GATT), International Trade, 1957–58. The shares of electrical and other machinery and transport equipment within Section 7, and of base metals, metal manufactures, and Section 8 in the total for Sections 6 and 8 less textiles, have been estimated from the totals for exports to all areas published in the United Nations, Yearbook of International Trade Statistics, 1956 and 1957, Volume I, supplemented by figures in Ministry of International Trade and Industry, Foreign Trade of Japan, 1957 and 1958.30

Statistics of exports of textiles and clothing from India and Hong Kong were taken from GATT, International Trade, 1957–58.

Some special problems arise in relation to the statistics of the United States. As stated above, the concept of “commercial exports” that is used corresponds, as far as is consistent with the need for an adequate distribution between markets, to the most reasonable estimate of the value of nonmilitary exports in each class. “Commercial exports” are taken to comprise the following:

(1) All exports of textiles and base metals (SITC Groups 65 and 68—there are no special category items in either group).

(2) All exports of chemicals, nonelectrical machinery, and groups in Section 6 other than textiles, base metals, and metal manufactures (SITC Section 5 and Groups 71, 61, 62, 63, 64, 66, and 67). Special category exports in each of these groups are not excluded as they are presumed not to comprise military shipments; the relatively small values involved are summarily allocated to the OEEC market.31

(3) Exports, excluding special category items, of electrical machinery, transport equipment, metal manufactures, and miscellaneous finished goods, mainly consumer goods (SITC Groups 72, 73, 69, and Section 8). Commercial exports thus defined are significantly smaller than non-military exports in these groups. Even if (as is implicitly assumed) there were no military shipments of items apart from these four groups of manufactures, the understatement would amount to some $250–300 million. However, as there is no way of assessing the markets or products in which the shortfall occurs, the assumptions made give the nearest practicable approximation to nonmilitary exports32 for the purposes of the analysis.

Another difficulty arises from the fact that the tables for the United States in OEEC, Statistical Bulletins, Series IV, do not allocate small value shipments by market areas after 1953. While it would be possible to exclude them altogether (as the Department of Commerce study does), to do so would introduce a significant bias in the present study, since the limit for exclusion was twice as high during the first half of 1956 as during the rest of the period from 1954 to 1958. Small value shipments go predominantly to neighboring countries. Thus, for classes where small value shipments are important,33 the series excluding small value shipments is likely to understate considerably the growth of U.S. exports to Canada and dollar Latin America from 1953–54 to 1956, and to overstate the growth from 1956 to 1958. In this paper, therefore, the known value of small value shipments in each class in each year has been allocated between the different markets in what seems the most reasonable proportion in the light of the Department of Commerce sample study undertaken in 1953,34 and of the subsequent development of trade in the group.35

The distribution by market areas is based on the divisions shown in OEEC, Statistical Bulletins, Series IV, Table of Exports by Geographical Areas. This table distinguishes South America and Central America. The figures for dollar and non-dollar Latin America were obtained by adding to the exports to Central America, and deducting from the exports to South America, the value of exports to dollar countries in South America as given in the Table of Exports by Monetary Areas.

The Series IV data do not distinguish exports to Japan. The values of exports to Japan from each of the other ten countries in three broad groups of manufactures (Sections 5, 7, and 6 and 8 combined) given in United Nations, Yearbook of International Trade Statistics, 1956 and 1957, Volume II, and Commodity Trade Statistics, January-December 1958, were used as a framework, and the composition within Section 7 and Sections 6 and 8 was estimated from data published in Foreign Trade of Japan, 1957 and 1958, and United Nations, Yearbook of International Trade Statistics, 1957, Volume I.

