12 November 1953 FACTORS DETERMINING UNITED STATES IMPORTS OF COFFEE The present paper is the first in a series which will offer analyses of the factors that account for the imports into the United States of several important commodities. The general purpose of these investigations is to test the current assumption that even minor fluctuations in United States business conditions give rise to wide swings in the volume, and particularly the value, r of United States imports, thereby creating a dangerous degree of instability in world trade. Coffee is the first item chosen for investigation since United States imports of it represent the largest single import flow of a fairly homogeneous commodity in international trade as a whole. In 1951, the value of imports into the United States account for nearly 2 per cent of world exports. Factors determining United States domestic consumption per person, 1923-1940 and 1946-1951. In order to explain fluctuations in imports, it is necessary to understand the forces which govern coffee consumption in the United States/ Without going into a detailed description of the method of analysis followed, it may be said that the quantity (expressed in terms of green beans) of coffee consumed per person can satisfactorily be accounted for by the level and fluctuations in income and in retail prices. During the years 1923 to 1940, a change in average real income per person in the United States by> say, 10 per cent had the net effect of changing coffee consumption per person by 4,1 per cent in the same direction. On the other hand, a change in the retail price of coffee (which relates to roasted and ground coffee), when taken in relation to the movement in general food prices? by, say, 10 per cent, gave rise to a change in ooffee consumption by 3.0 per cent in the opposite direction. * The retail price, thus adjusted, measures the amount of food in general, measured at 1935-39 prices, which consumers had to forego in order to buy a pound of coffee. This measure of the relative expensiveness of coffee is henceforth referred to as the "relative price". Sec/137/53 (U.S. - Coffee)
- 2 - Since the war the consumption of coffee per person in the United States seems to have been affected by variations in income and price, as before the war. But while the effect of changes in these two factors appears to have remained practically unaltered, the level of annual coffee consumption per person in these post-war years was systematically about 1^ pounds larger than would have been the case if pre-war relations still applied. The conclusion therefore is that there has been after the war an increase in consumers' preference, or desire, for coffee which gives rise to an increase of consumption per caput by about 10 per cent, after allowance is made for the post-war level of incomes and prices. Consumption per person, 1939 and 1951. The following table shows consumption per person, real income per person and relative price per pound of coffee, all expressed in comparable units, in 1939 and 1951. Coffee Consumption per person in the United States, 1939 and 1951. Consumption per person (lbs. green bean basis) Real income per person ($ at 1935-1939 prices) Relative retail price (cents at 1935-39 purchasing power over food) 1939 14.9 540 22.8 1951 Change 16.6 (+ 1.5 lbs = /* 10 per cent 786 + $246 = * 46 per cent 37.0 * 14.2 i = * 62 per cent From the coefficients given above, it is possible to calculate the theoretical csffee consumption which would have occured in 1951, if all factors influencing coffee consumption other than income and price had been the same in the two years considered. The 46 per cent increase in incomes would, if price and all other circumstances affecting the situation had been in 1951 as in 1939, have raised the coffee consumption by 19 per cent (46 per cent multiplied by 0.41), or 2.8 lbs. per person, as compared with 1939. The 62 per cent price increase, on the other hand, would, if incomes and all other relevant circumstances had been the same in the two years, have lowered coffee consumption by 19 per cent
- 3 - (62 per cent multiplied by 0.30), or 2.8 lbs. per person. In other words, if the pre-war relationships had still fully applied in 1951, coffee consumption would have been the same per person as in 1939, since the influence of a higher income would have been exactly compensated by the influence of a higher price. In fact, however, coffee consumption was 10 per cent or 1.5 lbs. larger per person. This means that consumers' preference for coffee has increased in such a way as to raise consumption, after elimination of price and income effects, by 10 per cent. National consumption and imports 1939 and 1951 Total national consumption in the United States is, of course, affected by the growth in population which was 17.6 per cent between 1939 to 1951. following table shows national consumption and imports in the two years, both in quantity (expressed in terms of green beans) and in value (expressed in current prices on a comparable basis, i.e. relating to the unroasted equivalent). Volume and Value of National Consumption and Imports in the United States, 1939 and 1951 1939 1951 Consumption per person (lbs. green beans) 14.9 16.6 National consumption (million lbs., green beans) 1,950 2,556 Retail price (cents per lb., green bean equivalent)* 17.9 69.4 Consumers' outlay ($ million) 349 1,774 Imports (million lbs., green beans) 2,014 2,693 Imports ($ million) 140 1,361 Average import price [f per lb., green beans) 6.9 50.5 The * i.e. with allowance for 20 per cent loss in roasting and grinding
