C~~~~~~~~~~~~r Lf) ~~~ ~ r a W 0 R L D B A N K C OM PA RAT I VE MA C RO E C ON OM IC STUD IES Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 P RE M A CHAND RA ATHUK O R A LA S I S I RA J AYA S U R I YA THE WORLD BANK, WASHINGTON, D.C. © 1994 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, N.W., Washington, D.C. 20433 All rights reserved Manufactured in the United States of America First printing August 1994 The World Bank Comparative Macroeconomic Studies series emerges from a research project that reviewed the macroeconomic experiences of eighteen developing countries over a period roughly from 1965 to 1990. So that the studies might be published with relatively little delay, the books have been edited outside the World Bank's Office of the Publisher by the Macroeconomic Research Department. The findings, interpretations, and conclusions expressed in this publication are those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to the members of its Board of Executive Directors or the countries they represent. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Dr., Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'1ena, 75116 Paris, France. Cover design by Sann Ferro. Premachandra Athukorala is a senior lecturer and Sisira Jayasuriya is a reader in economics in the Department of Economics at La Trobe University, Australia. Both were consultants to the World Bank for the writing of this book. Library of Congress Cataloging-in-Publication Data Athukorala, Premachandra. Macroeconomic policies, crises, and growth in Sri Lanka, 1969-90 / Premachandra Athukorala, Sisira Jayasuriya. p. cn. - (The World Bank comparative macroeconomic studies) Includes bibliographical references and index. ISBN 0-8213-2297-4 1. Sri Lanka-Economic policy. 2. Sri Lanka-Economic conditions. 1. Jayasuriya, S.K. 11. Title. 111. Series. HC424.A894 1994 339.5'0954893-dc20 94-31612 CIP Contents Foreword vii Acknowledgments ix Acronyms x Data Notes xi Chapter 1. Introduction I Chapter 2. Sri Lanka's Political and Policy History 4 The Arrival of Independence 6 The Political Economy of Consumer Subsidies 7 An Open-Market Economy: 1948-56 9 The Shift to Import Controls: 1956-65 10 Partial Liberalization: 1965-70 13 Regulation and Controls: 1970-77 17 The Turnaround in Economic Policy after 1977 20 The Rise of Ethnic Hostility 23 Chapter 3. A Profile of the Economy 25 Land and Resources 27 Growth and Welfare 28 The Structure of the Economy 29 Savings and Investment 31 Labor Force, Employment, Unionization, and Wage Formation 32 Trade Unions and Wage Formation 34 Commodity and Factor Markets 35 The Role of the State 46 Foreign Trade and Balance of Payments 51 Chapter 4. The First Crisis: 1973-75 56 Background to the Crisis 56 The Crisis 62 The Policy Response 65 The Pattern of Adjustment 71 vi Macroeconomic Policies, Crises, and Gro wth in Sri Lanka, 1969-90 ChapterS. The Second Crisis: 1978-82 76 The Background of the Crisis 76 The Investment Boom and the Terms of Trade Decline 79 Capital Flows, Income and Expenditure Effects, and Inflation 88 Chapter 6. Macroeconomic Policies and Long-Term Growth 94 Investment and Growth 94 Savings Behavior 96 Private Investment 98 The 1973-75 Crisis: Long-Term Effects 99 The 1978-82 Crisis: Long-Term Effects 103 Policy Responses, Income Distribution, and Political Stability 106 Chapter 7. Summary and Conclusions 112 Terms of Trade Fluctuations and Optimal Policy 112 Welfare Expenditures: Not the Key Issue 113 Macroeconomic Imbalances, and Trade and Exchange Rate Controls 114 Change and Continuity in the Post-1977 Period 115 Large Projects, Macroeconomic Stability, and Growth 115 Sri Lanka's Experience in a Comparative Setting 116 Conclusions 117 Appendix 1. 119 Determinants of Domestic Savings 121 Determinants of Private Investment 123 Short-Run Supply Response of Exports 124 Appendix II. Notes on the Statistics 129 Sources 129 National Accounts 130 Employment Data 130 Trade Data 131 Fiscal and Monetary Data 131 Wages 131 Price Data 132 Statistical Appendix Tables 134 Notes 158 Bibliography 161 Index 173 Foreword This volume is the product of a World Bank project on macroeconomic policy that reviewed the recent experience of eighteen countries as they attempted to maintain economic stability in the face of international price, interest rate, and demand shocks or domestic crises in the forms of investment booms and related budgetary problems. The project paid particular attention to the 1974-79 period (which in- cluded the first and second oil price shocks), the 1980-82 period of worldwide re- cession and external debt problems for many developing countries, and the 1983- 90 period of adjustment to economic difficulties and the resumption of growth. The objective of the project was to glean instructive lessons by analyzing the stabilization and adjustment policies pursued by these countries and assessing the outcomes. The authors of each country study were asked to deal with a common set of questions concerning the nature of the shocks or crises: their origin and de- gree of seriousness; the fiscal, monetary, exchange rate, and trade policies adopted in hopes of preventing permanent harm to the economy; and the results of the policies. No single computable macroeconomic model was used in the project, but the framework of the open-economy macroeconomic model was followed to ensure consistency in generalizing about results. This intensive study of many episodes generated ideas and suggested relationships showing the causes and effect behind policies, the nature of the shocks and crises, and the governmental response to them. The overall findings of the project are presented in a synthesis volume by I. M. D. Little, Richard N. Cooper, W. Max Corden, and Sarath Rajapatirana, Boom, Crisis, and Adjustment: The Macroeconomic Experience of Developing Countries. This study examines the macroeconomic policies of postindependence Sri Lanka, with emphasis on the management of shocks. In contrast to much of the literature on developing economies, which focuses on either long-term structural changes or short-term macroeconomic stabilization issues, this study links macro- economic policies to long-term growth. A major strength of the study is that eco- nomic policy changes are analyzed in the context of broader social and political circumstances. Some aspects of Sri Lanka's development policies have been unique among developing countries and have attracted worldwide attention. In particular, its remarkable achievements in advancing social welfare-achievements comparable vii viii Maeroeconomic Policies. Crises, and Growth in Sri Lanka, 1969-90 to those of many developed countries-have generated a large literature. Howev- er, there has been no comprehensive analysis of Sri Lanka's macroeconomic pol- icies and their implications for welfare and growth. This study fills this gap. The study offers an overview of the postindependence economic and political developments in Sri Lanka and a new perspective on the interaction between mac- roeconomic policies, political and social stability, and long-term growth. It chal- lenges widely held views on the relationship between welfare expenditures and growth. The study focuses on the policy experience of two periods (1973-75 and 1978-82), periods when the economy was subjected to major domestic and exter- nal stresses. Between these two periods there was a dramatic shift in government policies. A long period of extensive state intervention and inward orientation was followed by a commitment to economic liberalization. The analysis is of broader interest as a study of the reaction of a small open economy to major shocks under sharply contrasting policy regimes and external circumstances. Publication of this study was made possible in part by a generous grant from SIDA, the Swedish International Development Authority. Sarath Rajapatirana Director. "Macroeconomic Policies, Crisis, and Growth in the Long Run" Research Project Trade and Industry Policy Adviser Technical Department Latin America and the Caribbean Region Acknowledgments We accumulated many debts during the course of this research project, and we ac- knowledge some of them here. Our major debt is to Richard Cooper, Max Corden, Ian Little, and Sarath Rajapatirana, who provided guidance, advice, and support throughout the duration of the project. Max Corden and Sarath Rajapatirana de- serve special thanks for detailed comments and long discussions; they also extend- ed warm hospitality in Washington, D.C., as did Ian Little at Oxford. Vittorio Corbo and other participants at the Madrid and Mexico City workshops made many useful suggestions. Professor W. D. Lakshman generously shared his in- sights and made available valuable data and material, including unpublished manuscripts. Phil Sawicki gave much appreciated editorial assistance. Anita Bhatia insulated us from administrative and logistical problems, assisted by Mar- gert Kinzle and later by Charles Dade. Marianne Gierveld cheerfully transformed illegible scribbles into readable drafts; Sandra Barnes, Mercy Joseph, and Ann Loutitt also helped with secretarial services at various stages. Despite the pres- sures we placed on our families, they were generous with patience and love. This study is dedicated to them. ir Acronyms CBC Central Bank of Ceylon CBCPI Central Bank Consumer Price Index CCI Construction Cost Price Index CCPI Colombo Consumer Price Index CFL Ceylon Federation of Labor CFTU Ceylon Federation of Trade Unions CGE Computable General Equilibrium CMU Ceylon Mercantile Union CP Communist Party CRA Convertible Rupee Account CWC Ceylon Workers Congress DCS Department of Census and Statistics DFCC Development Finance Corporation of Ceylon DWC Democratic Workers Congress EDB Sri Lanka Export Development Board EPF Employee's Provident Fund FCBUs Foreign Currency Banking Units FEECS Foreign Exchange Entitlement Certificate Scheme FP Federal Party GCEC Greater Colombo Economic Commission GDFC Gross Domestic Fixed Capital GDP Gross Domestic Product GDPD Gross Domestic Product Deflator GDPR Real GDP Growth GNP Gross National Product GOBUs Government-Owned Business Undertakings ICC Insurance Corporation of Ceylon ICORs Incremental Capital/Output Ratios IDRC Canadian International Development Research Corporation ILO International Labour Organization IMF International Monetary Fund IPZ Investment Promotion Zones JVP Janatha Vimukthi Peramuna, or Peoples' Liberation Front LSSP Lanka Sama Samaja Party .Y Acronyms xi NSB National Savings Bank OGL Open General License PMB Paddy Marketing Board PSEs Public Sector Enterprises PSIP Public Sector Investment Program PPP Purchasing Power Parity REER Real Effective Exchange Rate SDR Special Drawing Right SLFP Sri Lanka Freedom Party TDR Trade Dependence Ratio TULF Tamil United Liberation Front UF United Front ULF United Left Front UNP United National Party WPI Wholesale Price Index Data Notes • Dollars are current U.S. dollars unless otherwise specified. • The symbol - in tables means not available. Chapter One Introduction In this book we examine the macroeconomic policies of postindependence Sri Lanka, with emphasis on the period from the mid-1960s to the late 1980s. We fo- cus on domestic and external shocks which subjected the economy to major stress- es during this period, and discuss and analyze the specific policy responses and their consequences. These episodes are placed in their historical, economic, and political contexts, and we examine the institutional and ideological influences on policy formation and the responses of economic agents. In contrast to most anal- yses, which concentrate solely on short-term stabilization aspects, we also attempt to determine the impact of macroeconomic policies on long-term growth. The overall study period encompasses two shorter periods (1973-75 and 1978-82) during which the economy was subjected to major shocks. Both periods show striking similarities and sharp contrasts in terms of the nature and origin of the shocks, and the policy responses. Although these two subperiods were remark- able for the intensity and magnitude of the shocks, in a sense the entire study pe- riod can be considered to have been an era of turmoil. Since the 1970s, Sri Lanka has experienced large fluctuations in its terms of trade, large domestic supply shocks, a dramatic investment boom, and huge capital inflows. Economic policies have ranged from strong state intervention in almost every important factor and product market to a major trade liberalization. Although the economy escaped se- vere recession, a major inflationary crisis, and a large external debt problem, eco- nomic and social developments were often erratic and unpredictable. And as the 1980s ended, Sri Lanka was in the grip of violent political conflict and its econo- my was in serious trouble. We believe, therefore, that this analysis holds lessons which have broad relevance for many other developing countries. Sri Lanka's economy shares many structural features with most low-income commodity exporting countries, but there are some important-even unique-dif- ferences that distinguish the Sri Lankan macroeconomic policy setting. The evo- lution of the economy up to the early 1960s was the subject of a widely cited study by Snodgrass (1966), and in the 1970s Sri Lanka was considered to be "one of the 2 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 most heavily researched of the developing economies of the world" (Pyatt and Roe 1977:23). In particular, Sri Lanka's record in achieving standards of literacy, health, and life expectancy comparable to those of many industrial countries de- spite being one of the world's poorest countries in terms of per capita income (US$470 in 1990) has generated a large and notable literature (for example, Isen- man 1980; Sen 1981; BhalIa and Glewwe 1986; Anand and Kanbur 1991). Recent studies have also included substantive analysis of trade and industrialization pol- icies (Lal and Rajapatirana 1989; Cuthbertson and Athukorala 1990). However, no studies of Sri Lanka's macroeconomic policies as such exist, despite their ob- vious importance to trade, industrialization, and welfare spending. We hope this study will help to fill the gap. In this study we analyze the policy responses to macroeconomic crises and their consequences, as well as why those, rather than alternative policies, were adopted. This involves an analysis of political economy factors as well as an ap- preciation of the perceptions and expectations held by policymakers and private agents as they confronted actual or potential crisis situations. The focus of the study, however, is on analysis of the changing economic situation, and no attempt is made to provide political economy explanations. In chapter 2 an overview of political and economic history since indepen- dence is presented. This review of the evolution of the postindependence economy is used to identify the 1973-75 and 1978-82 periods as two critical episodes for detailed analysis. The key structural and institutional features of the economy are described in chapter 3 to provide a background for the analysis to follow. Our choice of the two periods-which we will refer to as crisis episodes- was guided by the unprecedented stresses to which the economy was subjected at those times, rather than to any catastrophic collapse. As will be described in detail later, the 1973-75 period saw Sri Lanka confronted by a combination of acute do- mestic supply problems and a dramatic increase in the price of its major imports- oil, rice, wheat, and sugar. Limited trade liberalization in 1968-70, sluggish ex- port performance, and a 22 percent decline in the terms of trade during 1966-70 generated a mini-crisis in 1970 to which the government responded with stringent quantitative controls on imports. Thus, the 1973 shocks hit an economy that was already severely stressed. On the other hand, during the 1978-82 period the econ- omy was stimulated by a massive increase in investment, originating mainly in the public sector. Gross domestic investment rose from 16 to 32 percent of gross do- mestic product (GDP) within two years, and this investment boom coincided with a drastic 62 percent deterioration in the terms of trade between 1978 and 1982. But in contrast to conditions at the onset of the 1973-75 crisis, external reserves in the late 1970s were very healthy, and a major policy liberalization had opened up the economy. The response of the government to the investment boom involved no significant retreat from a liberal policy orientation, although the pace of liberaliza- tion slowed down. In chapters 4 and 5 the nature of the 1973-75 and 1978-82 macroeconomic crises, the policy responses, and their immediate impact are discussed. Chapter 6 Chapter One 3 presents a discussion of the consequences of policy responses to these crises for long-term economic growth. The key findings of the study and their implications are summarized in chapter 7. Chapter Two Sri Lanka's Political and Policy History Sri Lanka (commonly called Ceylon until 1972) is a tropical island located south of the Indian subcontinent; at its closest point to the mainland it is less than 50 ki- lometers from India. It has a land area of 65,610 square kilometers and in 1989 had a population of 16.8 million. Coastal plains and central mountains provide a wide range of environments. There is little seasonal variation in temperatures. A southwestern monsoon brings rain from April to June, and a northwestern monsoon visits from mid- October to mid-February. On the basis of variations in annual rainfall, the country can be divided into a wet zone, a dry zone, and an intermediate zone. The south- western region is the wet zone, with annual rainfall averaging 60 to 80 inches in coastal areas and 120 inches in the central hills. Traditionally, tea, rubber, coconut, and rice (paddy) have been the main agri- cultural crops. Until recently, the three plantation crops (tea, rubber, and coconut) were the major export earners. In the 1970s exports of gems and garments expand- ed rapidly, and tourism also became a major source of foreign exchange earnings. The reliance on primary exports has exposed the Sri Lankan economy to severe terms of trade fluctuations and, in the post-World War II period, to a long-term de- cline in its terms of trade. Recurrent balance of payments problems have plagued the economy in recent decades. Sri Lanka's recorded history reaches back at least 2,500 years. l Sinhalese and Tamil, the languages of the two largest ethnic groups, have a common Indian ori- gin, belonging respectively to the Indo-Aryan and Dravidian linguistic families. According to historical and archeological evidence, today's ethnic identities evolved over time through the intermingling of many waves of migrants from dif- ferent parts of India, who were absorbed into the dominant linguistic groups. However, myths about the racial origins of these two ethnic groups, which have become entrenched over time, play an important role in Sri Lankan political and social life. The Sinhalese trace their ancestry to a group of North Indian "Aryan" settlers who are believed to have arrived in Sri Lanka around 500 BC, while the 4 Chapter Two 5 Tamils trace theirs to the ancient Dravidian civilization of South India. The early settlements were located in the dry zone, with rice (paddy) as the main crop. Bud- dhism became the bedrock of Sinhalese culture and civilization, while Hinduism established itself as the religion of the Tamils. The north-central plains were the cradle of the ancient Sri Lankan civiliza- tion, which was based on a highly sophisticated system of rice irrigation. Thou- sands of man-made reservoirs dotted the landscape. Many were linked by canals and fed from large central reservoirs that sometimes covered many square miles. Indeed. Sri Lanka became. in time, one of the great irrigation-based civilizations of the ancient world (Farmer 1957). This civilization, which survived for over a thousand years, succumbed to a combination of factors. Repeated invasions from South India, civil wars, buildup of soil salinity, and the spread of malaria drove the bulk of the population to the wetter, hilly, and more densely forested southwest part of the country in the 13th century. A number of regional kingdoms, including a separate Tamil kingdom in the north, were established during this period. Western influence on the country began with the arrival of the Portuguese in 1504 and their establishment of rule over the coastal regions. In the historical past, Sri Lanka had traded in spices, gems, and elephants with the Greek and Roman empires in the West and the Chinese empire in the East. At the time of the arrival of the Portuguese, it had a significant trade in spices with Europe. Portuguese rule stimulated the spice trade and initiated a process of modernization of the economy. This process accelerated over the next three centuries as the Portuguese were dis- placed by the Dutch in the mid- 17th century, and the Dutch in turn were replaced by the British at the end of the 18th century. Throughout this period, much of the interior of the country remained independent, the capital having shifted to Kandy. Given the recent racial conflict in the country. it is worth noting that the royal families of Sri Lanka were closely linked by marriage and family ties to South In- dia, and that the last three "Sinhalese" kings at Kandy were ethnic Tamils. In 1815, Sri Lankan aristocrats, dissatisfied with the king, carried out a coup d'etat and signed a treaty whereby sovereignty was transferred to the British, who in turn promised to protect and maintain traditional customs and rights, including the Buddhist religion. This enabled British rule to be extended over the entire country and ushered in a period of rapid change that soon resulted in the emergence of a classic dualistic export economy (Snodgrass 1966). Coffee, and then tea and rubber cultivation spread rapidly during the 19th century, and modern plantations were established throughout the wet zone. Popu- lar resistance to the plantation economy. which encroached on subsistence agricul- ture, erupted into armed rebellion in 1847 and was brutally crushed. The way was then open for the unfettered expansion of the plantations. By the end of the 19th century the economy was dominated by exports of tea, rubber, and coconut prod- ucts.2 The development of the plantation economy was accompanied by the devel- opment of an extensive road network. South Indian labor was brought in to provide a resident labor force on the plantations. This, together with the break-up 6 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 of the traditional subsistence economy, made the country dependent on imports of its staple, rice. By the time it achieved independence from Britain in 1948, Sri Lanka was a highly trade-dependent country, exporting plantation crops and im- porting rice, wheat flour, and sugar, as well as manufactured consumer goods. The Arrival of Independence Sri Lanka became independent in 1948. Conditions at the time of independence appeared benign. The transfer of power was smooth and peaceful, the population had enjoyed universal suffrage since 1931, and an efficient administrative mech- anism was in place. Since 1932, Sri Lanka had been operating under the so-called Donoughmore constitution, which provided a large measure of self-govemment and provided a framework within which even avowedly revolutionary movements were able to carry out open political activities and participate in the electoral pro- cess. Despite considerable wartime stresses, the economy had not suffered any se- rious dislocations and provided a standard of living in 1948 that was well above that of neighboring countries. In contrast to the violence and turmoil that gripped the Indian