Some problems arise in using the OEEC figures for exports to Overseas Territories of Member Countries, Non-OEEC Member Countries in the Sterling Area, and Other Non-OEEC Member Countries (which comprise Japan and “Other Eastern Hemisphere” countries), since several overseas territories became independent between 1953 and 1958. Thus each of the territories listed below ceased to be included among the Overseas Territories and was added to one of the other two areas as indicated:

article image

In this paper, the Dependent Overseas Territories (DOT) have been defined as the territories included as such in the OEEC statistics for 1956. This involves excluding Northern Rhodesia and Nyasaland and adding these territories to non-OEEC sterling countries in 1953, and similarly excluding Indo-China and adding it to the “Other Eastern Hemisphere” area in 1953 and 1954. Estimates of the value of small exports to Northern Rhodesia and Nyasaland from the United States and other countries were based on import statistics. For Indo-China, the United Nations, Yearbook of International Trade Statistics gives details of exports for the United States and other countries by three broad groups, and this information was supplemented by study of the national import statistics. For 1957, it becomes necessary to add Ghana and Tunisia and Morocco to the figures of exports to Dependent Overseas Territories, and to reduce the figures for non-OEEC sterling countries and “Other Eastern Hemisphere.” Estimates of the values involved were obtained in the same way as for Indo-China. For 1958, it would have been desirable to make similar adjustments, and also to add Malaya to the Overseas Territories (and exclude it from non-OEEC sterling countries); but since, when this paper was prepared (early 1960), fewer data for a reliable estimate were available for 1958 than for 1957, it was decided to take a short cut. For each of the nine product classes, the growth between 1957 and 1958 of U.S. exports and of all countries’ exports to each of the three market groups defined as in 1958 was assumed to apply to the groups defined as in 1956.36 In most cases this approximation did not cause the growth of total exports to the three market areas together (which of course is known) to be seriously overstated or understated; but where it did, the figures were adjusted as seemed appropriate.

III. Mathematical Formulation of Concepts Employed

The concepts employed in this paper can be defined mathematically as follows:

Let r stand for the ratio of the value of manufactured exports from all countries covered (including the United States) to the world in the end year to their value in the base year. Then, let ri (i = 1,2,3, … . . 9) stand for the similar ratio for the ith product class of the nine classes of manufactures distinguished in the analysis, and let rij (j = 1,2,3, … . 9) stand for the similar ratio for exports of the ith product class to the jth market area of the nine market areas (other than the United States) distinguished in the analysis.

Let the symbols si and sij stand for the corresponding ratios for U.S. “commercial exports” of manufactures; and let x, xi, and xij represent the absolute values of those exports in the base year.

Then xix is the share in the base year of the ith product class in the total value of U.S. “commercial exports” of manufactures; and xijx is the share of exports of the ith product class to the jth market in the total value of U.S. “commercial exports” of manufactures in the base year.

The analysis of changes in the U.S. share of industrial countries’ exports of manufactures may then be set out as follows:

If the United States had maintained its share of world exports in each product class (taking all market areas together), the ratio between the value of its manufactured exports in the end year and the base year would have been

1 x Σ i = 1 9 r i x i ( shown in line B in Table 5 ) ( 1 )

instead of the observed value s.

Further, if for any particular product class the United States had maintained its share of exports in each market, the ratio between the value of its exports of that product class in the end year and in the base year would have been

t i = 1 x i Σ j = 1 9 r i j x i j ( 2 )

instead of the observed value si.

Finally, if the United States had maintained its share of exports in each product class in each market, the ratio between the value of its exports in the end year and the base year would have been

t = 1 x Σ i Σ j r i j x i j ( Line C in Table 5 ) = Σ i t i x i / x ( 3 )

The contribution of each product class to the total effect (i.e., to the change implied by the ratio t) is tixi/x, which is shown in columns 1 and 2 of Table 6.

The difference between (1) the actual ratio of total U.S. exports of manufactures in the end year to those of the base year and (2) what the ratio would have been if the United States had maintained its share of exports in each product in each market (st) is the net effect of changes in U.S. shares of exports of each product class in each market, shown in line H of Table 5.

It will be apparent that (siti)xi/x, shown in columns 3 and 4 of Table 6, is the part of this difference explained by changes in the U.S. shares of exports of a particular product class in all nine markets; and similarly, (sijrij)xij/x, shown in Table 12, is the part explained by the change in the U.S. share of exports of a particular product class in a particular market.

*

Miss Romanis, economist in the Special Studies Division, is a graduate of Cambridge University. She was formerly on the staff of the Oxford University Institute of Statistics, the Programmes and Plans Division, Ministry of Production, London, the Economic Directorate, Organization for European Economic Cooperation, Paris, and the Economic Survey Division, United Nations, New York. She has published several articles on economic subjects.

