V Global Effects of Changes in Exchange Rates of Program Countries
- Morris Goldstein
- Published Date:
- March 1986
Nowhere perhaps have “aggregation effects” been so long recognized and so much discussed as in the context of simultaneous exchange rate action by primary-producing countries.47 Just recently, for ex-ample, Please (1984) has criticized the World Bank and the Fund for taking too “piecemeal” an approach to exchange rate policy and has suggested that developing countries should collectively devalue against the currencies of the developed countries. On the other hand, as noted earlier, some observers have taken precisely the opposite tack—warning that exchange-rate-induced increases in production and in exports of primary commodities, if implemented simultaneously by many program countries, will merely depress the world price of these commodities and affect unfavorably the instigator’s terms of trade—and for little benefit since the demand for these goods is quite price inelastic.
In this section, the aggregate or global effects of multilateral exchange rate changes by a group of program countries are examined. The examination proceeds in three steps. First, conventional trade theory is reviewed for whatever light it can shed on the world price and export effects of multilateral exchange rate changes.48 Second, several empirical characteristics of primary-commodity trade and of the commodity structure of exports by developing countries are presented to demonstrate the practical relevance of the problem. Finally, the role played by non-program countries, and particularly the industrial countries, in conditioning the effectiveness of exchange rate action by program countries is discussed.
Differentiated versus Homogeneous Goods in International Trade
In thinking about the effects of multilateral exchange rate action by program countries, it is useful to distinguish between exports of differentiated goods (such as manufactures) and exports of homogeneous goods (primary commodities like wheat, sugar, tin, and so on).49 Two aspects of that distinction are particularly pertinent.
First, whereas producers of differentiated goods can charge a price that is different from those of their competitors, producers of homogeneous goods are constrained to price their goods at the world price. This world price, in turn, is determined by the inter-action of world supply and world demand for the good. This means that one program country, or even a group of program countries, will only be able to affect the prices of their primary commodity exports if they can influence either world supply or world demand. In general, it can be shown that (in the absence of inventory and order backlog changes), a country’s ability to influence the world price of a homogeneous good will depend (positively) on its shares of world production (or exports) and world consumption (or imports), and on the (absolute) value of its own price elasticities of supply and demand for the good.50 If a country or a group of countries is too “small” to affect the world price, then increases in domestic supply of the good will increase the volume of exports or reduce those of imports at the given world price.
The clear implication of the global character of price determination for primary commodities is that production (or export) shares and supply-price elasticities are crucial to an assessment of the likely price effects of multilateral exchange rate action by primary producing program countries. The number of program countries taking exchange rate action, by itself, is not likely to be a useful indicator of serious aggregation effects. For example, simultaneous depreciations by two rice producers that together account for say, 35 percent of world rice exports and have relatively high supply-price elasticities could well have larger effects on global prices than simultaneous exchange rate depreciation by ten program countries that either export quite different primary commodities, or collectively account for only a small share of world exports of a single primary commodity.
A second, related implication of the distinction between the two types of tradable goods is that any “beggar thy neighbor” effects associated with multi-lateral exchange rate action would probably show up as price effects for homogeneous goods but as a combination effect of volume and price changes for a producer of differentiated goods. In other words, if producers of primary commodities ignore aggregation and interdependence and devalue together they may all receive a lower price than they expected. If producers of manufactured goods ignore interdependence, by contrast, and simultaneously devalue, they may all receive a smaller market share and lower export volume than expected because the competitive price advantage initiated by devaluation tends to be less long lasting than expected.51 The same distinction also explains why the key parameters of interest in the case of differentiated goods are not global production or consumption shares but rather the elasticity of substitution in demand among export bundles of different producers, the pass-through effects of exchange rate changes onto local currency factor costs and export prices (or how much of the nominal devaluation can be converted to a real devaluation), and the difference between short-run and long-run price elasticities of demand.
