DI R ECTIONS I N -DEVELOP M ENT Boom, Crisis, and Adjustment The Macroeconomic Experience of Developing Countries, 1970-90 A Summary p3 ov0 I.M.D. LITTLE RICHARD N. COOPER W. MAX CORDEN SARATH RAJAPATIRANA p~~~ ~~~~~~ ,:f. x- ,'.'='ty',;I DIRECTIONS IN DEVELOPMENT Boom, Crisis, and Adjustment The Macroeconomic Experience of Developing Countries, 1970-90 A Summary Richard N. Cooper The World Bank Washington, D.C. © 1994 The International Bank for Reconstruction and Development/THE WORLD BANX 1818 H Street, N.W. Washington, D.C. 20433 All rights reserved Manufactured in the United States of America First printing November 1994 The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Library of Congress Cataloging-in-Publication Data Cooper, Richard N. Boom, crisis, and adjustment: a summary: the macroeconomiic. experience of developing countries, 1970-1990 / Richard N. Cooper. p. cmn - (Directions in development) ISBN 0-8213-3051-9 1. Structural adjustment (Economic policy)-Developing countries. 2. Economic stabilization-Developing countries. 3. Developing countries-Economic policy. I. Title. II. Series. IMI. Series: Directons in development (Washington, D.C) HC59.7.C637 1994 339-09172'4-dc2O 94-33974 CrI Contents A Note to the Reader iv A Brief Description of Events 2 Investnent Booms 4 Terms of Trade Shocks 6 The Debt Crisis 8 Painful Adjustment 11 Macroeconomic Policy 15 Crowing Fiscal Deficits 26 Increasing Debt Burden 17 Monetary Policy, Fiscal Policy, and Inflation 29 Exchange Rate and Trade Policies 21 Political Economy and Performance 23 A Brief Postscript 25 -9 Policy Advice 26 Notes 27 i.. A Note to the Reader Richard N. Cooper, Maurits C. Boas Professor of International Econom- ics at Harvard University, who wrote this summary of the book, Boom, Crisis, and Adjustment: The Macroeconomic Experience of Developing Coun- tries, is one of the original four authors. All are listed on the front cover of this booklet. Boom, Crisis, and Adjustment is the synthesis volume for the World Bank Comparative Macroeconomic studies, a series of country studies now published or forthcoming that cover Brazil, Colombia, Costa Rica, Cbte d'Ivoire, India, Indonesia, Sri Lanka, Thailand, and Turkey. in Boom, Crisis, and Adjustment The Macroeconomic Experience of Developing Countries, 1970-90 From the early 1950s the major focus of economic policy in developing countries was on economriic growth and on achieving the structural changes thought necessary for growth, notably the encouragement of manufacturing. From the iid-1970s to the late 1980s many countries were distracted from these long-term objectives by the need to adjust to a series of major shocks to their economies, some imported from the world economy, some generated at home. Perforce, governments had to tum their attention to short-run macroeconomic adjustnent. The burden placed upon them, and their skill in dealing with these largely unwanted tasks, varied considerably from country to country. How did countries respond to adverse, or favorable, changes in ex- trnal economic conditions? What can be learned from their variegated and diverse responses for appropriate macroeconomic management in the future? Where shocks were generated domestically, what were their origins and how can they be avoided in future? And what are the implications of poor or skilful macroeconomic management for long- run growth? Wlth such quesdions in mind the World Bank sponsored a large-scale comparative study of macroeconomic management in developing coun- tries during the turbulent period 1970-90, with reference to the late 1960s where relevant. The study was undertaken in two phases. The first phase involved an examination by economists who had some knowledge of the countries in question of macroeconomic disturban- ces and responses in each of seventeen countries. In the second phase generalizations from those country studies and other material were synthesized in the recently publshed volume Boom, Crisis, and Adjust- ment: the Macroeconomic Experience of Developing Countries.' Here we summarize the major findings and recommendations from that book The countries were chosen to be a representative (but not random) sample of noncommunist developing countries. They induded virtual- ly all of the large developing countries, measured by gross output, and a selection of small ones, accounting in all for over 60 percent of the 1 2 BOOM, CRISIS, AND ADJUSTMENT GDP of developing countries as a group. The total population was 1.8 billion and the countries included six from Latin America, six from Asia, five from Africa, and one (Turkey) from Eurasia.2 The countries covered by the study, with their major economic characteristics, are listed in table 1. The study was stimulated initially by concern that excessively broad generalizations were being drawn from the experiences of only a few developing countries. These were concentrated in South America, and we felt that explicit comparison among a substantial number of coun- tries would offer a firmer base for general conclusions. In addition, al- though we use standard economic concepts and categories, we had no strong preconceptions about the results to be obtained; the study was inductive in spirit. A Brief Description of Events There were two dominant trends in the world economy between 1965 and 1990. The first was the continued rapid growth of world trade, including especially exports of manufactured goods from developing countries; the second was the emergence in the early 1970s of a world money and capital market to which many developing countries had ready access, as in the 1920s. The princpal economic events of the period were the move from fixed to floating exchange rates among the major industrialized countries in March 1973; two sharp increases (1974 and 1979-S0) and one sharp fall (1986) in oil prices; dramatic movements in prices of several other primary products, such as coffee, cocoa, copper, and phosphates (all commodities of special relevance to our countries); a mild economic recession and two deep ones (1975 and 1982); a large appreciation of the U.S. dollar against other leading cur- rencies (198G-85), followed by an equally large and more abrupt fall (1985-86); and a debt crisis for many developing countries, during which voluntary external lending virtually ceased in the mid-1980s. Broadly speaking the 1960s and early 1970s were years of outstand- ing econoniic growth and, until 1973, relative price stability. This pat- tern broke in the mid-1970s with a commodity price boom, the first major oil shock, and the deepest recession since the 1930s. Consolida- tion in the late 1970s was interrupted by a second oil shock associated with the revolution in Iran. That episode, along with stringent efforts by several major industrialized countries to control inflation, led to a second major recession in 1981-82. The combination of heavy borrow- ing through earlier difficulties, high interest rates associated with high- er inflation and the anti-inflationary response, and 1982 recession produced the debt crisis of 1982-83. After a set-back in the late 1970s and early 19821s industrialized countries resumed their growth by the late 1980s. For developing countries the decade of the 1980s was one of Table 1. Basic Data on Eighteen Countries, 1965-90 GNP per capita GDP, Populatioln, Average annual Avrage annual External debt as Debts 1990 mid-1989 1990 growth rate, inflation rate, percent of GNP, rescheduled, Country (millions US$) (ttillions) (US$) 1965-90 1980-90 1990 1982-88 Argentina 93,260 32.3 2,370 -0.3 395.2 61.7 Brazil 414,060 150.4 2,680 3.3 284.3 22.8 Cameroon 11,130 11.7 960 3.0 5.6 56.8 Chile 27,790 13.2 1,940 OA 20.5 73.6 Colombia 41,120 32.3 1,260 2.3 24.8 44.3 Costa Rica 5,700 2.8 1,900 1.4 23.5 69.2 COte d'lvoire 7,610 11.9 750 0.5 2.3 203.9 India 254,540 849.5 350 1.9 7.9 25.0 Indonesia 107,290 178.2 570 4.5 8.4 66.4 (hi Kenya 7,540 24.2 370 1.9 9.2 81.2 Korea, Republic of 236,400 42.8 5,400 7.1 5.1 14.4 Mexico 237,750 86.2 2,490 2.8 70.3 42.1 Morocco. 