subcontinent in the immediate postwar years, Sri Lanka was an "oasis of stability, peace and order" (de Silva 1982). A leading Sri Lankan politician was expressing a widely held view when he stated that, of all postcolonial nations, Sri Lanka would prove "the best bet in Asia" (Jiggins 1976). But the peacefulness and order of the political and social scene in 1948 were deceptive. They masked deep underlying divisions among social ciasses and eth- nic groups. Politically, society was polarized between the conservative nationalist right and the marxist-oriented left to a degree almost unknown elsewhere in the Indian subcontinent, and there was much industrial strife that erupted into general strikes in 1946 and again in 1947. Ethnic antagonisms were also present, particu- larly between the Sinhalese nationalists who dreamt of a Sinhalese Sri Lanka and the minority Tamils, who feared that British rule would be replaced by Sinhalese rule. These ethnic tensions had been pushed to the background by political polar- ization and the anticolonial struggle. All the main political parties were multiracial and included members of minority ethnic groups in their leaderships. At the general election held in 1947 to elect members to the postindependence parliament, the conservative United National Party (uNP) emerged as the largest party. Yet, despite an electoral gerrymander which gave disproportionate weight to politically conservative rural areas, the UNP was unable to obtain a clear major- ity, winning less than 40 percent of the vote. But it was able to form a govemment led by Don Steven Senanayaka, with Solomon Bandaranaike as his deputy, in a coalition with other conservative groups. From the outset, however, the UNP faced a formidable challenge from the left, which was divided between Trotskyists and a smaller pro-Moscow communist party (cP). The Trotskyists themselves were split between the Lanka Sama Samaja Party (LSsP) and the Bolshevik-Leninist Chapter Two 7 party, a division that lasted for a number of years. Despite this division, the orga- nizational strength and political stature of the left effectively denied the UNP the kind of political dominance enjoyed by similar leaderships elsewhere, such as the Gandhi-Nehru alignment in India. The peacefulness of the transfer of power from the British to the conservative nationalists deprived the latter of the mass adulation accorded to nationalist lead- ers in other countries who had won independence through armed battle with colo- nial rulers. The UNP leaders had been part of the govemment during the war years and had fully supported the British. The Trotskyists had opposed the war, although the small pro-Moscow communist party had supported it and its members had suf- fered repression and imprisonment. This gave leaders of the Trotskyite left a high degree of popular support that enabled them to emerge as contenders for power even within the parliamentary framework. The Political Economy of Consumer Subsidies The economic problems of the 1930s provided fertile ground for the marxist left to grow rapidly in strength and influence, and leftist agitation over economic hard- ships compelled the conservative nationalists (later to become the UNP leaders), who were partners in governing, to exert pressure on the British colonial adminis- tration to enact measures to alleviate the worst effects of the depression. The im- plementation of a series of social welfare measures in the 1930s extended to the broader populace measures originally taken to help plantation workers and be- came the precursor of the extensive welfare system that developed in the postin- dependence period; this was a conscious response to political circumstances (Corea 1973). Health facilities were greatly expanded, as were educational insti- tutions. An extensive system of food subsidies was adopted during the war years and was intended in part as a way of meeting the leftist challenge (see de Silva 1982). This system of arrangements, which could already lay claim to being a wel- fare state, was inherited by the new government in 1948. Arguably, violent social conflict had been avoided in Sri Lanka because of the welfare system, and the in- fluence of the left had been contained. Hence, the govemment in the immediate postwar years had a strong desire to further expand welfare services and thereby deprive the left of popular support. But deteriorating economic conditions in the immediate postindependence period forced the new govemment to prune the wel- fare system, although the subsidized rice ration was maintained. Thus, the possibility of a potentially victorious electoral challenge from the left became a real possibility until the Korean war produced a commodity price boom, led by a surge in the price of rubber. Between late 1949 and early 1951, the prices of Sri Lanka's exports rose almost 40 percent and the country's terms of trade rose by over 30 percent. External reserves improved rapidly, as did the do- mestic money supply. The government removed certain import and exchange rate 8 Macroeconomic Policies. Crises, and Growth in Sri Lanka. 1969-90 restrictions and the newly established (August 1950) Central Bank began to grap- ple with the inflationary pressures arising from the export boom. The boom was temporary, but it had lasting effects through the changes it helped to bring about in government transfers. In October 1950 the government reintroduced a subsidy for consumer purchases of wheat flour and increased the subsidy for rice. The subsidy for rice was increased again in December. These were initially seen as measures to compensate for cost-of-living increases. By late 1950, rice was being marketed at less than half of its landed cost to the govemment (Karunatilake 1974). In 1950-51, subsidies were estimated to be about a fifth of all government expenditures (Corea 1965). By early 1952, expenditures on rice and flour subsidies exceeded expenditures on development projects (Central Bank of Ceylon 1951). In addition to these food subsidies, there were other transfers, which included the provision of free health care and free education up to the ter- tiary (university) level. Export prices declined rapidly in 1951. and in 1952 there was a trade account deficit of 297 million rupees, a sharp departure from the surplus of 238 million ru- pees the previous year. By the end of 1953 total foreign assets were only about half their 1951 level (Snodgrass 1966). At the same time, import prices were rising. By 1952, imported rice cost more than three times its subsidized price in Sri Lanka's markets. The government's budget deficit then expanded to reflect the increased cost of subsidies. These welfare expenditures brought about a massive electoral victory in July 1952 for the UNP, now led by Dudley Senanayaka, the son of Don Steven Senan- ayaka. The government had pledged during the election to maintain consumer subsidies, but in 1953 it found itself facing the choice of reducing food and other subsidies or sharply curtailing other fiscal expenditures. It opted to raise the rice price as well as the prices of publicly provided services, such as transport, postal, and electricity services. Other subsidies, including free meals for schoolchildren, also were cut. The political reaction was explosive. On August 12, 1953, the country was on the brink of revolution as the Left mobilized a massive public protest known as the Hartal, expressed in the form of a general strike. Despite the nonparticipation of the moderate opposition (now led by Solomon Bandaranaike, who had left the UNP), this mass protest shook Sri Lan- kan society to its roots. Originally conceived as a one-day protest, the Hartal erupted into a massive and violent demonstration of opposition to the government. Prime Minister Senanayaka resigned, and the price of rice was immediately re- duced and the ration was enlarged. But these actions did not save the UNP from a landslide defeat at the next general election in 1956, and "the political commit- mentto food subsidies was sealed once and for all" (Snodgrass 1966:194). For the next quarter century, whenever the issue of food subsidies came up, the specter of the Hartal haunted every Sri Lankan government. Chapter Two 9 An Open-Market Economy: 1948-56 During the UNP regime of 1948-56, Sri Lanka maintained an essentially open-mar- ket economy. The new government had begun with certain import and foreign ex- change restrictions originally imposed by the colonial government in the late 1940s to regulate a rundown of Sri Lanka's external balances in the United King- dom. In 1950 and 1951 most of these restrictions were removed, aided by an im- proved balance of payments positi6n due to the price boom. The remaining (and minimal) exchange and import controls were limited to transactions with coun- tries in the nonsterling area, particularly the United States. At the time of independence in 1948, Sri Lanka had a fixed exchange rate and its currency was fully convertible with the British pound (through its link with the Indian rupee). The Currency Board System of the British colonial government did not permit any independent monetary policy, and deficit financing was impossible. The money supply adjusted automatically to fluctuations in the balance of pay- ments (Corea 1965; Gunasekara 1962). One of the first acts of the new govern- ment in 1948 was to obtain advice on the setting-up of a Central Bank from John Exter, an official of the U.S. Federal Reserve System, who had also provided ad- vice on establishing central banks to the Philippines and South Korea. Sri Lankan exchange rate policies continued to be closely tied to those of the United King- dom, and Sri Lanka devalued the rupee against the dollar following a devaluation of the pound in 1949. The UNP had laid much stress on agricultural development. From the 1930s, its leaders, particularly Don Steven Senanayaka, were strong proponents of reviv- ing domestic food crop agriculture (based on irrigation schemes) in the now sparsely populated dry zone. This strategy involved large-scale public sector in- vestment in irrigation-cum-land development projects and the resettlement of landless people from the densely populated wet zone, who were primarily Sinha- lese (Farmer 1957). It was hoped that the outcome would be the creation of a pros- perous and contented farming population. Direct state intervention to help the development of manufacturing industries was not a policy priority. Indeed, a number of inefficient industries that had been set up during the war years were closed down. The UNP government saw its role mostly as one of encouraging private enterprise. Although it introduced a modest program to revamp some state industries set up during the war years and estab- lished some new factories to make Sri Lanka "less dependent on imported manu- factured goods" (Amarasinghe 1979:19), the commercial failure of most of the undertakings resulted in a gradual disengagement of the state from commercial and industrial ventures. In 1955 the government passed a law that provided for the conversion of government-owned industrial enterprises into public corporations as the first step in a long-term plan to transfer them to the private sector in the form of joint stock companies. Some state industrial undertakings were converted into public corporations under the law, and some were closed down, before the change in government in 1956 stopped further implementation of the privatization plan. 10 Macroeconomic Policies, Crises. and Growth in Sri Lanka, 1969-90 Two development "plans" were prepared during the early 1950s-a Six-Year Development Plan (1951-57) issued in 1950, and a Six-Year Program of Invest- ment (1954-55 to 1959-60). The former was merely a tied-up version of the bud- get speeches for 1947-48 and 1948-49, and was prepared only because the launching of the Colombo Plan required the government to submit a development plan to be used as the basis for aid negotiations. It was therefore "purely a pro for- ma exercise and probably had no influence on policy" (Snodgrass 1966: 110). The latter plan was nothing more than a synthesis of the proposals made by a World Bank mission that visited Sri Lanka in 1952 (World Bank 1953). Both of these documents placed heavy emphasis on investment in agriculture and infrastructure with a view to providing the basis for the private sector to play the leading role in the development process. Being merely investment programs of the public sector, they did not include overall growth targets for the economy. In the area of macroeconomic management, fiscal deficits (financed mainly by borrowing from the Central Bank) became the norm from the early 1950s on- wards, except for a brief period between 1954 and 1956. The deficit was the result of the rising cost of providing welfare services, in particular the supply of subsi- dized food, in the face of nonbuoyant government revenues. Government reve- nues came mostly from tariffs, and export duties moved upward after 1953. But these export duties were inadequate to bring about a substantial improvement in government revenues. Meanwhile, import duties changed little until the late 1950s. Although imports increased during the decade, the higher revenues from import duties were counterbalanced by larger government expenditures for im- ports (Snodgrass 1966). No serious attempt was made to develop other revenue sources. The Shift to Import Controls: 1956-65 The UNP suffered a landslide defeat in 1956 at the hands of a three-party coalition headed by the Sri Lanka Freedom Party (SLFP). The SLFP won the 1956 election by focusing its campaign on a promise to make Sinhala the only official language in Sri Lanka and by making an alliance of convenience with the left. Although the SLFP was not committed in ideological terms to socialism, it was not averse to adopting leftist rhetoric and promising to nationalize all foreign-owned plantations and to take over key industries. These promises did not prevent the coalition from gaining the support of many Sinhalese industrialists and small entrepreneurs who had not prospered under the UNP regime. The new SLFP-led government soon adopted a ten-year macroeconomic plan that was to cover the 1959-68 period and envisioned an average gross national product (GNP) growth rate of 5.9 percent a year. This growth was to be achieved through a policy of import-substitution industrialization and increased govern- ment participation in the manufacturing sector. Chapter Two 11 A year later, in 1957, the parliament adopted a comprehensive "Scheme of In- dustrialization" that established a state monopoly over basic and strategic sectors of production. This was accompanied by a State Industrial Corporation Act under which the government would provide the capital needed by new industrial under- takings. These attempts to restructure the economy, however, came on the heels of an election campaign whose emphasis on the question of language had aroused deep- seated ethnic hostilities, and for that and other reasons the 1956-59 period was a troubled one. The Central Bank sent warnings to the government about increased reliance on foreign borrowers to finance growing budget deficits, yet the deficit continued to grow because of sluggish export volume and adverse shifts in the terms of trade. And on the labor front there were prolonged strikes that were only settled by large increases in wages. These did little to pacify the unions. however, and the left-wing partners in the governing coalition withdrew their support. The SLFP itself was suffering from internal dissension, and this period of unrest culmi- nated in September 1959, when a Buddhist monk assassinated Prime Minister Bandaranaike. After several months of political turbulence, the UNP was able to regain power in March 1960 through alliances with smaller political parties. The UNP then promptly sought to solidify its position by enlarging rice subsidies for consumers. That was not enough. In July 1960, in a new election, the SLFP (now led by Bandaranaike's widow) regained control of the government. A month later, with the trade deficit growing ever larger, the Central Bank imposed selective credit controls to reduce imports of nonessential goods (automobiles, alcohol, cosmet- ics), and this was quickly followed by the imposition of steeper duties on other im- ports, including petroleum, tobacco, watches, and textiles. These anti-import tactics had various purposes. One, of course, was to try to keep the negative trade balance under some sort of control. Another was to protect domestic producers of these items from foreign competition. Even more drastic measures followed in January 1961-stringent controls on foreign exchange, a complete ban on imports of 49 luxury goods, and licensing re- quirements for most other imports that amounted in effect to the imposition of quotas. Along with its trade balance deficit the country also had a large deficit in its domestic budget. The government initially sought to reduce it by proposing reduc- tions in the national rice subsidy program. This proposed move by a nominally sympathetic government outraged the left and the trade unions, whose vociferous opposition prevented the proposed cuts and led to the ouster of the finance minis- ter who had suggested them. With revenues still far short of expenditures, the government then took anoth- er tack to obtain more income; it nationalized foreign-owned oil companies oper- ating in Sri Lanka. This action strained the relationship between Sri Lanka and those countries where the oil companies had their headquarters, particularly the 12 Macroeconomic Policies, Crises, and Growth in S-i Lanka, 1969-90 United States. In retaliation for the nationalization, the U.S. government cut off all financial aid to Sri Lanka under the so-called Hickenlooper amendment. Nonetheless, the SLFP continued to expand its intervention in the economy, a process it had begun in 1959 with the nationalization of the bus system and a take- over of cargo-handling in the port of Colombo. Between 1959 and 1964 the bank- ing and insurance systems both were taken over by the government. The Bank of Ceylon, the largest domestic commercial bank, was nationalized, and the opera- tions of foreign bank branches were restricted. These encroachments on the private sector were accompanied by the contin- ued expansion of government control over foreign trade. In 1963 a Foreign Ex- change Budget Committee was set up at the Ministry of Finance to allocate foreign exchange on the basis of "national priorities" (Kappagoda 1967), and fur- ther foreign exchange controls were imposed in 1964, along with a moratorium on repatriation of profits and dividends, and restrictions on overseas education and foreign travel. By 1965 these extensive exchange and import controls had trans- formed Sri Lanka's economy into a highly regulated, protectionist regime. But the economy continued to be highly trade-dependent. Even in 1965, despite the nu- merous trade restrictions, the combined value of imports and exports was equal to half of GDP. Why did the government resort to such extensive import-control measures during this period? If the rhetoric of the SLFP is taken at face value, it is possible to find "facts" which would be consistent with an explanation of these actions as resulting from a major ideological shift toward import-substituting industrial pol- icies. But the picture of an ideologically motivated SLFP government embarking on a policy of drastic import restrictions to foster industrialization caricatures the interplay of complex political and economic factors that fashioned government policy during this period. Further, these restrictions were by no means welcomed even by the left. The leading left-wing party, the LSSP, opposed trade restrictions on consumer goods, such as textiles, and attacked the government for protecting the interests of new domestic manufacturing firms at the expense of low income consumers. Pragmatism, rather than any deep ideological commitment, dictated the government's actions, even though some elements within the government were certainly ideologically inclined to support such policies. The import-substi- tution rhetoric provided a radical facade that was politically useful. These actions reflected the political weaknesses of the government. It was not willing, in the end, to run the political risk associated with making drastic expen- diture cuts in response to the balance of payments crisis, since that would have in- volved cuts in consumer subsidies. Thus, the government had little room to maneuver. If consumer subsidies could not be cut and "essential" imports had to be maintained, the government had few options left. Furthermore, the Central Bank, which was hardly a socialist bastion, strongly advocated import controls. While stressing that the most important task was to reduce the budget deficit, the Central Bank also viewed direct action to reduce imports as essential. Chapter Two 13 Meanwhile. the SLFP regime had to contend with opponents on both the right and the left. The regime survived an attempted right-wing coup d'etat by elements within the military in 1961, but faced a major challenge in 1963 when the three main leftist parties formed a United Left Front (ULF) and made demands that in- cluded substantial wage increases. In early 1964 the government was forced to back down in a confrontation with a militant left-wing trade union, and a political crisis was precipitated. In July 1964 Mrs. Bandaranaike offered partnership in a coalition government to the LSSP, the leading leftist party, and the other members of the ULF, to head off a trade union challenge. This offer was accepted by the LSSP and the CP, and was to be a move with profound implications. For the first time since the formation of the leftist move- ment in the 1930s, the left was no longer in the opposition. As junior partners in the coalition, the left had to take some responsibility for governing while allied to the SLFP, which the left had characterized from its inception as a capitalist party. From 1964 onwards the traditional left renounced its former revolutionary aspira- tions in practice, if not always in rhetoric. One of the results of this incorporation of the traditional left into the mainstream was the disappearance of a radical leftist opposition. The coalition government formed in 1964 was short-lived, however, losing a vote of confidence in parliament in November 1964 when some conservative members of the SLFP who opposed the link with the left joined the opposition. Partial Liberalization: 1965-70 The ensuing election in 1965 did not give any party a majority in the parliament. The UNP, with 39.3 percent of the votes, obtained sixty seats but lacked an absolute majority over the SLFP-led coalition, which obtained fifty-five seats. The UNP then formed a coalition with the Federal Party (FP) (the major Tamil party) and a num- ber of smaller groups to constitute a clear majority. The generally pro-Westem stance of the UNP created a more favorable envi- ronment for Western aid to Sri Lanka and also strengthened confidence among for- eign commercial banks and lending institutions. The government felt able to rely on greater external financing to overcome its payments problems without further restrictions on imports and entered (for the first time) into a standby agreement with the International Monetary Fund (IMF) to borrow up to US$30 million. The first US$15 million was borrowed in July 1965. A second step was an agreement with the United States on compensation for the nationalization of U.S. oil company holdings. This removed the barriers to a resumption of U.S. assistance, and an aid consortium was set up at a meeting con- vened by the World Bank. The countries in the consortium included Australia, Canada, Japan, the United Kingdom, and the United States, later joined by the Federal Republic of Germnany. 