II Industrial Country Income Growth and Non-Oil Developing Countries’ Exports
- Mohsin Khan, and Morris Goldstein
- Published Date:
- August 1982
In this section we begin our investigation into how the rate of growth of real income in industrial countries affects the income growth rate in non-oil developing countries by considering the relationship between income growth in the former and export growth in the latter.
As suggested in Table 1, the reduced-form relationship between real income in industrial countries and exports from non-oil developing countries can be subdivided into (i) the relationship between real income and total imports in the industrial countries; (ii) the relation-ship between industrial countries’ total imports and their imports from the non-oil developing countries; and (iii) the relationship between exports from the non-oil developing countries to the industrial countries and the former’s total exports. In addition, an examination will be made of how the reduced-form relationship itself between industrial countries’ real income growth and non-oil developing countries’ export growth has changed over time and how it differs among the four analytical subgroups of non-oil developing countries.
Total Imports of Industrial Countries
There are at least four important channels by which a slowdown in real income growth in industrial countries can affect the growth of the total imports of these countries. Two of these channels can be viewed as “direct” in the sense that the growth rate of industrial countries’ income is a proximate determinant of the growth rates of their trade volumes or trade prices. The remaining two channels are classified as “indirect” because real income growth must affect some other macrovariable (e.g., the rate of unemployment or inflation) in industrial countries before it affects imports.
Direct Volume Effect
In the absence of changes in either the relative price of imports or the non-price barriers to trade, the volume of imports demanded in industrial countries will be directly related to the level of real income in those countries. Hence, slower real income growth should translate directly into a slower growth in import volumes.
There is a considerable empirical literature on the determinants of industrial countries’ imports.15 Five conclusions that emerge from that literature are particularly relevant for this study. First, “consensus” estimates of the income elasticity of demand for total imports are greater than unity, usually falling in the 1.25 to 2.25 range. This means that, ceteris paribus, a fall in the rate of growth of industrial countries’ real income is likely to lead to a larger than proportionate fall in the growth of their import volumes. This proposition is supported by the data on (percentage) changes in those countries’ real GNP and their import volumes that are displayed in Table 4 for 1968–72 and 1973–80. In both periods, the implied income elasticities are larger than 1.5, even when the extreme import volume changes in 1975 and 1980 are excluded.16
|Average 1968–72||1973||1974||1975||1976||1977||1978||1979||1980||Average 1973–80|
|Volume of total imports||9.9||12.2||0.3||–8.0||14.4||3.1||5.4||8.4||–2.4||4.2|
|Real gross national product||4.5||6.3||0.7||–0.6||5.2||3.9||4.0||3.7||1.2||3.1|
|Implied income elasticity1||2.2||1.9||0.4||13.3||2.8||0.8||1.3||2.3||–2.0||1.62|
|Ratio of nominal imports to nominal gross domestic product||11.03||12.9||16.5||14.7||15.7||15.8||15.2||17.1||18.5||15.8|
|Ratio of real imports to real GDP||15.93||17.7||17.6||16.2||17.7||17.6||17.8||18.5||17.9||17.6|
Ratio of changes in import volume to changes in real GNP.
Excludes 1975 and 1980.
Ratio of changes in import volume to changes in real GNP.
Excludes 1975 and 1980.
A second conclusion of the empirical literature is that income elasticities are typically larger for “cyclical” than for “trend” changes in industrial countries’ real income, so that the falloff in their import volume is apt to be greatest when there are sharp declines in their real growth rates.17 The 8 per cent decline in import volume in industrial countries in 1975 when their real income declined by 0.6 per cent, and the 2.4 per cent import volume decline of 1980 when real income grew by only 1.2 per cent (see Table 4) are consistent with this larger cyclical-elasticity hypothesis.
A third conclusion is that the time lag with which industrial countries’ real income affects the volume of their imports is very short, with virtually all the effect taking place within one or two quarters. This short lag may be contrasted with the much longer time lag associated with the effect of relative price changes on the volume of imports into industrial countries, where available evidence indicates that perhaps only one half of the final effect takes place within one year and where lagged effects will still be occurring as much as two to three years after the initial change in relative prices. This implies that for periods up to one year, changes in industrial countries’ import volumes will be dominated by changes in cyclical real income movements.18
The fourth conclusion deals with the commodity composition of import demand. Specifically, there is evidence that the income elasticity of import demand is higher for manufactured goods than for nonmanufactured goods.19 Again, the raw data are supportive—with the average increase in the volume of manufactured imports into industrial countries being substantially larger than that for all imports in both 1968–72 (12.1 per cent versus 9.9 per cent) and 1973–80 (6.4 per cent versus 4.5 per cent).20 This higher-income elasticity for manufactured goods has implications both for export performance of the non-oil developing countries as a whole and for export growth differentials among them, because these countries as a group have significantly increased the share of manufactures in their total exports over the past 15 years,21 and because certain subgroups of non-oil developing countries have appreciably higher shares of manufactures in their total exports than do others.22
A fifth conclusion that emerges—but more from analysis of the aggregate figures than from existing econometric work—is that the income elasticity for imports into industrial countries appears to have fallen off in 1973–80, after growing rather briskly in 1962–72.23 Indeed, as suggested in Table 4, the rise in the nominal ratio of industrial countries’ imports to their GNP from about 13 per cent in 1973 to over 18 per cent in 1980 is almost entirely explained by the much faster rise in import prices vis-à-vis domestic prices over this period. In real terms, this ratio of imports to income has increased by only 0.2—0.3 percentage point over the 1973–80 period. Thus, while an elasticity greater than unity ensures that, ceteris paribus, the growth of real imports will exceed the growth rate of real income in industrial countries, the margin by which the former exceeds the latter is narrower now than it was in the 1960s and early 1970s.
Direct Effect on Import Prices and the Terms of Trade
When dealing with an individual country, the traditional assumption is that the country will have no influence over the price it pays for its imports, that is, the country is a “price taker” on the import side of its trade accounts. The industrial countries, however, constitute such an important share of world imports that this assumption is no longer tenable when discussing their total imports.24 The price-taker assumption is particularly inappropriate for primary commodities that typically trade at a “world price” determined by world supply and demand conditions. When supply conditions are relatively stable, a slowdown in real income growth in industrial countries can be expected to produce a significant decline in the rate of price increase for these primary commodities—and not infrequently an absolute decline in prices as well.
