World Bank Reprint Series: Number 383 Pradeep K. Mitra A Description of Adjustnent to External Shocks: Counry Groups Originally published in Stagflation, Savitgs, and the State: Perspectives on the Global Economy, edited by Deepak Lal and Martin Wolf. Chapter 4, pp. 103-114, February 1986. Published by Oxford Universityt Press. Chapter 4 A Description of Adjustment to External Shocks: Country Groups Pradeep K. Mitra THE 1970s will be remembered as a decade that witnessed serious convul- sions in the world economy. * Petroleum prices quadrupled in 1973-74, fell by a sixth between 1974 and 1978, and then increased by 80 percent in real terms during 1979-80. * The industrial market economies went into a recession in 1974-75 and thereafter recovered strongly before plunging into another reces- sion in 1979-080, from which a slow recovery is under way. Stagflation was born in the countries of the Organisation for Economic Co- operation and Development (OECD), with successive peaks of eco- nomic activity occurring at ever higher levels of unemployment. * The end of the decade saw an interest rate shock following the use of restrictive monetary policies to combat inflation in the leading indus- trial countries. In addition, in the early 1980s, the real prices of major primary products exported by developing countries-adjusted for rising prices of imported manufactures-fell to their lowest levels since World War II. Some perspective on the magnitude of these convulsions may be obtained from the following statistics. World trade in fuels increased from $29 billion in 1970 to $535 billion in 1980.' Paying for the 1970s' fuel price increases was equivalent to finding the money to buy all the exports of another United States or Federal Republic of Germany. The current account deficits of the oil-importing developing countries as a proportion of GNP doubled from about 2.5 percent in 1973 to 5 percent in 1980. Debt-servicing payments of all developing countries, deflated by their export unit values, rose nearly threefold between 1972 and 1979; interest The work reported in this paper has been undertaken in the context of World Bank research project no. 672-74, "Adjustment in Oil-Importing Countries." The author is deeply indebted to Hector Sierra for exceptional research assistance. 103 104 ADJUSTMENT TO EXTERNAL SHOCKS rates, deflated by export prices, rose from -10 percent in '979 to 20 percent in 1981. This chLpter is a first report on work on developing economies' adjust- ment to external shocks in the 1970s and early 1980s. It articulates a framework to impose analytical order on descriptions of shock and adjustment; to construct comparators that can place individual perform- ance in perspective; and to locate empirical regularities among growth performance, external shocks, modes of adjustment, and, when data permit, policy variables. The paper also develops the analytical framework with a number of examples. The analytical framework is used in the third section to classify thirty-three developing economies into five groups, according to certain features of their adjustment, which are then reviewed. The review pro- vides a convenient backdrop against which individual country adjustment may be viewed, a task to which two companion papers (Mitra 1985a and 1985b) are devoted. The conclusion, in a fourth section, is followed by appendix A, which lists data sources. Appendix B classifies thirty-three economies into five broad groups. Methodology The following methodology underlies the comparative analysis. (A more formal account of the model and decomposition method appears in Mitra 1984, 1985a, and 1985b.) An open economy macroeconomic model is estimated fcr each country over the 1963-81 period, with an assumed structural break after 1973. The output of the model during the 1974-81 period is then compared with the output for the same period as if the 1963-73 parameters had prevailed and under certain assumptions about the course of variables exogenous to the model. I refer to the hypothetical development as the "counterfactual." The changes in the principal macroeconomic aggregates between the two scenarios are then decom- posed into price and quantity changes. External Shocks External shocks comprise (a) international price effects, (b) recession- induced effects, and (c) net interest rate effects. International price effects measure the balance-of-payments impact of changes in an econ- omy's terms of trade