¹ See Viner, Jacob, Studies in the Theory of International Trade (New York: Harper and Brothers, 1937)Google Scholar, Chaps, viii and ix. Specific applications and formulas are discussed, pp. 558–63.

² Two series already exist. F. W. Taussig's well-known index numbers on net and gross barter terms of trade cover the period from 1880 to 1913. Taussig, F. W., International Trade (New York: The Macmillan Co., 1929), pp. 411–19Google Scholar. Werner Schlote covered most of the nineteenth century with a series on the net barter terms of trade from 1814 to 1933, but he changed his components in 1854 so that the scries is not strictly continuous and there are avoidable imprecisions which limit its usefulness. Schlote, Werner, “Entwicklung und Struckturwandlungen des englischen Aussenhandels von 1700 bis zur Gegenwart,” Probleme der Weltwirtschaft (Jena, 1938), Appendix, Table 15Google Scholar. See also below nn. 7 and 11.

³ Albert H. Imlah, “Real Values in British Foreign Trade, 1798–1853,” The Journal of Economic History, VIII (November), 133–52.

⁴ There are two limitations. Somewhat different price tables were used for England, Scotland, and Ireland in recording these values so that variations in the flow of goods to or from these sections could distort results. Fortunately the fluctuations seem to have been comparatively small, however. Secondly, the base year is remote from the period under analysis. It would be better, tending to ensure more cognate conditions, if there were, say, three series of absolute values, each with its own price base in a mid-point year, to cover this period from 1798 to 1864. Two such series, joined by means of overlapping years, are used for the index numbers presented below for the years from 1865 to 1913. It would be possible to construct such series to use in place of the official values from 1798 on, but it would be necessary to use the official values in order to do so and it is questionable whether there would be sufficient improvement in precision to justify the very considerable amount of labor involved.

⁵ Parliamentary Papers, 1909, [4954], CII, 53.

⁶ Ibid., 1914, [7432], LXXXIX, 30.

⁷ Sir Robert Giffen prepared absolute values for exports and gross imports, in terms of prices of 1873 and of 1883, for various years from 1840 to 1883 and these could be used to bridge the gap, but Giffen's figures omit many years and do not form a continuous series. “Report to the Secretary of the Board of Trade,” Parliamentary Papers, 1884–1885, [4456], LXXIGoogle Scholar, and “Third Report of the Royal Commission on the Depression on Trade and Industry,” ibid., 1886, [4797], XXIII, 329. Werner Schlote prepared tables of absolute values from 1801 on using six base-point periods for imports, seven for exports, and one for re-exports from 1801 to 1933. It is not quite clear whether he constructed these commodity by commodity or in lump, and whether he used Jevons’ general price index or individual commodity indexes. His price movements strongly suggest that he relied on Jevons’ general index, which would make the precision in using several base points rather more seeming than real.—Schlote, Probleme der Weltwirtschaft, Appendix, Tables 8–10 and 27.

⁸ One could, of course, follow the procedure used by F. W. Taussig for the years 1880 to 1899 and derive absolute values for net imports by applying a price index based on gross imports to the market value of net imports. But some very large changes occurred in the re-export trade through this longer period so that this easier method would produce rather unreliable results here.

⁹ The modification consisted in removing the commodities for which the absolute values of re-exports were separately computed by means of their own price indexes. From 1880 to 1899 the import index was used without modification since the component data in the Board of Trade series of absolute values for these years have not been published.

¹⁰ The differences in results from those of the usual type of index can be considerable, as Silverman has shown by comparing his export and import price indexes, constructed with monthly prices and fixed weightings, with those he prepared for Taussig by the method used here.—Silverman, A. G., “Monthly Index Numbers of British Export and Import Prices, 1880–1913,” The Review of Economic Statistics. XII (1930), 139–48CrossRefGoogle Scholar.

¹¹ This reverses Taussig's procedure by which a rise in the index numbers indicated “unfavorable” movement. The ratio used here is much less confusing, permits easier comparison with other indexes, and, as Jacob Viner has pointed out, does not involve any question of principle.—Studies in the Theory of International Trade. See note on p. 558. The complete formulas for the first three indexes (Columns I-K) are given by Viner with comment, ibid., pp. 558–63.