Evidence on the association between economic activity in the industrial countries and primary commodity prices has recently been supplied by Goreux (1980) in a study of the International Monetary Fund’s compensatory financing facility. Examining nominal and real price25 movements for the 11 most important primary commodities in world trade (excluding oil), Goreux found that for more than one half of these commodities, prices reached their peak in the first half of 1974 (shortly after the peak of the industrial countries’ business cycle), and were at their trough in the first half of 1975 (at the bottom of the recession in industrial countries).26 Taking a broader sample of 37 primary commodities and a longer time period (1962–79), Goreux also found that fluctuations in primary commodity prices could be explained to a large extent by movements in cyclical activity in the industrial countries and in prices of manufactured exports (taken as a proxy for inflation in world trade). Further, whereas each 1 per cent change in manufactured export prices is associated with a 0.7 per cent change in primary commodity prices, each 1 per cent change in the industrial country business cycle index is associated with a 2.2 per cent change in primary commodity prices.
This direct (primary commodity) price effect has three implications for the non-oil developing countries, each of which is unfavorable in a world of slow and highly variable industrial country real income growth. One is that because primary commodities (excluding oil) constitute a larger share of non-oil developing countries’ exports than their imports (Table 5), a slowdown in real growth in industrial countries, ceteris paribus, typically worsens the terms of trade of the non-oil developing countries. Consistent with this conclusion, Table 6 reveals that the terms of trade for non-oil developing countries as a group worsened by 1.2 per cent per annum on average during 1973–80 (average industrial country real income growth, 3.1 per cent per annum) versus an improvement of 1.3 per cent per annum in 1968–72 (average industrial country real income growth, 4.5 per cent per annum). Similarly, within the 1973–80 period, there were only three years (1973, 1976, and 1977) in which the terms of trade of the non-oil developing countries improved, and these three years were also three of the four years of highest real GNP growth among the industrial countries during this period. However, it should be noted that the effect on the terms of trade of slower growth in industrial countries depends very much on the types of primary commodities (and manufactures) exported by the non-oil developing countries. As illustrated in Table 7, whereas the world prices of metals, minerals, and cereals have done very poorly over the 1973–79 period, the world prices of beverages and timber have performed much better, to say nothing of the dramatic rise in petroleum prices. Partly reflecting this heterogenous nature of commodity price movements, Table 6 shows that there has been considerable diversity of experience among different subgroups of non-oil developing countries in terms of trade performance; thus, whereas net oil exporters had an average annual improvement of 5 per cent during 1973–80, low-income countries whose exports are more heavily concentrated in metals and minerals showed an average deterioration in their terms of trade of 3 per cent per annum.
|Other primary products||76||56||43||26||24||18|
|Average 1968–72||1973||1974||1975||1976||1977||1978||1979||1980||Average 1973–80|
|All non-oil developing countries|
|Market prices of primary commodities (excluding petroleum) exported1||2.2||51.2||30.3||–17.8||17.0||28.9||–7.9||14.6||11.2||15.9|
|Terms of trade||1.3||6.5||–6.9||–8.9||2.2||6.3||–5.6||–0.5||–3.1||–1.2|
|Purchasing power of exports||7.9||17.2||–5.4||–8.6||16.4||10.4||4.9||8.6||4.6||6.0|
|Net oil exporters|
|Market prices of primary commodities (excluding petroleum) exported1||2.4||63.2||32.4||–21.6||15.1||16.7||1.2||16.6||12.4||17.0|
|Terms of trade||–2.2||14.5||13.3||–13.3||1.5||7.0||–7.2||12.2||12.1||5.0|
|Purchasing power of exports||4.3||13.9||21.6||–14.6||15.0||11.0||4.0||25.6||20.7||12.1|
|Major exporters of manufactures|
|Industrial countries’ export prices of manufactures1||4.4||17.7||21.8||12.3||—||9.0||14.7||14.5||11.0||12.6|
|Terms of trade||–0.3||3.8||–9.9||–6.5||2.5||2.0||–3.6||–2.3||–4.9||–2.4|
|Purchasing power of exports||12.0||20.8||–8.3||–4.9||21.9||12.5||8.2||7.0||3.2||7.5|
|Market prices of primary commodities (excluding petroleum) exported1||1.2||47.3||33.3||–21.3||21.9||29.8||–10.2||9.8||6.4||14.6|
|Terms of trade||–0.3||2.0||–12.2||–13.6||8.0||16.5||–9.3||–6.4||–9.9||–3.0|
|Purchasing power of exports||8.6||5.7||–16.4||–3.0||18.2||12.8||–2.6||–3.2||–6.5||0.6|
|Other net oil importers|
|Market prices of primary commodities (excluding petroleum) exported1||1.9||46.8||4.1||–17.8||16.2||24.4||–6.2||15.0||8.8||15.2|
|Terms of trade||–2.6||7.8||–10.8||–9.5||0.1||9.4||–6.6||–1.9||–6.1||–2.2|
|Purchasing power of exports||5.2||14.5||–9.3||–15.3||11.7||7.2||1.6||3.8||–0.8||1.7|
|World trade prices1 for:|
|Non-oil primary commodities||54.0||27.5||–18.0||13.5||20.5||–4.5||16.5||9.5||14.9|
Price indices calculated in U.S. dollars.
|Time Period (Weights)||Petro-leum||33 Commodities (Excluding Petroleum) (100.0)||Total (66.2)||Food||Nonfood (15.0)||Timber (4.4)||Metals and Minerals (29.4)|
|Total (51.2)||Beverages (18.7)||Cereals (7.9)||Fats and oils (7.2)||Other (17.4)|
The second implication of the direct price effect is that, because it operates in the same direction as the direct volume effect, a decline in real income growth in the industrial countries usually induces a decline in the growth rate of non-oil developing countries’ export purchasing power (defined as the value of the latter’s exports deflated by their import prices). Again, a comparison of the 1968–72 and 1973–80 periods is instructive. With the exception of net oil exporters, each of the analytical subgroups of non-oil developing countries experienced a significantly lower growth rate of export purchasing power in the slow growth period (1973–80) than in the higher growth period (1968–72); see Table 6. Most conspicuously, the average annual growth of export purchasing power in low-income countries fell from 8.6 per cent in 1968–72 to only 0.6 per cent in 1973—80. In the cases of major exporters of manufactures and other net oil importers, the declines were from 12 per cent per annum to 7.5 per cent, and from 5.2 per cent to 1.7 per cent, respectively. In a similar vein, Rhomberg (1968) found that the elasticity of developing countries’ raw material export volumes with respect to industrial production in the industrial countries was 0.59 whereas the corresponding value elasticity was 1.56 (for the 1953–65 period). Thus, in the absence of offsetting world supply shifts, terms of trade considerations are apt to magnify the effects of an industrial country slowdown on export volumes from non-oil developing countries.27
Yet a third implication of the primary commodity direct price effect is that those non-oil developing countries whose exports are concentrated in a few primary commodities can be expected to suffer relatively high degrees of instability in export earnings. In this connection, Goreux (1980) found that countries with low shares of manufactures in total exports experienced the highest degree of export earnings instability over the 1959–76 period.