14 Macroeconomic Policies, Crises. and Growth inz Sri Lanka, 1969-90 The government also arranged some short-term financing from foreign com- mercial banks and announced the lifting of the 1964 moratorium on the repatria- tion of profits, dividends, and other investment income. Funds held prior to the rescinding of the moratorium were remitted in stages, and the government assured potential investors that no such moratorium would apply to new investments. For- eign investors were also advised that they would be protected against expropria- tion and discriminatory treatment (Fernando 1972). In addition, some attempts were made to overcome the clear antiexport bias of the economy. A bonus voucher scheme for nontraditional exporters, modelled after a similar scheme in Pakistan (Motha 1971) was introduced in 1966. It granted a bonus of 20 percent in the form of transferable import entitlement certificates to producers of certain nontraditional export items, provided there were net foreign exchange earnings of 25 percent of (FOB) value. The vouchers fetched a handsome premium on the market (Dahanayaka 1977). The 1967Devaluation Except for a partial devaluation against the dollar in 1949 (mainly to provide relief to the British pound) and some change in the rupee/dollar rate in the early 1950s, the exchange rate had remained stable for over fifteen years despite sustained pres- sure on the balance of payments. But when the British devalued the pound by 14.3 percent in November 1967, Sri Lanka followed by devaluing its rupee by 20 per- cent against the pound, a net devaluation of 5.7 percent. The devaluation, however, was followed by govemment-decreed wage in- creases for workers in the public sector and in the organized private sector, as well as an increase in taxes on traditional exports. Hence, the devaluation was insuffi- cient to prevent a further buildup of pressure on the balance of payments, and for- eign currencies continued to fetch a handsome premium in the black market. The Central Bank (1968) said that while the new rate was a good base rate, it was not enough to permit liberalization of a substantial range of imports. A unitary rate at a further depreciated level might have corrected the situation in some respects, but in other respects it might have created further problems. The 20 percent devalua- tion was felt to be adequate for major exports but not large enough to solve the problems of other exporters. The devaluation had a similar effect on the import side, producing a sufficient markup in the price of a wide range of essential im- ports, but failing to bring about a liberalization of all imports. Hence, some system of multiple rates appeared necessary. In May 1968, after consultations with the IMF, the government introduced a package of trade liberalization measures. The most important element was a For- eign Exchange Entitlement Certificate Scheme (FEECS). Under the scheme, all ex- ternal transactions were divided into categories A or B. Transactions in category A were to be conducted at the official exchange rate and were to include essential imports, such as rice, flour, sugar, drugs, fertilizer, and petroleum, as well as tra- ditional exports (the three plantation crops). Category B transactions were those Chapter Two 15 involving nontraditional exports and nonessential imports. These transactions were to take place through a certificate system at a lower (depreciated) rate, and it was expected that this rate (in the form of the prices of Foreign Exchange Entitle- ment Certificates) would be determined by market forces (as reflected by tender prices offered for certificates issued weekly by the Central Bank). A substantial range of category B imports (more than a thousand) was placed under Open General License (OGL) rules, and in 1969 (the first year of operation of the scheme) they accounted for 15 percent of the total value of imports. Export- ers of category B goods and services were entitled to receive certificates with a face value in rupees equal to the FOB value of exports when the foreign exchange receipts were surrendered to an authorized bank. Those who applied for foreign exchange for payments for imports in this category had to surrender certificates with an equal par value. A new tariff structure was also introduced as part of the scheme. Most essen- tial items were allowed free entry, while rates ranging from 10 to 300 percent were imposed on other items, depending on degree of nonessentiality.3 These trade lib- eralization measures were accompanied by a major effort to increase food produc- tion, particularly rice production. UNP governments had always sought rice self- sufficiency as a matter of national pride. and failure to achieve that goal had been intensified by the burden of paying for substantial rice imports. With bleak pros- pects for export products and the newly established import-substituting industries under strain because of shortages of imported inputs, the govemment envisaged import substitution in agriculture as a way out of its balance of payments crisis (Corea 1973). Policy measures in this sphere included giving high priority to food cultivation under the direct supervision of the prime minister's office, develop- ment of agricultural extension services, provision of subsidized inputs (fertilizer, other chemical inputs, credit), and steps to boost the prices received by farmers. The last involved drastically curtailing subsidiary food imports, increasing gov- ernment-guaranteed prices, and curtailing the subsidized rice ration. These poli- cies culminated in a formal plan to carry out the huge Mahaweli irrigation project.4 Despite the government's export promotion rhetoric, the traditional export in- dustries continued to suffer from the anti-export bias embodied in commercial policy (Cuthbertson and Athukorala 1990). Since export duties continued to be a major source of government revenue, the producers of these crops carried a large tax burden. The 1968 policy reforms added to this bias by excluding tea, rubber, and coconut from the FEEC scheme. It was argued that reducing dependence on the three traditional export crops would improve export earnings by stimulating crop diversification. Hence, such nontraditional export crops as cocoa, cinnamon, car- damom, and pepper were included in the scheme to give farmers an incentive to switch away from traditional plantation crops. Policy discussions generally ig- nored the potential efficiency losses involved in such a strategy. Indeed, given Sri Lanka's comparative advantage in the plantation crops, it is likely that substantial economic losses were incurred through this policy. Further, even for the nontradi- 16 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 tional agricultural exports the favorable impact of the FEEC scheme was largely offset by upward adjustments in export duties. The UNP government continued to favor the expansion of import-substituting industries. Many such industries were located in the public sector and had been started or planned under the previous government. The private sector industries that found conditions most conducive to expansion were those producing consum- er goods. particularly the nonessentials kept out by import restrictions. Government policies between 1965 and 1970 continued to be based on the view that the solution to Sri Lanka's external imbalances, at least in the short run, was to be found through adjustments on the import side. Efforts to increase the country's exports would take time. Gamani Corea, head of the Ministry of Plan- ning and Economic Affairs, expressed it this way (Corea 1971a:31): Increasing export earnings was particularly difficult since these earn- ings were not entirely the outcome of Ceylon's own efforts. In the case of tea, for example, an increase in productivity might even aggravate the problem by further dampening world prices. In the case of other items, both traditional and new, earnings could be increased by in- creased production, but this takes time. calling as it does for new ca- pacity. Such an objective must necessarily loom large in any development program for Sri Lanka, but it does not provide an over- night solution to the payments problem. Many policy advisers were aware of some of the limitations of an import-sub- stitution strategy (see, for example, Corea 1971 b), but the possibility that a policy of promoting import substitution would by its very nature tend to discourage ex- port-oriented industries because of general equilibrium effects was not seriously considered. Moreover, the efficiency losses associated with an import-substitution strategy were generally underestimated or ignored. Overall, the 1965-70 period can be characterized as one when a weak and hesitant attempt was made to liberalize the economy. The political determination to achieve a full-scale liberalization was lacking. since the government was quite aware that the political opposition could arouse strong anti-government agitation. Moreover, external economic conditions as reflected in the terms of trade contin- ued to be adverse. and low-cost external finance was not forthcoming. In this con- text the government did not, and probably dared not, contemplate a policy that would redress external imbalance through cutting consumer subsidies and thereby reducing domestic absorption. Attempts to increase output through regulation, production subsidies, and state investment could not go beyond the constraints im- posed by large consumer subsidies. Despite improved economic growth, the relatively successful food produc- tion drive, and the ambitious plan for the Mahaweli project, doubts about the gov- ernment's capacity to achieve sustained economic growth did not cease. The electorate did not forget that a UNP government had once again reduced the total amount of rice available at a subsidized price, even though the new scheme pro- Chapter Two 17 vided two pounds of rice free to each person per week. Rising inflation, youth un- employment, concern about foreign debt, and the electorate's perception that the government lacked conviction about its own policies produced a landslide victory for the SLFP-led United Front (UF) in the 1970 election. The UNP's hesitant attempt at liberalization was over. After the election, UNP leader Dudley Senanayaka commented that by tamper- ing with the rice subsidy for the second time in his political career he had paid the penalty for "disturbing the most cherished of Sri Lanka's sacred cows" (de Silva 1981). This probably understated the importance of many other sources of popular dissatisfaction with the government, but there was more than a grain of truth in it. Regulation and Controls: 1970-77 The new United Front government came into office at a time when economic cir- cumstances were worsening but popular expectations were high. The UF had at- tacked Senanayaka's UNP for reducing the rice subsidy and had pledged to restore it, and the inclusion of the LSSP and cp in the coalition added credibility to this promise. Many expected the new government to usher in an era of prosperity that would eliminate poverty, landlessness. and unemployment. The rhetoric and some of the actions of the new government during its first few months in power might have suggested that the UF coalition was set on a "so- cialist march." but the reality was more complex. The formation of the Front had not eliminated key differences between the three parties that formed the alliance. While the SLFP was willing to impose state controls and regulations, it was not, despite its occasional populist posturing, a party committed to overthrowing the existing order. The coalition with the lISSP and the cP was the product of special circumstances (Wilson 1979). Under Mrs. Bandaranaike the SLFP retained its dominance within the coalition when fundamental issues were at stake, even though some of the long-held views of the leftist parties influenced the course of economic policy in this period. In general, the LSSP and cP gave higher priority than the SLFP to equity issues, had a stronger commitment to the public sector, and did not want to be "dictated to by the IMF and the World Bank." Consequently. the two parties were reluctant to en- gage in large-scale foreign borrowing that involved conditionality. They also had much greater faith than Mrs. Bandaranaike in the efficacy of government controls. In April 1971, less than a year after coming to power, the UF government was faced with an armed insurrection led by the Janatha Vimukthi Peramuna (JvP, or Peoples' Liberation Front). an organization composed mostly of educated but un- employed rural Sinhalese youths. The JvP was the largest of several radical orga- nizations devoted to rectifying what was seen as a betrayal of socialism by the traditional leftist parties. The JVP's best-known leader. Rohana Wijeweera, had 18 Macroeconomic Policies. Crises. and Gr-owth in Sri Lanka, 1969-90 briefly been a member of a Maoist group that had split from Sri Lanka's pro-Mos- cow communist party in 1963. The jvp combined populist rhetoric with Sinhalese nationalism. It was a mea- sure of the high expectations that greeted the election of the UF government, and the rapidity with which disillusionment subsequently set in, that the JvP was able to attract thousands of youths to launch an armed challenge to the new government so soon after the election. But the JVP lacked wider support in the community, and its uprising was crushed by the government with the help of military and political support from both the Western and Communist blocs. Estimates of the dead ran to thousands. Abortive though it was, the IVP's uprising sent shock waves through Sri Lan- kan society and had far-reaching consequences. The government sought to as- suage its supporters on the Left while taking other steps to entrench its rule. These included a significant degree of land reform in 1972 and a takeover (with compen- sation) of foreign-owned tea and rubber plantations in 1975 (Fernando 1978 and 1980; Peiris 1977). Foreign policy also shifted. Even though the USSR and China had rushed in with support (including military equipment) during the uprising, Mrs. Bandaranaike chose to strengthen Sri Lanka's ties with the west. Defense ex- penditures began to rise, and the government moved toward more authoritarian forms of rule. This was facilitated by changes in the constitution that circum- scribed the power of the judiciary to restrain the activities of the government. When the UNP launched a civil disobedience campaign in 1974 to demand elec- tions in 1975, its public meetings and other activities were banned. Meanwhile, Sri Lanka was suffering from a combination of external and in- ternal shocks. These included the first oil price shock, large increases in the prices of the main food imports, and shortfalls in outputs of the major agricultural crops. The measures taken to deal with this crisis imposed economic hardships not expe- rienced by the Sri Lankan population since the 1940s. One of the UF'S objectives was the "laying of the foundation for the building up of a socialist society" (Budget Speech 1970:3). This meant the further exten- sion of state control over the economy. As soon as it came into power, the new government created an individual licensing scheme for all imports-in other words, a system of quantitative restrictions. The FEECS was retained, but exchange controls became more stringent. These measures were aimed at averting an imme- diate balance of payments crisis, given a rapid increase in the import bill and a rapid fall in external reserves during the last year of the previous government. Seven new industrial corporations were established by the government during the period 1970-77. A Paddy Marketing Board was established in 1971 and be- came the sole buyer of paddy from farmers. Most of the plantation sector also came under state ownership. Under the Land Reform Act of 1972 (and amended in 1975), 63 percent of the total area under tea, 32 percent of the area under rubber, and 10 percent of the area under coconut were nationalized. Earlier, in April 1971, the parliament had passed a Business Undertakings Acquisition Act that empow- ered the Minister of Finance to acquire any business firm if the acquisition was Chapter Two 19 considered to be in the national interest. By 1977, 26 business ventures had been converted into Government-Owned Business Undertakings (GOBUs). The focus of economic policy continued to be import substitution (Sri Lanka Ministry of Planning and Employment 1972), but some steps were taken to stim- ulate exports because of balance of payments pressure. The FEECS premium was raised to 65 percent and a convertible rupee account (CRA) was introduced to give further incentives to exporters. Permission to establish such accounts, limited originally to gem exporters, was later extended to other nontraditional exporters. It allowed a portion of export earnings to be used to finance imports and to pay for foreign travel, education, and other activities that had to be paid for with foreign exchange. Although a white paper on direct foreign investment was issued in 1972 to en- courage foreign investment, the climate for such investment (indeed, for private investment in general) was not congenial. Ceilings on incomes, land reform, limits on ownership of dwellings, and the 1971 acquisition law all created a climate of uncertainty (Athukorala 1984). Moreover, the Exchange Control Act of 1971 made it compulsory for resi- dents of Sri Lanka to declare any asset owned abroad within one month of acqui- sition. The act also prohibited residents from having overseas bank accounts and from disposing of foreign funds without the permission of the Exchange Control- ler. Under another provision of the act, proceeds from the liquidation of companies by nonresidents were placed in blocked accounts, the balances of which could be remitted abroad only under special circumstances. In the case of profits earned by nonresident shareholders, remittances were allowed only up to 12.5 percent of in- vested capital. The most unexpected changes were those involving welfare expenditures. The subsidized rice ration was increased to "two measures" (that is, four pounds) per person. But the terms of trade fell by 23 percent from 1970 to 1972 as import prices doubled, and because of the expansion of the rice subsidy and a drastic de- cline in domestic production because of bad weather, rice imports were 40 percent higher in 1971 than in 1969. Despite increased aid from the centrally planned economies, Sri Lanka found it impossible to finance its already very low level of imports. By the time of its second budget in November, 1971, the government had taken steps to reduce food subsidies and other welfare expenditures. A series of price hikes followed. The price charged to consumers for the rice ration was raised by 25 cents to I rupee. Sugar, which had been available at 72 cents per pound in unlimited quantities, was brought under rationing. Only three pounds per month were available at 72 cents a pound, while quantities above this were priced at 1.50 rupees per pound. In March 1972, the sugar ration was further reduced to two pounds per month, and the price was raised to 2.50 rupees per pound. In Novem- ber 1972 wheat flour was rationed (one pound per week). Income tax payers were declared ineligible for subsidized free rice in December 1972. The price of wheat flour was increased from 33 cents per pound to 38 cents per pound in January 20 Macroeconomic Policies. Crises, and Growth in Sri Lanka, 1969-90 1973, and again, to 48 cents in June 1973. In November 1973 the rice ration was reduced to one pound per week. These reductions in food subsidies were accompanied by other reductions in government outlays. Bus and rail fares, as well as postal, telephone, and telegraph charges, were increased. Not since 1953 had any government reduced welfare expen- ditures on such a scale, though the remaining subsidies were still very substantial. The political problems within the UNP in the immediate aftermath of the elec- tion and the crushing of the Jvp left no viable political opposition to the regime. The bulk of the trade union movement was under the control of the LSSP and the CP, and the trade union leaders had become partners and defenders of the regime. Moreover, certain measures could be justified to the electorate as the cost of deal- ing with the insurgency. By 1975, however. mass opposition to economic hardship could no longer be contained. Tensions within the government severely weakened the influence of the LSSP, and in 1975 the party was compelled to leave the coalition. Soon after, the killing of a university student by the police led to an upsurge of protests against the government, and the Cp, which had remained in the coalition after the LSSP'S departure, bowed out just before the 1977 general election. The electorate then de- livered a massive defeat to all the parties that had been associated with the United Front. During the campaign the UNP had urged liberalization of the economy, ar- guing that the lifting of trade restrictions would end the shortages of consumer goods and lead to rapid economic growth and full employment. This message had a great appeal to an electorate that had suffered from several years of economic hardship and authoritarian politics. The Turnaround in Economic Policy after 1977 The 1977 election was the first in which a majority of the votes was cast for a sin- gle party. The UNP. led by J. R. Jayawardene, won 140 out of a total of 168 seats. while the SLFP slumped from 90 to a mere 8. Not a single candidate from the Left was elected. Even allowing for the "first past the post" election system, this change suggested that political allegiances had become more volatile. The triumph of the UNP was welcomed by the Western powers, since Jayawar- dene had always shown a pro-Western orientation. When the new government be- gan an extensive reform of exchange rate and trade policies after taking office, the Western powers and the international agencies (which had long pressured for such reforms) responded with an unprecedented flow of aid (Levy 1985). The elections also marked a new phase in the Tamil community's growing separatist movement. The Tamil United Liberation Front (TULF) won a sweeping victory in the traditional Tamil areas of the North and Northeast with its pledge to strive for an independent Tamil state. TULF, with eighteen members in the parlia- ment, became the largest opposition party. Chapter Two 21 During its seven years in opposition, the UNP had seen how the UF coalition had prolonged its tenure, and the new government lost no time in capitalizing on its two-thirds majority to change the constitution in ways favorable to itself. Under the new rules, future parliaments would be elected on a proportional basis (almost guaranteeing that no party would have the power to reverse the UNP's changes in the constitution) and the new document also provided for simplified decisionmak- ing through a centralization of executive power in the office of the president. Jayawardene was, of course, the overwhelming choice for the job (Wilson 1980). The new administration initiated significant economic policy reforms soon after entering office.5 Quantitative restrictions on import trade were largely re- placed by tariffs under an open general licensing system. These tariffs usually pro- vided lower levels of nominal protection for domestic import-substitution industries. A number of measures to encourage direct foreign investment, includ- ing general tax incentives and the establishment of an export processing zone, were implemented. Most price controls were removed. Universal