relative to the counterfactual and are the sum of the export price effect and the import price effect. The export price effect PRADEEP K. M1TRA 105 measures the net impact of a fall in the purchasing power of exports over manufactures exported by the OECD countries relative to the counterfac- tual. The import price effect measures the net impact of a rise in the purchasing cost of imports in terms of manufactures exported by the OECD countries relative to the counterfactual. Both export and import price effects may each be subdivided into two components. First, the relative increase (decrease) in import (export) prices exerts an unfavorable impact on the balance of payments-the direct effect. Second, the price effect impoverishes the economy and, with unchanged policies, restrains imports, thereby exerting a favorable impact on the balance of payments-the indirect effect. It can be shown that, if the economy's savings propensity is positive, the di-ect effect dominates the indirect effect, so that a relative increase (decrease) in import (export) prices always exerts a damaging effect on the balance of payments When measured against a 1971-73 base as a percentage of GNP, interna- tional price effects averaged on an annual basis over the 1974-81 period ranged from an extremely unfavorable 7.5 percent in Chile and 5.9 percent in Uruguay through a somewhat less unfavorable 3.5 percent in Malawi, 2.9 percent in the Philippines. and 2.6 percent in Taiwan to a moderately favorable 3.5 percent in Malaysia and 3.7 percent in Tunisia to an extremely favorable 9.8 percent in Nigeria and 14.2 percent in Indonesia. Although import price effects were unfavorable in all cases, the magnitude of export price effects was extremely unfavorable in Chile and Uruguay on the one hand and very favorable in Nigeria and Indone- sia on the other. Recession-induced effects on the balance of payments are twofold. The export volume effect (a direct effect) is the shortfall in an economy's exports as a result of a slowdown in the rate of growth of GNP in principal trading partners. From this effect must be subtracted the import-saving effect (or indirect effect), that is, the restraint in the growth of imports, with unchanged policies, due to the slowdown in income growth induced by the export volume shortfall. Recession-induced effects were generally positive, ranging as a percentage of GNP from 0.1 percent in Spain and Uruguay through 1.4 percent in the Republic of Korea and 1.9 percent in Taiwan to 3.7 percent in Indonesia. Net interest rate effects are twofold. The payments effect measures the impact on the balance of payments of an increase in real interest rates (in terms of manufactures exported by OECD countries) payable on a coun- try's debt relative to the counterfactual. From this figure must be sub- tracted the receipts effect, that is, the impact on the balance of payments 106 ADJUSTMENT TO EXTERNAL SHOCKS of an increase in real interest rates (in terms of manufactures exported by OECD countries) earned by a country's interest-bearing assets relative to the counterfactual. Net interest rate effects, when measured vis-a-vis real interest rates prevailing in 1971-73, ranged as a percentage of GNP from -0.6 percent in Mali and -0.2 percent in Kenya to 2 percent in Korea and 2.5 percent in Bolivia. Payments effects were particularly important in Bolivia, Korea, and Singapore; the payments effect on short-term debt was important in Singapore and, to a lesser extent, in Portugal. Modes of Adjustment Economies unfavorably affected by external shocks had four basic ways (and combinations thereof) of responding to external shocks: (a) trade adjustment, (b) domestic resource mobilization, (c) investment slowdown, and (d) additional external financing. To avoid unnecessary repetition the reader is asked to remember that, as with shocks, all modes of adjustment are measured as deviations from the counterfactual. The examples provided below have been drawn from a list of the economies that suffered rather than benefited from external shocks during the 1974-81 period. Trade adjustment is the sum of export expansion and import substitu- tion. Export expansion is the increase in the responsiveness of exports to changes in GNP growth in principal trading partners. It has a twofold effect. The direct effect measures the favorable impact on the balance of payments of boosting exports. From this figure must be subtracted the indirect