As a final comment on the direct price effect, it should be recognized that the overall deterioration of net oil importing countries’ terms of trade that took place during 1973–80 also reflected the large increases in their cost of oil imports. In this regard, Cline (1980) has shown that, whereas 15 of 25 sample non-oil developing countries showed a deterioration in their terms of trade between 1971–72 and 1975, 19 among these countries showed an improvement in their terms of trade once the cost of oil imports was excluded from the import price data.28 For this reason, actual terms of trade changes of the non-oil developing countries during 1973–80 obviously ought not to be viewed as solely reflecting changes in real economic activity in the industrial countries.
Indirect Protection Effect
To the extent that slower real growth in industrial countries is associated with higher unemployment rates (as was dramatically the case in 1973–80)29 and particularly when higher unemployment occurs in “sensitive” domestic import-competing industries, slower growth is likely to bring forth pressures for increased protection against imports. Obviously, unless such pressures are resisted by authorities in the industrial countries, the overall growth of the volume of their imports will be less than that suggested by real income and relative price considerations, and the slowdown will have an import-depressing effect beyond that already indicated.
A review of trade policy developments in industrial countries during 1974–80 (Anjaria et al. (1981), and Nowzad (1978)) uncovers both positive and negative elements. On the positive side, despite increased pressures for protection in industrial countries, governments refrained from resorting to new across-the-board trade restrictions. In this sense, the decline in the average annual growth of world trade volume from 8.5 per cent in 1963–72 to 5.3 per cent in 1973–80 ought not to be viewed as reflecting an upsurge in protectionism. In fact, international efforts toward greater liberalization of tariffs and trade barriers proceeded in the context of the Tokyo Round of Multilateral Trade Negotiations which were concluded in 1979. These negotiations produced agreements on both tariff reductions and some nontariff measures. When implemented over the 1980–87 period, the Tokyo Round tariff reduction will yield an average ad valorem tariff of only 5 to 6 per cent in industrial countries.30
On the negative side, there were some disquieting commercial policy actions in a number of industrial countries after 1973. These trade actions were concentrated in industrial sectors producing textiles, clothing, footwear, steel, ships, electrical consumer goods, and a variety of other manufactured goods.31 Most recently, trade in petrochemicals and automobiles was also subject to restrictive measures. While these trade actions affected trade among the industrial countries themselves, their principal effect was on exports originating in non-oil developing countries (and in Japan). Indeed, it is important to note that, even when sectoral trade restrictions are applied in a nondiscriminatory manner (across countries), their de facto impact is almost always quite uneven across exporting countries simply because the commodity and geographic distribution of exports differ widely from country to country. The main reasons why the above-cited protectionist actions are particularly worrisome to non-oil developing countries are because they involve sectors (usually those with high labor/ capital ratios) where comparative advantage has shifted, or is likely to shift in the future, to the non-oil developing countries and because the industrial countries still constitute such an important market for developing countries’ exports.32
Although the effects of existing trade barriers on exports from non-oil developing countries are difficult to quantify, there are strong suggestions from the recent empirical literature that the effects are by no means trivial. For example, Cline et al. (1978) calculated that a 60 per cent cut in industrial countries’ tariffs and agricultural nontariff barriers would probably increase exports of the non-oil developing countries, excluding exports of oil and textiles, by roughly 3 per cent (using 1974 values). Further, a similar liberalization of textile trade would produce another 3 per cent increase in non-oil export earnings. They also found that most of the induced expansion in non-oil developing countries’ exports would occur in manufactured goods, and that a relatively small number of export-oriented Asian states (Hong Kong, Korea, and the Philippines) would be the main beneficiaries.
Indirect Competitive Effect
For imports of differentiated goods (primarily manufactures) into industrial countries the volume of imports demanded is a function not only of their real income but also of the price of imports relative to that of similar domestically produced goods. To the extent that a slowdown in real growth in industrial countries reduces both their inflation rates and the prices of domestic goods relative to those of imports, there would be a depressive effect on imports that would be additional to the three other effects already mentioned.33
Although existing empirical work on the determination of inflation in industrial countries reveals large uncertainties about the form, size, and timing of the relationship between inflation and excess demand (e.g., see Spitaller (1978), Fellner (1978), and Okun (1978)), there would seem to be little doubt that any “indirect competitive effect” of an industrial country slowdown would take longer and would be smaller than the direct import effects discussed earlier in this section. For example, Cline (1980), using an inflation equation developed by Modigliani and Papademos (1978), found that, starting from a base of 7.7 per cent unemployment and 6.6 per cent inflation, holding the unemployment rate at 9 per cent for five years would reduce the inflation rate to only 4 per cent in the United States. Okun (1978), summarizing the results of six Phillips curves for the United States, similarly concluded that an extra percentage point of unemployment maintained for a year would reduce the inflation rate by 0.3 percentage point. Somewhat more optimistically, Cagan (1978) estimated that maintaining the unemployment rate one percentage point above the full employment rate for a typical four-year business cycle would reduce the inflation by 1.5–3.0 percentage points (depending on whether rational or adaptive expectations are assumed). This evidence, coupled with a “consensus” long-run (two to three years) price elasticity of demand in industrial countries for imports of manufactures of, say, 1 to 1.5 (see Stern et al. (1976), Deppler and Ripley (1978), Goldstein (1980), and Goldstein and Khan (forthcoming)) produces the result that it may take as much as three or four years before a slowdown in industrial countries would generate a significant competitive effect on imports into those countries.