food subsidies were replaced by a food stamp scheme whose beneficiaries were to be those who earned the least income. As it turned out, however, the scheme actually covered more than half of the population. The exchange rate system was also reformed. The exchange rate was unified by eliminating the FEEC and CRA schemes, the rupee was devalued by 45.5 percent (from 9.1 rupees per U.S. dollar to 16.7 rupees per dollar), and a managed float was adopted with a view to making the exchange rate an active policy instrument. The lifting of many restrictions on capital transactions was an important step toward greater integration of the domestic capital market with the international market. However, the rupee was not made fully convertible. Repatriation of pro- ceeds from sales of shares in Sri Lankan companies was permitted without prior approval of the Controller of Exchange, and foreign exchange dealers were per- mitted to approve applications for such remittances. Moreover, foreign investors were allowed to repatriate their capital contribution to the equity of a company upon sale or liquidation of the investment. Transfer of shares was freely permitted if it did not increase the ratio of nonresident shareholdings originally approved. Stringent restrictions on the release of foreign exchange for foreign travel and study abroad were relaxed. As in the past, however, residents of Sri Lanka were not permitted to invest abroad, to maintain accounts with banks abroad, or to raise capital in foreign markets for domestic investment without prior approval. Financial repression was eased. Measures aimed at permitting interest rates to attain realistic levels were introduced in 1977 (Hettiarachchi 1978). Then, in 1979, the Finance Act (promulgated in 1961) was amended to allow foreign banks to once again set up branches in Sri Lanka. This was accompanied by a foreign currency banking scheme under which commercial banks would be authorized by the Central Bank to operate foreign currency banking units. Export expansion was also accorded high priority. The main elements of the reform, such as trade liberalization, relaxation of controls on foreign exchange transactions, a more realistic value for the rupee, and the elimination of various 22 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 controls on private sector activities, were intended to generate an economic cli- mate conducive to export expansion. However, it was recognized that these poli- cies alone could not be expected to generate the desired export push in a country whose traditional exports were a small number of primary products and which re- tained a strong import-substitution bias (Santiapillai 1981). To provide additional direct support for exports, a Sri Lanka Export Development Board (EDB) was es- tablished in 1979. Unlike its predecessor, the EDB was a semi-governmental body with indepen- dent legal and financial status. Its activities were financed by an Export Develop- ment Fund made up of direct contributions from the Treasury and receipts from a 10 percent cess on imports subject to duties of more than 50 percent and a 2 to 10 percent cess on certain minor agricultural exports. The EDB set in motion a wide range of export promotion schemes, including a revised duty rebate scheme, direct cash subsidies to exporters of designated products, medium- and long-term credit schemes to finance export-oriented investment, annual presidential export awards, and a number of schemes aimed at enlarging the volume of nontraditional exports. In addition, a preshipment refinance scheme operated by the Central Bank for non- traditional exporters got underway in the second half of 1978. Another important aspect of export policy was the establishment of a Greater Colombo Economic Commission (GCEC) to organize and operate export process- ing zones (investment promotion zones or IPZ) that would attract direct foreign in- vestment. The first investment promotion zone, at Katunayake near the Colombo International Airport, was opened in June 1978. The incentives offered to IPZ in- vestors included allowing 100 percent foreign ownership in investment projects, a tax holiday for up to ten years on the salaries of foreign personnel, royalties, and dividends, and a duty exemption for imported machinery and other production in- puts (Ramanayake 1984). The liberalization of 1977-78 was accompanied by a massive public sector investment program whose chief components were the Accelerated Mahaweli De- velopment Project and a housing construction program sponsored by Prime Min- ister R. A. Premadasa. Initially, the Mahaweli project was intended to be spread over thirty years, but the new government decided that it was necessary to com- plete most of the project within five or six years. The housing program included a Public Sector Housing Program with a con- struction goal of 100,000 urban dwellings and a Rural Housing Program with a goal of one million dwellings. Other government projects given priority included construction of a new administrative capital (Sri Jayawardenepura) and a new par- liament house. The huge fiscal stimulus and the policy reforms-which were ac- companied by large capital inflows from aid donors, remittances sent by a growing number of migrant workers in the Middle East, and larger tourist revenues- brought about rapid growth. Not everything was favorable, however. The tea price boom of 1976-77 col- lapsed in 1978, and from 1979 onward the effects of the second oil shock began to be felt in a steady downward movement in the terms of trade. The economy had Chapter Two 23 begun to show signs of major macroeconomic strain in 1980, when rapid growth in the money supply led to strong inflationary pressures. In July 1980 these pres- sures led to a major strike by govemment employees. The govemment then used its emergency powers to dismiss thousands of govemment workers. Meanwhile, as the terms of trade decline started to affect the current account, official aid proved insufficient to finance necessary imports. Loans from commercial sources had to be obtained to meet payment obligations. The 1978-82 period was unique in many ways. The unprecedented capital in- flows, huge surge in investment, and sharp fluctuations in terms of trade created enormous adjustment pressures. All the ingredients for a crisis were there, but no crisis occurred. Nonetheless, we call this the second crisis episode and analyze it in more detail below. Extemal pressure had eased by 1984, but the country was then confronted by a major domestic crisis. The 1971 JVP uprising had signalled that the long period of peace and stability in Sri Lanka might be ending, and simmering ethnic tensions eventually escalated into large-scale violence in 1983, opening a new chapter in Sri Lanka's history. The Rise of Ethnic Hostility A series of incidents that culminated in police action in 1979 against Tamils in Jaffna led to the buming down of a library housing Tamil literary manuscripts. Af- ter widespread anti-Tamil rioting in 1979 in predominantly Sinhalese areas (in- cluding the capital, Colombo), large numbers of Tamils began moving northward. These events also resulted in a major change in Tamil political leadership. Militant armed groups emerged as the new leaders, replacing an older generation of moderates. Years of frustration with the policies of successive Sinhalese gov- emments had led these Tamil militants to the conclusion that the only acceptable alternative was the establishment of a separate Tamil state that would be called Ealam. The ethnic conflict then took a dramatic tum for the worse in 1983, when a group of Tamil political prisoners being held in jail in Colombo was massacred and no action was taken by the authorities to punish those responsible. This pre- cipitated widespread armed conflict in the Tamil dominated northem and eastem provinces, and a full-scale civil war got under way. Large numbers of Tamil refugees fled across the water to Tamil Nadu, an eth- nic Tamil state in South India that was also used by various militants as a base for conducting the armed conflict in Sri Lanka. Meanwhile, Sri Lanka's predominant- ly Sinhalese armed forces moved into the northem and eastem provinces, and in the ensuing battles Tamil civilians suffered heavy casualties. These events strained relations between Sri Lanka and India, but the two govemments agreed to allow Indian armed forces to occupy the northem and eastern provinces in 1988. 24 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 The goals of both countries were to disarm the Tamil militants and prepare the ground for a settlement within the context of a unified state. The Indian army was unable to disarm the main Tamil group, however, and eventually withdrew in 1990.6 In the meantime a rejuvenated JvP found itself able to exploit the grievances of segments of the Sinhalese population and the widespread anti-Indian feelings of Sri Lanka's Sinhalese majority. The jvP launched a campaign of terror in Sri Lanka's southern Sinhalese-dominated areas that was only halted after two years of savage conflict. Economic growth slowed to a crawl, and exports stagnated. Unemployment, which had declined in the early post-1977 period. started to increase, and by the end of the 1980s was estimated to be around 15 percent. There were continuing pressures on the balance of payments, and a major crisis was averted only because of substantial assistance from Japan and Western donors. In the midst of a deteri- orating situation J. R. Jayawardene resigned as president in early 1989 and was re- placed by Premadasa. A program to alleviate poverty was launched, and after remaining unchanged for a decade the nominal value of food stamps was in- creased. Although economic policy continued to be dictated by military impera- tives as the decade ended, moves to implement a series of large-scale privatization initiatives were put back on the political agenda. As Sri Lanka entered its fifth decade of independence, the "best bet of Asia" in 1948 was being tragically transformed. By 1990 Sri Lanka was a simmering volcano of political, social, and ethnic conflicts despite the restoration of a sem- blance of political stability and a degree of recovery in economic performance. Even this hesitant recovery was threatened by the Iraqi invasion of Kuwait, which dealt new blows to the battered economy by disrupting a major market for exports of Sri Lankan tea and by eliminating large numbers of jobs in the Middle East for migrant workers from Sri Lanka. Chapter Three A Profile of the Economy Sri Lanka's major ethnic group is the Sinhalese, who at the time of the 1981 census accounted for 74 percent of the population (Table 3.1). Sri Lankan Tamils, Indian Tamils (descendants of immigrant workers from South India in the late 19th and early 20th centuries), and Moors (descendants of the Arab traders who settled in Sri Lanka centuries ago) accounted for 12.6 percent, 5.6 percent, and 7.4 percent, respectively. Over the past forty years the Sinhalese and Sri Lankan Tamil propor- tions have increased slightly while the proportion of Indian Tamils has declined, mainly because of a ban on further immigration in 1957 and the repatriation of Tamil laborers under a, 1964 agreement between the governments of Sri Lanka and India. Most of Sri Lanka's Sinhalese are Buddhists, most of its Tamils are Hindus, and most of its Moors follow Islam. Christians, particularly Roman Catholics, constitute an important religious minority. The population of Sri Lanka increased from 6.7 million in 1946 to 16.8 mil- lion in 1988 (table 3.2). This rapid increase reflected a spectacular decline in the Table 3.1 Ethnic Composition of the Population, Selected Years, 1946-81 (percentage of total population) Ethnicity 1946 1953 1963 1971 1981 Sinhalese 69.4 69.4 71.6 72.0 74.0 Tamil 22.7 23.0 21.6 20.5 18.2 Ceylon Tamils 11.0 11.0 11.0 11.2 12.6 Indian Tamils 11.7 12.0 10.6 9.3 5.6 Muslims 7.1 6.3 6.4 6.7 7.1 Othersa 5.8 1.3 0.4 0.8 0.7 a. Burghers and Eurasians, Malays, Europeans, and Veddhas. Source: Government of Sri Lanka, Department of Census and Statistics, Statistical Abstract (annual). 25 26 Macroeconomic Policies, Crises. and Growth in Sri Lanka. 1969-90 death rate (particularly in the 1940s and the 1950s) and occurred despite a moder- ate decline in the birth rate. The crude death rate declined from twenty-three per thousand in 1946 to seventeen per thousand in 1953, and then to six per thousand in the 1980s. The birth rate showed little change up to about the mid-1960s but then trended downward, reflecting the success of a national family planning pro- Table 3.2 Population and Vital Statistics, 1946, 1953, and 1963-88 Net Population Population Birth Death migration growth mid-year rate rate rate rate Year (millions) (perl,000) (perl,000) (per 1,000) (percent) 1946 6,657 19.9 23.1 +1.6 2.55 1953 8,099 31.0 16.8 -0.3 3.62 1963 10,651 34.1 8.6 -1.0 2.55 1964 10,889 33.2 8.8 -1.0 2.44 1965 11,133 33.2 8.2 -0.5 2.50 1966 11,439 32.3 8.3 -0.5 2.40 1967 11.703 31.6 7.5 -0.6 2.41 1968 11,992 32.0 7.9 -0.7 2.41 1969 12,252 30.4 8.1 -0.9 2.23 1970 12,516 29.4 7.5 -0.8 2.19 1971 12,608 30.4 7.7 -2.7 2.27 1972 12,861 30.0 8.1 -3.2 2.19 1973 13,091 28.0 7.7 -3.8 2.03 1974 13,284 27.5 9.0 -4.0 1.85 1975 13,496 27.8 8.5 -2.3 1.93 1976 13,717 27.8 7.8 -3.8 2.00 1977 13,942 27.9 7.4 -3.7 2.05 1978 14,190 28.5 6.6 -2.8 2.19 1979 14,472 28.9 6.5 -3.0 2.24 1980 14,747 28.4 6.2 -4.6 2.22 1981 15,011 28.2 5.9 -3.3 2.23 1982 15,189 26.8 6.1 -6.0 2.07 1983 15,416 26.2 6.1 -6.8 2.01 1984 15,599 24.8 6.5 -4.9 1.83 1985 15,837 24.3 6.2 -0.6 1.81 1986 16,117 22.3 5.9 -2.7 1.60 1987 16,586 20.7 5.8 -3.1 1.49 1988 16,806 21.3 6.2 -1.2 1.51 Source: Government of Sri Lanka. Department of Census and Statistics, Statistical Abstract (annu- al); and Central Bank of Ceylon. Sri Lanka: Socio-Economic Data (1991). Chapter Three 27 gram and a continuing rise in the average age of marriage. The rate of overall an- nual population growth, which had been around 2.5 percent up to the mid-1960s, had declined to about 1.5 percent by 1988. One consequence of rapid population growth in the immediate postwar period was a population structure with a large proportion in the younger age categories. In 1981, 46 percent of the population was under 20. As a result of improvements in basic health care, average life expectancy had risen to 70 by 1988. With the exception of the return of plantation workers to India, Sri Lanka did not experience significant outmigration until fairly recently. The first wave of out- migration occurred when considerable numbers of the Burgher community left the country to settle in Australia, Canada, and the United Kingdom, particularly after the change in the official language after 1956. The early 1970s saw the beginning of the emigration of professionally qualified personnel from all communities for foreign employment, an outflow which accelerated as new employment opportu- nities developed in the Middle East. Such opportunities also emerged for relative- ly unskilled labor. Between 1976 and 1984, approximately 318,000 Sri Lankan workers emigrated, mainly to the Middle East, for temporary employment. The estimated stock of workers abroad increased from 740 in 1976 to 38,100 in 1980 and 129,609 in 1984 (Athukorala 1990). Nearly 80 percent of the country's population lives in rural areas, and there has been no significant trend toward urbanization. Colombo is the only large urban center; a few smaller cities have developed as regional administrative and com- mercial centers. It should, however, be noted that describing the population as "ru- ral" is somewhat misleading. Thanks to a well-developed and cheap transport system, large numbers of workers commute to paid employment in urban centers while living in rural areas. About two-thirds of the population lives in the wet zone, which comprises about a third of the land area. The Sri Lankan Tamils have traditionally lived in the northern and eastern parts of the country, though signifi- cant numbers are settled elsewhere, including Colombo. Most of the Indian Tamils live in the wet zone regions where the tea and rubber plantations are situated. Land and Resources Sri Lanka is a densely populated and largely agricultural country. It does not pos- sess any major mineral resources, with the exception of semi-precious gems and graphite.7 About 60 percent of the land area of 64,454 square kilometers is culti- vable, and about 35 percent is under cultivation. Paddy (rice), the main staple crop, accounts for about 45 percent of the land under cultivation, while tea, rubber, and coconut occupy about 42 percent. These proportions have changed somewhat in the past four decades, the main change being a doubling in the size of the rice area. Widespread adoption of modern high-yielding varieties and chemical inputs, as well as the bringing of new land under cultivation, has raised rice production 28 Macroeconomic Polic ies. Crises. and Growth in Sri Lanka. 1969-90 considerably, and Sri Lanka is now close to self-sufficiency in rice. The cultivation of rice, as well as of most other food crops. is primarily a smallholder activity; a majority of the farms are less than two hectares in size. (After land reform was car- ried out in the 1970s, large private holdings ceased to exist.) The overall population density is close to 250 per square kilometer, while ag- ricultural population density exceeds 700 per square kilometer. Fast-flowing riv- ers originating in the central hills provide some hydroelectric power, but Sri Lanka imports oil to meet most of its energy requirements (about 71 percent in 1989). Growth and Welfare Between 1950 and 1989. Sri Lanka's real GDP increased at an annual compound rate of 3.07 percent in absolute terms and 1.05 percent in per capita terms. This growth was far from smooth. If the 1968-70 period (during which Sri Lanka ex- perimented with trade liberalization reform) and the five years following the 1977 policy reforms are excluded. these rates fall to 2.5 and 0.7, respectively. A marked decline in growth after about 1983 reflected the damaging impact of ethnic con- flicts. As a result of the decline in the population growth rate, the annual growth rates of absolute and per capita GDP have moved closer to each other since the late 1970s. The World Bank now classifies Sri Lanka as a low-income country. In 1988, per capita annual income was US$375 (World Bank 1989). At the time of inde- pendence (and until about the 1960s), however, per capita income was comparable to, and often higher than, that of several Southeast Asian countries even when measured in conventional terms. According to purchasing power parity (PPP) ad- justed GDP estimates of Kravis, Hetson. and Summers (1982) (table 3.3), Sri Lan- ka's per capita income in 1950 was almost twice that of India, higher than those of Thailand and South Korea (henceforth Korea), and only marginally lower than that of Malaysia. Even in 1960, Sri Lanka's income level was similar to that of Korea as conventionally measured; by PPP adjusted measures, it was similar to that of Colombia, and well above those of Brazil and Korea. Since then it has slipped below these, and many other countries, by both yardsticks. Even now, however, Sri Lanka performs quite well in terms of most social in- dicators. In 1988 its adult literacy rate was 85 percent, almost all children of pri- mary school age were enrolled in schools, and 56 percent of the eligible age group was enrolled in secondary schools. Infant mortality was 37 per thousand, well be- low the average of 56 for upper-middle income countries. Because of these achievements. Sri Lanka has often been cited as an example of a country that has been able to provide its people with a high standard of living despite relatively slow growth and low GDP (Sen 1981). It has also been argued, however, that the picture is much less impressive when one shifts from a single-period cross-country comparison to a time series Chapter Three 29 Table 3.3 Per Capita GDP of Selected Asian Countries as a Percentage of the Per Capita GDP of the United States (purchasing power adjusted) Country 1950 1960 1970 1979 India 7.1 7.5 6.5 5.7 Korea 7.6 8.2 11.8 24.8 Malaysia 14.6 16.7 15.6 23.2 Pakistan 9.0 7.8 8.4 7.6 Philippines 10.3 11.4 11.7 13.3 SriLanka 11.4 10.2 9.4 9.4 Thailand 9.9 9.5 11.7 13.4 Source: Kravis. Heston, and Summers (1982), table 1.4. analysis that takes into account Sri Lanka's initial relative position and its capacity to improve upon (or at least to maintain) that position (Myint 1985; Bhalla and Glewwe 1986). In the first two decades after independence, Sri Lanka's standard of living, measured by the usual indicators, remained well above those of other de- veloping countries. Since then, however, Sri Lanka has slipped. In 1950, for in- stance, Sri Lanka's school enrolment rate (54 percent) was higher than that of Korea (43 percent) and comparable to that of the Philippines (59 percent). By 1979, however, both Korea (at 94 percent) and the Philippines (at 85 percent) had surpassed Sri Lanka (at 74 percent) in terms of this measure. In 1950, Sri Lanka's infant mortality rate (number of deaths per 1,000 live births) was 77, compared with 102 for the Philippines, 84 for Thailand, 85 for Korea, and 91 for Malaysia. By 1985, Sri Lanka, with a rate of 36, ranked behind Malaysia (28) and Korea (27), and both Thailand (43) and the Philippines had improved more rapidly than Sri Lanka. Nevertheless, the overall performance of Sri Lanka in terms of social welfare indicators has been quite outstanding. Anand and Kanbur (1991), in a de- tailed analysis of the Sri Lankan experience, have demonstrated that these achievements can be attributed to government expenditures on food, health, and education. The Structure of the Economy At the time of independence, Sri Lanka was a prime example of an economy heavily dependent on a limited range of agricultural export commodities. In 1948, the earliest year for which economywide national account data are available, the direct contribution of the three plantation crops (including processing) to GDP was 32 percent. In addition to this direct contribution, a host of activities in the services 30 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 sector embracing trade, transport, and finance depended on the plantation econo- my. Local produce accounted for only 25 percent of domestic absorption of goods, and export earnings from the three plantation crops covered over 95 percent of the import bill (Dasgupta 1949). Data on sectoral growth and the changing composition of GDP between 1950 and 1989 are summarized in tables 3.4 and 3.5. The production pattem exhibited only limited structural changes during the immediate postindependence period up to the early 1960s; the pattern then began a profound shift. The GDP share of the primary sector (agriculture, forestry, and fishing) declined from 46 percent in 1960-64 to 37 percent in 1970-77 and 27 percent in 1984-89. By the late 1980s, the share of plantation agriculture was only one-third of the average share in the 1950s, while the share of domestic agriculture had come to exceed that of export agriculture. The share of the industrial sector, which was steady at around 10 per- cent up to about the mid-I 950s, expanded rapidly thereafter, surpassing 20 percent by the late 1980s. This was mostly a reflection of rapid growth in the manufactur- ing sector, the share of which increased from 5 percent in 1960-64 to 13 percent in 1984-89. During the 1960s and until the mid- 1970s, the sector expanded in re- sponse to the impetus given by import restrictions and public sector investment. Since the late 1970s the most dynamic sector within manufacturing has been the garment industry, which has emerged as a major earner of foreign exchange. The share of the services sector increased steadily until the mid-1970s and then ex- panded very rapidly following the 1977 policy reforms. In general, the data show Table 3.4 Gross Domestic Product of Sri Lanka, 1950-89, Sectoral Composition at Constant Factor Costs (average GDP share) Sector 1950-54 1955-59 196044 1965469 1970-77 1978-83 1984-89 Agriculture, 46.6 45.5 46.1 41.4 36.8 24.8 27.2 forestry, fishing and hunting Export agriculture 29.8 27.7 26.7 22.7 17.0 10.2 8.7 Domestic agriculture 14.1 14.6 8.3 15.6 17.5 11.8 14.9 Industry 11.2 11.0 10.3 12.9 16.5 19.0 21.4 Construction 4.2 4.9 2.4 4.7 4.9 4.8 7.3 Manufacturing 5.0 5.1 5.6 7.5 9.1 9.7 12.9 (excluding export processing) Services 42.3 43.8 43.6 45.6 46.7 56.2 51.4 Total GDP 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: Estimates for 1950-70 are at 1959 factor cost; estimates for 1971-83 and 1984-89 are at 1970 and 1982 factor cost. Source: For 1950-78. Savundranayagam (1983); for the rest of the period, Central Bank of Ceylon, Annual Report (annual). Chapter Three 31 Table 3.5 Gross Domestic Product of Sri Lanka, 1950-89, Growth Rates at Constant Factor Costs (average growth rate) Sector 1950-54 1955-59 1960-64 1965-69 1970-77 1978-83 1984-89 Agriculture, 3.9 2.5 2.5 2.4 2.1 4.1 1.1 forestry, fishing and hunting Export agriculture 2.9 1.4 1.5 0.5 -1.7 0.5 1.7 (including export processing) Domestic agriculture 10.2 4.2 7.0 5.6 3.5 5.3 0.6 Industry 3.4 1.7 7.6 12.0 1.0 4.2 3.9 Construction 11.8 2.4 5.0 13.1 -2.6 8.8 1.0 Manufacturing -1.4 1.6 10.1 11.6 2.3 6.4 5.7 (excluding export processing) Services 6.5 2.5 3.7 3.8 3.7 7.2 2.9 Total GDP 4.5 2.5 4.1 4.8 2.9 6.0 2.6 Note: Estimates for 1950-70 are at 1959 factor cost; estimates for 1971-83 and 1984-89 are at 1970 and 1982 factor cost. Source: For 1950-78. Savundranayagam (1983); for the rest of the period, Central Bank of Ceylon, Annual Report (annual). a nontradable bias in sectoral growth performance after 1977. The combined share of nontradable production (services plus construction), which varied between 46 and 50 percent during 1960-77, increased to about 60 percent during 1978-89.8 Savings and Investment Average gross capital formation during the latter half of the 1960s rose to 16 per- cent from around 14 percent in the earlier period (table 3.6), but then fell sharply during 1970-77. This was followed by a massive increase in the investment level, which averaged more than 27 percent of GDP from 1978 to 1983 and then declined marginally to 24.6 percent during the 1984-89 period. The massive increase in do- mestic investment during the post-1977 period required large amounts of foreign capital because of slow growth of domestic savings. The domestic savings ratio changed little until the mid-1980s, when a significant increase was recorded. The public sector share (including public corporations) in total annual gross capital formation increased from 48 percent to 52 percent between 1960-65 and 1971-77. Despite market-oriented policy reforms in 1977, this ratio further in- creased to 58 percent in 1978-83 because of aid-funded public sector infrastruc- 32 Macroeconomic Policies. Crises, and Growth in Sri Lanka. 