effect, that is, the boost in import growth due to the expansion in income growth induced by the direct effect. Of the thirty-three econo- mies to which the analysis underlying this chapter has been applied, those in which export expansion played a prominent role include Singapore, Korea, the Philippines, Chile, and Thailand as well as Taiwan. Import substitution is the reduction in the responsiveness of the econ- omy's import demand to income. The direct effect measures the bal- ance-of-payments impact of restraining imports. From this figure must be subtracted the indirect effect, that is, the boost in import growth due to the expansion in income growth induced by the direct effect. Examples of adjustment through significant import substitution are Brazil, Yugosla- via, and Malawi. Both export expansion and import substitution improve the trade balance and boost GNP growth. Domestic resource mobilization measures the import-restraining effect PRADEEP K. MITRA 107 of a slowdown in income growth induced by improved savings perform- ance as defined below. It may be broken down into its private and public components. "Private resource mobilization" is the reduction in the responsiveness of private consumption to income. This was important in Honduras, Morocco, Singapore, Yugoslavia, Jamaica, and Korea. "Pub- lic resource mobilization" .ias two parts, "public consumption restraint," or the reduction in the responsiveness of public consumption to income, and "tax intensification," or the increase in the responsiveness to income of indirect taxes less subsidies. This term therefore ignores any changes in the direct tax effort, an omission that may be justified on grounds of their relative unimportance in developing economies. El Salvador, Singapore, and Honduras favored this mode of adjustment. Investment slowdown measures the import-restraining effect of a slow- down in income growth brought about through a reduction in the ratio of investment to income relative to the period 1971-73.2 This was a domi- nant mode of adjustment in Jamaica, Singapore, Mali, and Kenya. Net additional external financing measures changes in gross additional external financing (defined as capital flows, reserves, and transfers and services net of interest payments, deflated by a price index of manufac- tures exported by OECD countries) less changes in net interest payments resulting from changes in real net debt relative to the counterfactual.3 This measure played an important role in a large number of countries, for example, Mexico, El Salvador, Honduras, Morocco, Mali, Portugal, Spain, Guatemala, Turkey, the Philippines, Uruguay, and Kenya. Patterns of Adjustment An analysis of the experience of thirty-four developing economies over the period 1974-81 reveals that twenty-five of them suffered adverse external shocks. Their responses to these shocks varied considerably, a feature that is worth bearing in mind in the following discussion. To impose a measure of analytical order on the richness and diversity of experience, however, it is convenient to divide the economies into five groups, according to the sign of external shocks and the degree of reliance on different modes of adjustment.4 Group 1 (Chile, Costa Rica, Philip- pines, Singapore, Korea, and Taiwan) adjusted principally through ex- port expansion and public resource mobilization. Group 2 (Argentina, Brazil, Guatemala, Honduras, India, Kenya, Malawi, Mali, Thailkind, Turkey, and Uruguay) relied on either export expansion or public re- source mobilization., whereas Group 3 (Jamaica, Portugal, and Yugosla- via) was characterized by import substitution and negative public resource 108 ADJUSTMENT TO EXTERNAL SHOCKS mobilization, Group 4 (El Salvador, Mexico, Morocco, and Spain) resorted to financing without domestic adjustment. Finally, Group 5 (Benin, Bolivia, Colombia, Indonesia, Ivory Coast, Malaysia, Niger, Nigeria, and Tunisia) experienced favorable external shocks. The (un- weighted) average shock adjustment fig -;-,s for the 1974-81 period are shown for the five groups in table 4-1. Export expansion and public resource mobilization. The average shock was highest for Group 1 at 3.98 percent of GNP. International price effects accounted for roughly 60 percent of total shocks, with the recession- induced and net interest rate effects contributing equally to the remain- der. All economies of the group resorted heavily to export expansion, which exceeded external