25,220 25.1 950 2.3 7.2 97.1 Nigeria 34,760 115.5 290 01 17.7 117.9 * Pakistan 35,500 112.4 380 2.5 6.7 52.1 Sri Lanka 7,250 17.0 470 2.9 11.1 73.2 Thailand 80,170 55.8 1,420 4.4 3.4 32.6 Turkey 96,500 56.1 1,630 2.6 43.2 46.3 Note: The 'technical notes' in World Development Report 1992 explain the meaning and methods of calculation of the figures in the first five columns. The average annual inflation rate (column 5) is measured by the growth rate of the GDP implicit deflator, while other figures of inflation rates in this booklet refer to the growth rate of the cost-of-living index-but the two measures nonnally tell very similar stories. External debt in column 6 refers to the total extemal debt stock, long-and short-term. For column 7, note that Cameroon became rescheduled in 1989. Source: World Development Report 1992 (for the first five columns). Column 6 is from the World Bankes World Debt Tables 1991-92, Volume 2. Column 7 comes from the International Monetary Fund and chapter 4 of LCCR. 4 BOOM, CRISIS, AND ADJUSTMENT Table 2. Growth and Inflation, 1965-90 (percent a year) Item 1965-73 1973-80 1980-90 GcP growthl Industrial countries 4A 2.5 3.1 Developing countries 6.2 5.1 3.2 Per capita 3.7 3.0 12 Inflation Industrial countries GDP deflator 5.4 9.5 4.5 Fxports (U.S. dollars) 5.6 12.0 2.6 Deve.oping countries cDP deflator 9.7 24.9 61.8 Nonoil exports (U.S. dollars) 4.5 12.5 02 Source: Calculated from Intemational Financial Statistics Yearbook (1990), and World Bank, World Development Report, 1992. adjustment, relative stagnation, and for some, accelerating inflation (see table 2). Investment Booms In the early 1970s development planning was fashionable among both officials and academics. Govemrnments took on the roles of entrepre- neur, investor, and manager of economic activities in addition to their more traditional functions. As conditions permitted, tiis tendency pro- duced remarkable investment booms in virtually all our countries dur- ing the 1970s and early 1980s. As shown in table 3, investment as a share of gross domestic product (GDP) rose substantially during this period-by an astonishing 18 percentage points in Sri Lanka and not much less in Morocco and Nigeria. In Nigeria the rise occurred over a span of only two years. Except in Chile, India, and Kenya, the invest- ment booms were led by the govemrnent, although the private sector also played a significant role in Brazil, Cameroon, Indonesia, and Sri Lanka. The Korean investment boom was undertaken mainly by pri- vate firms under dose government supervision and inducement as part of its big push into heavy and chemical industries. In all of the countries investments were often undertaken hastily and with too little serious evaluation, and they failed to make a proportionate contribu- tion to economic growth. Only in Indonesia, Kenya, and Nigeria were the investment booms generated by sharp improvements in the foreign terms of trade (on which more below). In all other cases the investment booms were do- mestically conceived and launched, although improvements in the A B3RIEF DESCRIPTION OP EVENTS 5 Table 3. Investment Booms (percentage of CDP) Inivestment Cou ntnj Period Plblic Private Total Argentina 1973-77 8.4 PA 13.3 10.7 21.7 24.1 Brazil 1972-75 5.1 7A 16.3 19.8 21.4 27.2 Cameroon 1974-79 4.5 4.6 11.1 17.4 15.6 22.0 Chile (F) 1976-81 5.1 4.2 7.7 16.8 12.8 21.0 Colombia (F) 1978-83 5.5 9.1 9.9 8.1 15.4 17.2 Costa Rica (F) 1975-79 7.0 8.9 15.0 16.4 22.0 25.3 C6te d'Ivoire 1974-78 11.3 21.0 8.1 8.7 19.4 29.7 India 1977-81 8.4 10.0 11.4 15.7 19.8 25.7 Indonesia 1974-61 8.0 132 8.6 16.7 16.6 29.9 Kenya (F) 1977-80 9.0 107 9.1 13.0 18.1 23.7 Korea, Republic of 1976-79 5.5 7.5 18.8 25.8 24.3 33.3 Mexico (1) 1971-75 4.7 8.9 13.3 12.5 18.0 21.4 Mexico (2) 1977-81 7.7 11.7 11.9 14.0 19.6 25.7 Morocco 1973-77 4.7 20.7 12.2 13.5 16.9 34.2 Nigeria 1974-76 4.8 17.3 11.9 14.1 16.8 31.4 Pakistan 1973-77 7.1 15.3 7.1 5.9 14.2 21.2 Thailand (F) 197441 3.7 92 19.7 16.5 23.4 25.7 Sri Lanka 197742 6.5 16.7 7.3 15.2 13.8 31.9 Turkey 1973-77 8.5 12.5 9.6 10.1 18.1 22.5 Note: F indicates gross fixed capital formation. The figues for total investmnent differ from those derived from the World Bank data base (which are expressed as a percentage of GNC, not GDP) given in the appendix tables in LCCIL But there is no serious conflict. The figures for Korea are from Sentdn and Solimano (1991). Source LCCR, table 33, and IMP, Supplement on Government Finance Statistics, 2986. terms of trade facilitated, and perhaps augmented, the booms in Cameroon, Costa Rica, COte d'Ivoire, and Morocco and the second Mexican boonL Specific political factors as well as general "develop- mentalism!' were behind most of the investment booms. But in all cas- es other than the first three mentioned above, they could not have been undertaken without ample sources of extemal credit. This credit came partly through foreign assistance programs (especially important in the case of Sri Lanka) but mainly through the revived intemational fi- nancial market, which was given a strong boost by the first oil shock as oil-exporters temporarily placed their much higher export earniings in the London-based financial market. Even those countries with favor- able movements in their terms of trade soon began borrowing in the international market, encouraged in part by the relatively low interest rates prevailing in the mid-1970s, but mainly by the ample availability of funds. Thus the widespread character of the investment boom was made possible by revival of the world capital market. 6 1300M, CRISIS, AND ADJUSTMENT Terms of Trade Shocks The first oil shock occurred when OPEC ministers agreed in December 1973 to raise crude oil prices from under $4 a barnel to over $11 a barrel. This promptly redistributed about $65 billion anually from oil- importing countries to oil-exporting countries and imposed an adverse movement in the terms of trade on oil-importing countries that varied in magnitude according to the importance of oil in their total imports. This action, of course, also irmproved the terms of trade of oil-exporting countries (among our countries Indonesia and Nigeria, joined later in the 1970s by Cameroon and Mexico). Combined with the anti-infla- tionary responses of the major industrialized countries, especially the United States and Germany, the sharp increase in oil prices led to the recession of 1974-75, which in turn depressed earnings on most exports of industrial materials. Table 4 reports the magnitude of terms-of-trade shocks in terms of GDP for the years 1974 and 1975. Chile experienced the biggest adverse shock over the two years (copper prices dropped sharply in 1974), while, as noted, Indonesia and Nigeria experienced significant tin- provements. For Morocco the negative effect of higher oil prices was more than offset by higher prices for phosphates, its most important ex- port. Argentina, Colombia, and Mexico were virtually oil self-sufficient, so movements in their terms of trade were dominated by other factors. Since countries differ substantially in their openness to foreign trade, a given effect of changes in the terms of trade on GDP can represent sub- stantially different burdens in terms of their claim on export earings. For example, in India, a large and heavily protected country, the adverse movement of its terms of trade amounted to 25 percent of Indian ex- ports, while a roughly equivalent change (in temis of GDP) in Thailand's terms of trade amounted to only 5 percent pf Thailand's exports. Oil prices rose less than inflation over the period 1974-78, but other prices moved more substantially. In particular, a major frost in Brazil led