1969-90 Table 3.6 Savings and Investment, 1960-89 (as a percentage of GDP) Sector 1960-65 1966-70 1971-77 1978-83 1984-89 Domestic savings 12.0 11.8 13.1 13.0 16.6 Investment 13.8 16.1 12.0 27.8 24.6 Foreign savingsa 0.7 4.3 -1.1 14.8 8.0 Composition of investment Private sector 52.5 54.5 48.2 (42.5) Public corporations 15.5 21.0 13.2 (33.5) 76.5 Govemment and public enterprises 32.0 24.5 38.6 24.0 23.5 a. Represents domestic savings-investment gap. b. From 1984 on the figure for the private sector and the figure for public corporations was combined. Source: Appendix tables SA-2 and SA-3. ture development projects. By far the largest share of public sector investment after 1978 went to the Mahaweli project. Labor Force, Employment, Unionization, and Wage Formation The census of 1981 showed an estimated labor force of about 5.0 million (or 34 percent of the total population). The pattern of labor force growth has reflected the combined impact of population growth and the resulting changes in the age struc- ture, the increasing monetization of labor markets, and patterns of labor migration. The labor force grew at the average rate of 1.5 percent per year in 1953-63, 3.3 percent during 1963-7 1, and 2.7 percent in 1971-80. The slowdown in the last pe- riod was partly due to outmigration (Korale 1987). The distribution of the labor force continued to reflect the dominance of the agricultural sector. Some 46 per- cent of the labor force was still employed in agriculture in 1981, compared with 53 percent in 1953. Employment in industry increased from 12 percent to 16 per- cent during the same period (see table 3.7). Yet a large proportion (about two-thirds) of the labor force is engaged in wage employment, a high figure for a developing country with a predominantly rural and agricultural population. One reason is that almost all employees in the planta- tion sector are paid employees. Furthermore, extensive cultivation of cash crops by smallholders and the relatively long exposure of Sri Lankan rural society to commercialization have made wage employment pervasive even outside the plan- tation sector. Labor force participation rates have shown a gradual decline in the 10-19 age group because of improved access to secondary and higher education. In the 15-19 age group the figure fell from 59.2 percent in 1946 to 42 percent in Chapter Three 33 Table 3.7 Sectoral Distribution of Employment, Selected Years, 1953-81 (percent) Sector 1953 1963 1971 1981 Agriculture, forestry, fishing and hunting 52.9 52.6 50.4 45.9 Export agriculture 28.6 24.6 20.1 18.5 Domestic agriculture 22.9 26.6 28.7 27.2 Industry 12.1 12.4 13.4 16.1 Construction 1.9 2.7 3.1 4.2 Manufacturing 9.7 9.2 9.6 12.0 Services 28.4 29.5 28.0 31.3 Activities not adequately described 6.6 5.5 8.2 6.2 Source: Govenurnent of Sri Lanka, Department of Census and Statistics,Statistical Pocket Book (1983). 1981. On the other hand, female participation as well as male participation in the 25-54 age group has steadily risen. In general, participation rates are somewhat higher in the rural areas. This is markedly so in the case of women, due to high rates of female participation in the plantation sector (Korale 1987). Some labor market segmentation arises from the significant differences in the ethnic composition of the work force. Caste barriers to labor mobility, on the other hand, are almost nonexistent, in contrast to the situation in other parts of the Indian subcontinent. Beginning in British colonial times, Sri Lanka's plantations em- ployed a large proportion of immigrant South Indian laborers of Tamil origin. The majority of them were stripped of citizenship rights in 1948 and, through both le- gal and nonlegal barriers, effectively excluded from the wider labor market. Al- though some plantation laborers subsequently gained citizenship, in general their occupational mobility has been low. Linguistic and cultural barriers, as well as low levels of formal education, have restricted the movement of even those who have obtained citizenship rights into other areas of employment. By the mid- 1960s, concern was being expressed about the emergence of open unemployment. But it is difficult to assess the degree of the problem, since unem- ployment figures from underdeveloped and predominantly agricultural countries are known to be unreliable. Estimates of unemployment in the early 1960s ranged from around 10 to nearly 14 percent. Figures for the early 1970s suggested a de- terioration of the employment situation. There was a marked improvement in the early 1980s following the trade liberalization reforms, but more recent figures sug- gest that a noticeable upturn in unemployment may have taken place in the latter part of the decade. Recorded unemployment tends to be higher among the relative- ly well-educated, many of whom seek employment in the public sector. 34 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Trade Unions and Wage Formation Sri Lanka has a long history of trade union organization and worker militancy (Jayawardena 1972). Trade unions first became associated with political parties in the 1920s. After the emergence of Marxist-influenced parties in the 1930s, trade union activities became closely associated with left-wing politics, giving a distinct political favor to much trade union activity. This close association with political parties influenced the evolution of the trade union movement, which often split along party lines. In the 1960s, most of the trade unions were dominated by the LSSP or the cP. The Ceylon Federation of Labor (cFL) and the Ceylon Mercantile Union (cMu), both linked to the LSSP, and the Ceylon Federation of Trade Unions (CFTu), linked to the CP, were the major umbrella organizations. However, their in- fluence among plantation workers was relatively weak. After the disenfranchise- ment of Indian Tamil workers in 1948, they organized themselves into "nonpolitical" unions with a distinct ethnic character, of which the Ceylon Workers Congress (cwc) and the Democratic Workers Congress (DWC) were the largest. The trade unions have been a powerful force in Sri Lankan political and eco- nomic life, but this strength has not always been reflected in traditional expres- sions of union power, such as collective bargaining agreements. This has led some observers, such as Fields (1986), to conclude that unions in Sri Lanka have had little influence on wage-setting. This view is based on an ignorance of the major channel through which trade union power was expressed. In Sri Lanka the effects of union power were most directly and strongly felt in their impact on the "social wage" in a country where consumer subsidies were of enormous significance until the early 1980s. The extension and defense of these subsidies were achieved through the mobilization of union strength. A major change in the post- 1977 pe- riod has been the marked reduction in the industrial and political power of the trade unions (Rodrigo 1988). Collective bargaining has not been a common system of wage adjustment in Sri Lanka, except in some of the highly skilled trades in the private sector. Major wage revisions in the government sector have usually occurred as a result of the recommendations of pay commissions appointed to inquire into the wage struc- ture (Rodrigo 1983). Reviews of this kind took place in 1955, 1969, 1975, and 1982. In addition to these reviews, changes have also been made as a result of pe- riodic cabinet decisions. Trade unions in the public sector have tended to be par- ticularly well-organized, and union agitation over wages was undoubtedly a factor influencing the creation of the commissions and the implementation of their rec- ommendations. Many cabinet decisions on wage increases may have been intend- ed to forestall intensified union activity. In the nongovemment sector the major institutional mechanism in wage de- termination has been Wages Boards, first established in 1941 under minimum wage legislation. The operations of these boards, comprising equal numbers of employer, government, and union representatives, have covered a large proportion of the workers in the nongovernmental sector. In industries that are not covered by Chapter Three 35 Wages Boards the commissioner of labor frequently has the power to fix minimum wages. Institutional mechanisms of arbitration and conciliation, with decisions binding on all parties, have been of particular importance in the urban sector. Col- lective agreements arrived at through direct bargaining between employer and em- ployee organizations are often found in the banking and urban mercantile sectors. The government has been the effective employer of a large proportion of the work force even in the so-called nongovernmental sector (due to the large number of public corporations, including those in the plantation crop sector). In general, the formal sector wage determination process has reflected the role of the state as the country's largest employer as well as the philosophy that state intervention to maintain industrial harmony is a legitimate government activity. Nonetheless, the pattern of wage changes has also reflected changes in the strength of the organized trade union movement. For example, a substantial wage increase in the plantation sector in 1984 was influenced significantly by the govemment's perception of a "need to avoid a restless situation among another minority com- munity" (Rodrigo 1988:xiii). Government policies on wages in the unorganized sector of the labor market, on the other hand, have been quite ineffective. Government legislation on mini- mum wages, for example, has had little or no effect on wages in the large informal services sectors, or in smallholder agriculture or small-scale industry. Commodity and Factor Markets As previous discussions have indicated, Sri Lanka's commodity markets have been subjected to extensive government intervention. This intervention has taken the forms of foreign trade taxes, various implicit and explicit subsidies, quantita- tive controls (most commonly on imports but sometimes on exports), price con- trols, controls on the physical movement of particular commodities, and regulations specifying compulsory sales to government procurement agencies. The most extensive state presence in the primary commodity markets has been in the rice sector, where the government has been involved in the distribution of inputs, the purchase of paddy from farmers under a guaranteed price scheme, and distribution of the rice ration. The latter two functions had their origins in the war period but were continued in subsequent years. The distribution of subsidized inputs became particularly important after the rise in fertilizer prices in the after- math of the oil price increases in the 1970s. The continuation of the paddy pur- chasing scheme in subsequent years was aimed at maintaining producer incentives by counteracting the downward pressure on rice prices through the provision of subsidized rice to consumers. Until the post- 1977 reforms opened up the rice mar- ket, government purchases were made at a higher (guaranteed) price than the price in the open market, and generally at a price above the world price. Such purchases were used to provide some of the rice available to consumers. 36 Macroeconomic Policies. Crises. and Growth in Sri Lanka. 1969-90 Although the government's purchase price was usually higher than the market purchase price, there was large private sector involvement in the domestic paddy trade, since many farmers found it difficult to meet the quality standards set by the state procurement agency. Then, in 1972, the government granted monopoly pro- curement rights to the Paddy Marketing Board (PMB), which had been set up in 1971 to replace the Agrarian Services Department. This was an attempt to reduce the costs of providing the subsidized rice in the face of domestic supply setbacks, price rises in the open market, and steep increases in international prices. The cre- ation of the PMB dislocated traditional marketing channels and created dissatisfac- tion among both producers and consumers because there was no increase in the purchases of the state agency (Central Bank of Ceylon Annual Report-1975). The monopoly was abolished in late 1975. Since then, the role of the state has diminished, but domestic rice prices have continued to be influenced by the government through the release of imports into the market. On the supply side, too, government intervention has been important through the provision of research and extension facilities, subsidized inputs and credit, and investments in infrastructure, particularly irrigation. The input subsi- dies have been scaled down in recent years, but given the electoral strength of the country's rice farmers, it is unlikely that a withdrawal of the government from the rice market is on the horizon.9 The role of export taxes in the production of plantation crops is a continuing theme in this study and will be discussed in detail later. Here, we simply note that they were the most important source of govemment revenue for a long period of time, and that the progressive tax structure enabled the government to obtain a large share of any world price improvements. The state has also intervened in the domestic markets to a limited extent through the licensing of traders, and through the nationalization of the foreign-owned Agency Houses in 1971. In general, there have been few direct price interventions in the domestic tea or rubber markets, but the situation has been different in the coconut sector. Co- conuts are an important commodity in Sri Lanka, where they are consumed in the form of fresh nuts and as coconut oil. the most widely used cooking oil in the country. Coconut exports have usually been processed products, especially copra, desiccated coconut, and coconut oil. Government intervention in the coconut sec- tor has been motivated by factors similar to those operating in the rice sector-that is, they have been aimed at maintaining "reasonable" prices for both producers and consumers. In the case of coconuts there have been the additional objectives of securing government revenues and export earnings. Such interventions were also based on recognition of significant monopsonistic elements in the markets- for example, the relatively small number of oil millers. Until the 1970s, interven- tion was mainly aimed at stabilizing producer prices. A floor price scheme estab- lished in 1968 banned exports below a stipulated minimum, and in 1972 the largest private oil mill was commandeered by the government in an attempt to en- sure better producer prices in the copra market. This was followed in late 1972 by a scheme whereby the government bought oil at a specified price from the millers Chapter Three 37 and sold it to exporters at a price that was based on the world price. Throughout this period, coconut production was relatively stagnant, and rising domestic con- sumption was eroding the exportable surplus. A supply shortfall in 1973 due to drought and insect attack resulted in a steep rise in the prices of fresh nuts used for domestic consumption. The government then banned oil exports to dampen pric- es. In subsequent years such bans were resorted to by the government whenever domestic nut prices rose steeply, but the bans did not extend to desiccated coconut exports. (For an analysis of government policies during this period, see de Silva 1979.) During the mid-1980s a substantial liberalization of export controls was implemented. Supply-side intervention in the plantation crop sector has included provision of subsidized inputs, research and extension facilities, and replanting grants, the latter two activities funded by levies on exports. Similar interventions have been made in the case of many minor agricultural crops, particularly those designated as nontraditional export crops. However, their overall impact is likely to have been relatively minor. The extensive array of quantitative restrictions on imports, the establishment of state-controlled domestic marketing monopolies to distribute rationed goods at regulated prices, and the concomitant development of black markets affected most consumer goods markets from the early 1960s until the late 1970s. The conse- quences are analyzed in greater detail in later chapters. The Financial Sector In the years immediately following independence, Sri Lanka's commercial bank- ing sector was dominated by foreign banks.10 Of the twelve commercial banks, only two were indigenous. Of these two, the Bank of Ceylon was far more impor- tant than Hatton Bank Ltd. The commercial banks were mostly engaged in financ- ing the plantation sector and foreign trade, and were concentrated in Colombo. Under these circumstances, it is hardly surprising that the informal sector was the main source of funds for most of the population to the extent that they utilized cap- ital markets at all. Typically, the informal sector provided short-term credit at con- siderably higher rates of interest than those prevailing in the organized sector (see Tambiah 1963 and Tilakaratna 1963). The major turning points in the evolution of commercial banking were the na- tionalization of the Bank of Ceylon and the launching of a new state-sector bank, the People's Bank, in 1961. These actions coincided with the imposition of ex- change rate controls which almost completely insulated the domestic financial sector from international capital markets. The nationalization inaugurated a period in which credit allocation became a major instrument of government investment policies, particularly in terms of achieving sectoral investment goals. Interest rates were sector-specific, with priority sectors having substantially lower rates. Gov- ernment initiatives taken during this period set the stage for the financial repres- sion that prevailed until the late 1970s. There was also a drive to expand banking 38 Macroeconomic Policies, Crises, and Growth in Sri Lanka, /969-90 services, particularly to the rural sector, and between 1960 and 1965 the number of branch offices rose from 45 to 97. All of this expansion was carried out by the two state banks, which by 1965 accounted for two-thirds of all bank deposits (table 3.8). In addition to the commercial banks, there were three savings banks at the time of independence-the Ceylon Savings Bank (established in 1932), the Post Office Savings Bank, and the National Savings Movement (1944). These three in- stitutions were amalgamated in 1972 to form the National Savings Bank (NSB), which by the mid-1970s accounted for over 50 percent of total savings and time deposits mobilized through institutional sources. There was also an Employee's Provident Fund (EPF), established in 1958 and administered by the Central Bank. The NSB and EPF have been useful sources of captive funds for the government.11 Sri Lanka's first development bank, the Development Finance Corporation of Ceylon (DFCC), was established in 1955 to assist in the establishment of private enterprises in industry and agriculture by providing medium to long-term financ- ing. Its objective was to encourage greater equity financing rather than direct loans, with a view to encouraging companies to meet their capital requirements by issuing shares. Other important financial institutions in the 1960s included private finance companies and insurance companies. The finance companies utilized share capital and bank borrowings, and acceptance of deposits from the public was rare until the late 1960s. Their major activity was the financing of purchases of consumer durables. Up until 1961 the insurance industry was composed of some 122 private Table 3.8 Financial Market Institutions, Percentage Distribution of Assets Institution 1965 1970 1977 198/ /984 Commercial banksa 48.7 51.5 58.8 62.4 57.3 Other deposit-taking institutions 18.4 21.5 17.8 16.4 21.3 FCBULs - - - 3.1 4.6 National Savings Bank 18.4 19.3 16.0 11.6 13.0 Finance companiesb - 2.2 1.9 1.7 3.7 Development finance institutions 2.8 1.3 3.1 3.1 - National Development Bank' - - - 0.4 0.6 Development Finance Corporation 0.8 1.0 0.7 1.9 1.7 Nonbank financial institutions 29.2 24.2 22.1 18.1 18.3 State-owned provident funds 18.9 17.6 15.2 12.2 12.4 State-owned insurance companies 0.9 1.2 2.3 1.8 1.2 Total 100.0 100.0 100.0 100.0 100.0 (millionsofrupees) 3,718 6.112 17,217 50,598 95,901 a. Excluding Foreign Currency Banking Units (FCBUs. b. This figure relates to only those companies which report their assets to the Central Banks. c. Figures for 1965 and 1970 show the combined deposits of Post Office Savings Bank, Ceylon Sav- ings Bank, and the Savings Certificate Fund, which were amalgamated in 1973 to form the National Savings Bank. Source: Khatkhate (1980), supplemented with various issues of Central Bank of Ceylon, Monthly Bulletin of Statistics. Chapter Three 39 companies in a highly competitive market (Karunatilake 1986). Then a state- owned Insurance Corporation of Ceylon (icc) was established and given a monop- oly on life insurance. In 1964 it was granted a monopoly on all other forms of in- surance as well. If the financial system was at a relatively early stage of development in the 1960s, the equity market was practically nonexistent. Although company shares and stocks had been traded in Sri Lanka since the late 19th century, the business was tra- ditionally carried out by commodity brokerage firms as a part of their general com- modity trading under an institutional set-up devised by the Share Brokers Association. The activities of the share market took the form of direct dealings among brokers. Historically, plantation companies accounted for more than 90 percent of all shares listed for trading, but this began to change in the early 1960s as the available shares of industrial, financial, and commercial companies started to increase. Nonetheless, the market was in the doldrums during much of the period be- tween 1960 and 1977, and annual share turnover declined from 19.8 million ru- pees in 1960 to 7 million rupees in 1976 (Ceylon Brokers' Association-1977). Constraints on foreign exchange remittances dissuaded foreign companies from selling shares, and only a tiny minority of Sri Lankans owned any shares at all (Karunatilake 1986). Issuance of bonds became an important source of long-term financing for the govemment in the mid-1950s, but these bonds accounted for a relatively small fraction of total public debt until the late 1970s. Sales of government securities were handled by the Central bank, and the bulk of the purchases was made by in- stitutional investors. Treasury bills were mainly absorbed by the Central Bank. As with government securities, there was little or no private investor participation in this market, and the rate of interest was generally below market rates. There was no secondary market. Financial sector reforms in 1978 (chapter 2) had a major impact on the sys- tem. Fourteen foreign banks opened branches in Colombo in 1979, bringing the total number of commercial banks operating in Sri Lanka to 21. These new banks aggressively pursued customers, and their share of total bank deposits increased from 5 percent in 1980 to 15 percent in 1983. The share of commercial banks in total assets of capital market institutions declined after 1981 because of rapid ex- pansion in the operations of nonbank financial institutions, notably finance com- panies, foreign currency banking units (FCBUs), leasing companies, and the Development Finance Corporation. Although the domestic capital market's ties to world capital markets have improved via the foreign banks and the FCBUs, controls on capital movements have interfered with market integration. The share market gained strength in the aftermath of financial and trade lib- eralization in 1977. During the period 1980-84 there were eighty-seven new is- sues of company shares with a total value of 2,307 million rupees. By contrast, there were only twelve new share issues during the period 1960-77, with a value of about 230 million rupees. By the latter part of the 1980s there were about 170 companies listed on the Sri Lankan stock exchange. 