This does not mean, of course, that the inflation rate in industrial countries is of little concern to non-oil developing countries.34 Inflation in industrial countries affects, inter alia, import prices for the non-oil developing countries, the real interest rate on their debt (particularly if the inflation rate was unanticipated), their capital requirements, and the real value of their international reserves (see IMF, World Economic Outlook (1981), Colaço (1980), and Cline (1980)). Thus, the rate of inflation that is associated with a given industrial country slowdown is important to non-oil developing countries but for reasons not primarily related to its effects on imports into industrial countries.
Industrial Country Imports from Non-Oil Developing Countries
The previous section discussed the likely response of industrial countries’ total imports to a fall in real growth rates in those countries. What is, however, of greater concern to the non-oil developing countries is how their exports to industrial countries will change in response to changes in growth in the industrial countries. To the extent that the industrial countries’ imports from the non-oil developing countries are less affected by an industrial country slowdown than are their total imports, the repercussions for growth rates in the non-oil developing countries are, of course, less serious.
Table 8 documents that the volume of industrial countries’ imports from the non-oil developing countries in fact grew faster (5.6 per cent per annum) in 1973–80 than total industrial country imports (4.2 per cent per annum). Virtually all of this higher growth rate of import volumes from non-oil developing countries took place during 1977–80 when these imports grew at a 6.8 per cent average annual rate (versus 3.6 per cent for the volume of all imports into industrial countries). In contrast, the 1968–72 period was marked by a slower average growth rate for imports from non-oil developing countries (7.7 per cent) than for total industrial country imports (9.9 per cent per annum). As such, it seems legitimate to conclude that the implied income elasticity for industrial countries’ imports from non-oil developing countries was roughly constant, or even rising, over the 1968–80 period while that for their total imports was probably falling.
|Average 1968–72||1973||1974||1975||1976||1977||1978||1979||1980||Average 1973–80|
|Volume of industrial countries’ imports from non-oil developing countries||7.7||7.8||—||–0.8||10.7||3.6||9.8||11.0||2.9||5.6|
|Volume of total industrial countries’ imports||9.9||12.2||0.3||–8.0||14.4||3.1||5.4||8.4||–2.4||4.2|
|Implied income elasticity for industrial countries’ imports from non-oil developing countries||1.81||1.2||—||1.3||2.1||0.9||2.4||3.0||2.4||1.71|
|Share of non-oil developing countries in total imports into industrial countries (value terms, level in per cent)||…||15.1||14.4||14.0||14.3||15.0||14.7||14.9||14.9||14.7|
|Non-oil developing countries’ export prices (U.S. dollars)||2.0||32.4||36.3||–1.3||4.1||14.2||4.1||17.6||17.1||15.5|
|Share of oil exporting developing countries in total industrial country imports (value terms, level in per cent)||…||8.3||15.6||15.2||15.8||15.6||14.3||14.3||6.8||14.5|
|Industrial countries’ share of total industrial country imports (value terms, level in per cent)||…||73.0||65.8||66.9||65.7||65.3||67.9||66.5||64.0||66.9|
|Volume of industrial countries’ imports from oil exporting developing countries||…||6.7||—24.6||—9.1||10.7||7.8||—2.1||—1.6||—16.8||—3.6|
|Volume of industrial country imports from industrial countries||…||20.2||2.4||–9.6||12.3||4.5||7.7||7.9||2.4||6.0|
|Oil exporting developing countries’ export prices (U.S. dollars)||…||40.2||229.1||4.9||6.4||9.5||0.4||43.9||67.7||50.3|
|Industrial countries’ export prices (U.S. dollars)||…||20.6||25.2||11.1||0.6||7.9||12.9||15.8||13.2||13.4|
Elasticity calculated as average of annual elasticities.
Elasticity calculated as average of annual elasticities.
Despite the relatively high growth rate for the volume of their imports from the non-oil developing countries during 1973–80, the share of the value of industrial countries’ imports coming from the non-oil developing countries was almost constant over this period at about 15 per cent; see Table 8. The main reason why the relatively high volume growth was not translated into a large gain in non-oil developing countries’ market share of industrial country imports was that the prices of the former’s exports were not increasing as rapidly as those of some other suppliers to the industrial country markets, for example, the major oil exporters.35 Thus, while non-oil developing countries’ export prices increased at an average annual rate of 15.5 per cent, oil exporting developing countries’ export prices, reflecting the large increase in world oil prices, increased by about 50 per cent per annum (see Table 8). As a result, the oil exporting developing countries were able to increase their share of imports into industrial countries from 8.3 per cent in 1973 to 16.8 per cent in 1980, despite a negative average growth rate in import volume over this period. Thus, while the relatively high growth rates of industrial countries’ import volumes from the non-oil developing countries are encouraging for the latter’s growth prospects, this volume growth will need to be accompanied by relatively favorable export price performance (i.e., terms of trade developments) if the non-oil developing countries are to increase their share of imports into industrial countries’ markets in the future.
Taking into account the trends shown in Table 8, the question arises as to what factors are important in shaping the share of the industrial countries’ imports that come from the non-oil developing countries. In this respect, the following three factors deserve attention: (i) the commodity composition of both industrial countries’ total imports and non-oil developing countries’ total exports; (ii) the relative price competitiveness of the latter’s exports and the price elasticity of the former’s import demand for their goods; and (iii) any particular tariff preferences or other trade measures granted to, or imposed against, non-oil developing countries.
Commodity Composition of Industrial Countries’ Imports and of Non-Oil Developing Countries’ Exports
Even when the overall demand for imports in industrial countries is growing slowly, demand for particular imports will be growing substantially faster than the average; for example, the import market in industrial countries for small cars was stronger after 1973 than before, despite the slower overall growth in their import demand in the latter period. If the commodity structure and ultimately the production of exports from non-oil developing countries permit rapid response to the faster growing categories of import demand in industrial countries, it is apparent that the adverse consequences of slower overall import growth can be reduced, if not eliminated.