- 4 - It will be seen that the volume of national consumption and imports is closely similar in both years. The slight excess of imports over consumption in 1939 and the somewhat larger increase in imports than in consumption between the two years are due to fluctuations in stocks held by dealers in the United States. Comparing, on the other hand, consumers' outlay and the value of imports, a large change in their relation, has occured since pre-war. Imports accounted for 40 per cent of consumers' outlay in 1939, and for 77 per cent in 1951. In other words, the margin between the import price and the retail price which had been 61 per cent in 1939 had dwindled to 27 per cent of the retail price in 1951. This change is not only quite extraordinary, since dealers' and processors' margins usually remain fairly constant*, but also obviously important from the point of view of international trade, since the consumption of coffee in the United States does depend, inter alia, on the retail price of coffee, while the dollar earnings of supplying countries do depend on the United States import prico. Importance of processors' and dealers'margins within the United States for world trade in coffee ' In order to estimate the importance of the change in processors' and dealers' margins within the United States for the supplying countries, a calculation has been undertaken on two extreme assumptions. In both cases, an attempt has been made to measure the difference between the actual value of United States 1951 imports f coffee and the value such imports would have had, if the change in dealers' and processors' margins had not occurred. The first estimate is drawn The reduction in processors' and dealers' margins since the war is not a phenomenon peculiar to coffee. According to the United States Department of Agriculture, the farmer's share in the consumer's dollar rose from 38 cents in 1939 to 50 cents in 1951 (Agricultural Statistics 1952, p. 669) - a tendency probably due to the growth of chain stores in food distribution. What is remarkable, however, is the size of the reduction in the margin in the case of ceffee. This is an indication that coffee has come to be regarded by retailers as a commodity such as sugar or flour, which must be carried as a matter of course. in spite of the small benefit that it brings. The fact the United States consumers spend a larger proportion of income upon food than before the war, in spite of reduced margins on comparable items, is due to more advanced manufacuring of many foodstuffs which now incorporate a larger amount of services.
- 5 - up on the assumption that the pre-war margin was added to the actual 1951 import price and consists of studying the effect of the resulting increase in the retail price upon the quantity of coffee consumed in, and therefore imported into, the United States. The second estimate is drawn up on the assumption that the 1951 volume of coffee imports into the United States was unaffected and that the maintenance of margins took entirely the form of a decrease in the import price.. On the first assumption, the higher retail price of coffee would,.have curtailed 1951 consumption by 6.4 lbs. per persen* representing a decrease in national as well as individual consumption, and hence in the volume and value.of imports, by 38.6 per cent. The value of imports in 1951 would have been lower by $525 million. If, on the other hand, the pre-war margin had been maintained through a reduction in the import price, without a change in either the domestic retail price or the volume of imports and consumption, the import price would have been 26.8 cents per lb. instead of 50.5 cents per lb. - a reduction by 46.9 per cent in both the import price and the value of imports. The decline in the 1951 value of imports on the second assumption would have been $638 million. Summary of findings The investigation of coffee imports leads to the following results: 1) both before and since the war, the net effect of income variations in the United States upon coffee consumption and imports is of the * The increase in the relative retail price of coffee which actually was 62 per cent between 1939 and 1951 would have reached 205 per cent. The influence of price with an increase of 62 per cent was a decrease in consumption by 2.8 lbs. per person (see above); with an increase of 205 per cent the decrease in the quantity consumed per person would be 9.2 lbs. (It goes without saying that this figure can only be approximate, since the extrapolation of the price elasticity beyond the range of past experience necessarily involves a certain degree of inaccuracy.) Hence, the difference between the actual consumption in 1951 and the consumption which would have taken place, had the maintenance of pre-war margins taken the form of an increase in the retail price, without a change in the import price, would have been 6.4 lbs. per person.
- 6 - order of 4 per cent for every 10 per cent change in (real) income. Hence, fluctuations in United States business conditions would have only a moderate effect upon coffee imports; 2) Post-war consumption and imports have risen by 10 per cent, after allowance for the effect of higher income and higher price, as a result of an increase in consumers' preference, or desire, for coffee. Without this increased preference, the value of 1951 coffee imports would have been lower by $124 m.; 3) finally, the heavy reduction in processors' and dealers' margins as compared with pre-war has raised the 1951 value of import by more than half a billion dollars. While this latter result is significant in itself, it also draws attention to the fact that international trade cannot be studied in isolation. To understand its fluctuations, conditions of internal trade in both the importing and the exporting countries must be taken- into consideration. The reason why this obvious conclusion has hitherto received only limited, if any, attention is that a country's foreign trade is often studied by means of general index numbers of volume, value and unit price, and not, or only rarely, in terms of individual commodities.