IV Summary and Conclusions
- Mohsin Khan, and Morris Goldstein
- Published Date:
- August 1982
This paper has examined the relationship between the rate of economic growth in the non-oil developing countries and that in industrial countries, with the intention of appraising the effects of slower industrial country growth on the non-oil developing countries during 1973—80. At the risk of oversimplifying the arguments and the evidence examined in the preceding sections, the principal conclusions emerging from our analysis can be summarized as follows:
(1) There is a difference between saying that the rate of economic growth in industrial countries has an important and positive effect on the rate of economic growth in non-oil developing countries and saying that the former’s growth rate is the overriding determinant of the latter’s growth rate. The first proposition is true; the second is not.
There is evidence that slower (faster) industrial country growth rates are associated, ceteris paribus, with slower (faster) non-oil developing country growth rates, particularly via the effect of real income growth in industrial countries on the growth of non-oil developing countries’ exports. But there are also clear indications that a host of other factors strongly affect non-oil developing country growth as well, including, inter alia, the commodity composition and relative competitive position of their exports, the existing tariff and nontariff barriers to trade with industrial countries, the share of non-oil developing countries’ exports going to industrial countries, the scope of demand and production linkages between the export and domestic sectors in non-oil developing countries, the availability and cost of external finance to them, the flow of migrants’ remittances to the non-oil developing countries, the cost of their imported inputs (particularly oil), the quantity and quality of human and physical capital, and the stance and general orientation of their economic policies.
In sum, the rate of economic growth in industrial countries sets broad limits on the range of feasible growth rates in non-oil developing countries but the actual growth rates achieved by individual non-oil developing countries, or groups of them, would seem to be the outcome of other factors, including their economic policies and those of other groups of countries (e.g., the recycling of the surpluses of oil exporting countries).
(2) During periods when other growth-inducing factors act to offset, or to partially compensate for, the effect of slower industrial country growth, the observed link between industrial country growth rates and non-oil developing country growth rates will be weak, even though the independent effect of the former’s growth on the growth of the latter may be quite strong. A good illustration of this is the rather modest decline in the average non-oil developing country growth rate (of real GNP) from 5.8 per cent per annum in 1968–72 to 5.2 per cent in 1973–80, in the face of a much sharper decline (from 4.5 per cent per annum in 1968–72 to 3.1 per cent in 1973–80) in average industrial country growth. This apparent resiliency of non-oil developing country growth was attributable not to the absence of a significant adverse industrial country income effect but rather to the compensations that were made for it by the non-oil developing countries, oil exporting countries, multilateral financial institutions, and by industrial countries themselves. These compensations took the form, among others, of an increased (volume) share of non-oil developing countries in total imports into industrial countries; a rerouting of some non-oil developing country exports to faster growing import markets; a continuous increase in the share of high-income elasticity products (i.e., manufactures) in their total exports; tariff preferences for developing countries in industrial country markets (e.g., the GSP); an increased availability of external finance to non-oil developing countries at attractive real interest rates and an increased flow of migrants’ remittances—both of which enabled these countries to support import growth and investment rates at higher levels than would otherwise have been possible; and a reorientation, in some non-oil developing countries, of economic policies that contributed to higher agricultural production, higher export supply, and higher overall economic efficiency. By the same token, prospects for non-oil developing country growth in the medium and long term depend not only on the likely pace of growth in industrial countries but also on the scope, strength, and durability of the compensative actions that can be taken if industrial country growth turns out to be slower than desirable. For this reason, projection and scenario exercises for non-oil developing country growth need to pay particular attention to these potential compensative factors. Failure to do so could clearly result in overestimating the impact of slower industrial country growth on non-oil developing countries.