Table 9 shows how total industrial country import demand over both the 1968–72 and 1973–80 periods varied by commodity group (i.e., manufactures, raw materials, agricultural goods, and fuels).36 The major conclusions emerging from Table 9 can be summarized as follows. First, the fastest growing component of industrial country import volume in both periods was manufactures. Specifically, when the volume of total imports into industrial countries was growing at an annual average rate of 9.9 per cent (1968–72), the growth rate for manufactured imports was 12.1 per cent. More important, during 1973–80, the average annual growth rate was 6.4 per cent for manufactured imports and 4.5 per cent for total imports. Each of the other three commodity groups displayed slower volume growth than total imports in both periods. The second conclusion is that, with the exception of fuels, the faster volume growth of manufactures was sufficient to overcome a poorer relative price performance so that in value terms manufactures also grew faster than the other import categories, that is, an average annual growth of 17 per cent versus 14.8 per cent for raw materials and 15.7 per cent for agricultural goods. The third and final conclusion is that, with manufactures making up more than one half (54 per cent in 1980) of the value of all industrial countries’ imports, and with no other commodity group accounting for more than 29 per cent (fuels) of the total imports, future changes in their imports are likely to be dominated by movements in manufactured goods. That is, the fastest growing import category (in real terms) is also the largest one.
|Value share of 1980 total imports (In per cent)||Average 1968–72||1973||1974||1975||1976||1977||1978||1979||1980||Average 1973–80|
Refers to 14 industrial countries (including the seven largest).
Refers to 14 industrial countries (including the seven largest).
Moving from the demand to the supply side of the industrial countries’ import market, it is necessary to identify the groups of countries that were best placed to take advantage of the relatively high growth rates of manufactured imports during 1968—80. Unfortunately, data on the commodity composition of merchandise exports are not available for country groupings used by the International Monetary Fund. However, such data do exist for World Bank country groups, and the composition of these groups is similar enough to the Fund’s classification to at least permit some broad conclusions to be drawn.
Table 10 shows the commodity composition of merchandise exports in 1960 and 1978 for industrial market economies, capital-surplus oil exporters, and for three groups of non-oil developing countries as classified by the World Bank—low-income countries, middle-income oil exporters, and middle-income oil importers. For the purposes of this paper, it is useful to view the figures for low-income countries as indicative of those for the Fund subgroup “low-income non-oil developing countries”; those for the middle-income oil exporters as indicative of those for the Fund’s “net oil exporters” category, and those for middle-income oil importers as a composite indicator for the Fund subgroups “major exporters of manufactures” and “other net oil importers.” Similarly, industrial market economies can be viewed as the Fund’s industrial country group and capital-surplus oil exporters as the Fund group “oil exporting countries.”
|Fuels, Minerals, and Metals||Other Primary Commodities||Manufactures|
|Industrial market economies||11||8||23||15||66||77|
|Capital-surplus oil exporters||96||98||4||–||–||2|
|excluding China and India||15||49||79||40||6||11|
|Middle-income oil exporters1||46||78||50||14||4||8|
|Middle-income oil importers1||16||11||67||37||17||52|
Considering the broadest country groups first, it is clear from Table 10 that the country group with the largest share of manufactures in total exports is the industrial countries themselves, with a 77 per cent share in 1978. At the other end of the spectrum are the capital-surplus oil exporters with only 2 per cent of total exports in manufactured goods. The non-oil developing countries as a group lie in the middle of this spectrum but the differences within this group are what is most interesting. Going from the low to high, the middle-income oil exporters had a share of manufactures in total exports of only 8 per cent in 1978—not a surprising statistic given the concentration of this group’s exports in fuels, minerals, and metals (78 per cent in 1978).37 The low-non-oil developing countries had by 1978 increased the share of manufactures in total exports to 30 per cent, but if China and India are excluded, the figure falls to only 11 per cent. Finally, there are the middle-income oil importers with a share of manufactured exports of 52 per cent in 1978. What these figures suggest is that any short-term expansion of industrial countries’ markets for manufactured imports is likely to be among the industrial countries themselves, and the benefits for the non-oil developing countries are likely to accrue mainly to the middle-income oil importing countries (particularly major exporters of manufactures).
Equally intriguing as the intergroup differences in the level of the share of manufactured exports are those in the changes in this share over time. As Table 10 illustrates, the most dramatic expansion in the export share of manufactures took place for middle-income oil importers whose share more than tripled from 17 per cent in 1960 to 52 per cent in 1978. Next in line were the low-income countries who increased their share from 18 per cent in 1960 to 30 per cent in 1978; if China and India are excluded from the low-income group, the increase is from 6 per cent to 11 per cent. Middle-income oil exporters also doubled their share of manufactures over this period from 4 per cent to 8 per cent. Finally, the industrial market economies saw their share of manufactures grow more modestly, from 66 per cent in 1960 to 77 per cent in 1978.
Given the relative dynamism of manufactures in world trade in general and in industrial countries’ imports in particular, it is tempting to ask whether those subgroups of non-oil developing countries with the highest and fastest growing shares of manufactures in total exports also witnessed the most rapid growth in total export volumes. Again, the available data are not precisely suited for investigating this issue and there are obviously other factors (e.g., internal economic policies in the exporting countries) that would also affect export performance. Nevertheless, it is relevant to note that the single subgroup within the broad category of non-oil developing countries with the highest average annual percentage increase in the volume of total exports over both the 1968–72 period (12.4 per cent)38 and the 1973–80 period (10.0 per cent) was the major exporters of manufactures. Similarly, the subgroup with the lowest average volume growth in total exports over the 1973–80 period (3.9 per cent) was low-income countries. Hence, it would seem that, ceteris paribus, those non-oil developing countries that were able to diversify their exports most rapidly away from primary commodities (other than oil) and toward manufactures were best able to counteract the slowdown in overall industrial country import growth in 1973–80.
Price Competitiveness of Non-Oil Developing Countries’ Exports
Even if non-oil developing countries were to continue to diversify their structure of production in order to increase the share of high-income elasticity products (i.e., manufactures) in their total exports, their success in actually penetrating markets in the industrial countries would still depend in large part upon the price competitiveness of their exports vis-à-vis both domestic producers of similar goods in the industrial countries and suppliers of these goods from other exporting countries.