40 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 The share market in Sri Lanka remained essentially a primary market in the 1980s, and more than 90 percent of sales consisted of purchases of new issues. Hence, the total volume of transactions in a given period was only a small fraction (about 6 to 10 percent) of the total value of listed shares. Debentures and notes were not listed or traded. It has been estimated that only about 7,000 people in Sri Lanka hold shares in companies listed on the stock exchange (Karunatilake 1986). A large proportion of all shares is in the hands of closely held companies, and the tendency of these companies to withhold profits from shareholders makes it diffi- cult for a secondary market to emerge. The issuing of securities has been an important source of long-term financing for the govermnent since the mid- 1950s, and reliance on such securities took a sharp upturn after 1977. By 1985 the value of government securities outstanding was 36,570 million rupees, compared with 6,963 million rupees in 1976. The value of securities as a percentage of total public debt increased from 16 to 28 between these Table 3.9 Interest Rates, 1965-87 National Savings Bank Commercial bank Year deposit rate Deposit rate lending rate 1965-70 3.7 (-2.4) 3.7 (-2.4) 9.0 (-2.9) 1971-76 7.2 (-8.4) 5.0 (-11.1) 8.5 (-7.1) 1977-83 14.9 (1.3) 13.8 (0.2) 16.3 (2.5) 1984-87 14.2 (5.2) 9.8 (0.1) 20.2 (12.9) 1977 11.1 (8.1) 10.6 (7.7) 11.3 (8.4) 1978 12.6 (2.3) 11.4 (I.1) 12.8 (1.5) 1979 12.1 (-5.1) 11.8 (-5.3) 13.8 (-3.2) 1980 17.0 (2.3) 17.5 (2.9) 17.2 (2.6) 1981 17.2 (0.3) 17.8 (1.0) 8.4 (1.6) 1982 17.8 (-5.1) 13.6 (-9.3) 20.0 (-2.9) 1983 16.5 (5.1) 14.1 (2.7) 20.6 (9.2) 1984 16.7 (4.6) 12.6 (0.5) 20.5 (8.4) 1985 14.3 (3.3) 11.2 (0.2) 20.1 (9.1) 1986 12.8 (8.1) 7.5 (0.3) 20.3 (13.1) 1987 12.8 (4.8) 7.4 (-0.5) 19.8 (11.9) a. In each column, the first figure indicates the nominal rate and the second (bracketed) figure the real rate. The deposit rate is the weighted average of savings deposit and one year fixed deposit rates. The lending rate is the weighted average rates of loans classified by securities. The real rate is the dif- ference between the current nominal rate and the inflation rate as measured by the GDP deflator (non- services) for the previous year. b. Annual average. Source: Columns I and 2: compiled using data from Central Bank of Ceylon, Review of the Economy (annual). Column 3, Central Bank, Monthlly Bulletitn of Statistics (various issues). Chapter Three 41 two years. This may have been due in part to the fact that the maturity period on gov- emmnent issues was shortened from ten or twelve years to three years, and the rate of interest on the securities was increased from below-market level to 16 percent. In April 1981 the Central Bank inaugurated a secondary market in Treasury bills. Under this arrangement the Central Bank participates in the market by buy- ing and selling bills with varying maturities at interest rates that are determined day to day in light of market rates. Secondary market transactions increased from 7,542 million rupees in 1981 to 39,587 million rupees in 1984, the commercial banks being the major participants. Participation in the secondary market allowed the Central Bank to prevent sharp fluctuations in short-term interest rates and to siphon off excess liquidity in the banking system (Hettiarachchi 1986). Financial Intermediation and Credit Throughout the period 1960-77, pervasive restrictions on the financial system and negative real interest rates produced a process of financial disintermediation (ta- bles 3.9 and 3. 10).12 The degree of financial intermediation, as measured by the M3/GDP ratio, declined or remained stagnant in most years. By the mid- 1970s, this ratio was below the level recorded in the early 1960s. Financial market reforms in 1977 were followed by a remarkable improve- ment in financial intermediation. The M3/GDP ratio rose from an average level of Table 3.10 The Behavior of Real Monetary Aggregates Mi/GDP M21GDP M31GDP Year MIla M2a M3a (percent) (percent) (percent) 1960 61 37 37 17 22 29 1965 87 55 53 20 26 33 1970 83 65 62 14 22 29 1975 59 43 47 12 17 25 1977 69 54 55 15 24 32 1978 82 7 1 73 14 26 34 1979 94 87 89 15 29 38 1980 100 100 100 14 30 38 1981 96 109 108 12 31 40 1982 107 131 133 12 31 40 1983 116 137 141 12 31 40 1984 113 138 145 10 28 38 1985 127 154 165 12 30 41 1986 136 154 168 12 28 40 1987 146 161 174 13 30 41 a. At constant 1980 prices (]980=100). Source. Athukorala and Rajapatirana ( 1990); data in appendix table SA-2 1. 42 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 29 percent in the 1970-77 period to 34 percent in 1978 and stabilized at around 40 percent in the 1980-87 period. While M I continued to increase in real terms, how- ever, its ratio to GDP declined. This suggests a shift in household asset portfolios from cash holdings to savings and time deposits in response to favorable interest rates (Athukorala and Rajapatirana 1990). Table 3.11 shows the distribution of credit in the public sector, public corporations, and the private sector, and the distribution of credit to the public sector by lending insti- tutions. The govemment share of total domestic credit increased from 58 percent in 1970 to 64 percent in 1979, and then declined to about 40 percent in 1985. Throughout the period, almost half of the government's domestic credits came from the Central Bank, and most of the remainder from the two captive lenders, the NSB and the EPF. Table 3.12 presents data on the sectoral distribution of commercial bank credit. Commerce, including foreign trade, accounted for about 50 percent between 1965 and 1984. Credit to industry, the second largest category, stabilized at around 25 percent from 1970 to 1984. Several trends can be observed in the pattem of sectoral credit al- location after 1977. First, a decline in agriculture's share of credit, second, a rise in housing's share of credit, and third, a fall in consumers' share of credit. Loans to the industrial sector accounted for the bulk of loan approvals of the long-term credit institutions. Table 3.11 Total Credit Extended By Non-Central Bank Financial Institutions, Selected Years, 1977-85 (1980 prices) Sector 1977 1979 1981 1983 1985 Total credit 15,361 25.669 33,990 43,963 47,798 (millions of rupees) Govemment, net (percent) 49.1 40.7 33.0 32.5 32.0 Public corporations (percent) 13.4 15.1 12.3 7.3 6.0 Co-operatives (percent) 10.0 8.7 3.5 3.1 1.8 Private sector (percent) 27.5 35.5 51.2 57.1 60.2 Govemment credit sources, net (millions of rupees) 7,426 10,465 11,241 14,331 15,291 Commercial banks (percent) 0.7 -2.3 0.5 -0.9 -1.9 National Savings Bank (percent) 40.9 46.7 44.7 51.7 57.9 Employee Provident Fund (percent) 40.7 43.1 46.2 45.8 56.7 Insurance companies (percent) 16.6 14.6 12.2 7.0 6.2 Source: Central Bank of Ceylon, Review of the Economy (annual). Table 3.12 Commercial Bank Loans to Private and Public Sector Corporations Distribution 1970 1975 1977 1978 1979 1980 1981 1982 1983 1984 1985 Total loansa 6,226 6,341 7,222 11,491 13,769 17,136 18,916 21,780 23,273 22,241 24,476 Index(1980=100) 36 37 42 67 80 100 110 127 136 130 143 Distribution by sector (percent) Tradable sectors 40.0 36.5 40.6 43.1 40.7 35.6 35.2 36.6 36.5 33.8 33.0 Agriculture 13.1 14.1 17.2 16.0 14.8 13.2 13.2 12.1 10.9 9.2 9.3 Industryb 26.9 22.4 23.4 27.1 25.9 22.4 22.0 24.5 25.6 24.6 23.7 ,4 (21.6) (18.6) (21.1) (24.3) (22.3) (18.4) (17.6) (19.3) (20.7) (19.8) (19.0) Nontradable sectorsC 60.0 63.5 59.4 59.4 59.3 64.4 64.8 63.4 63.5 66.2 67.0 Maturity structure (percent) Short-termn 58.9 74.6 77.5 74.0 75.0 70.2 68.4 68.9 71.0 72.2 70.6 Medium-term 28.7 19.3 16.5 20.4 19.4 21.2 20.4 22.2 20.4 18.6 19.4 Long-term 12.3 6.1 6.0 5.5 6 8.6 10.2 8.9 8.6 9.2 10.0 a. Millions of rupees at 1980 prices. b. Percentages of loans to manufacturing are given in parentheses. c. Sum of loans for commercial, financial, housing and consumption. Source: Central Bank of Ceylon, Monthly Bulletin of Statistics (various issues). 44 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Informal Money Markets In Sri Lanka, as in many other developing countries, the informal money market has played a very important role in the provision of credit, though its importance has been declining over time because of the gradual expansion of the formal market. In 1981-82, about 60 percent of total household borrowing (which includes borrowing by informal businesses and cottage industries) was provided by the informal market, as against 80 percent in 1973. The use of informal sources of credit is by no means confined to households and small businesses. It is believed that the Pettah money market (the informal money market in Colombo) has been an important source of funds to many commercial enterprises needing very large amounts of money. The available evidence suggests that the informal money market in Sri Lanka is segmented, noncompetitive, and not well-integrated with the formal market. In- formal loans come from a wide range of sources, in contrast to the general tenden- cy to think of informal credit as synonymous with pure money lending. The share of pure moneylenders in total informal borrowing of households seems to be less than 25 percent. About one-third of the borrowed money comes from friends and relatives. The rest is mostly related to land tenure relations and the systems of product marketing and consumer credit. Most of the credit supplied by money- lenders is extended at interest rates that are much higher than those of institutional credit, while credit from the other sources is granted either free of interest or at rates much below institutional rates. The data indicate that informal credit is ob- tained mostly for "nonproductive" purposes. A major failing of the informal sys- tem has been the almost complete absence of lending for long-term investment, such as the planting of long-gestation crops (Jayasuriya, Barlow, and Shand 1981). Price Movements Table 3.13 indicates variations in the annual rate of inflation in terms of four price indices: the Colombo Consumer Price Index (CCPI) as adjusted by Bhalla (1987), the GDP deflator (GDPD), the Wholesale Price Index (wPi), and the Central Bank Consumer Price Index (CBCPI). 13 Up until the early 1960s the annual inflation rate remained around two percent. During this period the domestic money supply in- creased rapidly because of the increased reliance of the government on the Central Bank for deficit financing. Yet the economy was able to meet the resultant addi- tional demand through an increased supply of imported goods, drawing upon for- eign reserves accumulated during the Korean War boom (1950-52) and the tea price boom (1954-55) (Gunasekara 1963). By the early 1960s, when a stringent import-control regime evolved in response to depletion of foreign reserves, mon- etary expansion began to bring about increases in the price level. Inflationary pressure was further aggravated in the early 1970s by domestic supply bottlenecks resulting from a scarcity of imported intermediate goods and increases in import prices brought about by the first oil price shock. After 1977, significant depreciation of the external value of the rupee and domestic demand Chapter Three 45 Table 3.13 Changes in the Price Level, 1950-89 (annual percentage changes) Year CCP/ GDPD WPI CBCPI 1950-59 0.7 2.2 - - 1960 -1.9 7.3 - - 1961 1.6 -1.5 - - 1962 1.4 -0.4 - - 1963 2.4 1.3 - - 1964 3.1 1.1 - - 1965 0.3 -2.2 - - 1966 -0.2 3.5 - - 1967 2.2 2.9 - - 1968 5.8 14.7 - - 1969 7.4 8.9 - - 1970 5.9 -0.2 - - 1971 2.6 3.5 - - 1972 6.2 7.1 - - 1973 9.9 15.6 - - 1974 12.1 22.9 - - 1975 6.7 8.0 3.0 3.0 1976 1.2 6.3 8.6 3.8 1977 1.2 20.3 20.9 15.0 1978 11.8 7.6 15.8 8.9 1979 10.9 8.8 9.5 18.6 1980 26.2 21.5 33.7 37.1 1981 17.9 16.0 17.0 23.8 1982 10.8 17.1 5.5 11.1 1983 13.9 14.5 24.9 11.3 1984 16.6 17.1 25.6 - 1985 1.4 0.9 -15.2 - 1986 7.9 5.8 -2.9 - 1987 7.7 7.0 13.3 - 1988 13.9 11.5 17.7 - 1989 11.6 9.7 9.0 - Notes: ccPI-Colombo Consumer Price Index; GDPD-Gross Domestic Product Deflator; wpi- Wholesale Price Index; cBcpi-Central Bank Consumer Price Index (unpublished). Source: Appendix table SA-7. pressure generated by the massive investment in the public sector and migrant worker remittances added to inflation. Whatever the price index used, the average annual inflation rate during the 1980s was almost twice that in the 1970s. Sri Lan- ka, however, has never experienced a period of hyperinflation. Even allowing for the deficiencies of the indices and the acceleration of the inflation rate in the 1978- 84 period, strong inflationary tendencies did not become entrenched in the econo- 46 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 my. Thus, it can still be considered a low-inflation country by average developing country standards (Corden 1991). The attitudes of policymakers of varying political persuasions toward infla- tion have been remarkably similar, and strong anti-inflation attitudes have also been entrenched in the Central Bank, which inherited a legacy of monetary con- servatism from the British. The Role of the State The direct role of state enterprises in the economy has been quite extensive, in terms of both policy formation and direct impact on fiscal revenues and expendi- tures. So far, however, no attempt has been made by official Sri Lankan agencies to determine the components of total value added in the public and private sectors. But value-added disaggregations are available for 1961, 1966, and 1974, thanks to a research project sponsored by the Canadian International Development Re- search Corporation (IDRC) in the mid-1970s (Lakshman 1979). As shown in table 3.14, the IDRC found that public enterprises accounted for 5.7 percent, 7.3 percent, and 12.2 percent of Sri Lanka's GDP in those three years. If allowance is made for the second stage of the takeover of plantations in 1975 and the takeovers that oc- curred under the Business Undertakings Acquisition Act during 1972-76 (dis- cussed in chapter 2), the proportion would no doubt have exceeded 15 percent by the end of the United Front regime in 1977. The importance of the public sector has been greatest in the finance and in- surance sector, where the public enterprise share rose from 50 percent in 1960 to 94 percent in 1974. This took place during a period when the activities of foreign- owned banks had been severely restricted through discriminatory legislation. Ac- cording to Central Bank data, the share of the two state-owned banks (the Bank of Ceylon and the People's Bank) in total commercial bank assets increased from 35 percent in 1965 to 74 percent in 1974. The very high public sector share in the util- ity services sector was due to the monopoly position of the Electricity Board. The public enterprise shares of the transport and manufacturing sectors increased from 36.5 to 45.6 percent, and 1.8 to 33.9 percent, respectively, between 1961 and 1974. Although the public sector share in wholesale and retail trade declined from 5.4 per- cent to 4.1 percent between 1961 and 1966, it then increased to 10.6 percent in 1974. The public sector share of manufacturing value added declined from 33 per- cent in 1974 to 29 percent in 1981 (Athukorala 1986). Judging by the public sector manufacturing output index (table SA-4), we surmise that this declining trend per- sisted in subsequent years. There was also a significant decline in state involve- ment in foreign trade. The liberalization of import trade and the removal of various restrictions on internal trade in 1977 resulted, in all likelihood, in a significant de- cline in the public sector share of wholesale and retail trade as well. With regard to the other sectors, though, there is no evidence that suggests any decline in public sector involvement. The liberalization measures that expanded the Chapter Three 47 Table 3.14 Percentage Contribution of Public Enterprises to Sectoral Value Added and Total GDP Sector 1961 1966 1974 Agriculture, forestry, hunting and fishing 0.1 0.1 0.5 Mining and quarrying 0 0 5.5 Manufacturing 1.8 7.6 33.9 Electricity, gas and water 88.9 90.9 93.5 Construction 0 4.2 6.8 Wholesale and retail trade 5.4 4.1 10.6 Transport, storage and communication 36.5 39.0 45.6 Finance and insurance 49.6 52.4 94.5 Other services - 1.2 1.0 Gross domestic product 5.7 7.3 12.2 Source: Lakshlnan (1979), table 4.3. role of private banks in commercial banking and the activities of private sector fi- nance companies seem to have been counterbalanced by government involvement in the services and construction sectors. In short, the "open economy" strategy adopted in 1977 did not produce a significant reduction in the role of public enterprises. Such enterprises appear to have accounted for about 6 percent of total em- ployment by 1973 (Lakshman 1979, table 7.6). With the nationalization of plan- tations, this would have increased to about 8 percent by 1976. The public sector share of total employment in the "organized" manufacturing sector increased from 25 percent in 1971 to 51 percent in 1979, but then declined to 39 percent in 1982. At least a part of this relative decline can be attributed to the growth of the mainly privately owned gannent industry. By the mid-1980s, more than 40 percent of all paid employees in the country were in the public sector (Fields 1986). This figure is unlikely to have changed much in the latter part of the decade. Fiscal Operations Table 3.15 presents summary data on government fiscal operations. Up until 1977. current expenditures accounted for about 75 percent of total expenditures, but the share then varied from 45 percent (in 1978) to 59 percent (in 1985). This was due to the rapid expansion of public investment referred to earlier, rather than to any significant decline in the absolute level of current expenditures. The composition of current expenditures demonstrates the major role of wel- fare expenditures in government fiscal operations. Combined expenditures (cur- rent and capital) on social welfare (mostly health, education, and food subsidies) accounted for about 40 percent of total government expenditures annually 48 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table 3.15 Summary of Fiscal Operations, 195089 (percentage of GDP) Expansionary impact of Governmenit fiscal operations savings (bank borrowing Total Recurrent (current account Budget for deficit Year Revenue expenditures ex-penditures surplus/deficit) deficit financing) 1950 21.8 23.2 - 1.4 - - 1955 21.6 18.9 2.7 - - 1960 22.5 29.6 - 7.1 - - 1966 23.6 32.4 24.0 -0.3 8.8 1.5 1970 20.7 29.4 20.1 0.5 8.7 3.2 1971 21.5 30.2 21.8 -1.2 9.7 1.6 1972 22.2 31.5 23.0 -0.7 9.2 0.8 1973 22.5 30.4 21.5 0.9 7.9 0.3 1974 20.5 27.4 19.3 1.2 6.8 0.1 1975 19.7 30.2 20.0 0.2 10.3 0.8 1976 20.4 33.2 19.8 0.6 12.7 2.3 1977 19.2 28.1 17.1 1.5 8.8 -2.3 1978 28.8 46.5 25.7 3.1 17.7 0.4 1979 25.5 43.2 23.1 2.4 17.6 1.2 1980 22.6 48.7 21.2 0.3 21.6 10.3 1981 20.4 39.1 20.1 0.2 18.7 4.6 1982 19.4 41.3 21.9 -2.5 21.0 5.2 1983 22.6 42.0 21.5 1.1 19.4 0.4 1984 26.9 38.3 20.7 6.3 11.4 0.5 1985 25.0 44.7 22.4 2.5 19.7 0.8 1986 22.8 33.0 18.9 1.8 12.2 1.7 1987 23.8 32.5 20.1 1.3 11.1 1.8 1988 21.7 34.3 20.7 -2.0 15.6 4.7 1989a 24.1 31.6 22.0 -0.4 -10.0 -0.4 a. Provisional. Source: Appendix table SA-10. between 1960 and 1977 (table 3.16). Food subsidy expenditures were the largest single item, accounting for about 40 percent of all welfare expenditures. In 1978, however, the food subsidy was abolished for nearly one-half of the population, and the system of providing specified quantities of food at subsidized prices was replaced by a food stamp scheme targeted at families with less than a specified amount of monthly income (initially set at 300 rupees). Because of these changes, food subsidy expenditures as a share of total expenditures declined from 16 percent in 1977 to a mere 0.2 percent in 1985. The real value of food stamps fell steeply, since their nominal value was not adjusted for inflation (Ravallion and Jayasuriya 1988). In recent years the fastest-growing category of government ex- penditures has been transfers (mostly capital transfers) to public enterprises. As Lal and Rajapatirana (1989) have pointed out, the increase in these transfers out- weighed the expenditure reduction resulting from curtailment of food subsidies. Chapter Three 49 Table 3.16 Social Welfare Expenditures Percentage of total government Percentage Category Year expenditures of GDP Education Health Food 1955 33.9 6.4 46.6 31.4 10.2 1960 33.3 9.8 45.3 24.3 18.3 1965 44.9 13.6 35.1 16.5 44.2 1970 37.5 12.4 43.2 18.1 42.0 1977 30.4 8.1 32.9 16.7 37.8 1980 8.5 3.8 37.6 18.1 40.1 1982 8.1 3.2 45.3 20.3 28.1 1985 7.8 3.1 46.7 21.6 23.5 Source: Central Bank of Ceylon, Review of the Economy (annual). Government expenditures in the 1950s amounted to about 20 percent of GDP and then rose to around 30 percent during the period 1960-77. After 1977 the figure rose to about 45 percent. Government revenues also increased, but at a slower rate. Throughout the period 1959-85, total revenues as a ratio of GDP remained in the narrow range of 20 to 25 percent. The outcome, naturally, was increasingly larger budget deficits. Table 3.17 sets out data showing the changing pattern of government reve- nues for the period 1959-85. During this period, taxation accounted for 80 to 89 percent of total revenues, the balance coming from a variety of other sources, such as receipts from trading enterprises (five to eight percent), income from invest- ments, sales of capital goods, and land rents. As in most other developing coun- tries, a striking feature of Sri Lanka's tax structure is the dominance of indirect taxes. The contribution of these taxes to total tax revenue has varied from 78 to 83 percent. Among indirect taxes, export and import duties have historically been the largest single source of revenue. In the 1950s they contributed 56 to 62 percent of all revenues and then declined to about 25 percent by the mid- 1970s. Export taxes Table 3.17 Source of Government Revenues, 1950-85 Source 1950-59 1960-69 1970-77 1978-83 1984-85 Export duties 29.3 15.3 13.1 19.3 7.9 Import duties 28.5 27.5 14.7 18.5 22.3 Othertaxes 38.3 47.2 51.5 58.0 63.3 Nontax revenue 3.9 l0.Oa 20.7a 4.2 6.5 a. Includes receipts from Foreign Exchange Entitlement Certificates (1968-77). Source: Central Bank of Ceylon, Review of the Economy (annual). 