1

It may be argued that fluctuations in economic aid also have direct repercussions on the level of U.S. exports of manufactures (more especially of capital goods), but the connection is not, in general, nearly as close as that between military aid and military shipments, and in the absence of any comparable series for “economic aid financed exports,” it is not possible to make an over-all allowance for the effect of fluctuations in U.S. economic aid upon U.S. exports of manufactures.

2

Excluding U.S. military shipments.

3

It is not possible to analyze changes in nonmilitary exports directly, since no data concerning the composition of military exports are published. See Appendix II for description of concept of “commercial exports.”

4

The nine product classes distinguished (with their numerical codes under the Standard International Trade Classification for reference) are as follows:

article image

The last class comprises manufactures of wood (SITC Group 61), paper (Group 62), rubber (Group 63), nonmetallic minerals (Group 64), precious metals (Group 66), and leather (Group 67). U.S. exports of this class consist largely of semimanufactured products.

See Appendix II, Notes on Statistical Sources and Methods, for a description of the statistical problems involved in the analysis.

5

Nine if, as in most sections of this study, the U.S. market is excluded. The nine other areas are listed in Table 4, p. 251.

6

In this study “the non-dollar world” is to be understood as excluding the communist countries.

7

U.S. trade statistics show all military shipments as “finished manufactures,” but some may consist of petroleum products, etc., which are included under the U.S. definition but are excluded under the SITC definition of manufactures.

8

For description, See Appendix II.

9

See also Appendix III for mathematical definitions of concepts employed.

10

Albania, Bulgaria, Czechoslovakia, Eastern Germany, Hungary, Poland, Rumania, U.S.S.R., and Mainland China.

11

It may be noted that the exclusion of exports to the United States reduces the growth of other countries’ exports only very slightly between 1953–54 and 1956, and has no effect between 1956 and 1958 (see Table 2). The continued rapid growth of exports to the United States between 1957 and 1958 was, however, responsible for about half the expansion in other countries’ manufactured exports; thus the decline in the U.S. share of exports to the world outside the United States between 1957 and 1958 was not as great as the over-all figures would suggest.

12

As may be seen from a comparison of column 2 of Table 11 with the final column of Table 6, this accounts for about three sevenths of the fall in the U.S. share of chemical exports between 1956 and 1958, about one sixth of the fall in nonelectrical machinery, and one seventh in base metals.

13

In dollar Latin America, the Dependent Overseas Territories, Japan, and “Other Eastern Hemisphere” countries. Changes in the very small share of the United States in exports to the communist countries are of no significance.

14

The rise or fall referred to here comprises the sum of losses (or gains) of shares in each of the nine product classes of exports within each market as shown in Table 7. In every case except one (the non-dollar Latin American market from 1953–54 to 1956), changes in the commodity composition of demand tended to increase the U.S. share of total exports of manufactures to each market. Consequently, in each market except non-dollar Latin America its share fell by a smaller percentage (or rose by a larger percentage) than is shown in Table 7. The U.S. shares in exports of manufactures to the nine market areas in 1953–54, 1956, and 1958 are given in Table 8.

15

See page 261.

16

When the effect of Japan’s postwar recovery of its former textile markets in the Middle East and Far East between 1953–54 and 1956 is excluded.

17

Part of the rise in the U.S. share over this period occurred in base metals, miscellaneous finished goods, and “other” manufactures (as Table 12 shows); this seems to have been due to a drop in the value of exports from several European countries, notably France, Belgium-Luxembourg, and Sweden. The subsequent recovery of European exports largely explains the fall in the U.S. share of exports in SITC Sections 6 and 8 to non-dollar Latin America between 1956 and 1958.

18

The major part of the gain seems to have been in industrial machinery, but the U.S. share in exports of railway equipment also increased (see page 262).

19

The rise in disbursements by the Export-Import Bank more than offset reductions in U.S. private direct investment in Brazil and Chile, as shown by the following figures, in millions of U.S. dollars:

article image

The increase in the value of U.S. exports implied by the increases in the U.S. shares in total exports of machinery and transport equipment to Argentina, Brazil, and Chile, from 33, 48, and 57 per cent, respectively, in 1956 to 51, 59, and 65 per cent in 1958 was in each case roughly commensurate with the increase in the total of Export-Import Bank disbursements and U.S. direct investment combined.