Perhaps because the share of primary commodities in total exports is larger for the non-oil developing countries, the conclusion is sometimes drawn that export price competitiveness is of only minor importance to these countries. Two bodies of empirical evidence contradict such a conclusion. The first is the empirical evidence on the determinants of the demand for exports of individual non-oil developing countries. In a study using annual data for 1951–69, Khan (1974) found that the quantity of exports for each of 15 individual non-oil developing countries could be well explained by the level of industrial countries’ real income and by the ratio of the developing country’s export price to an average of export prices of industrial countries. Further, the relative export price elasticity of demand was statistically significant for 9 of the 15 non-oil developing countries in the sample. The average value of the price elasticity over the period of one year was close to unity (0.94).
A second and more recent study of import competition in the U.S. market by Grossman (1981) provides the best evidence to date about both how non-oil developing countries’ export price performance affects industrial countries’ imports from them and on the sources of competition for non-oil developing countries’ exports to industrial markets. Grossman estimates quarterly import demand equations for 11 representative manufactured commodities that entered the U.S. market during 1968–78 (e.g., leather products, tires and tubes, iron and steel wire rods, television receivers, and still cameras). An interesting feature of Grossman’s study is that separate equations are estimated for imports from industrial countries and those from non-oil developing countries, thereby permitting one to determine the substitutability between 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 (over one year) of 1.7 and with estimates for individual commodity groups ranging from 0.52 (chinaware and earthenware) to 3.38 (television receivers) and 3.69 (leather goods).
An implication of Grossman’s (1981) results is that relative price performance was an important element in the ability of the non-oil developing countries to increase their share of manufactured imports into the U.S. market from 13.1 per cent in 1968 to 24.7 per cent in 1978.39 The other major conclusion of the study is that while imports from both industrial countries and non-oil developing countries are close substitutes for domestically produced goods, they are imperfect substitutes for one another, at least in the U.S. market. Grossman’s explanation for this finding is that whereas U.S. industries produce a wide spectrum of goods within each commodity category, non-oil developing countries and other industrial countries export smaller and minimally overlapping subsets of the goods consumed in the United States. For example, U.S. imports of tires and tubes from other industrial countries consist largely of tires for trucks, buses, and cars, and the larger pneumatic tires for other uses. In contrast, U.S. imports from non-oil developing countries are dominated by bicycle tires and tubes, motorcycle tires, and the smaller pneumatic tires. The domestic U.S. tire industry (particularly the larger firms) produces a wide variety of tires, thus facing competition from both import sources.
Tariff Preferences for Non-Oil Developing Countries40
The relative competitive position of non-oil developing countries in industrial countries’ markets depends not only on their export prices and exchange rates vis-à-vis other export suppliers and domestic producers but also on the tariffs (and other trade measures) imposed against their exports in the industrial countries. To the extent that non-oil developing countries receive tariff preferences that are not accorded to other suppliers, their relative competitive position should improve, which in turn should increase their market share and thereby partially offset any slowdown in overall import growth in the industrial countries.
The rationale for granting tariff preferences to developing countries has been that these countries face obstacles not shared by industrial countries, namely, small local markets that limit economies of scale, and tariffs in industrial countries, which while not high on an average basis, are high on those products that are most important in developing country exports.
In an effort to meet these concerns, industrial countries introduced, in stages from 1971, tariff cuts on developing country exports that have come to be known as the Generalized System of Preferences (GSP). The eligibility requirements, product coverage, and special provisions of the GSP differ from one industrial country to another but World Bank estimates indicate that in 1976, the GSP covered about 18 per cent of all industrial countries’ nonfuel merchandise imports from developing countries (see Table 11).
|Nonfuel Merchandise Imports from Developing Countries|
|Country or Trading Group||Total value||Value||Percentage of total|
Although the aims of the GSP are clear enough, its actual effects on developing countries’ exports have been subject to question. To begin with, there are those (e.g., Murray (1977)) who argue that non-oil developing countries may not really need preferential access to industrial country markets because they possess a comparative advantage in many manufactured products. Stein (1981), for example, cites the increased market shares of developing countries in total manufactured imports into industrial countries and in total world trade in manufactures before the implementation of the GSP in 1971 as evidence in support of this argument.41 Similarly, the fact that developing countries’ exports to industrial countries have typically increased fastest in those labor-intensive manufactures subject to trade restrictions (e.g., cotton textiles) has been interpreted as indicative of a significant comparative advantage for those countries in the production of these goods.
Whatever the need, or lack of it, of non-oil developing countries for tariff preferences, there is as yet little agreement on the effect of such preferences on their exports. On the one hand, some studies (e.g., Iqbal (1976)) argue that preferences can be an important stimulative factor in export growth and that actual export growth of the non-oil developing countries, within some GSP plans, has been faster for eligible than for noneligible items. On the other hand, there are those studies (e.g., Baldwin and Murray (1977) and Cline et al. (1978)) that purport to demonstrate that developing countries would have gained considerably more in additional exports from substantial (50–60 per cent) most-favored-nation tariff cuts than from GSPs. At the same time, these latter studies also imply that most-favored-nation tariff cuts (e.g., Tokyo Round) will gradually diminish the competitive advantage that the non-oil developing countries obtain from GSPs. Finally, there is evidence that the export gains from GSPs have been concentrated in a small number of non-oil developing countries.42
Non-Oil Developing Countries’ Exports to Industrial Countries and Their Total Exports
Thus far, we have focused our discussion exclusively on how a slowdown in real income growth in industrial countries will ultimately affect their imports from non-oil developing countries. Industrial countries, however, are not the only market for non-oil developing countries’ exports and a rational response to a slowdown in industrial countries would be for the non-oil developing countries to try to shift some of their exports to faster-growing and faster-importing regions (including perhaps among themselves). To the extent that non-oil developing countries are able to bring about this shift, they will weaken the link between the growth of their exports to industrial countries and the growth of their total exports and hence, in turn, that link between industrial countries’ real income growth and non-oil developing countries’ real income growth.