to a threefold increase in coffee and cocoa prices in 1976-77, sharp- ly improving for several years the terms of trade of major exporters of those products-Cameroon, Colombia, Costa Rica, Cote d'Ivoire, and Kenya among our countries. As table 4 also shows, growth rates gener- ally were unaffected by the first large oil shock: growth in the post- 1974 period was actually higher in many countries than it was in the early 1970s, and it declined substantially only in Kenya, Brazil, Mexico, and (paradoxically) Nigeria. The study finds no correlation between the magnitude of the terms- of-trade shock and the change in growth rates. Countries experiencing an adverse movement either increased their exports (Korea, Thailand, and Pakistan) or borrowed (Chile, Costa Rica, Brazil, and Argentina), or both, to cover their increased import bill. Of course, heavy borrow- A 13RIIW DESCRIPTION 01 UVENTS 7 Thble 4. Terms of Thade Shocks, 1974 and 1975, and CNP Growth porcen t) Percentage of Gor CNP growdl/ Colrtn: Jn274 7975 2970-73 7974-79 Chile -12.4 -5.1 1.7 2.8 Cameroon -3.8 -3.7 2.7 8.6 C6te d'Woire -0.9 -6.4 6.4 5.8 Kenye 4.1 -3.0 9.7 5.6 Cost}s Rica -5.1 -1.0 7.4 5.2 Pakistan -3.1 -2.8 4.7 4.8 Korea, Republic of -4.0 +0.9 9.1 9.6 Sri Lanka -3.1 -0.6 3.1 5.2 Thailand -0.8 -3.0 7.0 7.3 Brazil -2.6 -0.5 11.9 6.1 Mexico -0.9 -0.6 8.5 5.8 Turkey -1.1 -0.3 6.3 5.3 Argentina -0.9 -0.3 3.6 2.4 India -0.9 +0.0 2.4 3.4 Colombia +0.9 -1.1 7.2 5.3 Morocco +3.0 +1.3 3.9 5.6 Indonesia +17.0 -3.0 8.5 6.6 Nigeria +23.1 -2.6 9.8 3.9 Note: The terms of trade must be regarded as giving an order of magnitude only: there are considerable divergences in several cases between World Bank data and those given in the country studies. (But only in the case of Kenya were World Bank data clearly wrong.) Source: LCCR, tables 3.1 and 3.5. ing abroad increased external debt, laying the basis in some countries for later difficulties. But it also mitigated the world recession which would have been deeper had not so many countries borrowed to con- tinue their investments and their irmiports of oil. Iran's revolution in 1979 led to a marked decline in Iran's oil exports and set off the second large oil shock. This was not so concentrated in time as was the first, but over two years the price of crude oil increased from roughly $12 a barrel to $33 a barrel. This led to another major transfer, this time over $100 billion anrually, from oil-importing to oil- exporting countries. Once again, the oil price increase itself and the anti-inflationary policy responses it provoked in all the major industri- alized countries led to an unexpectedly sharp world recession in 1981- 82. Concems with the increase in inflation during the 1970s led coun- tries, especially the United Kingdom (which by 1979 had become an exporter of North Sea oil) and the United States, to pursue much tighter monetary policies than they had done earlier, and short-term interest rates rose sharply in 1980 and 1981, with the three-month dollar Lon- don interbank offered rate, LIBOR, exceeding 17 percent in 1981. The 8 BOOM, CRISIS, AND ADJUSTMENT sharp increase in interest rates imposed a heavy debt-servicing burden on those countries that had already acquired significant external debt subject to floating interest rates (generally adjusted at six-month inter- vals), a feature of many bank consortium loans. Thus, many developing countries again experienced an adverse movement in their terms of trade, composed partly of higher oil prices and partly of higher interest rates on their outstanding external debt. Table 5 reports the cuxrLulative terms-of-trade effect on our eighteen countries arising over the three years from 1979 to 1981, in terms of an- nual GDP. C6te d'Ivoire suffered the most severe effect, at 13 percent in- duding interest rates, and half of our countries experienced an adverse movement of more than four percentage points. In terms of interest rates alone, Chile and Korea experienced the most serious effects. As in 1974, Indonesia and Nigeria benefited from the high-2r oil prices, as did Cameroon and Mexico thls time, although Mexico's gain on oil was lost on interest rates.3 The Debt Crisis The combination of investment boom, worsening terms of trade, and deep world recession resulted in great enlargement of current account deficits between the early 1970s and 1982. Table 6 reports current Table 5. Terms of Trade and Interest Rate Effects, 1979-81 (cumulative percentage of annual GDP) Couny Term of tmde effect InWeres' rate 5.6% in 1979. Big shock-total negative effect/CDP > 43.% in 1979-81 (see table 5). Inaction-see LCCR texL Lack of fiscal control-see LCCR text Recession-actual all in annual nuIG. Real exchange rate appeciation-percentage appreciation of the real exchange rate from 1979(1) to 1982(2) > 15% a. These countries all had periods of rapid real appreciation. In Argentina, tht. e was a huge appreciatioo from 1978(1) to 1981(1), followed by a fal back to the levels of 1978. In Brazil the real exchange rate fell from 1978(1) to 1980[1), but then rose by 45% to 1982(3). In Costa Rica, the real exchange rate appreciated from 1978W() to 1980(4), before faling back to 1978 levels. b. The lack of real apprecation in the period 1979-82 does not imply that the currency was not overvalued during that periodin the case of Morocco our author argues that the dirham was overvalued relative to 1970, taken to be an equihbrium year forthe exhange rate. c. In Cameroon the terms of trade shock was offset by a rising volume of oil exports. d Fiscal stimulus until 1984, at cost of worsening extemal position- e. In 1980 due to poor harvesL Sour: LCCR, table 4.4, modified slightly in light of new infonnation. MACROECONOMIC POLICY 15 negative shock than countries that performed much better. Rather, the differences lie largely in the speed with which they reacted to adverse developments and in the polices they pursued, to which we now tum. Macroeconomic Policy Clearly there is a wide range of experience among our eighteen coun- tries, even tough they were aLl part of the same world economy and to varying degrees were buffeted by the same shocks and offered the same opportunities. Why did experience vary so much from country to country? Was it due entirely to differences in circumstance, or did dif- ferences in policy play an important role? It is difficult to avoid the judgm.;ent that differences in policy played the major role in determining performance. It is true that countries dif- fered in their circumstances, and in particular in the magnitude of the shocks that they experienced. But surprisingly we found no correlation between the magnitude of the shocks and subsequent economic per- formance, measuring the latter by growth in per capita income during the 1980s and by rates of inflation. Even more surprising, there is no correlation between economic performance and the sign of the shock: countries that experienced positive external shocks-oil-exporting countries in the mid-1970s and early 1980s, coffee-exporting countries in the late 1970s-did not, on average, have superior performance compared with countries experiencing negative shocks. The potential benefits from positive shocks were too often squandered. What is the possible explanation for this astonishing finding? The answer seems to be that a sharp rse in export eamings causes a kind of euphoria in govemment and perhaps also in private circles. This euphoria weakens the normal financial disciplines and entices government agencies and enterprises to go on a spending spree. The extent of the euphoria varies from country to country, with some coun- tries, Nigeria in 1979-80 for example, essentially having lost cental budgetary control for a period. State enterprises involved in the export boom, Pertamina in Indonesia in 1974, for example, and Pemex in Mexico in 1980-81, are