If the commodity structure of trade matters for assessing the global effects of exchange rate changes, it should also be pointed out that some conclusions on the inefficiency of exchange rate measures for producers of primary commodities seem to be of dubious validity. This applies particularly to the notion that if a country’s terms of trade are fixed (if they can affect neither the foreign-currency price of its exports nor of its imports), exchange rate changes can be of no value to it. Such an analysis ignores the point that exchange rate adjustments can still “work” on the supply side by increasing the relative price of tradable goods vis-á-vis nontradable goods, that is, by altering the internal terms of trade. Indeed, no matter what the commodity structure of a country’s trade, one of the more robust lessons of experience seems to be that a sine qua non for successful export performance is that exporting be consistently profitable relative to other activities in the economy. Put in other words, the relative price variable in the export supply function may well have a different denominator in the case of homogeneous products than in the case of differentiated ones (the price of nontradables, for instance, rather than the domestic price of tradables), but this does not change the basic message that the supply response to an exchange rate change hinges on engineering an improved domestic rate of return to exporting activities.52 Even when a country or country group is too “small” to affect the world price for a given primary commodity, exchange rate changes can still improve their export earnings by increasing their export volumes.
Finally, in the real world, neither countries nor tradable goods fit neatly into categories like “price-takers” versus “price-setters” or “homogeneous” versus “differentiated” goods; instead, they fall on a spectrum between these poles, and have more market power over prices of some commodities and less over others. This caveat should be recognized in interpreting the summary empirical characteristics that follow.
The previous discussion suggests that the global effects of multilateral exchange rate changes by program countries cannot be assessed without information on such empirical parameters as the share of primary commodities in the exports of program countries, the shares of program countries in the global production and consumption of primary commodities, supply-price and demand-price elasticities for primary commodities in program countries, and the demand-price elasticities for manufactured exports. In this subsection, some empirical evidence on those characteristics of trade is presented. Because the country composition of the program country group changes so much from year to year, most of the calculations use figures for non-oil developing countries, or sometimes even for all developing countries.
Changing Commodity Composition of Trade in Developing Countries
One of the more fundamental but still relatively unappreciated changes in international trade over the past two decades is the large increase in the share of manufactured goods in the exports of non-oil developing countries.53 As shown in Table 20, the share of manufacturing in the total exports of middle-income oil-importing developing countries has more than tripled (from 17 percent to 59 percent) over the past twenty years. The share of manufacturing has also increased sharply in the exports of low-income countries and of middle-income oil exporters, although the level of that share was still quite low (less than 10 percent). Table 21 gives further information on this shift, using a regional disaggregation of non-oil developing countries, concentrating on the difference between 1968–70 and 1979–81, and disaggregating primary commodity exports into energy and non-energy components. The salient points emerging from Table 21 are that: (1) by 1979–81, fully 59 percent of the total merchandise exports of non-oil developing countries were manufactures; (2) only in non-oil developing countries in the Western Hemisphere did manufactures account for less than half of total exports in 1979–81; (3) the rise in the share of manufacturing between 1968–70 and 1979–81 was greatest among non-oil developing countries in Asia but substantial increases were recorded in the other four regional groups as well; and (4) the falling share of primary commodities in the exports of this group reflected a decline in non-energy commodities.54 Indeed, the share of energy primary commodities in the total exports of these countries actually rose from 1 percent in 1968–70 to 7 percent in 1979–81.
(In percentage share of total merchandise exports)1
|Middle-income developing countries|
|Low-income developing countries|
|China and India||…||54||…||46|
|1983 program countries||—||44||—||56|
|Commodity Groups||Total||Africa||Asia||Europe||Middle East||Western Hemisphere|
|Value Share Total|
This increased importance of manufactured goods in the exports of developing countries has two implications. First, in assessing the global price effects of multilateral exchange rate changes by developing countries, it would be misleading to assume that they by and large fit into a primary-commodity framework. As indicated in Table 20, the dominance of primary commodities on the export side still is accurate for low-income developing countries (apart from China and India) and for middle-income net oil exporters, but it is decidedly not valid for middle-income oil importers. On a weighted-average basis, as a group non-energy primary commodities now are only a third of total exports of non-oil developing countries (Table 21) and primary commodities as a whole are only two fifths. Interestingly enough, the 1983 class of program countries had a primary-commodity export share significantly above that average, namely 56 percent (see Table 20).