shocks by more than one-third and to public resource mobilization, of which the principal component was tax inten- sification. Together, export expansion and public resource mobilization accounted for 154 percent of external shocks. Import substitution was significantly negative everywhere except in Costa Rica, especially during the later years of the period. Whereas Chile, the Philippines, and Taiwan relied on substantial additional external financing and stepped up their ratio of investment to GNP, Korea sustained an investment boom with comparatively limited recourse to additional external resources. In con- trast, Singapore adopted a somewhat contractionary package, with a cut in the share of investment and real repayment of borrowed funds; Costa Rica had a similar adjustment profile as well. The ratio of external financing to external shocks was higher in 1974-81 than irn 1974-78 but was nevertheless quite modest for this group in relation to the others. Export expansion or public resource mobilization. International price effects accounted for roughly 80 percent of external shocks for Group 2. This group occupies a position between Groups 1 and 3 in terms of adjustment characteristics. Three broad patterns of adjustment may be distinguished. First, Argentina, Guatemala, India, Mali, and Uruguay resorted to export expansion while exhibiting negative import substitu- tion and negative public resource mobilization, which was significantly worse in the years 1979-81 than in 1974-78. Second, and in quite a contrast, Honduras and Kenya adjusted through a combination of import substitution and public resource mobilization, with export expansion turning negative. Third, the remaining countries-Brazil, Malawi, Thai- land, and Turkey-relied on a combination of export expansion and import substitution, with negative public resource mobilization aggravat- ing the balance-of-payrnents impact of disturbances from the inter- national environment. For the group as a whole, negative public resource mobilization added 40 percent to external shocks. There was significant additional external financing, especially in countries such as Honduras, Table 4-1. Balance-of-Payments Effects of External Shocks and Modes of Adjustment, 1974-78 and 1974-81 Averages (percentage of local currency GNP) Group 1 Group 2 Group 3 Group 4 Group S Effect 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 External shocks 1. International price effects a. Export price effect i. Direct effect - 1.97 -2.87 -0.63 -0.45 -3.86 -3.24 -3.16 -2.31 -7.59 -9.26 ii. Indirect effect -2.38 - 3.05 - 0.37 -0.37 -2.87 -2.34 - 1.75 -1.06 -2.93 -3.57 Difference ( = i-ii) 0.41 0.18 - 0.27 - 0.08 -0.99 -0.89 - 1.41 - 1.25 -4.66 -5.69 b. Import price effect i. Direct effect 6.08 8.06 3.16 :..71 4.55 4.98 ), 03 1.28 2.03 2.41 ii. Indirect effect 4.81 5.80 1.44 1.72 3.20 3.47 1.03 0.53 0.90 1.24 Difference (=i-ii) 1.27 2.25 1.72 2.00 1.34 1.51 1.00 0.75 1.13 1.17 Sum (= la + lb) 1.68 2.43 1.45 1.91 0.35 0.61 -0.41 -0.50 -3.53 -4.52 2. Recession-induced effect a. Export volume effect 1.97 2.04 0.60 0.69 1.18 1.30 1.22 1.46 0.73 1.27 b. Import saving effect 1.27 1.28 0.30 0.39 0.84 0.91 0.65 0.81 0.08 0.33 Difference (= 2a - 2b) 0.70 0.76 0.30 0.30 0.34 0.39 0.57 0.66 0.65 0.93 3. Net interest rate effect a. Payments effect i. Medium and long term 0.11 0.68 -0.09 0.18 0.05 0.72 0.06 0.45 0.10 0.75 ii. Short term -0.01 0.87 -0.01 0.16 0.00 0.40 -0.03 0.22 -0.01 0.15 Sum (= i + ii) 0.10 1.54 -0.10 0.34 0.04 1.12 0.03 0.68 0.09 0.90 b. Receipts effect 0.01 0.76 -0.01 0.09 -0.10 -0.15 0.00 0.06 0.00 0.28 Difference (= 3a - 3b) 0.10 0.78 -0.09 0.25 0.14 1.27 0.04 0.62 0.09 0.63 Total shock (= 1+2+3) 2.48 3.98 1.66 2.47 0.83 2.27 0.20 0.77 -2.79 -2.96 (Table continues on the following page.) Table 4-1. (continued) Group 1 Group 2 Group 3 Group 4 Group 5 Effect 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 1974-78 1974-81 Modes of adjustment 1. Trade adjustment a. Export expansion i. Direct effect 12.79 17.05 0.75 1.66 -7.60 -7.31 0.63 -0.02 -0.02 0.25 ii. Import augmenting effect 9.09 11.60 0.18 0.55 -5.41 -5.23 0.32 -0.13 -0.91 -0.58 Difference (= -ii) 3.70 5.45 0.57 1.11 -2.19 -2.08 0.31 0.15 0.89 0.83 b. Import substitution i. Direct effect 0.97 -4.20 0.87 0.85 4.68 4.43 -3.32 -3.28 -3.88 -5.04 ii. Indirect effect 1.45 -2.59 0.36 0.38 3.38 3.13 -1.55 -1.28 -0.17 -0.36 Difference (=i-ii) -0.48 -1.61 0.50 0.48 1.31 1.30 -1.77 -2.00 -3.71 -4.68 Sum (= la+lb) 