20

Demand factors probably explain the slight net improvement in the U.S. shares of different markets for nonelectrical machinery, while changes in the degree of discrimination may explain the slight improvement for transport equipment when the special effect of transfers of ships to Liberia in 1956 is excluded (see Table 9).

21

Changes in the degree of discrimination are unlikely to have been very important in the Middle East. In the Far East, the position of U.S. exporters may actually have weakened as a result of the growing balance of payments difficulties of some countries which were important U.S. markets, and of the decline of exports to the territories of French Indo-China (which are included in this area) from their abnormally high level in 1953–54.

22

Analysis of Changes in the United States Shares of Export Markets for Manufactures, 1954–1958 (an unpublished study prepared by International Economic Analysis Division, November 1959).

23

See page 259.

24

The somewhat delayed recovery of Germany in this field may reflect the effect of restrictions placed on German production in the immediate postwar years, in conjunction with a required planning and production period of several years for large-scale heavy electrical plants. Between 1953–54 and 1956, exports from the Federal Republic rose by nearly 70 per cent; between 1956 and 1958, by less than 25 per cent.

25

See pages 256–58.

26

While exports shown as going to OEEC countries rose from $72 million to $91 million, the value of special category items in this group fell from $253 million to $204 million.

27

Japanese and French exports declined sharply, and those of Belgium-Luxembourg increased only slightly, between 1955 and 1956.

28

Exports from the United States, Sweden, and Canada, for example, consist largely of semifinished manufactures of wood and paper (i.e., newsprint), while these products are of negligible importance for most OEEC countries and Japan. In general, a large part of U.S. exports of this group (as of the exports of Canada and Sweden, but not of those of most OEEC countries and Japan) probably consists of less highly finished products, and their value is consequently more likely to vary with raw material prices.

29

Special category exports of rubber products have been imputed to OEEC countries, which probably took most of them. Practically all rubber products were removed from the special category classification by 1958. But the drop in special category exports of rubber products was not matched by any similar increase in the value of U.S. rubber exports to OEEC countries, as shown by the following figures, in millions of U.S. dollars:

article image

30

The 1957 volume was published by the Japan External Trade Recovery Organization, and the 1958 volume by the Japan Export Trade Promotion Agency.

31

The coverage of special category exports in these groups was drastically reduced in 1957. This fact (which may itself be an indication that the exports were not essentially military in character) makes it desirable not to introduce an artificial element in the growth of exports by excluding special categories. The inclusion of special categories results in a considerably less favorable trend for the U.S. share of exports of such items as tires and nonelectrical machinery to OEEC countries than that presented in the Department of Commerce study, referred to in footnote 22.

32

Though there are also some changes in the coverage of special category items, they are not very significant in relation to the large totals of special category exports in each of these groups.

33

Notably miscellaneous finished goods, metal manufactures, textiles, “other” manufactures (Section 6), and electrical machinery. In 1958 the share of exports (excluding special category shipments) accounted for by small value shipments amounted to 20, 18, 13, 12, and 10 per cent of these five groups, respectively.

34

See Foreign Trade Statistics Notes, September 1953.

35

To give one example of the improvement which this adjustment achieves, the Series IV figures for U.S. exports of textiles to both Canada and Central America show a distinct fall from 1955 to 1956 and a very marked increase between 1956 and 1957. The whole of this dip is explained, according to estimates made for purposes of the present paper, by the larger exclusions in 1956. U.S. exports actually appear to have grown more between 1955 and 1956 than between 1956 and 1957.

36

Thus the growth of exports to non-OEEC sterling countries including Malaya and Ghana between 1957 and 1958 was assumed to be representative for that area less Malaya and Ghana. The growth of exports to “Other Eastern Hemisphere” including Tunisia and Morocco between 1957 and 1958 was assumed to be representative for that area less Tunisia and Morocco. The growth of exports to the Overseas Territories excluding Malaya, Ghana, Tunisia, and Morocco between 1957 and 1958 was assumed to be representative for the area including these four territories.

  • Collapse
  • Expand