The second implication of the growing importance of manufactures is that because both the price and income elasticities of demand for manufactures are generally higher than those for other commodity groups, there is a strong likelihood that the aggregate price and income elasticities of demand for (non-oil) developing country exports are now considerably higher than they used to be.55 Consistent with this proposition, Goldstein and Khan (1982) found that the income elasticity of demand for export volumes from non-oil developing countries was higher during 1973–80 than 1965–72. Similarly, there is increasing evidence that exports by non-oil developing countries do respond significantly to competitive prices on the demand side. For example, Khan (1974) found that the quantity of exports could be well explained for 15 individual non-oil developing countries by the level of real income in the industrial countries and by the ratio of the developing country’s export price to an average of export prices of industrial countries; for the 9 of those 15 non-oil developing countries where this price elasticity was significant, its average value over a year was 0.94.56 More recently, Grossman (1982) estimated quarterly import demand equations for 11 representative manufactured commodities that entered the U.S. market during 1968–78. Since separate equations are estimated for imports from industrial countries and those from non-oil developing countries, Grossman is able to determine the substitutability among three classes of goods—domestically produced goods, imports from industrial countries, and imports from non-oil developing countries. One of Grossman’s main conclusions is that U.S. imports of manufactures from the non-oil developing countries are quite price sensitive, with a mean (own) price elasticity of demand of 1.7 (for a one-year period). These results do not point toward any “elasticity pessimism” for exports from non-oil developing countries.
Primary Commodity Trade
Although manufactures are now clearly much more important than they used to be in the exports by non-oil developing countries, this does not alter the facts that primary commodities are still more than half the exports of (1983) program countries, and that such commodities constitute the mainstay of foreign exchange earnings in many individual program countries. For example, again using World Bank data for 1981, primary commodities accounted for 90 percent, 88 percent, 80 percent, 70 percent, and 60 percent of total exports in Chile, Kenya, Argentina, Uruguay, and Brazil, respectively. This is reason enough to examine those factors identified earlier (such as shares in world production and supply elasticities) as crucial in deter-mining the world price effects of multilateral exchange rate changes by primary-producing countries.
Table 22 shows the share of world exports of primary commodities accounted for by developing countries in both 1968–70 and 1979–81. Shares for oil exporting developing countries and for non-oil developing countries, and for energy and non-energy primary commodities, are given separately. Two conclusions stand out. First, the weight of developing countries as a whole, and of non-oil developing countries in particular, in world exports of primary commodities is much lower than is often supposed. Specifically, by 1979–81, non-oil developing countries accounted for only 19 percent of world exports of all primary commodities and for only 29 percent of exports of non-energy primary commodities. Clearly, the dominant exporters of non-energy primary commodities are not the non-oil developing countries but the industrial countries (which account for 68 percent of world exports), whereas for energy, the dominant exporters are, of course, the oil exporting developing countries (who account for 75 percent of world exports). If anything, the figures in Table 22 probably understate the dominance of the industrial countries in the exports of non-energy primary commodities because they do not take into account exports of synthetic substitutes for these commodities. The second conclusion is that the share of non-oil developing countries in world exports of primary commodities has fallen rather significantly over the past decade or so. Whereas their share of all primary commodities was 31 percent in 1968–70, it was, as noted earlier, only 19 percent in 1979–81; similarly, the decline for non-energy primary commodities was from 38 percent of world exports in 1968–70 to 29 percent in 1979–81. More disaggregated figures reveal further that the declines in the export shares of these developing countries were most pronounced for agricultural raw materials, in metals and minerals, and food; the decline was smallest in beverages and tobacco.
|Non-Oil Developing Countries|
|(In percent of world total)|
|Commodity Groups||World Value (In billions of U.S. dollars)||World||Industrial Countries||Oil Exporting Developing Countries||Total||Africa||Asia||Europe||Middle East||Western Hemisphere|
More information on the potential of developing countries to influence world commodity prices can be obtained by examining the production, consumption, and import shares (as well as export shares) of individual commodities. Table 23 shows the shares of developing countries in the world production, consumption, exports, and imports of 23 of the more important (non-oil) primary commodities in international trade. Separate share figures are given for all developing countries and for the top three and top five developing country aggregates where data were available.