especally difficult to controL Even when bud- getary control remains the government may take the view, or behave as if it takes the view, that the improvement in export earings is permanent rather than temporary, and it will budget accordingly. This seems to have occurred following the 1976-77 coffee boom in Costa Rica, Cote d'Ivoire, and Kenya. During this time recurring public expenditures were increased in response to enlarged revenues that turned out to be temporary and should have been known to be tempo- rary. Other countries, however, managed to maintain budgetary disd- 16 BOOM, CRISIS, AND ADJUSTMENT pline in the face of an export boom. The Cameroort government adopted the simple expedient, not easily replicated elsewhere, of keep- ing its oil export revenues a secret and sequestering most of them abroad. This way they entered neither government revenues nor the domestic income streamt. So there is clearly a risk that an export boom, leading to higher gov- emment revenues, will create a euphoric optimism about future reve- nues and slacken budgetary discipline over the ever-present pressures to enlarge public spending. Long desired investment plans may be brought forward and implemented with little or no serious analysis of the likely returns! When revenues drop, the govemment is faced with the prospect of replacing them or cutting the recently increased expen- ditures, both of which are politically difficult. Cuts tend to be concen- trated on investments, which are discrete and do not involve commitments of indefinite duration to employees as operating pro- grams do. Governments under these circumstances are lkely to post- pone the difficult decisions if they can, and cover their deficits by borrowing abroad. Growing Fiscal Deficits Government expenditures rose in most countries over the 1970s, and more rapidly than revenues. As a result, budget deficits for all nonoil developing countries grew by 2.8 percent of GDP, from 3.5 percent in 1972 to 6.3 percent in 1982; in the industrialized countries deficts grew even more, by 3.2 percent of GDP. Over the subsequent five years from 1982 to 1987 deficits declined sharply in the industrialized countries but continued to rise in developing countries as expendihtres rose slightly and revenues declined slightly Deficits must be financed, either by borrowing or (temporarily) by drawirng down cash balances. Borrowing can be drawn from foreign- ers, public or private; or from domestic lenders, nonbanks or banks; and among banks, from commercial banks or from the central bank. It is noteworthy that in 1980, a year in -which extemal funding was gener- ally available, only four of our countries, COte d'Ivoire, Morocco, Indo- nesia, and Kenya in order of relative size of external borrowing, financed more than half of their central government budget deficits ex- temally. Even Sri Lanka, with its huge deficit, financed only a quarter of it abroad. There was very little direct external borrowing by the federal gov- emments of Argentina, Brazil, and Mexico, countries that subsequent- ly became the three largest public debtors. Those countries later became large debtors because they had given guarantees on borrowing by state enterprises (SoEs) and because they assumed large amounts of private debt during the large devaluations and debt crises after 1982 MACROECONOMIC POLICY 17 Indeed, SOEs were major sources of credit expansion in some of our countries, sometimes with direct access to foreign funds or to the central bank, and not always under close financial control (Argentina, Brazil, Costa Rica, Mexico, Turkey). In some federal countries state or provincial govemments were also a source of extemal borrowing (Argentina, Brazil, Nigeria). The remainder had to be borrowed from domestic sources. Financial markets are largely nonexistent in most developing countries. (Brazil and Mexico were the exceptions during much of our period.) But that did not stop them borrowing from nonbank financial institutions, such as savings banks, insurance companies, pension funds, and specialized lending agencies. Thus, a substantial fraction of govemment obli- gations are held by the nonbank public, especially in Costa Rica, Pakistan, and Sri Lanka, among our countries. At the margin, however, almost all governments must turn to the banking system for incremental financing of budget deficits that can- not be financed overseas. This financing is sometimes done almost ex- elusively by the central bank, as in Mexico, Turkey, and Indonesia; sometimes, as in Korea and Thailand, it is done mainly through depos- it money banks (DMBs), which in turn have access to the central bank. Such financing of course is inflationary, except insofar as the banks re- duce their lending to private borrowers correspondingly, or output can grow sufficiently to satisfy the increased demand. Increasing Debt Burden Government borrowing accumulates over time as outstanding debt, which can come to represent a substantial charge on the budget8 Interest payments for nonoil developing countries as a group rose from 7 percent of total central government expenditure in 1977 to 10 percent in 1982 and furither to nearly 18 percent in 1987. Amorization of past debt, insofar as it could not be refinanced, increased the burden fur- ther. A budget defict is sustainable only if there is some prospect that it can be financed in the future as well as in the current period. The long-run dynamics of budget deficts, which are all but universally ignored in the budgeting process, are such that the growth of future revenues must exceed the average interest rate paid on outstanding debt so long as the govemment is runnng a prmary deficit (that is, a deficit exdusive of interest payments). The notion of primary deficit is especially useful in a highly infla- tionary environment, where some portion of interest payments in ef- fect amortizes the debt, since a given nominal debt is declining in value in real terms. Amortization is present both for external debt and for internal debt It is especially important for internal debt willingly held by the public, such that holders will wish to preserve the real val- 18 BOOM, CRISIS, AND ADJUSTMENT ue of their claims on the government, and therefore will willingly rein- vest the relevant portion of their interest earnings. In most of our high inflation countries, however, domestic government debt was compul- sorily acquired, for example, by financial institutions or by pension funds.9 Such holders of debt will, if possible, allow the real value of their claims to erode with inflation, preferring investments in other as- sets. Under these circumstances, the inflation-corrected deficit under- states the financing problem of the government. The notion of the primary deficit has another value as well. To the extent that governument debt is held by foreigners, or by institutions (such as pension funds or life insurance companies) whose current dis- bursements are unrelated to their current eanings. interest payments by the government do not have the expansionary impact that other ex- penditures do. In developed countries where government bonds are widely held by the public, interest payments by the government in- crease the effective purchasing power of the public. But in many devel- oping countries the primar deficit may give a better measure of the impact of govemment spending on the economy. Although overall budget deficits on average were virtually un- changed during 1982-87, primary budget deficits declined dramatical- ly, the difference being the sharp relative increase in interest payments, both to foreigners and to domestic holders of government debt (see table 7 and LCCR, table 10.11, for the rise in interest pay- ments). indeed, by 1987 half of our countries ran primary surpluses. The primary deficits grew significantly only for Brazil, Cameroon, Costa