Schlote constructed his index on net barter terms of trade by dividing a price index of exports of manufactured goods by the price index of gross imports for the years 1814–1853, and by the price index of, imported raw materials from 1854 on. His index numbers show the general trends but are not very precise on short-term variations, and the series is not strictly continuous. Reduced to the same base, his index numbers for net barter terms of trade for 1814 is 136 to my 149; for 1815, 141 to my 153; for 1824, it is 144 to my 164; and for 1825, 129 to my 138. Schlote.—Probleme der Weltwirtschaft, Appendix, Table 17.

¹² Taussig, F. W., “The Change in Great Britain's Foreign Trade Terms after 1900,” The Economic Journal, XXXV (1925), 1–10CrossRefGoogle Scholar.

¹³ For discussion of its limitations see Gottfried von Haberler, The Theory of International Trade (New York: The Macmillan Co., 1936), pp. 163–66Google Scholar; and Viner, Theory of International Trade, pp. 562–63.

¹⁴ Viner, Theory of International Trade, pp. 563–64.

¹⁵ The net barter terms moved a little less unfavorably in the postwar period, however, than could hitherto be supposed, because average prices of the goods actually imported dropped a little more than is shown in any of the wholesale price indexes on which we have formerly had to rely. From 1816–1818 to 1838–1840 the numbers in Silberling's unweighted index, the best available, fell 25 per cent, and in Jevons’ 28 per cent. The net import index above, in effect a weighted one, shows a decline of 33 per cent for the same years.

¹⁶ Cottons constituted 40 per cent of the total value of exports in 1816, and in spite of the phenomenal fall in prices, the same in 1850. By 1880 they had fallen off to 34 per cent. Woolens were 22 per cent in 1816, 15 per cent in 1850, and 9 per cent in 1880.

¹⁷ Since the tariffs on cotton and wool were specific duties, the relation to value varied with price movements but, unlike the duties on most other goods, they were reduced to very moderate proportions and finally repealed in this period. The duty on foreign cotton, for example, figures at about 7 per cent ad valorem in 1826 and, though the levies were cut in half in 1833–1834, at about 6.2 per cent in 1844 on the eve of complete repeal. East Indian cotton was admitted at lower duties. In the case of wool, the growing needs of the industry were placed ahead of domestic sheep raising. The duties on foreign wool were reduced substantially in 1824–1825 and, though the ad valorem effect also fluctuated with price changes, the average rate could not exceed 8 per cent and rarely exceeded half that figure on the better grades until repeal in 1844. Wool from British colonies was admitted free after 1825 and Australia soon became a major source of supply.

These light duties stand in contrast to an average rate of 43 per cent on the value of other net imports in 1844.

¹⁸ The import values, and therefore the import price movements, cited throughout this paper arc measured before tariff duties were collected. Duties do not enter into the international accounts directly affecting the terms of trade. They are elements in the manufacturer's costs, of course, and as such have some bearing on export prices. The fact that the duties on cotton and wool were reduced- and finally repealed in this period means that the cost of the imported raw material to the British manufacturer fell still more rapidly than is indicated by the index numbers in Table II or by the percentages stated below.

¹⁹ This method, by averaging all kinds of cotton exports together, could minimize price decline (or articles which, requiring much fabrication, should show the results of improved machine processes more markedly, and exaggerate that in yarn which requires little processing, when, as was the case until the mid-century, the proportions of the yarn exported increased considerably. Actually the difference in the rates of price decline from 1816–1818 to 1849–1851 is almost negligible, however. Prices of cotton-yarn exports fell 71.4 per cent, while those of other cotton exports fell 72.4 per cent. That the decline in the price of yarn should have so closely matched that in articles more strongly affected by labor-saving machinery was undoubtedly due largely to the corresponding decline in the price of the raw material.

²⁰ To reconcile these numbers with those given for the terms of trade in Table I it is necessary to take three factors into account: the more precise method used for calculating net import prices of cotton and wool, and not only the proportions of cotton and woolen exports to the total (see above, note 16), but also the proportions of the imports of each to the total figure. Net imports of raw cotton constituted 13.8 per cent of the total value of British imports in 1816, 18.2 per cent in 1850, and 10.7 per cent in 1880. Net imports of wool rose from just under 3 per cent in 1816 to slightly more than 3 per cent in 1880. Their price movements consequently exerted much less influence on average import prices than did those of cotton and woolen exports on average export prices.