|World||All||Top Three||Top Five||World||All||Top Three||Top Five||World||All||Top Three||Top Five||World||All||Top Three||Top Five|
In brief, two results deserve mention. First, not surprisingly, developing countries have considerably more weight in the world supply of primary commodities (in production and export shares) than in the world demand (consumption and import shares) for them. Of the 23 primary commodities shown in Table 23, developing countries had a 50 percent or greater share of world production of 16 of these goods; the corresponding figure for world consumption was only 8 goods. This means of course that their influence on the world prices of these commodities typically comes from their role as suppliers. Second, there is considerable variation across commodities in the concentration of production and of exports by developing countries. Whereas production was highly concentrated for coffee, cocoa, tea, rice, jute, natural rubber, tin, and manganese, it was much less so for beef, nickel, primary aluminum, iron ore, lead, and zinc. The potential is therefore greatest for this former group of primary commodities for collective exchange rate action to influence world prices; where the concentration in production is relatively low, as in the latter group of primary commodities, it would take simultaneous exchange rate action by many producing countries to achieve the same impact on world supply.
As indicated earlier, the effects on world prices of exchange-rate-induced supply shifts depend not only on the global production shares of the developing countries but also on their supply-price elasticities for the relevant primary commodities. High supply elasticities, ceteris paribus, increase the influence of supplying countries on the world price and low elasticities lower it. Table 24 presents some representative estimates of supply elasticities for a variety of agricultural and mineral primary products. While such estimates are known to be quite sensitive to the choice of country and commodity, to the specification of the model, and to the estimation period, the results seem to point to three general conclusions.
|Commodity||Country||Period||Author of Study||Short-Run|
|Nigeria||1952–72||Olayemi and Olayide||0.04||1.75|
One is that the short-run (one-year) price elasticities are usually much smaller than long-run ones. This suggests that the danger of significant effects on world prices being generated by multilateral exchange rate action is much less in the short run than in the long run. Second, for some of the commodities with relatively high producer concentration, such as copper, even long-run elasticities are low enough to dampen the world price effects of exchange rate changes. For other concentrated commodities, however, such as cocoa and perhaps coffee, the potential for induced long-run price effects is larger because the supply elasticities themselves are larger. Third, the sample of significant supply-price elasticities in Table 24 supports the earlier proposition that even with fixed external terms of trade, the capacity exists to boost exports by real exchange rate depreciation if such depreciation can be translated into an increase in the real returns for producing exportables.
Thus far, we have made a preliminary identification of those primary commodities where the potential is greatest for multilateral exchange rate action by producers to affect the world price. It would also be useful to know which countries have a relatively high potential to affect world prices by their own supply or demand actions. For this, not only the relative degree of producer concentration by commodity must be known, but also the shares of those commodities in each country’s exports or imports. Clearly, a country whose exports are highly concentrated in a few high-producer-concentration commodities will have more market power in its export markets than one whose exports are more diversified among low- and high-producer-concentration commodities.
Branson and Katseli (1980) have recently constructed just such an index of market power for the exports and imports of 101 countries, both developing and developed. Specifically, they define export market power (Zx) as:
where λi is the country’s export share of commodity i in total world exports of i, and δi is commodity i as a proportion of the country’s total exports, so λiδi is the country’s export share of commodity i weighted by the relative importance of commodity i in the country’s exports. The index of import market power (Zm) is defined symmetrically. The results, based on 1974 United Nations trade data, are shown in Table 25. Three conclusions stand out.
|Market Power Indices|
|Countries||For exports (Zx)||For imports (Zm)|
|Central African Republic||0.0016||0.0001|
|Germany, Fed. Rep. of||0.0646||0.0272|
|Iran, Islamic Rep. of||0.1244||0.0061|
|Papua New Guinea||0.0427||0.0005|
|Syrian Arab Republic||0.0094||0.0012|
|Trinidad and Tobago||0.0571||0.0083|
Data refer to 1972.
Data refer to 1972.
First, consistent with our earlier results on the shares of developing countries in world production and consumption of individual primary commodities, Table 25 suggests that developing countries have appreciably more market power over exports than over imports. For 1983 (Group A) program countries, for example, the index of export market power was 0.073, while that for imports was only 0.003—only
Observe also that, on this index of market power, global export or import shares alone are not necessarily a fail-safe predictor of market power; for example, Table 25 indicates that Japan, not the United States, has the highest import market power of the 101 countries in the table (presumably because high-concentration consumption goods represent a larger share of its imports). Third, at least for 1983 program countries, the degree of import power of program countries was about the same as for all non-oil developing countries. Their export market power was significantly higher. This suggests that the risk of aggregate price effects from multilateral exchange rate action is somewhat greater for program countries than for all developing countries. This conclusion however seems to be quite sensitive to the year’s country-composition of program countries. In 1982, for example, program countries had slightly lower average export market power (0.034) than all non-oil developing countries (0.044).