The importance to non-oil developing countries of being able to diversify the geographic distribution of their exports is forcefully illustrated by three statistics. First, these countries exported over twice as much to industrial countries as to any other country group in 1973–80. As shown in Table 12, exports to industrial countries comprised 64 per cent of non-oil developing countries’ total exports in 1973–80, versus 6.3 per cent for exports to oil exporting countries and 24 per cent for exports to other non-oil developing countries. Second, and more striking, the average growth of total import volume in industrial countries over the same period (4.2 per cent) was less than one fourth of that for oil exporting countries (18 per cent) and only about two thirds of that (6.5 per cent) for non-oil developing countries. In short, non-oil developing countries were exporting the bulk of their goods to the slowest growing import market in 1973–80. Third, the average growth of the volume of imports from non-oil developing countries showed even greater disparity across these three country groups during 1973–80 than did the figures for total import volume growth. Specifically, as shown in Table 12, the relevant volume figures were 5.6 per cent for industrial countries, 17.6 per cent for oil exporting countries, and 11.8 per cent for the non-oil developing countries themselves. Thus, the volume of imports from the non-oil developing countries was growing two to three times as fast in the regions where about one third of their exports were going as in the regions where two thirds of these exports were going. It is also worth observing from Table 12 that a faster growth in the volume of imports from the non-oil developing countries than for total imports did not guarantee that the non-oil developing countries’ share in the value of total imports would increase. Only in their own import market were the non-oil developing countries able to increase their market share during 1973—80. This lack of close correspondence between volume and value changes constitutes another reminder of the importance of relative export price developments in determining changes in import market shares. On the more optimistic side, it is important to note that the non-oil developing countries did recognize and respond to the import growth disparity among their different export markets. Whereas exports to industrial countries accounted for 64 percent of non-oil developing countries’ total exports in 1974, this percentage had dropped to 62 per cent by 1980. By the same token, exports to the fastest growing import market (namely, the oil exporting countries) rose from 4.9 per cent of non-oil developing countries’ total exports in 1974 to 6.7 per cent in 1980. Also, moving in the preferred direction, exports to other non-oil developing countries increased from 23 per cent in 1974 to 26 per cent of total exports in 1980 (Table 12).
|Destination of non-oil developing country exports (in per cent of total exports)|
|Oil exporting countries||…||4.9||6.3||6.0||7.0||6.8||6.3||6.7||6.31|
|Non-oil developing countries||…||23.0||23.5||22.4||22.7||23.1||24.6||25.6||23.61|
|Volume of total imports (average percentage change)|
|Oil exporting countries||20.3||38.5||41.4||20.4||14.6||5.2||–11.6||15.4||18.0|
|Non-oil developing countries||13.9||7.5||–4.s0||2.4||6.7||7.8||11.9||5.8||6.5|
|Volume of imports from non-oil developing countries (average percentage change)|
|Oil exporting countries||28.9||20.2||34.0||19.8||16.6||1.6||2.1||17.6||17.6|
|Non-oil developing countries||27.5||12.8||2.1||6.9||5.3||12.4||15.2||16.2||11.8|
|Imports from non-oil developing countries as a share of the value of total imports (in per cent)|
|Oil exporting countries||16.4||17.4||14.5||14.9||14.7||13.7||15.1||15.7||15.3|
|Non-oil developing countries||15.7||17.0||15.4||16.6||17.5||17.5||18.9||…||16.92|
If import growth rates in oil exporting countries and in the non-oil developing countries themselves were to continue to exceed those in industrial countries by a wide margin, there would, of course, be a case for further diverting exports from industrial countries to these groups. Because, however, there are large differences in the commodity composition of both imports and exports across different country groups, and because some non-oil developing countries are better placed in terms of location to take advantage of rapid growth in certain import markets, it is to be expected that some non-oil developing countries will be better able than others to make such a switch in export destination. In the case of the oil exporting countries, the structure of imports is now more heavily weighted toward manufactures than that of any other country group. In 1978, manufactures accounted for 85 per cent of total imports for the four capital-surplus oil exporters and 73 per cent for the oil exporting middle-income countries versus 58 per cent for industrial market economies, 52 per cent for low-income countries, and 61 per cent for oil importing middle-income developing countries.43 Furthermore, imports of oil exporting countries from non-oil developing countries appear to have grown fastest during 1973–77 for manufactured goods (particularly capital goods).44 These two facts clearly gave the major exporters of manufactures an advantage not shared by other non-oil developing countries, and this was reflected in the unusually rapid expansion of exports to the oil exporting countries achieved by countries such as Greece, India, and Korea over the 1973–77 period. By geographic region, Asian non-oil developing countries were able to increase their share of exports going to capital-surplus oil exporting countries from 2 per cent in 1973 to 7 per cent in 1977; in contrast, this share remained constant for African non-oil developing countries and expanded from only 1 to 2 per cent for non-oil developing countries in Latin America.
Turning to trade among the non-oil developing countries, Havrylyshyn and Wolf (1981) have identified a number of developments that bear repeating. First, in line with the well-known (Heckscher-Ohlin) factor endowments theory of trade flows, exports from some non-oil developing countries to other non-oil developing countries are more capital-intensive than their exports to industrial countries. Table 13 provides some information on the commodity composition of both exports and imports for a sample of 33 non-oil developing countries in 1977. Note that the weight of capita! goods in intra-non-oil developing country exports is more than twice as high as for exports from the non-oil developing countries to industrial countries. Similarly, in line with Hecks-scher-Ohlin expectations, the weight of capital goods is more than twice as high for non-oil developing countries’ imports from industrial countries as for non-oil developing countries’ exports as a whole. To some observers (e.g., Krueger (1978)), this difference in factor intensities implies that intra-non-oil developing country exports are less beneficial than their exports to industrial countries because the former are less employment-generating than the latter. To others (e.g., Hughes (1980)), trade among the non-oil developing countries is still seen as conferring great benefits to them because of the opportunities it provides for economies of scale and learning by doing, and for the competitive pressures which it generates.
|Non-Oil Developing Countries’ Exports||Non-Oil Developing Countries’ Imports From Industrial Countries|
|Commodity Category||To industrial countries||To other non-oil developing countries|
|Food and beverages||35.1||32.3||10.1|
|Metals and minerals||7.8||9.9||3.0|
A second development is that, although exports of manufactures absorbed an increasing share of total nonfuel trade among developing countries in 1963–77 (from 41 per cent in 1963 to 54 per cent in 1977), this rise in the share of manufactures was much smaller than in developing countries’ trade with industrial countries (where it rose from 17 to 50 per cent over the same period). As a consequence of this latter development, the share of developing countries in developing-country trade in manufactures fell from 40 per cent in 1963 to 25 per cent in 1977; see Table 14, Operating in the opposite direction, developing country markets became more important as an outlet for nonfuel primary commodities where their share of total trade rose from 16 per cent in 1963 to 23 per cent in 1977 (Table 14). Much of this expanded primary commodity trade among the non-oil developing countries occurred in food and nonfood agricultural commodities whereas trade in metals and minerals was relatively constant. In short, despite the relative dynamism of world trade in manufactures during the 1960s and 1970s, primary commodity trade among the non-oil developing countries would seem to be a promising avenue in the face of sluggish growth in industrial countries’ imports.45 In this respect, it is worth mentioning that the subgroup of non-oil developing countries known in the World Bank as the “newly industrialized countries”46 and in the Fund as major exporters of manufactures47 has constituted a particularly important market for non-oil developing countries’ exports of primary commodities, accounting for roughly one half of all intra-non-oil developing country exports of these goods in 1977.
|Food and beverages||16||19||18||19||21||23|
|Metals and minerals||11||9||10||11||11||12|
|Total nonfuel primary||16||19||18||20||20||23|
Based on sample of 33 non-oil developing countries.