²¹ The net import index for cotton fell one point below the number for 1845 only because a higher proportion of cheaper East Indian cotton was taken in 1848.

The case of wool is not so clear cut. The price series used for Australian wool, which was duty free, reached its low point in 1843, but improvements in quality and grading confuse the matter, and Australian wool continued to offer strong competition to the higher priced wools hitherto taken from other sources. A major factor in the decline of the net import index from the mid-thirties on was the increasing proportions of the cheaper wools taken. This accounts also for the greater part of the sharp drop in the net import price index in 1847–1849.

²² See Imlah, Albert H., “The Fall of Protection in Britain,” Essays in Honor of George Hubbard Blakeslee (Worcester: Clark University, 1949), pp. 306–20Google Scholar.

²³ This change in price trends is usually attributed to three main factors coming into operation in the fifties and sixties: the Crimean, American, and Prussian Wars; new and large supplies of gold; and export of capital goods. These are conveniently reviewed in Rostov, W. W., British Economy of the Nineteenth Century (New York: Oxford University Press, 1948), pp. 20–24Google Scholar.

²⁴ In weighing the effects of British protective duties on world prices it should be kept in mind that British exports and imports constituted an enormous share of all international trade throughout this period. Together they were over 30 per cent of the estimated world total in 1800, and they were still approximately 25 per cent in 1850. Furthermore, British net imports were regularly substantially larger than exports (see “Current Values” in Table I). The British market was paramount and exerted a profound influence on world prices of the staples imported. British tariffs on these goods could aggravate price fluctuation by depressing world prices in years of ample supply in relation to British demand, though in years of short supply the burden was probably borne by British consumers.

The gradual reduction in the direct toll taken by British customs duties is indicated by the following figures. In 1841 when virtually every article of import was subject to tax, net customs collections constituted 31.8 per cent of the current value of net imports. In 1847, with the suspension of the Corn Laws, the average rate dipped to 21.6 per cent. Ten years later it had fallen to 14.1 per cent, and by 1867 it was down to 9.7 per cent with spirits, wine, and tobacco contributing half the revenue, sugar and tea, more than a third.

²⁵ The estimates of market values of imports may tend to be a little high in years of rapidly rising prices—See Imlah, “Real Values,” p. 145. The import price index numbers may, therefore, be a little too high, and the net barter numbers a little too low, in such years as 1818 and 1825. Allowance for this does not alter the fact of sharp fluctuations though it may moderate the degree in some cases.

²⁶ In the one case (1835–1836) where the volume of imports did not rise appreciably more than that of exports, the improvement in prices, and large investments abroad, strengthened foreign purchasing power for British goods while very good harvests reduced domestic costs of foodstuffs and, with good conditions of employment, left a larger margin for purchase of manufactured goods in Britain itself.

²⁷ The tendency of foreign loans to raise export prices in the borrowing country is very clearly illustrated in 1835–1836 when British loans were largely channelled to the United States. British imports from America rose in price while those from other areas were relatively little affected and even went down. Indian cotton prices did not rise with American, for example.

²⁸ This circumstance is quite consistent with the price trends. It is interesting that in terms of values also, Britain was somewhat more heavily an importing than an exporting country at the beginning and in the middle than at the end of the series, though the growing magnitude of her adverse balance of visible trade tends to obscure this fact. In 1798–1800 the adverse balance constituted 14.9 per cent of the total value of Britain's visible exports and net imports; in 1816–1818, 8.9 per cent; in 1849–1851, 14.0 per cent; and in 1911–1913, 12.1 per cent. Thus Britain's “invisible” credits, whether from foreign investments, shipping, or other sources, were not less important in the beginning and middle than at the end of the period. Possibly they were as important in 1816–1817 also, but were left to accumulate for such a burst of imports and foreign loans as occurred in that boom year 1818.

²⁹ The average of the gross barter index for these years was 138.5. In the years 1900–1904 another peak was reached at 129.1.

³⁰ Jenks, Leland Hamilton, The Export of British Capital to 1875 (New York: Alfred A. Knopf, 1938), p. 61Google Scholar.