Role of Industrial Countries
So far, the analysis of the likely impact of multilateral exchange rate changes by program countries has concentrated almost exclusively on the “producer” or “exporter” side of the market. It is abundantly clear, however, that the impact of exchange rate action will be no less conditioned by what happens on the “consumer” or “importer” side of the market. Since, as previously documented, the majority of exports from non-oil developing countries and program countries go to industrial countries, the effectiveness of exchange rate action by program countries cannot be evaluated without discussing the economic policy scenario and underlying trade behavior of those industrial countries.
Two basic points are relevant. First, if non-oil developing countries are to further reorient their export structures toward manufactures to benefit from the relatively high price and income elasticities of demand for these goods, industrial countries will have to be willing allies in that transformation. As shown in Table 26, both price and income elasticities of demand appear to be higher for manufactures than for non-manufactures, at least in industrial countries. This, coupled with the fact that manufactures have been the fastest growing component of exports of non-oil developing countries over the past twenty five years, has been an incentive for developing countries to orient their production and export patterns more and more toward (labor-intensive) manufactures. But even if many primary-producing non-oil developing countries were convinced of the wisdom of such a change in export structure, and could obtain the funds to finance it, their efforts could come to naught unless industrial countries were willing to provide satisfactory access to their own markets. Without such access, the greater volume of manufactures could not be sold at above “dumping” prices (or in sufficient volume if quantitative restrictions applied). Also, without such access, it would be impossible to induce those non-oil developing countries that are already major exporters of manufactures to move up the “chain of comparative advantage” and produce more skill and capital-intensive manufactures to make room for the “new” producers of manufactures.57 This is why a Fund staff report on Exchange Rate Policies in Developing countries, concluded that:
… from an optimal standpoint, the speed of adjustment of the non-oil developing countries exporting primary products should in part be related to the speed at which industrial countries are able to adjust to a larger and more diversified flow of imports from the rest of the world.
|Commodity Categories||Deppler-Ripley1 (1978)||Taplin2 (1973)||Basevi3 (1973)||Clark4 (1977)||Ball-Marwah4 (1962)||Houthakker-Magee4 (1969)||Barker5 (1976)||Stern etal.6 (1976)||Theil and Clements7 (1978)|
|Foods and beverages||−0.34||−0.57||−0.55||−1.14||−0.47||−0.18 to||−0.13||−0.78||−0.58|
|(SITC 0+1)||−0.28 to||−0.18|
|(SITC 2 + 4)||—||−0.63||−0.13||−1.25||−0.83||−0.50||−0.95|
|Fuels (SITC 3)||—||−0.63||−0.04||—||…||−0.96||…|
|Manufactures and miscellaneous goods|
|(SITC 5–9)||−0.97||−1.23||−0.71||−4.72||−1.98||−1.8 to||−1.37||−1.34||−1.16 to|
|Foods and beverages||2.83||0.84||1.08||0.38||0.96||0.30 to||1.12||…||0.14|
|(SITC 2 + 4)||0.32||0.75||0.83||0.96||1.15||…||0.25|
|Fuels (SITC 3)||1.22||0.96||1.40||—||…||—||…||…|
|Manufactures and miscellaneous||1.27||1.44||1.46||2.60||2.07||1.11||1.99||…||0.28 to|
|goods (SITC 5–9)||0.32|
Unweighted average, 14 industrial countries.
Unweighted average, 25 industrial or semi-industrial countries.
Unweighted average, 10 industrial countries.
Median estimate, all industrial countries.
Pooled, cross-section for 13 industrial countries.
Unweighted average, 14 industrial countries.
Unweighted average, 25 industrial or semi-industrial countries.
Unweighted average, 10 industrial countries.
Median estimate, all industrial countries.
Pooled, cross-section for 13 industrial countries.
It is also why projections for exports by non-oil developing countries in the Fund’s World Economic Outlook analyses are always made conditional on the stance of protectionism in the industrial countries.