50 Macroeconomic Policies. Crises, and Growth in Sri Lanka, 1969-90 declined because of unsatisfactory revenue growth for exports of tea and rubber after the mid- 1950s. while import taxes fell because of slow growth of imports and the virtual elimination of "nonessential" items from the import list (Athukorala and Huynh 1987). However, revenues from taxes on foreign trade continued to re- main larger than those from domestic sources, owing to the FEECS. Following the 1977 reforms, the combined share of export and import duties in total revenue rose sharply, reflecting the immediate impact of exchange rate depreciation and a sud- den surge of imports to meet pent-up demand. Since then, the percentage has come down to around 25 to 30 percent. Export duties have been levied mainly on tea, rubber, and coconut products. As mentioned earlier, the duty lists were based on a sliding scale that substantially reduced gains to producers in periods of high prices. The immediate impact of the devaluation of 1977 was to raise the rupee prices of these exports, and this resulted in large revenue gains to the government. Substantial changes in the export crop taxation system were implemented in the 1980s, reducing the tax burden on those industries. The reduction in tax rates, combined with the stagnant volume of the main export crops, reduced the share of these taxes in total government revenue in the latter part of the 1980s. As a part of export promotion, duties on exports of most of the minor agricultural products were either sharply reduced or abolished during the period 1978-80. As a result, the combined share in total export tax rev- enue of taxes on minor exports declined from 10.4 percent in 1977 to 5.5 percent in 1985. There were no taxes on manufactured exports during the study period. Revenues from taxes on domestic production and trade as a share of total tax revenues expanded rapidly in the past two decades. Since the early 1980s this share has exceeded that of taxes on foreign trade. The Business Turnover Tax (in- troduced in 1964) has been the most important domestic production tax. Revenues from direct taxes ranged from 17 to 22 percent of all revenues dur- ing the period under study. This relatively low figure reflected the underdeveloped state of the economy and the weaknesses of the tax collection system. Among di- rect taxes, the corporate income tax is the most important. Its share of total tax rev- enues has remained in the range of 10 to 13.4 percent over the past decade. The relative share of taxes on personal income showed a mild declining trend over the same period. from five percent in 1974 to four percent by 1985. Despite the resur- gence of private business activity after 1977, corporate income taxes failed to show buoyancy. That may be partly due to the wide array of tax exemptions that were offered to private enterprise, but tax evasion may have been equally respon- sible. A study conducted during the early 1970s estimated that the loss of income tax revenue due to evasion could be as high as 60 percent of the total collected an- nually. There is little evidence that there has been a significant improvement in the efficiency of tax collection since then. Chapter Three 51 Foreign Trade and Balance of Payments The modem Sri Lankan economy has been heavily trade-dependent. At the time of independence the trade dependence ratio (TDR), or total imports and exports of goods and nonfactor services as a share of GDP was 70 percent and remained vir- tually unchanged throughout the 1950s (table 3.18). From then up to 1977 it de- clined secularly, reflecting the combined impact of sluggish export performance and stringent import controls. Following trade liberalization in 1977 it increased dramatically, reaching 83 percent in 1980, and then declined to about 65 percent on average during the 1980s. Throughout the postindependence period the level of the TDR has been dominated by the behavior of imports; the exports-to-GDP share has been either stable or declining. In particular, the increased openness of the economy during the 1977-89 period showed itself almost entirely in a surge in imports rather than in export dynamism. Table 3.19 provides data on the changing composition of foreign exchange re- ceipts. In the 1950s and 1960s, commodity exports accounted for more than 90 percent of total foreign exchange receipts. Earnings from tourism and remittances by migrant workers then began to make a significant contribution. The Ceylon Tourist Board was established in 1966, and tourism then became one of the fastest- growing sectors of the economy until ethnic conflict in 1983 brought its growth to a halt.'4 Even in 1985, when the industry was already in decline, the share of re- Table 3.18 Trade Dependence: Imports and Exports as a Percentage of GDP Ratio of Ratio of Ratio of imports and imports exports exports Year to GDP to GDP to GDP 1950 31.8 38.3 70.1 1955 35.7 35.5 71.2 1960 31.5 28.2 59.7 1965 25.6 25.5 51.1 1970 17.6 15.3 32.9 1975 20.7 15.2 35.9 1977 18.1 19.1 37.2 1980 54.5 28.3 82.8 1983 45.2 23.0 67.1 1986 35.4 23.8 59.2 1989 37.2 27.3 64.5 Note: Exports and imports include both goods and nonfactor services. Ratios are estimated using cur- rent price data. Source: Central Bank of Ceylon. Review of the Economy (annual) and Annual Report (annual). 52 Macroeconomic Policies. Cr ises, und Growth in Sri Lanku, 1969-90 Table 3.19 Composition of Current Account Receipts in the Balance of Payments (percent) Commodity Services Private Public Total Year exports total Travel transfersa transfers (SDR mn) 1950 94.3 4.2 - - 314 1955 94.3 4.8 - - 421 1960 93.8 5.4 - - - 402 1970 83.3 13.5 1.9 -1.2 4.4 341 1975 77.7 11.2 3.1 0.3 10.7 597 1980 71.2 11.1 6.1 9.3 8.4 837 1985 57.1 18.9 8.3 14.9 9.1 1,609 1987 63.1 15.7 3.9 12.8 8.4 2,053 1989 60.6 17.3 3.1 13.4 8.7 1,776 a. Composed mostly (95 percent or more) of remittances of Sri Lankan workers working overseas. Source: Compiled from Central Bank of Ceylon, Annual Report (annual). ceipts from tourism in total current account receipts was eight percent, as against two percent in 1970. But during the late 1980s the government was compelled to provide considerable relief to the tourist industry, which suffered from massive excess capacity and found itself unable to service the debts incurred earlier in the hotel construction boom. As the tourist industry lost its momentum, remittances from migrant workers became a major source of foreign exchange receipts. The share of these remittanc- es in total current account receipts increased from two percent in 1980 to 13 per- cent in 1989. By mid-1985, "exported workers" had become the single most important source of foreign exchange. surpassing earnings from tea exports. Re- flecting the changing economic circumstances of the Middle East oil exporters, growth in remittances slackened toward the end of the decade. Until the early 1970s, the three plantation crops accounted for over 90 percent of total merchandise exports, with tea alone accounting for nearly two-thirds of the total (table 3.20). Since then, gems and garments have also become important sources of foreign exchange. By 1989 the combined share of the "traditional tri- ple" in merchandise exports had declined to 40 percent, while manufactured ex- ports accounted for 46 percent (garments alone accounting for 30 percent).15 As can be seen in table 3.21, the composition of imports also changed over time, with a distinct shift toward producer goods (investment and intermediate goods). The decline in the share of food imports reflects the expansion of import- substitution production, both in manufacturing and domestic (food-production) agriculture. Sri Lanka's manufacturing sector is heavily dependent on imported inputs. Out of the total real output of this sector, about 60 percent is directly Chapter Thiree 53 Table 3.20 Export Structure, Selected Years, 1974-89 (percent) Sector 1974 1977 1980 1984 1989 Major agricultural products (tea, rubber and coconut) 71.6 72.8 54.4 57.6 39.4 Minor agricultural products 8.5 7.7 7.2 5.1 4.3 Minerals 3.9 5.0 4.6 2.5 5.6 Manufactured goods (excluding petroleum) 3.6 5.6 14.4 25.5 46.7 textiles and garments 0.8 2.2 10.4 20.5 31.4 Petroleum productsa 10.1 9.1 17.8 8.9 4.0 Total exports (SDR million) 433 629 812 1,431 1,216 a. Sales of aviation fuel by the state-owned petroleum refinery (import content of which is over 90 percent). Source: Appendix table A- 14 and Central Bank of Ceylon, Review of the Economy (annual). dependent on use of imported inputs (Cuthbertson and Athukorala (1990), table 2.24). The direct import content of gross domestic fixed capital formation varied between 30 and 40 percent over the period 1970-84 (Cuthbertson and Athukorala (1990), table 2.25). Exchange rate policy played little or no role as an instrument for achieving external balance until the reforms of 1977. A fixed rate with almost no change in par value was maintained until the late 1960s, and even the introduction of a mul- tiple rate in 1968 did not affect the commitment to a fixed rate. Plans to make the value of the FEECs reflect market forces were soon aborted. The switch to a more active policy after 1977 permitted a steady depreciation of the nominal rate in the subsequent period under the managed floating regime, but, as will be discussed later, did not prevent overvaluation of the real rate. The achievement of full con- vertibility was not seen as a realistic goal within the planning horizon, thus reflect- Table 3.21 Import Structure, Selected Years, 1950-89 (percentage shares) Sector 1950 1960 1970 1975 1980 1984 1989 Consumer goods 72.2 61.0 56.0 50.5 29.9 23.2 24.5 food 45.0 38.3 46.3 48.0 18.9 10.4 14.5 Intermediate goods 15.2 20.1 20.0 36.0 47.7 50.0 56.4 Investment goods 12.6 18.1 23.0 12.4 24.0 25.6 17.6 Note: Textile and clothing imports previously classified as consumer goods were classified as inter- mediate goods in 1989. Source: Appendix table SA-13 and Central Bank of Ceylon. Annual Report (1989). 54 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 ing the incomplete nature of the capital market reforms and concern about the consequences of a market-based exchange rate. Sri Lanka's balance of payments position has been characterized by widening deficits in the merchandise and current accounts, with the exception of small sur- pluses recorded in 1965 and 1977 (table 3.22). This has been the outcome of ad- verse movements in the terms of trade and stagnation in export volume, which together reflected a continuing erosion in the import purchasing power of export earnings. After the boom in tea prices in the mid- 1950s, Sri Lanka experienced a sharp long-term deterioration in its commodity terms of trade because of adverse world market conditions (table SA- 18). Escalation of import prices, notably after the on- set of the first oil crisis, intensified the terms of trade decline. To make matters worse, there were no compensatory movements in export volume. The total export volume index (1978=100) increased from about 80 in the early 1950s to 110 in the mid- 1960s, and then fluctuated between 110 and 90 without showing any trend un- til the late 1980s. In the absence of export volume growth, the index of import pur- chasing power of export earnings (income terms of trade) moved in tandem with, or declined at a faster rate than, the terms of trade. The slow growth (or stagnation) in export volume, in turn, reflected not so much a failure to develop new export items as stagnation in exports of the traditional export products. In the 1950s, current account deficits were almost totally financed by foreign reserves accumulated during World War 11, the Korean War commodity boom, or the 1954-55 tea boom. After these reserves were depleted, import restrictions be- came the basic tool for managing the balance of payments up to 1977. However, given persistent adverse movements in the import-purchasing power of export earnings, some reliance on foreign financing was unavoidable. Sri Lanka's reliance on foreign savings increased to an unprecedented extent after 1977. In addition to the increase in "normal" import flows triggered by lib- eralization, the rise in external dependence reflected the increased use of foreign finance to implement the public sector investment program. The heavy reliance on foreign finance did not lift the debt service burden to unmanageable levels, since nearly 75 percent of the foreign financial assistance received during this period consisted of grants and long-term loans. The debt service ratio (debt service pay- ments as a ratio of earnings from exports of goods and services), which was 18 percent in 1970-77, declined to 16 percent in 1978-83 but then increased to 24 percent in 1984-89. While such a debt service ratio would have been considered to be a very serious problem in earlier times, this figure was not unduly high by developing-world standards. Nevertheless, the rising debt burden was becoming a matter of concern in the early 1990s. Chapter Three 55 Table 3.22 Balance of Payments, 1950-89 (SDR millions) Exports Imports Trade Current Nonmonetary Basic Year (FOB) (CIF) balance account capitala balance 1950 296 248 48 28 -6 221 1951 374 324 50 18 -9 9 1952 296 358 -62 -93 6 -87 1953 314 343 -29 -33 -8 -41 1954 362 290 72 64 4 68 1955 397 310 87 67 -10 57 1956 372 331 41 17 -2 15 1957 350 370 -20 -40 -3 -43 1958 341 359 -18 -32 4 -28 1959 372 411 -39 -43 4 -39 1960 377 421 -44 -46 3 -43 1961 358 376 -18 -19 8 -11 1962 370 400 -30 -29 18 -21 1963 358 392 -34 -36 18 -18 1964 371 411 -40 -33 11 -33 1965 401 403 -2 12 3 15 1966 352 423 -71 -60 25 -35 1967 340 417 -77 -60 28 -32 1968 332 395 -63 -69 36 -33 1969 321 446 -125 -133 44 -89 1970 339 391 -52 -58 30 -28 1971 314 361 47 -36 68 32 1972 284 322 -38 -29 39 10 1973 309 348 -39 -21 57 36 1974 425 583 -158 -113 69 -44 1975 464 632 -168 -91 48 -43 1976 484 557 -73 -5 61 56 1977 651 622 29 117 32 149 1978 675 819 -144 -75 132 57 1979 675 1,037 -362 -117 164 47 1980 818 1,576 -758 -507 306 -201 1981 901 1,596 -695 -381 341 -40 1982 918 1,808 -890 -516 477 -39 1983 993 1,794 -801 -441 420 -21 1984 1,427 1,890 -463 -51 335 284 1985 1,296 2,007 -711 -417 328 -89 1986 1,029 1,678 -649 -363 279 -84 1987 1,077 1,602 -525 -265 234 -31 1988 1,098 1,666 -568 -290 194 -96 1989 1,206 1,748 -542 -290 252 38 a. Direct foreign investment and other private capital (long- and short-term) plus central government loans received. Source: Central Bank of Ceylon, Review of the Economy (annual) and Annual Report (annual). Chapter Four The First Crisis: 1973-75 In this chapter we focus on the 1973-75 crisis, the policy responses, and the short- run economic impact of crisis management measures. A similar treatment of the second crisis episode is undertaken in the next chapter. Since both episodes in- volved actual or potential balance of payments crises, our evaluation concentrates on the management of extemal sector disequilibrium. This chapter begins with an overview of the macroeconomic performance of the Sri Lankan economy during the period 1960-73. Background to the Crisis Between 1960 and 1973, Sri Lanka's GDP (at constant 1970 prices) grew at an an- nual rate of 3.4 percent (table 4.1). Growth was not uniform, however. For in- stance, throughout the period 1965-69 the annual growth rate was higher than the average figure of 3.2 percent during the period 1960-64. Growth accelerated after 1966 and reached a peak of 7.5 percent in 1968, and then slowed down to 3.5 per- cent in 1970. In 1971 the growth rate was negative, mainly because of turmoil in the wake of the jVP uprising. In 1972, as the economy recovered, the rate was 2.1 percent. Performance improved considerably in the following year, when an in- crease of 4.1 percent was recorded. The lower growth record during 1970-72 is largely attributable to the weak performance of the agricultural sector. The youth uprising had its harshest impact in the rural areas. Moreover, land reform in 1972 disrupted production, and annual crops like rice were particularly affected by bad weather. In addition to these fac- tors, coconut production suffered from an insect pest attack in 1973. Meanwhile, the manufacturing sector, which grew at an annual rate of more than eight percent in the 1960s, recorded slower growth, mainly because of shortages of imported inputs. 56 Chapter Four 57 Table 4.1 Movement of Major Economic Variables Before and During the 1973-75 Crisis Episode Annual average Economic variable 1960-69 1970 1971 1972 1973 1974 1975 Economic growth (annual percent) Real GDP 3.9 3.9 -0.8 2.1 4.1 3.3 2.3 Per capita GDP 1.4 1.7 -3.1 0.1 2.1 1.4 0.4 Sectoral growth: Plantation agriculture 2.2 1.0 0.9 5.4 -13.3 -2.3 7.5 Agriculture (other) 10.3 5.1 -6.7 -2.1 13.4 19.8 -2.5 Industry 9.8 17.5 1.9 3.3 3.2 -5.2 6.3 Services 4.5 2.4 -0.7 2.9 3.0 5.2 2.9 Employment growth (annual percent) Total employment 2.6 3.3 4.0 3.3 -1.7 5.5 2.7 Manufacturing employmenta 15.6 14.2 8.5 12.5 -14.0 -7.6 4.0 Consumption, saving and investment Real per capita consumption (1970=100) 90 100 102 89 89 93 94 Real investment (1970=100) 80 100 84 82 77 64 70 Private 75 100 90 91 88 69 73 Public corporations 61 100 76 63 49 38 42 Government 105 100 81 81 76 72 86 Savings/GDP ratio (percent) 12.2 15.5 15.1 15.7 12.5 8.3 8.1 Private savings 12.1 16.4 17.0 17.0 11.8 7.2 8.6 Investment/GDP ratio (percent) 15.4 17.2 15.2 14.5 13.5 12.5 13.9 Private investment 8.5 8.9 9.2 8.6 7.7 8.2 7.9 Incremental capital/ output ratio (percent) 3.5 4.3 4.7 4.1 4.9 3.4 3.3 Prices and wages General price level (percent change) BACCPt 2.6 3.3 4.0 12.9 20.5 25.3 3.7 GDPD 3.5 -0.2 3.5 7.2 15.6 22.9 8.1 Terms of trade Agriculture-manufactufing (1970=lO00)' 95 100 93 89 103 128 98 Plantation sector 97 100 91 83 91 108 73 Domestic agriculture 91 100 94 94 112 142 119 Real wage index (1978=100) Private sectorc 93 87.2 85.0 80.1 68.4 64.1 71.0 Public sectord 157 62.1 62.1 62.3 65.5 73.7 81.5 Real interest rate (annual percent) NDB deposit rate 0.5 -2.0 4.0 1.2 1.6 -14.1 -0.2 Commercial bank deposit rate -2.6 0.7 1.0 0.2 -12.9 -21.5 -0.3 Commercial bank lending rate -2.9 1.2 1.4 0.4 -8.7 -16.3 0.8 (continued on next page) 58 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table 4.1 (continued) Annual average Economic variable 1960-69 1970 1971 1972 1973 1974 1975 Exchange rates Nominal exchange rate (rupees/US dollars) Official rate 5.2 6.0 6.4 6.4 6.5 6.6 7.2 Official rate with FEEC premium 5.9 9.2 9.2 9.9 10.6 10.9 11.6 Curb-market rate 12.6 15.2 18.5 14.4 11.25 12.1 14.8 REER index (1980=100) Total exports 55 51 52 48 55 57 54 Traditional exports 72 64 63 61 58 60 57 Nontraditional exports 29 36 41 47 53 54 59 Manufactured goods 27 131 51 51 56 66 74 Trade and balance of payments (sDR millions) Trade balance -47 -52 -47 -38 -39 -158 -168 Current account balance -48 -58 -36 -29 -21 -113 -91 Basic balance -30 -28 32 10 36 -44 -43 Foreign reserves in import-month equivalent 2.2 1.9 2.5 3.7 3.6 2.2 1.9 Debt-service ratio 7.5 21.7 21.8 21.8 22.9 17.8 22.8 Ratio of outstanding debt to foreign reserves 7.1 7.5 3.5 5.2 4.8 5.3 6.8 Trade indices (1978=100) Export volume 81 107 104 102 103 89 107 Import volume 112 77 68 67 60 42 53 Terms of trade 144 106 99 94 82 72 58 Import purchasing power of exporteamings 170 113 102 95 84 64 62 Gross producer margin of export production (percent) Tea 16.5 7.2 15.6 18.8 -5.3 5.9 -8.2 Rubber 45.1 31.9 33.5 7.9 37.4 14.6 -9.5 Fiscal operations (petrcent of GDP) Govemment expenditure 31.6 29.4 30.2 31.5 30.4 27.4 32.5 Govemment revenue 23.0 20.7 21.5 22.2 22.5 20.5 19.7 Budget deficit 8.6 8.7 9.7 9.2 7.9 6.8 10.5 Expansionary impact 1.5 3.2 1.6 0.8 0.3 0.1 0.8 Real money supply and financial deepening Ml (1980=100) 74 84 94 89 87 64 59 M2 (1980=100) 46 63 65 71 61 47 43 M3 (1980=100) 45 62 66 72 64 50 47 M2/GDP 24 22 24 26 22 19 18 M3/GDP 31 29 31 34 30 26 25 a. Employment in organized manufacturing sector. b. Wages of workers in Wages Board trades. c. Wages of all central govemment employees. d. Ratio between national account deflators for agriculture and manufacturing. e. Commercial bank borrowings and money creation for the purpose of financing the budget deficit. Source: Related tables in the text and statistical appendix. Chapter Four 59 Investment averaged 15.4 percent of GDP during the period 1960-69 and reached 17.2 percent in 1970. It started to fall in 1971 and slumped to 13.5 percent of GDP in 1973 (table 4.1). Land reform, income ceilings, limits on ownership of dwellings, and the Business Undertakings Acquisitions Act, together with the shock of the jvP uprising, created an environment that discouraged private sector invest- ment. Since access to foreign savings also narrowed, investment declined too. While the available data on employment are fragmentary, estimates suggest a decline in the growth rate of total employment as well as that of employment in the organized manufacturing sector in 1972 and 1973 (table 4.1). According to data from the Consumer Finance Survey of the Central Bank, the unemployment rate increased from 14 percent in 1964 to 24 percent in 1973. Unemployment was particularly acute among educated youth seeking employment in the modem sec- tor (Jones and Selvaratnam 1971 and 1972). As seen in table 3.7, it was the do- mestic (food-production) agricultural sector that contributed most to labor force absorption between 1963 and 1971. This arrested the decline in agriculture's share of total employment despite considerable labor shedding in the plantation crop sector. Trade, Balance of Payments, and Budget Deficits Weakened economic performance in the years leading up to the 1973 shock was the most visible indicator of deeper problems. The protectionist trade regime, the bal- ance of payments situation, and the government's fiscal problems were other mark- ers. These circumstances limited the policy options available to the government when a combination of external and domestic shocks precipitated a crisis in 1973. Trade Dependence Policymakers in Sri Lanka, like their counterparts in other developing countries, had expected the growth of domestic production to reduce the heavy dependence of the economy on foreign trade. In particular, it was believed that the domestic economy would be less vulnerable to adverse movements in the terms of trade. As noted in chapter 3, the direct contribution of export agriculture to total domestic output and employment declined from the early 1960s. At the same time, the out- put and employment shares of domestic manufacturing and agriculture increased significantly. The ratio of imports in final consumer expenditures fell from 32 per- cent in 1958-60 to 17 percent in 1970-73. These data seem to suggest that the economy had become less exposed to the vagaries of international economic cir- cumstances. The reality was quite different. The structural transformation expressed in the above figures had been achieved by placing the economy in a new and more pre- carious posture of trade dependence. The reduced importance of imports in final goods masked greater dependence on imported intermediate and investment goods. This is shown by the data in table 4.2. 