(3) Statements about the impact of slower industrial country growth on non-oil developing countries can be misleading because of the striking differences of this impact across the different subgroups of non-oil developing countries. Specifically, on the basis of some simple reduced-form tests, it was found that the rate of industrial country growth had the greatest impact on domestic growth rates in middle-income oil importing countries, that is, for major exporters of manufactures and for other net oil importers. In the case of the other two analytical subgroups, namely net oil exporters and low-income countries, this relationship was much weaker.87 This lower sensitivity to industrial country real income growth by the latter two subgroups of non-oil developing countries cannot be traced to a single factor but seems to be explainable in terms of (a) the low share of exports in GNP (at least for low-income countries); (b) the importance of terms-of-trade changes relative to export volume changes over the 1973–80 period; (c) the higher proportion of primary commodities in total exports (with the attendant greater frequency of supply shocks relative to industrial country income-induced demand shocks); (d) the higher share of agriculture in GDP (in low-income countries) with the associated vulnerability of agricultural production to local or regional exogenous events (e.g., droughts); (e) the lack of complementary factors (e.g., infrastructure, financial markets) to translate an increase in export earnings into higher growth; and (f) the significant share of foreign exchange receipts that originated outside the industrial countries over the 1973–80 period (e.g., migrants’ remittances from oil exporting countries).
This greater sensitivity of economic growth in middle-income oil-importing developing countries to industrial country growth implies, of course, that they will be most vulnerable to a slowdown in growth in industrial countries. Consistent with this proposition, it was noted that major exporters of manufactures and other net oil importers displayed the largest declines in real GNP growth rates between 1968–72 and 1973–80 (i.e., 2.3 percentage points and 1.0 percentage point per annum, respectively). In contrast, the average annual growth rates for low-income countries and for net oil exporters were both marginally higher in the later slow industrial country growth period. But it is equally important to note that the middle-income oil importing developing countries also displayed relatively high levels of economic growth in both periods, and there are good reasons for believing that this superior growth performance was in part attributable to their more outward-looking economic structure and economic policies. Even though their vulnerability to sharp changes in industrial country growth rates is greater than in more inward-looking, low-income non-oil developing countries, the former’s vulnerability to other types of external shocks is probably lower. For example, as noted in several World Bank studies, when imports already compete with a wide range of domestically produced goods, import growth can be reduced in the wake of a payments crisis without eliminating many crucial imports; as such, the repercussions on growth are likely to be smaller.
In short, a relatively high degree of integration with the world economy carries both benefits and costs. Relatively high sensitivity of domestic growth with respect to industrial country growth is beneficial when industrial countries grow rapidly and costly when they grow slowly. However, the gains from trade extend far beyond the direct income effect associated with changes in the rate of export growth, and these other gains are not likely to be forthcoming unless a country maintains a consistent outward-looking policy. Turning inward as a response to projected slow industrial country growth rates might reduce the short-run fluctuation in non-oil developing country growth rates but would also likely reduce the medium-term level of their growth.
(4) Examination of the sublinks that lie beneath the reduced-form relationship between industrial country growth rates and non-oil developing country growth rates suggests that the link between real income growth in the former and export growth of the latter is considerably stronger and quicker than that between the non-oil developing countries’ export growth and real income growth. For all non-oil developing countries, a consensus estimate of the former elasticity might be on the order of 1.3 whereas the latter is closer to 0.1. In other words, industrial country real income growth is considerably more important relative to other factors in explaining non-oil developing countries’ exports than it is in explaining their real income growth.
This analysis of the association between real income growth in the industrial countries and export volume growth of the non-oil developing countries showed that a slowdown in the former’s economic growth typically produced a rapid and larger than proportionate slowdown in their total import growth as well. During the 1973–80 period, however, there were two factors that helped to minimize the consequences on non-oil developing countries’ exports of this slower growth of industrial countries’ imports, albeit primarily for those non-oil developing countries with higher shares of manufactures in their total exports. One was that the (implied) income elasticity for industrial countries’ imports from the non-oil developing countries was larger than that for their total imports, so that non-oil developing countries captured a larger part of the industrial countries’ slow-growing import volume. Since econometric studies reveal that the demand in industrial countries for manufactured imports from non-oil developing countries is quite price elastic, an implication follows that favorable relative price performance by the non-oil developing countries aided in this increased market share. The contribution of tariff preferences to the non-oil developing countries is more difficult to evaluate; they probably spurred exports of non-oil developing countries but less so than would have a larger reduction in average industrial country tariffs on a most-favored-nation basis. A second factor was that non-oil developing countries were able to increase their total exports (in volume terms) faster than their exports to industrial countries over the 1973–80 period by sending a larger share of their exports to oil exporting countries and to other non-oil developing countries—both of which increased their total imports and their imports from non-oil developing countries faster than did the industrial countries. Behind the long-term increase in the share of manufactures in total non-oil developing countries’ exports was the increasing importance of trade among the non-oil developing countries themselves as an outlet for their exports of primary commodities.