Rica, and Pakistan. So the niid-1980s was a period of general budgetary retrenchment-the inevitable reaction to a crisis of exces- sive spending External debt poses another problem. Not only must future revenue be increased to service the debt, but in effect the revenue must be raised in foreign currency, since the debt is overwhelmingly denomi- nated in U.S. dollar or another foreign currency. A few governments enjoy direct ownership of exportable products such as petroleum or other minerals; but most governments are unable to raise revenues di- recdly in foreign currency. Instead, insofar as they can not borrow fur- ther abroad they must buy the foreign currency directly or indirectly from exporters. Large currncy devaluations can play havoc with the budget, since the local-currency interest payments on foreign debt must rise abruptly by the amount of the devaluation.0 Even gradual depredations of the currency as under a crawling peg regime, increase external debt service requirements in terms of local currency; and when these depreciations go beyond merely compensating for domes- tic inflation, they raise the budgetary burden of extemal debt servicng in real terms. MACROECONOMIC POLICY 19 As noted above, most central governnents did not borrow exten- sively abroad. We found, however, that extensive external borrowing was often done by soEs, both nonfinancial and financial, and by pri- vate firms. Debt problems in many countries arose because these bor- rowers' debts were often guaranteed by the government or by the central bank. So if a currency devaluation led to a negative cash flow for the domestic enterprise, the government had to help resolve the problem. Typically, the extemal debts were directly or indirectly as- sumed by the central bank, which then ran the losses associated with application of the new exchange rate to debt servicing. Central bank deficits became especially large in Chile, Costa Rica, and Argentina and added to the fiscal pressures created by the budget deficits. Among our countries, eleven experienced increases in the ratio of public or publicly guaranteed extemal debt to GDP in excess of ten per- centage points between the end of 1979 and the end of 1982 (table 7), an as.. ncnding rate of increase that was dearly not sustainable in the long run. The increases were due partly to extensive borrowing, partly to gov- emment absorption of private external debt, but partly also to currency deprecation, which raised the value of extemal debt relative to GDP. Monetary Policy, Fiscal Policy, and Inflation Since on average about half of budget deficits in developing coun- tries are financed by the banldng system, there is a dose connection be- tween budgetary policy and mnnetary policy. A budget deficit leads to an expansion of the money supply. The typical developing country has a rapidly growing demand for money as the economy becomes more monetized and as households and firms increasingly hold assets in fi- nancial formL The rise in demand for money-more rapid than the growth in GDP, unless discouraged by high inflation-permits govern- ments to ear substantial seigniorage, the difference between the face value of money and the cost of issuing it. Monetary expansion at a rate greater than this growing demand in an economy fully using its re- sources is inflationary. A rise in prices zeduces the real value of out- standing money balances, whidh the public will want to replenish, at least in part. That desire leads to more seigniorage, this additional por- tion being an "inflation tax." Governments can command real resources by inflating, and thereby inducing the public to spend less on goods and services, in order to re- plenish their cash balances in real terms. This process acts very much like a tax on cash balances. On average our eighteen countries raised 2.9 percent of CDP in seigniorage during the late 1970s. The key point for stabilization policy is that the inflation tax is an im- portant source of real resources for governments. A reduction in infla- 20 BOOM, CRISIS, AND ADJUSTMENT tion therefore requires that govemrment expenditures must be cut correspondingly, or that alternative sources of revenue be found. Thus, tducing average inflation may be costly in terms of public finance. (Average seigniorage for our countries fell to 2.1 percent of GDP in the late 1980s.) This fiscal inhibition to reducing the average rate of infla- tion does not, of course, bear on the desirability of reducing fluctua- tions in inflation. It has been noted that on average inflation accelerated in developing countries during the 1980s but that the diversity of experience was great. We found it useful to distinguish three types of inflation accord- ing to their duration and to some extent their motivation. Adjustment inflation shows a sharp increase followed by a decline in the rate of in- flation (in other words, a once-for-all increase in the price level); seigniorage inflation is where the government relies on the inflation tax as a source of revenue; and spiral inflation is where wages and other fac- tor costs adjust fully to the rise in prices. In the last case, inflation may accelerate over time, perhaps even reaching hyper-inflation (conven- tionally defined as inflation over 1000 percent a year or 47 percent a month, reached by Argentina and Brazil among other countries), and induce a sharp reduction in the demand for real money balances. Adjustment inflation was common during the 1970s, as can be seen in table 9. Thirteen countries experienced a sharp increase in inflation, followed by a decline, during 1973-75, and ten during 1980482. These were the periods of large increases in crude oil prices. hiflation contrib- utes to and eases the adjustment to the new crcumstances (here, much higher energy prices), because it is difficult and often costly in modem economies to reduce the general level of prices. While inflationary bub- bles are probably inevitable in the presence of external shocks, they need not become spiral inflation. With proper management, the infla- tion will be temporary; once the adjustment has occurred, prices need not continue to rise. Several countries succeeded in restoring single digit inflation after the two-digit inflation of the mid-1970s, and even more after the inflation bubble of the early 1980s. Some countries (Argentina, Colombia, Mexico, for none of which was the first oil shock significant) found themselves at a new level of inflation after the shock, and still others (Brazil, Argentina after the second oil shodk, Turkey) found themselves with accelerating inflation. But these cases of extremely high inflation were exceptional, not typicaL Inflation is difficult to control when it gets built into public expecta- tions; and where it provides significant revenue to the govemment thrugh seigniorage, it is difficult to reduce without finding an altema- tive source of revenue, as noted above. Spiral inflation, which might be triggered by a currecy devaluation, arises when the competing claims on a country's output exceed its output, and institutional arrange- ments do not prevent the central bank from financing one or more of MACROECONOMIC POLICY 21 the parties whose claims exceed their incomes, usually government, sometimes enterprises. inflation reconciles the conflicting claims ex post by reducing the real values of outstanding assets and liabilities and by reducing the real value of spending by some parties. An excess demand for output can be satisfied by borrowing abroad to pay for imports, and that recourse was taken by many countries when their terms of trade deteriorated during the 1970s. When foreign credit became much scarcer following the debt cri.sis of the early 1980s, they had to face the need to reduce demand to fit total production. Some countries succeeded in reducing budget deficits, in