3.22 3.84 1.07 1.59 -0.88 -0.78 -1.46 -1.86 -2.82 -3.85 2. Resource mobilization a. Private 1.08 0.54 -0.61 -0.44 -1.53 -0.96 0.72 0.65 0.98 1.27 b. Public i. Publicconsump.restraint -0.09 0.19 -0.69 -0.88 -2.93 -4.04 -0.61 -0.87 0,25 0.16 ii. Tax intensification 0.49 0.49 -0.10 -0.12 0.28 0.39 -0.25 -0.24 -0.86 -1.14 Sum (=i + ii) 0.40 0.68 -0.79 -1.00 -2.65 -3.64 -0.86 -1.11 -0.61 -0.98 Sum (= 2a+ 2b) 1.48 1.22 -1.39 -1.44 -4.18 -4.61 -0.14 -0.46 -0.37 0.29 3. Investment slowdown -1.13 -1.91 -0.46 -0.69 2.48 2.78 -1.60 -0.84 -1.31 -1.74 4. Net additional ext. financing -1.09 0.83 2.45 3.01 3.41 4.88 3.39 3.93 0.97 2.34 Total(= 1+2+3+4) 2.48 3.98 1.66 2.47 0.83 2.27 0.20 0.77 -2.79 -2.96 Note: Definitions of Groups 1-5 are in Appendix B. Source: See text and Appendixes A. PRADEEP K. Mr1RA 111 Mali, Guatemala, Turkey, Kenya, and Thailand, with this mode of adjustment exceeding external shocks by more than 20 percent for the group as a whole. There was some increase in the share of investment in GNP in all countries except Mali, Kenya, and Malawi. Import substitution and negative public resource mobilization. Although the shocks experienced by Group 3 were less unfavorable than those affecting Groups 1 and 2, their composition was rather different. International price effects accounted for less than 30 percent of external shocks, whereas net interest rate effects exceeded 55 percent of shocks, largely because of their relative importance in Jamaica. The adverse balance-of-payments impact of negative public resource mobilization was more than one and one-half times as large as that of external shocks in this group, with the effect being extremely strong in Jamaica. Import substitu- tion played a dominant role in all of the countries in Group 3; export expansion was significantly negative. External financing was much more important than in Groups 1 and 2 but much less so in the later years of the period. The average, however, conceals marked intercountry differ- ences: although it played a prominent role in Portugal, it was much less important in Jamaica and was virtually negligible in Yugoslavia. Financing without adjustment. External shocks averaged 0.77 percent of GNP for Group 4. Recession-induced and net interest rate effects accounted for one and two-thirds times this figure, principally because of their overwhelming importance relative to external shocks in El Salvador and Mexico. Table 4-1 clearly indicates the virtual lack of domestic adjustment across the board. Export expansion was negative except in Mexico (because of petroleum) and especially in Morocco and El Salva- dor. A major import and investment boom was under way in Morocco and, in relation to external shocks, in El Salvador. Public resource mobilization was positive in El Salvador but was more than offset by worsening performance in the other countries, especially Morocco and Spain. Additional net external financing was extremely important in all countries and was more than five times as important as external shocks for the group as a whole. Favorably affected countries. The countries of Group 5 experienced favorable shocks usually because they had been exporters of petroleum or of other primary commodities, so that the boom in prices in the mid-1970s allowed them to benefit over the period as a whole. Inter- national price effects alone exceeded total shocks by more than one-half in absolute terms. Export price effects, as a proportion of shocks, were extremely favorable in the nonfuel primary producers (the Ivory Coast, Bolivia, Tunisia, and Malaysia), followed by petroleum exporters (Ni- geria and Indonesia), which were in turn succeeded by Colombia.' Im- 112 ADJUSTMENT TO EXTERNAL SHOCKS port price effects, though significant in the Ivory Coast, were distinctly less important. Differences in the relative price movements of primary commodities during the 1970s accounted for variations in the pattern and timing of adjustment among members of the group. On average, how- ever, adjustnment to favorable shocks took the form of an import boom that intensified in 1979-81 as compared with 1974-78, a stepping up of the share of investment in GNP, a slackening of public resource mobilization efforts, and substantial additional external financing at the end of the period under review. With respect to particular countries, there was an import boom in Bolivia, Colombia, Malaysia, and Tunisia and, to a somewhat lesser extent, in Indonesia