Turning to the association between export growth and real GNP growth in non-oil developing countries, the authors, like earlier investigators, were able to identify only a weak positive relationship. We also found that this relationship or elasticity was: (i) small relative to those for growth of the labor force and growth of capital; (ii) significant only for major exporters of manufactures and other net oil importers;88 and (iii) stronger over longer time periods than for year-to-year changes or for three-year moving averages of the year-to-year changes. Put in other words, export growth is a positive source of economic growth in non-oil developing countries, but it is not as important as other traditional and primarily domestic sources of growth; it seems to make the strongest contribution to growth once countries reach some minimum stage of development; and it is sustained export growth rather than short-run fluctuations that exerts the greatest influence on the growth process.
In examining the growth experience of non-oil developing countries during 1973—80, four factors were identified that helped to protect real GNP growth in the face of a harsh external environment characterized by low industrial country growth rates, high global inflation rates, and large oil price increases.
(a) First, there was a significant increase in the flow of migrants’ remittances into non-oil developing countries, particularly into those in the low-income subgroup. By the late 1970s, these remittances amounted to 20–50 per cent of merchandise exports in quite a few of the labor exporting countries. We also found that while migrants’ remittances were quite sensitive to the level of and change in real economic activity in the host country, a slowdown in industrial country growth affects only a limited number of the non-oil developing countries receiving the remittances because many of them received their remittances from oil exporting countries or other non-oil developing countries.
(b) A second mitigating factor was the increased availability of external finance to non-oil developing countries, and at real interest rates that were negative (relative to the increase in their export prices) until 1981. For the low-income subgroup, the bulk of this finance came from official sources while for the middle-income non-oil developing countries it derived from the private sector. This external financing, in turn, permitted higher growth rates of imports and of investment in non-oil developing countries than would otherwise have been possible.
(c) A third factor weakening the link between non-oil developing country growth and industrial country growth was the substantial share of agriculture in GDP, especially for the low-income subgroup. Since agricultural production depends mainly on factors other than industrial country real growth, a substantial share of GDP is likewise only weakly related to real growth rates in industrial countries. In those non-oil developing countries where policies toward agriculture promoted favorable relative incentives and provided for significant investment in complementary inputs, agricultural production did well; the converse was true for those non-oil developing countries that neglected agriculture in their development policies. Still, partly as the result of exogenous, uncontrollable factors (e.g., droughts), agriculture showed slower growth in non-oil developing countries over 1970–80 than did either industry or services, and hence slowed overall growth in those countries during 1973–80.
(d) Fourth, it was observed that the growth performance of non-oil developing countries was strongly affected by the orientation and quality of their own economic policies. Specifically, outward-looking development strategies (as characterized by, inter alia, provision of similar incentives for domestic and export production, realistic real exchange rates and public utility prices, tolerance of wide-ranging import competition for domestically produced goods, and the operation of an automatic and stable incentive system) produced better results for economic growth and for adjustment to external shocks than did inward-looking strategies.
(5) Finally, the findings about the changing strength of the growth linkage between the industrial countries and the non-oil developing countries as one moves from the 1960s to the 1970s should be considered preliminary in view of the limited number of subperiod observations. Nevertheless, there were persistent indications that industrial country growth rates had a stronger impact on non-oil developing country growth rates in 1973–80 than in 1965–72, and this for both the underlying link between industrial country real income and non-oil developing country exports and that between non-oil developing country exports and their real income. It appears that this greater sensitivity to industrial country growth in the later period is traceable to changes in the structure of production and exports which have been gradually taking place in non-oil developing countries over at least the last two decades. Specifically, the large long-term increase in the share of manufactures in their exports is probably at the heart of this heightened sensitivity—both because of the higher income elasticity for manufactured exports and because of the larger growth spinoffs that are thought to derive from exports of manufactures. Similarly, the long-term fall in the share of agriculture in real GNP in non-oil developing countries has reduced the influence of the most indigenous growth sector, and this has not been offset by any appreciable rise in the share of the other (relatively) indigenous or nontradable sector.