practice a necessary component of stabilization programs, both to reduce de- mand and to influence expectations favorably. Others instead replaced foreign credit with credit from the banking system, and this substitu- tion increased the inflation rate. In this respect inflationary pressures were a direct consequence of the debt crisis and the associated decline in net inflow of resources, which as shown in table 8 was significant for several countries. Inflation, or at least moderate inflation, did not seem to have an ad- verse effect on growth. But high inflation is disruptive of both econom- ic and social relations. Once high inflation takes hold it is difficult to reduce without a real reduction in demand. A reduction in demand of- ten produces a recession and a rise in unemployment. This is particu- larly so if real wages rise as a result of formal or informal indexation based on lagged inflation or as a result of a devaluation associated with the stabilization program. In practice, successful stabilization pro- grams have generally involved conscious govement effort to restrain money wage increases; in some countries such "heterodox" measures may be necessary to avoid growing unemployment during a stabiliza- tion program, which if uncontrolled would result in political backlash against the program. Selected price controls may in turn be necessary to adhieve wage restraint, but they create serious allocative problems; if associated with government enterprises, as they often are, they lead to increased subsidies (or reduced profits) and thereby work against the required fiscal restraint Exchange Rate and Trade Policies In pursuit of price stabilization several countries have fixed their exchange rate, either directly to another currency or to a pr- announced "tablita" of devaluations at rates lower than the rate of inflation. The record has been mixed. Two noteworthy cases were Argentina and Chile in the late 1970s; both failed, although for quite different reasons. Argentina failed to exercise fiscal restraint, leading to capital outflow and a payments crisis. Chile ran a surplus in its budget and experienced substantial inflows of capital for a while, but suffered 22 BOOM, CRISIS. AND ADJUSTMENT from a strong rise in real wages because of lagged indexation, followed by a sharp worsening in its terms of trade, which also created a pay- ments crisis. On the other hand, Mexico's exchange rate stabilization in the late 1980s and Argentina's in 1991, both combined with severe fiscal re- straint and other measures, may have played an important role in al- tering expectations about future inflation in those countries. The move to generalized floating among major currencies in 1973 compeUled developing countries to choose between letting their curren- cies float or fixing them to one or another of the major currencies. Most chose initially to fix their exchange rates, usually to the U.S. dollar but sometimes to another currency or to a basket of currencies. A few countries, notably Brazil and Colombia, had already adopted a crawl- ing peg, fixed from day to day but changed frequently. The terms of trade shocks of the mid-1970s, and again in the early 1980s, led many countries on fixed exchange rates to devalue, and led several to shift to a regime of managed floating exchange rates. By the end of the 1980s most of our countries had adopted more flexibility in the management of their exchange rates and on the whole seemed to benefit from it. The most important exceptions were Cameroon and COte d'Ivoire; these were in multinational currency unions with their currency, the CFA franc, fixed to the French franc and hence, loosely, to all the currncies in the European Monetary system. Both countries were suffering in the early 1990s, partly because their export products were not fully com- petitive in the world market and indeed the CPA franc was devalued for the first time in early 1994. Table 11 reports the major devaluations and changes in exchange rate regime in the eighteen countries. Fixing the currency to a relatively stable major currency, if it can be done credibly, has the advantage of creating expectations within the country that inflation will be kept low by virtue of the exchange rate commitment But it has two important disadvantages. First, developing countries have been, and no doubt will continue to be, subject to sub- stantial shocks in their terms of trade, or to other major economic distur- bances. Adjustment to these shocks will require changes in factor prices. Unless wages and other factor prices are quite flexble, these shocks will lead to unemployment and economic distess. Allowing the exchange rate to change will not eliminate the pail, but will ease the process of adjustment. Real wages appear to be quite flexible in developing coun- tries; in particular, currency devaluation does not get rapidly translated into corresponding incTeases in money wages. Thus, nominal deprecia- tion generally leads to real depredation, and that stimulates exports. The experience of our counties suggests that flexible exchange rates need not lead to higher inflation. Indonesia, Korea, and Morocco all switched from fixed exchange rates in the 1970s to flexible rates in the 1980s, yet managed to lower their rates of inflation (LCCR, table 8.5). MACROECONOMIC POLICY 23 Table 11. Devaluations and Regime Changes in Sixteen Countries Country 1965-73 7974-79 1980-83 1984-90 Cameroon a a a a Chile R (1974)a R (1982) (1985-86) Colombia a (1985-86) Costa Rica a (1g74)a R (1981) (1986-87) Cote dlvoire a a a a India (1966)' R (195-N88) Indonesia (1970)a (1978)a R (1983) (1986) Kenya a (1974)a R (1982) (1985-87) Korea, Republic of R (1974)a R (1980) (195-86 Mexico a (1976)a R (1982) R (1985K)b Morocco a a a R (1984-85) Nigeria a a a R (1985-88) Pakistan (1972)a a R (1982) (1985-8 SriLanka (1967)a R (1977) (1985-87) Thailand a a a R (1984) Turkey (1970)a a R (1980) (1985-87) Note: R is regime change during period at heading; from fixed to flexble, or opposite. Dates of nominal devaluations in the first thuee periods (if more than 10 percent) and of real devaluations in last period. a. Fixed rate regime at end of period. All others have flexible rate- b. Tablita at end of period. Source LCCR, table 82. Second, over the years many developing countries have adopted substantial restictions against imports, which remain high even after the import liberalizations of the 1980s. Further liberalization will al- most surely require currency devaluation to stimulate exports enough to cover the enlarged import bilL Fixing the exchange rate makes cur- rency depreciation more difficult. Historically, countries often responded to balance-of-payments cri- ses by imposing restrictions on imports. We found that such restric- tions usually did not solve the payments problem, and they often sbtved domestic production of critical inputs or parts, leading to a de- cline in production. Trade liberalization in such cases is highly desir- able to permit ineases in production and to increase the flexibility of economies to respond to unforeseen disturbances. Politial Economy and Performmnce It is sometimes suggested that difficult economic reforms or stabiliza- tion, especially tight fiscal discipline, can only be carried out by authoritarian govemments. The experience of our eighteen countries does not support this view. The form of govemment-democratic or 24 BOOM, CRISIS, AND ADJUSTMENT authoritrian-did not determine success or failure in macroeconomic policy. It is true that some of the most successful economies, notably Korea and Indonesia, were authoritarian; but sc were some of the least successful, such as Argentina in the late 1970s, C8te d1Ivoire, and N-ige- ria. Transitions to democratic rule may create or intensify macroeo- nomic problems as in Argentina and Brazil in the 1980s but