and Nigeria. It was accompanied by an investment boom, which was particularly marked in the Ivory Coast and Benin. There was a slackening of public resource mobilization efforts in the Ivory Coast and less of one in Malaysia and Tunisia. Net real additional financing was important in Tunisia, Colombia, and Bolivia, was negligible in Indonesia and Nigeria, and was negative in Malaysia and the Ivory Coast. Conclusions The framework developed in this chapter serves to impose a measure of analytical order on the richness and diversity of individual experience. It has been applied to thirty-three economies, and the results have been aggregated to describe the broad contours of group adjustment, both as an end in itself and with a view to placing individual performance in perspective. It is against this background that the experience of individual countries is discussed in two companion papers (Mitra 1985a and 1985b). Appendix A. Data Data on national accounts, price deflators, and exchange rates are taken from the World Bank's World Tables. The index of international inflation is the unit value index of manufactured exports f.o.b. from developed countries and is taken from various issues of the United Nations Monthly Bulletin of Statistics. Export and import trade weights are taken from the International Monetary Fund's Direction of Trade Statistics. The calculations distinguish public and publicly guaranteed medium- and long-term debt from short-term debt. The latter has a maturity of less than one year. Outstanding medium- and long-term disbursed debt be- PRADEEP K. MJTRA 113 longs to different vintages and carries different interest rates. Data on interest payments therefore reflect such terms and conditions. In the absence of a detailed breakdown, the nominal interest rate on medium- and long-term debt has been calculated as interest payments outstanding and disbursed debt Both numerator and denominator are taken from the World Bank's Debtor Reporting System, which, however, reports only public and pub- licly guaranteed medium- and long-term debt. It is assumed that the rate payable on short-term debt as well as that earned by the country's interest-bearing assets, equals the London inter- bank offered rate (LIBOR). This rate has been understood to correspond to six months' maturity (source: Salomon Brothers until 1978 and the Inter- national Financial Statistics [IFS] of the International Monetary Fund thereafter). Short-term debt data is derived from the Bank for Interna- tional Settlements' Maturity Distribution of International Bank Lending. Interest-bearing assets are defined as follows: Total Reserves minus Gold (line 1 l.d. in the IFS) less Use of Fund Credit (line 2 e.s. in the IFS), expressed in dollars. Gold has not been included as part of reserves. Appendix B. Composition of Groups Group 1: Chile, Costa Rica, Philippines, Republic of Korea, Taiwan, Singapore. Group 2: Argentina, Brazil, Guatemala, Honduras, India, Kenya, Malawi, Mali, Thailand, Turkey, Uruguay. Group 3: Jamaica, Portugal, Yugoslavia. Group 4: El Salvador, Mexico, Morocco, Spain. Group 5: Benin, Bolivia, Colombia, Indonesia, Ivory Coast, Malaysia, Niger, Nigeria, Tunisia. Notes 1. "Billion" means "thousand million." 2. This measure could be broken down, data permitting, into its p.ivate and public investment components. 3. See equation (A.26) in Annex 1 of Mitra 1984 or Mitra 1985a for an algebraic statement. 114 ADJUSTMENT TO EXTERNAL SHOCKS 4. The members of each group are listed for easy reference in appendix 2. 5. Export price effects were extremely favorable in Niger as well, but here external shocks were positive in 1974-78, with the terms of trade improving sufficiently thereafter to yield negative shocks for the period 1974-81 as a whole. References Mitra, Pradeep. 1983. World Bank research on adjustment to external shocks. World Bank research news 4:3 (Fall/Winter). 1984. A description of adjustment to external shocks: Country groups. Develop- ment Research Department discussion paper 85. Washington, D.C.: World Bank. . 1985a. Adjustment to external shocks in selected semi-industrial countries. In G. Szego, ed. Studies in banking andfinance. Amsterdam: North Holland. Forthcoming. . 1985b. Adjustment to external shocks in selected less developed countries. Washington, D.C.: World Bank, Country Policy Department. Processed. World Bank. 1981. World development report 1981. New York: Oxford University Press, chap. 6.