The importance of increased access to markets in the industrial countries would not have to be emphasized so much were it not for evidence: (1) that the postwar momentum toward trade liberalization came to a halt in 1979 and has suffered some reversal since then;58 (2) that the piecemeal protectionist measures adopted by industrial countries have tended to fall hardest on sectors (particularly labor-intensive manufacturing) where non-oil developing countries have, or are likely to have, a comparative advantage;59 and (3) that industrial countries have recently (1983) begun to rely more on quantitative controls and quotas—on measures that cannot be offset by exchange-rate induced gains in price competitiveness.60
While the effects of existing trade barriers in industrial countries on exports by non-oil developing countries are notoriously difficult to quantify, estimates from the recent empirical literature strongly suggest that these effects are far from trivial.61 For example, Cline et al. (1978) have estimated that a 60 percent reduction in industrial countries’ tariff and agricultural nontariff barriers would increase exports by non-oil developing countries, exclusive of oil and textiles, by approximately 3 percent (using 1974 values); a similar liberalization of textile trade would, according to the same authors, produce perhaps another 3 percent increase in exports earnings by these countries. More recently, Whalley (1984) has used an applied general-equilibrium trade model to estimate the income effects associated with, inter alia, the abolition of tariff and non-tariff barriers in industrial countries. Interestingly enough, he finds that removal of protection in industrial countries would increase income in non-oil developing countries by roughly $30 billion—a figure larger than the approximately $20 billion annual aid flow from the North to the South.62
The second basic point about the role of industrial countries is that their own business cycle, inflation, and exchange rate developments can often—in fact usually do—swamp the effect of exchange rate changes by non-oil developing countries on world commodity prices. Goreux (1980), analyzing a sample of 37 (non-oil) primary commodities over 1962–79, found that each 1 percent change in an industrial-country business-cycle index was associated (ceteris paribus) with a 2.2 percent change in primary commodity prices, while each 1 percent change in the prices of manufactured exports by the industrial countries associated with a 0.7 percent change in primary commodity prices.63 More recently, Chu and Morrison (1983) undertook a more extensive analysis of the determinants of non-oil primary commodity prices over 1958–82. They found that much of both the long-term and short-term movement in these commodity prices could be explained by changes in industrial production, and in inflation rates adjusted for changes in major currency exchange rates (vis-á-vis the U.S. dollar) in the industrial countries. The point is not that exchange-rate induced production changes by non-oil developing countries do not matter for primary commodity prices; like other supply shocks, surely they do. But they are by no means all that matters, or even probably what matters most. For this reason, it is important not to conclude that because the prices of non-oil primary commodity exported by non-oil developing countries were weak in 1981–82, and because these countries were depreciating their exchange rates during that period, the latter necessarily was responsible for the former.64 The roles played by the recession, the high inflation, and interest rates in the industrial countries, and the appreciation of the U.S. dollar vis-á-vis other industrial-country currencies also need to be taken into account.65
This section has explored the possibility that simultaneous exchange rate action by program countries could have serious aggregate effects on the prices of exports by program countries. It has been argued that this proposition should be applied mainly to primary commodities and that the potential for significant aggregate price consequences depended mainly on the ability of program countries to affect world supply. While this potential is clearly much higher for some commodities (cocoa, coffee, natural rubber, or tin) than for others (wheat, citrus fruits, or iron ore), and in the long run rather than the short run (because supply elasticities were much larger in the long run than in the short run), the risks are much reduced in practice because (1) primary commodities now rep-resent a significantly smaller share of exports by non-oil developing countries than they did two decades ago; (2) the share of non-oil developing countries in world exports of primary commodities is now also considerably smaller than even a decade ago; and (3) not all program countries change their exchange rates at the same time and those that do usually do not export the same products.66 Still, such interdependence and aggregation effects associated with multilateral exchange rate action need to be closely monitored and this section has identified the individual non-oil developing countries and primary commodities where “market power” seems to be relatively high. A case has also been made for the view that exchange rate adjustment can be useful to protect the profitability of exporting even for those non-oil developing countries who face fixed external terms of trade. Last but not least, the role played by industrial-country policies in conditioning the effects of exchange rate action by program countries has been emphasized.