not, however, in Chile in 1990. Both authoritarian Paldstan and democratic India followed conservative macroeconomric poLicies. Sound or cau- tious policies can be pursued under many forms of government Table 12 organizes our countres according to whether they were economi- cally successful, measured by per capita economic growth during the 1980s; whether they had stable or unstable governments during the late 1970s and early 1980s; and whether their cizens enjoyed a high or a low degree of cvil liberty, as measured by Freedom House311 There is no evident correlation between economic performance and form of govemment or degree of authoritaianism. It is also dear that tradition and history-notably traumatic experi- ences of chaos or inflation-influence current attitudes of elites and the general public. Some countries have long establshed traditions of fis- cal conservatism. In others, to simplify, things may have to get worse- possibly very bad indeed-before they get better. The stabilization suc- cess of Mexico after 1988 and the possble success of Argentina since 1991 seem to bear this out But there is no real policy lesson here, since one would not advocate creating chaos or a hyperinflation in order to Table 12. Political Regime and Economic Growth Economic pefbrmanzce Stabl Unstabl Good kidiaa 3.0 Thailanda 53 Indonesia 33 Turkeya 2.9 Korea 72 Pakistan 29 Sd lanjJa 2.6 Poor Braila 02 Argentiraa -15 Cameroon -0.8 Nigeria -13 Costa Rica 0.6 Cote dIvoire -32 Kenya 0.4 Medcoa -1.0 a. Countries in upper half of Freedom House list. Not& Trend annual growth rabe in GDP per capita over the period 1980-1990 follows each country. Source LCCR, table 122. MACROECONOMIC POLICY 25 teach the community the need for macroeconomic discipline or com- mon sense. Those who determine govemment policy must be con- vinced of the correctness of responsible macroeconomic policies. In a democracy the relevant group may be the larger community or a limit- ed elite, and in an authoritarian state it may be just one or few individ- uals or perhaps influential members of the military. If high inflation is to be avoided, they must believe that inflation is undesirable and they must understand that financing budget deficits by monetization is likely to be inflationary; they must also understand that continuous debt-financing whether at home or abroad, wfll lead to trouble even- tually if the debt grows too rapidly. In short they mast apprecate the need for financial controL In those countries where macroeconomnic policy has, on the whole, been a success, there have usually been particular individuals in high government positions-above all the Ministry of Finance-who have been well qualified and who have been able to carry weight with the ultimate decisionmakers. They have been the actual makers and imple- mentors of policy. In ecent years these have often been professional economists. Such persons are increasingly available in many countries. Whether they hold high positions or not, their ability to persuade and to carry weight has been crucial. This has, among other things, de- pended on tradition, historical experiences, and the strength of pres- sure groups. In authcritarian states the personal views of the key individuals have obviously been decisive. A Brief Postscript History does not stop with the publication of a volume. A number of important eve itS have occunred since the events reported here. Some have already neen mentioned: Argentina in 1991 introduced a sifnifi- cant new stabilization program, accompanied for a change by tight fis- cal discipline. It was still hnlding three years later The CFA fianc was devalued by 50 percent in early 1994, affecting both Cameroon and Cobe d'Ivoire among our countries. The latter country also embarked on serious structural reforns in 1991. India in 1991 introduced a pro- gram of liberalization of trade, and of industril and financial controls. Colombia, Korea, and Mexico continued their previous programs of trade liberalization, and Mexico joined a free trade area with Canada and the United States in early 1994. Turkey, except for its high inflation rate a moderately good perform- er during te 1980s, ran into a serious fiscal csis in 1994. After a bold run at liberalization and stabilization by President Color in 1990, Bra- zils inflation resumed its acceleration, reaching over 2000 percent dur- 26 DOOM, CRISIS, AND ADJUSTMENr ing 1993. Another attempt at stabilization occurred in mid-1994. Other countries generally continued on the courses that had been established by the late 1980s. Policy Advice We can sum up our policy advice, derived from our findings in this study, in nine injunctions to policymakers: 1) Assert firm overall budgetary control and budgetary accountabil- ity including control over access of governmental entities or agencies to the central bank and to inteniational capital markets. 2) Be aware of the implications for the future of both external and domestic government debt; relate borrowing to returns on investment, to borrowing rates, and to the prospective growth of exports and of government revenues. 3) Be sensitive to the maturity of debt, both extemal and internal; maintain creditworthiness. 4) PFtrsuit of the first thee points will normaly permit the use of monetary and fiscal policy to help stabilize nonagricultural output around its long-nm trend. 5) Resist euphoria when export prices rise exceptionally, new re- sources are discovered, or new borrowing opportunities open. 6) Avoid jerky movements m the real exchange rate and thus pei- ods of over-valuation, which can have adverse effects on tradable goods industries. it is generally better to have a flexifble exchange rate regime, with frequent adjustments. Do not commit to a fixed exchange rate if there axe still extensive trade restrictions. 7) Avoid using import controls, except in extremis, and then remove them as soon as macroeconomic ircumstances permit; if pursued credi'bly, this stance will enhance the effectiveness of imnport controls m the rare occasions they are used. 8) Avoid building rigidities into the economy, such as persistent im- port controls or extensive wage indexation. 9) Maintain flexibility in policy, witiin an articulated medium-term framework, and in particular correct mistakes quicldy. NOTES 27 Notes 1. By Ian MID. Little, Richard N. Cooper, W. Max Corden, and Sarath Raja- patirana, Oxford University Press, December 1993, hereafter LCCL I am grate- ful to my co-authors for comments on this summary. Several of the first phase studies have also been published, or are in the process of being published. 2. Korea was added to the original list, making eighteen countries in all, because much material was available and Cooper was working independently on a book on Korea. 3. Table 5 shows that Cameroon experienced an adverse movement in its terms of trade in 1979-81. This is seriously misleading Oil, with rapidly rising prices from early 1979, was a new export for the country, while the statistcs reflect export-weighting of several years eariier: All the statistics reported here should be interpreted as rough indicators rather than accurate measurements. 4. Two-thirds of tiis extraordinary incease was caused by the 1981 devalua- tion of the colon, which reduced GDP in dollar terms without reducing the extemal debL The domestic burden of servicing external debt rises correspond- ingly follwing currency devaluation. 5. Three-fourths of the sharp increase in Chile's debt service, relative to CDP, was due to a deval-aation of the peso in 1982 6. Globally, net new private capital flows to developing countries fell from $74 billion in 1981, the peak year to a low of $27 billion by 1986. 7 Indeed, too little serious cost-benefit analysis of major public investments is undertaken even in normal times. 