Based on sample of 33 non-oil developing countries.
A third notable feature of intra-non-oil developing country trade is that a relatively small number of countries, most of which are members of the subgroup major exporters of manufactures account for a large proportion of intra-non-oil developing country exports. As shown in Table 15, ten exporters accounted for over 83 per cent of total nonfuel exports and for more than 93 per cent of intra-non-oil developing country exports of manufactures in 1977. The only non-oil developing countries in the “top ten list” who were not in the major exporters of manufactures subgroup were Spain, Malaysia, and India. The principal products involved in intranon-oil developing country trade are coffee, crude rubber, motor vehicles, rice, and cotton.48
|Total Nonfuel Exports||Manufactures||Capital Goods—Narrow Definition||Food and Beverages|
A fourth development which has been quite important for trade among the non-oil developing countries during the past two decades, and which looms even more important for future trade among these countries, is their apparent lack of trade liberalization. Indeed, Hughes (1980) cites the high level of protection in developing countries as the “principal constraint” to more trade among non-oil developing countries. Also, protection is not only high in non-oil developing countries but is concentrated in products that other developing countries typically export. To a minor extent, this trade-restricting effect of protection has been dampened by the formation of customs unions and free trade areas among the non-oil developing countries, such as the Latin American Free Trade Area, but doubts exist about the viability and benefits (i.e., trade creation versus trade diversion) of such arrangements. In the end, it would seem that attempts to minimize the export consequences of an industrial country slowdown by resorting to greater trade within the non-oil developing country group will require some liberalization of trade barriers in the non-oil developing countries themselves.
Industrial Country Real Income and Non-Oil Developing Countries’ Exports: The Reduced-Form Relationship
The preceding three subsections reviewed the factors and structural relationships that determine the growth of non-oil developing countries’ exports vis-à-vis a given growth of industrial country real income. However, the evidence on the reduced-form relationship itself between non-oil developing country export volume and industrial country real GNP has not yet been examined. Given the focus of this paper, it will be useful to know (i) if there is a significant relationship between the growth of industrial country real income and the growth of non-oil developing countries’ total exports; (ii) if this relation-ship has become stronger or weaker over time; and (iii) if this relationship differs within the subgroups of non-oil developing countries.
Table 16 seeks to provide answers to these three questions by presenting estimates of the elasticity of the volume of exports of non-oil developing countries with respect to real GNP of industrial countries.
|Industrial||Time Period||Level of Aggregation||Industrial Countries’ Economic Activity Variable||Estimated Elasticity|
|Rhomberg (1968)||1953–65||Total exports, developing countries||Industrial production||0.371|
|Hicks et al. (1976)||1965–72||Non-oil developing country exports of manufactures||Real GDP||1.7 (Japan/Oceania)|
|2.0 (Western Europe)|
|4.1 (North America)|
|Deppler and Ripley (1978)||1968–76||Total exports, non-oil developing countries||Manufacturing real output||0.501|
|Goldstein and Khan2||Total exports, all non-oil developing countries||Real GNP||1.331|
|Total exports, major exporters of manufactures||Real GNP||1.791|
|Total exports, net oil exporters||Real GNP||0.76|
|Total exports, other net oil importers||Real GNP||0.08|
|Total exports, low-income countries||Real GNP||0.05|
Indicates statistical significance at the 5 per cent level.
Variables defined in terms of percentage changes.
Indicates statistical significance at the 5 per cent level.
Variables defined in terms of percentage changes.
First, there is a significant positive association between changes in industrial country real GNP and changes in the volume of non-oil developing countries’ exports.49 Second, there are indications that the size of the elasticity itself has been increasing over the past 20 to 30 years. Thus, whereas the authors’ estimated elasticity for the 1973–80 period (for total non-oil developing countries) was 2.30, the corresponding (insignificant) elasticity for the 1963–72 period was negative. Similarly, the authors’ estimated elasticity for 1963–80 (1.33) is substantially larger than Rhomberg’s (1968) estimate (0.37) for 1953–65 using a virtually identical estimating equation. One explanation for this finding is that the demand elasticity of non-oil developing country export volume is considerably higher for manufactures than for primary commodities, and that non-oil developing countries have been steadily increasing the share of manufactures in their total exports over the past two decades. In this respect, it is interesting to note Lewis’s (1980) observation that the “implied” export volume elasticity for developing country exports of primary commodities was roughly 0.8 for both the 1873–1913 and 1953–73 periods. As seen in Table 16, the export volume elasticity for non-oil developing country manufactured exports appears to be substantially higher—on the order of 1.7 to 4.1 according to Hicks et al. (1976). The third conclusion, and one that emerges from the authors’ own estimates, is that the sensitivity of export volume to real economic activity in industrial countries is highest for middle-income net oil importing developing countries—that is, the major exporters of manufactures and other net oil importers. These subgroups are the ones with the highest estimated elasticities over the 1973–80 period (2.63 and 2.30, respectively), and they are also the only two subgroups for which the estimated elasticity itself is statistically significant. One potential explanation for this is that, since export volumes in the other two subgroups (i.e., low-income countries and net oil exporters) are more heavily concentrated in primary commodities, they are more constrained or affected by “supply” shifts.50 In this case, shifts in export demand induced by changes in industrial country real economic activity would be a less powerful explanatory factor for observed changes in export volumes. Finally, it should be emphasized that the models generating the results shown in Table 16 are simple reduced-form relationships. As such, the above conclusions should be viewed as tentative rather than definitive.