8 This formulation treats borrowing from the central bank as debt, on which interest should in princple be paid, even though under favorable circum- stances most of that will be retumed to the government in the form of earnings on seigniorage, discussed below. In fact, most goverrments do not pay inberest on tir debts to the central bank, so that portion of the debt should be excluded from this disumssion. 9. The notable exception was Brazil, where domestic debt was indexed, first (roughly) to the rate of inflation, then after 1982 to the exdchange rate, and a substantial portion of the govermmert deficit was financed by sales of securi- ties to the nonbank public. In this case the notion of operational (inflation-cor- rected) deficit made some sense, as the Brazilan government persuaded officials of the Intenational Monetary Fund. 10. Where oil or other export revenues accrue directly to the government, by the same token, currency devaluation results in a proportionate increase in domestic currency revenue. 11. Not reported in table 12 are tire- countries with mixed performance over the 1980s: Colombia (with growth in per capita income of 1.1 percent a year), a stable democracy; Morocco (1.9 percent), a stable authoritarian country; and Chile (1A percent), authoritarian until 1990. For use of Freedom House rank- ing, see LCCR, p. 363. World Bank Publications Order Coupon CUSTOMEPS IN TlHE UNITED STATES: CUSTOMERS OUrSIDE THE UNITED Complete this coupon and run to STATES: The World Bank Concact your local World Bank Publications Box 72474619 distnbutor for information on prices in local Philadelphia, PA 19170-8619 curmrncy and paymes terms. (See apposite page U.S.A. fora cmpletce list of distnbutors.) If no distnbtor is lised foryourcomuy, use this order form and Charge by acdit card by calling (202) 473-1155 retun i Lo the US. address. Orders tbatare sent or fax this completed order coupon to C202) 676- to the U.S. addres from countries with 0581. distributors wil be returned to the customer. Oyanriy Title Price Total Boom, Crisis, and Adjustment The Macroeconomic $42.99 Experience of Developing Countrlas (Full Text) Subtotal USS Postage and hmndling* USS _ TOt USS * If apurchase order is used. actual postage will be charged. If payment is by check or credit card, postage amd handling charges are USS5.00 per ordcr. For air mall delivery outsidee U.S, include USS8.00 for the first item and USS6.00 for each additional item. CHECK METHOD OF PAYMENT O Enclosed is my check payable toc Te World Bank. O Charge my 0 VISA O MasterCard 0l American Express Cedi card accoUnt nmober Expiratin daue Signatur O Bill me. Onsritiona! customers oly. Purchase order must be included) PLEASE PRITrCLEARLY Nlae FLnn Addrs City Statc Postal code__ Count_y Telephone Distributors of World Bank Publications ANCSNTINA DOMIICAN NUULC WYA SWGMEIUTAtAN, C..l_HuchSRL E;nTaCponA. AhflEoo.StMAz1AJLS. uYANIKuinRUNm I CobdaCawom R_bhowi..bsWb1C6la 0-au KIM uf Qt okAadPsdBdIEatI. Flaws 2041A RrO-C.4U/43 DS Pf.DSC45 GcSdlWhO k1dh UNIummeAl dowtmdffcz219>S Ndr 4a g cmmdna ddLb. kknaSoe xonk urnvucwrsgw ADfl4 PMEClook3UCOP Ir e Cipi SOUI AH CA. SOTSWANA ion u1s tAb AArma D.. D ,mKam agbtas Frsbghhlsm M.CAUSue Sal wind Untty Fra. AUSTRALL%PArUA NEW C*u SusIn Alu GIINRA. FIL FWMON KmSl.k ba Cale PJD. O 1141 ISLAIJS.VAMIIATII ThitMWikat&!s m PANOKa Caps?... mm AM WuSiM SAMOA. 41.96u .l YS DA.halandlaSavm Cair seat r _r_a*ais wr E41W11f.IeNbe tansieS IubelpdafSavlin Xi s _3I FINLAND MALAYS PJD. sn 4 Whims Ahanurnsm Ki4.haappe usW.saysl a.tq.ea.w lu ros-1 AUSrnnA Sw4Ot HtddakflO wnua.nieinva Was CAdat da"it SMM Kal Lap SPAIN Cmj FRANC Mm.PrUUS lbA S5. A1021 Win Weld D&NWa-Ls mU CSo37 Mia Cib; IOC 231 yid ZANGLADGSI 7Sn5P15i ApinindoP1bat n4V kUu"bdshu.dwdqnt400lIaImn.MawDZ LIflwlr'u.mdandAEDS MamaS. Dad 16 LHNO.Yukg PNan*L&Ais DMn Beanclo Oha..edl 74Arm Npp&odafsAI.eS IhUS*a/mkdkflf Oaha tW 9115OMM FJD.se3 SRI LANtKA AED TEM MMADIYS 7WAE Hasbobe LuleHamook"aha Bmk*a&w Gnmica ~~~~~~~~~~~~~~~PXD. Rm 241 Flavkw.20s urs FIpSSdSIJLS NEWZALAJD 10S SI CaiBUp&.ImA. 1W'6mbd GmmWn 35.91ina,u Sh. USWPL6 Cutlas MeanIla OSSsgmg4iW IOlI2AlsPrlVaMOB3ag994 tadomAlz 75. KDAAvts 110GN KNr. MACAC AnOa SWEDNN 45-tWyudam Sun Nwcaa Foibsftkradmisqs I=ICUN WkiarnI caitn USeeflyhPai L.-bo 3qfgwqpqpip fE gm356 ]ms DeL army 7th Hlow ThinCzmm6IuOd1SSh S.ml &32Sw&dAS AwAv.dotlZ CauulM Kegrmg hive"i BEqSUS FWCARY4 WmwtpewM1aAB UKAZI FwadSanfeMwbS bamony NORLwAy P.0.11. IN PuMIaco..T.micd Do.ubov=IUt 4-9 Nmnmmnns1ad CuSs 5.271 255dm N.. ieahoCood,.fl FaDsZ antmu SWITZERLAND OIS. salamuS? INDI samW.'F~a-jhm AXIS NUbds. Pies LiLmaiit Ub.PqUL CANADA 132 MastSm nAruue Caepofimsiar LaDJuas MMorrn-G Mina*Am CHtWZtamwaw Uaacb.v1b.Qaa ~ INDONEUKAPO a. 5fabfia.. 343 SM FL~~~~~P 6dbita IsE [e51 Lbrirlftyat uin.anuaagc.. P0.1m Bam i CapesSw4mAbon t UN4 Auma. Road J&amb1 UAdSIU CH2ISIZljuen Oftaw*lObnoflRt3W Apnd.324N OWEA tubal ItalilILAN Cabal THIADeps'S CH.A Haadl Famnc aD. IflS.Sfs p,uwrnas Ro PuS'Ham Tarn hamadm gUka&CI Ia'gi.k UPSg ItSLAND CnSdamiisuaTwel TRINIAD &TDUACO. CwmaqsAg-cy AySmAiuma Car FLV. ANTIsncArBAUDA. COLONSIA l-Haacct Rod disC=s. anda sNAADat,0voIiu.n M lx a. CRUNADA. GUYANAl. Ap.tadmAwin03WD JAMAICA.MONTSUBRAr, Sr. Iqeis Di. ISRAELL POLAND KfTTIA; NDVISr.S LUCIA. YotanisabnuLid. blepsia dm Swvim MVINCf ; &CfNADI?U5 co olzn FAD.. 105.AUma.2/37 Syd=MmShidSdL1ua ld`Edt i DLmm TSmIr 5S60 3.07 sWmu #g9 Wit. Su Abbe.- ~ ~ ~ ~ ~ ~ ~~~7A EE-IHIUi 043.7.5O ITALY FThiat We ladke AbldJmOIFIueu CX dSa.al. Ssm-o Iflj 9 SPA 1LLa3 UNTW XSNCOO CYPRUS viODUBDmCea inm.ernwaawa Mdab(E L CS dApplIw Dadi% Cmal&baj.fl Fnl u33 C1.aCdllq SM25FIra, POErUGAL Aw. Hevpda .Lh342PC 6.Uqsnw Souse lrqgmd lAwS. PSuaga !4a5 PAX Boa flD JAMAICA lo DoCa.e7O-4 Pike.l Z Inl RS.puMMuesIJtiHoMs Z WI OUNkLMAKKrSn SAUDIAARABA.QATAR 'Teak RmdArdbmvde Srnlwad.UIimSr J~~~~~~~~~lrfumkakSin PFIX gm Sr113 RinmoemADifl JAPAN P.OA DaO. LItS pCM D dU490FdebgC Nas DeoSueHn2 Hohgo 3.Omm.Sawank2Aa 113 3 Takyo Wor1d wBank Comparative Macroeconomic Studies Boom, Crisis, and Adjustment: The Macroeconomic Experience of Developing Countries (Fuli text) 1993. Stock No. 60891. 454 pages. $42.95. Other related titles: Macroeconomic Policies, Crises, and Long-Term Growth in Indonesia 1965-90 1994. Stock No. 12212. 230 pages. $13.95. India: Macroeconomics and Political Economy 1964-1991 1994. Stock No. 12652. 416 pages. $24.95. Macroeconomic Policies, Criszs, and Growth in Sri Lanka, 1969-90 1994. Stock No. 12297. 192 pages. $11.95. Economic Crises and Long-Term Growth in Turkey £993. Stock No. 12298. 152 pages. $8 95. To o(-tlei; -;please utse tlhefiolo ing>> cou0pon/1 World Bank Publications Order Coupon I CUSTOMIERS IN THEUSA: CUSTOMERS OUTSIDE THE USA: Complete this coupon and return to Contact your local World Bank Publications The World Bank distributor for information on prices in local Box 7247-8619 curnency and payment terms. If there is no Philadelpha, PA 191704619 distributor for your country, use this order form USA. and return it to tie US. address. Orders that are sent to the US. address from countries with Charge by credit card by calling (202) 4734155 distributors will be retured to the customer. or fax this completed order coupon to (202) 676- 0581. Qty Stock# Tide Price Total * If a purchase order is used, actual postage will Subtona USS be charged. If payment is by check or credit card, postage and handling charges are Postage & handling* US$ US$5.00 per order. For air mail delivery TOnl US$ outside the U.S., include US$8.00 for the first item and US$6.00 for each additional item CHECK METHOD OF PAYMENT O Enclosed is my check payable to The World Banlk O Charge my 0 VISA 0 MastreCard O American Express Ciedit card aiountt nuniber Expirationdate Signature CO Bill me. (Institutional customers only. Purchase order must be included.) PLEASE PRINT CLEARLY Name Firm Address city State Postal code CoWIWy Telephone