60 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table 4.2 Capacity Utilization of Manufacturing Industry and the Importance of Imported Inputs Imported inputs Capacity as a percent of utilization Total Gross Year (percent) inputs output 1964-66 62 23 1967-69 - 68 25 1970 - 75 28 1971 43 74 26 1972 44 68 32 1973 42 67 32 1974 40 72 40 1975 54 61 40 1976 64 69 40 1977 60 66 40 1978 70 79 49 1979 72 82 53 1980 73 89 59 1981 74 87 57 1982 76 78 50 1983 74 89 35 1984 75 86 30 1985 74 80 28 1986 78 86 - Source: Central Bank of Ceylon, Annual Report (1984), and Revtiew of the Economy (annual). Imported inputs used in domestic manufacturing increased from 62 percent of all raw materials imports in 1964-66 to 75 percent by 1970-72. The "direct" import content of gross manufacturing output increased from 23 percent to 30 per- cent between these two periods. Imported machinery and equipment accounted for over 33 percent of gross domestic capital formation during 1970-73 (Central Bank of Ceylon, Annual Report-1974). Import compression, in short, strengthened import dependence (figure 4.1). The most notable change was a dramatic decline in nonessential consumer goods imports. These accounted for almost one-fifth of total import expenditure in the late 1 950s, but by the early 1 970s had declined to less than five percent. On the other hand, as domestic production activities expanded, the demand for intermediate and investment goods grew. The combined share of these imports was 52 percent in 1970-72, compared to 28 percent in 1950-51. The balance con- sisted mostly of essential food imports. During 1970-72, rice and wheat flour im- ports alone accounted for 25 percent of the total import bill.16 Chapter Four 61 Figure 4.1 Sri Lanka - Percentage composition of imports by major use categories, 1960 - 86 Index 70 Consumer goods 60 50/ \ \ 50 \ F _ < /\\ // \p \ / 40 Intermediate goods / \\I (including oil) I / / Intermediate 30 - / (excluding oil) - 20 ' ,_ -,. ''8 ' Investment 10 1960 1965 1970 1975 1980 1985 Year Source: table SA-13. In sum, nonessential imports were gradually converted into the essential im- ports needed to maintain domestic output and living standards. The upshot was that there was no longer any compressible import fat to cushion unexpected short- ages of foreign exchange. By 1973. the import-substitution policy pursued by a highly restrictionist trade regime had made the Sri Lankan economy extremely vulnerable to extemal shocks. Surpluses in the Basic Balance via Import Restrictions At first glance, the balance of payments figures for 1971 and 1972 suggest that problems in the external sector were diminishing. Despite recurring current ac- count deficits, a surplus in the basic balance was achieved in 1971 and main- tained in 1972. Moreover, external reserves had risen from their low 1970 level. This picture of relatively benign external sector conditions, however, was misleading. The improvement in the basic balance was partly due to an increase in exter- nal assistance but was primarily the result of stringent import controls. Import vol- ume in 1971-72 was almost 40 percent below the average for the 1960s and had fallen 15 percent since 1970. Export volume had declined marginally from the 62 Macroeconomic Policie5, Crises, and Growth in Sri Lanka, 1969-90 1970 level, and by 1972 the import purchasing power of exports had fallen to 84 percent of the 1970 level. These import restrictions were the response to a mini-payments crisis that had loomed in 1970. The proximate causes were the surge in imports that had followed partial trade liberalization in 1968, unfavorable trends in exports, and the inade- quacy of external assistance to deal with the consequent payments gap. The larger incentives given to nontraditional exports did increase their volume but had little impact in the short run. Their volumes during 1970-73 were 20 percent higher than the 1967 level, and their share in total commodity exports was 10 percent, having risen from an average six percent during 1960-66. The continuing deterioration in the terms of trade and stagnation in overall export volume growth created a payments gap that was financed by increased re- course to suppliers' credits and commercial bank credit, thus raising the external debt level. Such borrowings were necessary because (a) there was no significant private capital inflow; (b) IMF drawings, which had been an important part of ex- ternal financing, were not available after 1968 (Sri Lanka had fully utilized avail- able concessionary funds); (c) loans from the World Bank were ruled out because of its insistence on cuts in consumer subsidies as a loan condition; and (d) aid flows, though they did rise appreciably during 1965-70, were inadequate to plug the gap. By 1970 the debt service ratio had passed the 20 percent mark and the new UF government, which considered it a dangerously high level, found it hard to ob- tain further credit (Hewavitharana 1975). As described in chapter 3, restrictions on nonfood imports were not enough to handle the payments problem. The government was forced to cut back some of its consumer subsidies to stem the pressure on foreign exchange reserves caused by the promise to restore the rice subsidy. The link between the budget deficits, imports, and the balance of payments, which had been a recurrent theme in the Central Bank's annual review of the economy since the 1950s, could not be ig- nored any longer. But the situation in 1973 was such that little could be done to improve revenues in the short term, while any significant expenditure cuts were bound to be politically difficult or impact adversely on short-run economic activ- ity. Thus, at the time of the first oil shock in 1973, the Sri Lankan economy was already fragile. The Crisis The 1973 increase in world oil prices was only one of a series of adverse external and internal shocks. The combined impact of those shocks was such as to make Sri Lanka one of the world's worst-affected economies during this period. The impact of the oil price increase itself can be seen in table 4.3. Despite a 23 percent reduction in the volume of gross imports, the net import bill increased more than threefold. In addition, fertilizer prices increased as a direct result of the Chapter Four 63 Table 4.3 Major Imports: Quantity, Price, Volume, and Share in Total Imports Item 1972 1973 1974 1975 1976 Petroleum products Volume-gross imports (millions of barrels) 13.2 12.7 11.0 10.8 10.6 Price (1967/69=100)a 147 208 757 826 888 Value-gross imports (millions of dollars) 28 46 136 124 136 Percent of total imports 7.6 10.6 19.0 16.9 23.7 Value-net imports (millions of dollars)b 15 26 84 74 77 Percent of total impons 4.2 6.3 12.7 10.9 14.8 Rice Volume (thousand tons) 262 338 297 450 419 Price (dollars/ton) 90.7 148.8 403.9 303.7 204.6 Value (millions of dollars) 23.7 50.3 120.0 136.7 85.6 Percent of total imports 6.5 11.2 16.8 18.5 14.6 Flour Volume (thousand tons) 324 365 442 455 380 Price (dollars/ton) 111.2 195.8 319.5 319.1 243.1 Value (millions of dollars) 36.3 71.5 141.1 145.2 92.3 Percent of total imports 10.2 16.0 19.8 19.7 15.8 Sugar Volume (thousand tons) 214 190 42 61 45 Price (dollars/ton) 194.6 273.6 824.9 619.9 330.2 Value (millions of dollars) 42.5 52.1 34.7 38.0 15.1 Percent of total imports 11.6 11.6 4.9 5.2 2.6 Fertilizer Volume (thousand tons) 304 307 358 183 225 Price (1967/69=100)c 89 142 465 290 164 Value (millions of dollars) 11 17 33 33 12 Percent of total imports 3.1 3.9 5.4 4.4 2.3 Combined share of above in total import bilid 35.5 49.1 59.2 58.3 37.0 a. Index of world market price of petroleum (World Bank). b. Value of total (gross) imports minus value of exports by Sri Lanka Petroleum Corporation. c. Index of world market price of urea (World Bank). d. Net of export earnings of Sri Lanka petroleum corporation. Source: Compiled from tables 2.4 through 2.7 in Kappagoda and Paine (1981). 64 Macroeconomic Policies, Crises, and Growt i in Sri Lanka, 1969-90 oil price rises, and Sri Lanka's fertilizer import bill doubled. Oil and fertilizer im- ports rose from just over 10 percent of all imports to more 18 percent. External shocks did not cease after the oil price rise. In the wake of the oil price increase, there was a worldwide commodity price boom. For many develop- ing countries, these increases in the prices of nonoil commodities helped to cush- ion the impact of the oil price increase itself. But, given the nature of its imports, this commodity boom exacerbated the adverse impact of the oil price rise on Sri Lanka. The prices of the country's other major imports-rice, wheat, sugar-had started to rise even before the oil price shock. By 1974 the effective import prices of rice and sugar had risen more than four times above 1972 prices, while the price of wheat rose by almost three times. In fact, the rise in the bill for these imported food items exceeded the rise in oil import costs. Unfortunately, the general price boom affected the prices of Sri Lanka's major exports much less than it did its imports. True, the world price of natural rubber began to rise in 1973, partly in response to the rising costs of petroleum-based feedstocks in the synthetic rubber industry, and tea prices benefited in 1974 from a global shortage of tea due to bad weather in the other main producing countries and a sharp rise in coffee prices in 1976-77. The prices of coconut products also were more favorable, but the cumulative effect of these export price increases was much less than the increase in the import price index. As a result, the terms of trade fell by almost 13 percent in 1973, by 12 percent in 1974, and by a massive 19 per- cent in 1975. The cumulative impact was a drop of 38 percent from the 1972 level. The impact of this decline was aggravated by adverse developments on the supply side. Rice production fell in 1973 (table SA-6). It recovered somewhat in 1974 (when output reached the 1970 level) but remained well below the 1970 lev- el through 1977. The beneficial effects of the price rises of Sri Lanka's traditional exports were largely nullified by reductions in output due to management disloca- tions caused by land reform, bad weather, and (in the case of coconut) insect at- tacks (Fernando 1980a; Peiris 1978; Thorbecke and Svejnar 1984). Tea output in 1974, 204 million tons, was the lowest in twenty years. After a slight improvement in 1975 to 214 million tons, output then plunged to 197 tons (the lowest production figure since 1950) in 1976. Rubber output in 1974, 132 mil- lion tons, was the lowest since 1960. Coconut output also showed a substantial fall during the period 1972-74 in comparison with any three-year period during the two previous decades (Peiris 1978). The drop in 1973 meant higher domestic pric- es and led to a two-year ban on exports of coconut oil and copra. As a result of the decline in export volume of the three traditional exports and the drastic fall in the terms of trade, the index of import purchasing power of export earnings fell by nearly 35 percent between 1973-75 (table 4.1). Sri Lanka's share of the total ex- ports of nonoil developing countries dropped to 0.45 percent in 1973-75, com- pared to two percent in the early 1960s. The result was a balance of payments crisis. Despite a 30 percent reduction in import volume, the trade deficit quadrupled in 1974 (table 4.1). The current ac- count balance rose from less than one percent of GDP in 1973 to four percent in Chapter Four 65 1974. According to one study (Jayatissa 1982), the effect of the change in terms of trade worsened the trade deficit by US$176 million in 1974 and US$250 million in 1975, while the decline in export volume worsened the deficit by an extra US$35 million in 1974 and US$37 million in 1975. Kappagoda and Paine (1981) compared the "possible normal production levels had normal conditions pre- vailed" with actual production of the three major export crops and estimated that export earnings foregone in 1974 and 1975 were US$165 million and US$52 mil- lion, respectively. These were approximately equivalent to the current account deficit in 1974 and to about 30 percent of the current account deficit in 1975. The low rice harvests during these years (with the exception of 1974) led to larger rice and wheat imports. Since their prices in the international markets had risen, these imports were a massive additional cost. The rice and wheat import bill, which was US$60 million in 1972, jumped to US$222 million annually during 1973-75. The Policy Response In theory, the optimal policy response to an unanticipated economic shock de- pends on whether the shock is perceived as permanent or temporary. A country can attempt to finance an adverse shock believed to be only temporary-that is, if permanent income is not affected. If the shock is deemed to be long-term, howev- er, it represents a cut in perrnanent income to which it is necessary to adjust. The domestic supply shortfalls clearly fell into the temporary category, since the un- usually severe agroclimatic conditions that caused the harvest failures were not likely to persist indefinitely. The duration of the terms of trade deterioration was harder to predict. Throughout the world there was great confusion about the likely duration and intensity of the oil price hike, and this uncertainty affected the per- ceptions of Sri Lanka's policymakers. The dominant view was that external conditions were likely to worsen even more. In other words, the terms of trade decline was seen as permanent. Many in the government, including finance minister N. M. Perera, felt that a global reces- sion (or even a depression) could not be ruled out. Perera, who had lived through the 1930s, was inclined to expect a similar economic collapse in the 1970s. The oil price shock had come soon after the partial breakdown of the Bretton Woods system in 1971 and a resurgence of inflationary pressures in the major industrial economies. In the Budget speech in November 1974, Perera talked of ... the impending disaster that hangs over the world if the perilous course of inflation is allowed to continue unchecked. No country out- side the socialist world will go unscathed. The unemployment and frustration, the hunger and disease, and the suffering and degradation that will accompany such a global economic collapse can be better imagined than described. 66 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 A successful financing strategy could be pursued only if the government was prepared to make major changes in policy, but there were formidable obstacles to such a course of action. The 1970 election, the increased state control in key sec- tors of the economy, and the country's foreign policy stance had distanced it from the major Western donor countries. Borrowing under the IMF quota was limited by previous borrowings in 1968-70, while the World Bank had decided in 1971 that "... as long as the new policies were in place it would make no further commit- ment to Sri Lanka" (Levy 1985:11-12). Meanwhile, traditional commercial sources had become wary of lending to a govemment which did not enjoy the confidence of the IMF or the World Bank. By 1971, overseas exporters were refusing to accept letters of credit on Sri Lankan banks unless they were confirrned by foreign international banks, but the interna- tional banks were disinclined to provide such confirmation (Hewavitharana 1975). The government was not especially popular with Westem governments either. During the 1970 election campaign the SLFP-led coalition had espoused a great deal of socialist rhetoric to assure their leftist constituencies that the coalition would not adopt pro-capitalist policies. Hence, dealing with international financial institutions like the imF and the World Bank would expose the government to charges of selling out. Although the support extended to the government by the Western powers dur- ing the jvP uprising in 1971 led to some warming of relations, the West was not in- Table 4.4 Sri Lanka's Account Position at the International Monetary Fund, 197075 (SDR millions, end of year) 1970 1971 1972 1973 1974 1975 Stand-by arrangements: Amount drawn - 14 - - 7 - Undrawn balance - 10 - - 17 - SDRs - - 13 13 14 11 Percentage of allocation - - 38 40 43 32 Use of IMF credits 79 72 75 74 102 125 Compensatory financing - - - - - - Oil facility - - - - 34 65 Credit tranche 29 38 37 18 5 4 Extended IMF facility - - - - - - Quota 98 98 98 98 98 98 Fund's holding of rupees 177 171 173 172 200 223 Percent of quota 180 173 176 176 204 227 Source: International Monetary Fund, International Financial Statistics Yearbook (1985). Chapter Four 67 clined to provide large-scale financial assistance, and aid from the Eastern bloc was quite inadequate to adopt a financing strategy. The government therefore opted to adjust by reducing domestic absorption while attempting to ease the blow through a strategy that relied heavily on direct controls, particularly import restrictions. External Finance The 1973 oil price shock softened the attitudes of external lenders. In response to the pleas of the government, there was a substantial rise in aid flows from Westem and Eastern bloc nations, and from the multilateral donor agencies. Aid commitments rose from an annual average of US$1 Imillion in 1970-73 to US$162 million in 1974, and US$377 million in 1975; aid disbursements followed a similar pattern. This aid came on somewhat easier terms compared to previous years, though the grant component declined in 1974. Despite the fact that drawings from the JMF tranches were con- strained by previous borrowings, the RMF again became an important source of finance (table 4.4). The drawings, however, came largely from the oil facilities program, which had a longer repayment period but a higher interest rate. Discussions in 1975 between Sri Lanka and the IMF on another stand-by agreement broke down, allegedly due to IMF demands for further subsidy reductions (Kappagoda and Paine 1981). Import Restrictions The data reported in table 4.5 demonstrate the severity of the import compression policies and their implications for the allocation of imports. In volume terms, im- Table 4.5 Basic Indicators of Import Compression Before and During the 1973-75 Crisis Episode Commodity 1960-69 1970-72 1973 1974 1975 Import volume indices (1978=100) 82 71 60 42 52 Consumer goods 130 122 114 77 83 Food 126 133 125 75 87 Intermediate goods 134 116 64 62 76 Investment goods 63 48 39 23 45 import composition (percent) 100 100 100 100 100 Consumer goods 56 55 53 47 51 Food 43 46 46 43 48 Intermediate goods 24 24 30 42 36 Oil 7 2 11 20 17 Investment goods 20 21 17 11 13 Import/GDPratio(percent)' 25.2 15.2 13.2 12.4 13.3 a. Estimated using constant (1970) price data. Source: Relevant tables in the appendix. 68 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 ports in 1973-74 were 41 percent below the 1970-72 level, and the 1970-72 fig- ure itself was only 60 percent of the average during the 1960s. Imports had not fallen to such a low level since the 1940s. The sharpest cuts were absorbed by in- vestment goods imports, whose volume fell by 52 percent. The volume of con- sumer goods fell by 37 percent, and that of intermediate goods 47 percent. The achievement of cuts of this magnitude in such a short time was only possible through the most stringent direct controls, leading to rationing, shortages, and black markets in almost all important commodity markets. Fiscal Policy The fiscal response to the crisis was the adoption of orthodox austerity measures, including severe cuts in current expenditures. Government revenues declined in real terms and as a proportion of GDP, mainly because of the reduced volume of trade. Meanwhile, higher import costs reduced the weak profitability of the gener- ally inefficient public sector enterprises. This necessitated large budgetary trans- fers to these enterprises, though in some cases the higher costs were passed on in part to the public. For example, petroleum prices and transport charges were raised. The cuts in food subsidies were extensive (chapter 2) and politically painful, but the government felt that they were unavoidable. According to Kappagoda and Paine (1981), the cost of rice imports required to maintain per capita availability of rice at the 1968-71 level would have amounted to an average outlay of 6.1 per- cent of annual GDP during the 1974-75 period. To a limited extent, increased imports of relatively cheaper wheat compensat- ed for reduced imports of rice, but in absolute terms wheat too had risen in price. Cuts in the sugar subsidy caused consumption to fall from an annual average of 262,000 tons in 1970-71 to a mere 49,000 tons in 1974-76. The savings produced by reduced imports of sugar alone during the 1974-75 period amounted to US$294 million, almost equal to the total value of 1974 exports. Despite the cuts, the government's expenditures on food subsidies actually rose in absolute terms because of higher food prices, though only by a small frac- tion of the total amount of the price rise. Through such restraint, the government's current expenditures in 1974 were kept below current revenues, and the budget deficit as a proportion of GDP was only 6.8 percent, well below the average for the 1960-73 period (table 4.1). Current expenditures in 1975 were also kept below current revenues, but the overall budget deficit increased because both food subsidy expenditures and trans- fers to public sector enterprises rose. Capital expenditures, too, increased some- what. The deficit was financed by the larger volume of foreign financing that became available in that year and by the savings held in captive funds (the Em- ployees' Provident Fund and the National Savings Bank). Hence, the inflationary impact of fiscal policy was relatively mild. Chapter Four 69 Monetary Policy Monetary policy played a subordinate role in the govemment's response to the cri- sis. There was no significant attempt to increase interest rates to influence domes- tic demand for money. In fact, real interest rates in the formal sector were negative in 1974 and (despite some increase in nominal rates in 1975) continued to be very low or negative (table 4.1). There is some reason to believe that ex ante rates may have been somewhat higher because inflation was anticipated. Direct controls in the form of ceilings on commercial bank credit for domestic use were introduced in 1974, while selective measures were used to encourage credit for the tradable sectors (exports and import substitutes). The credit ceilings had no impact, however. Sluggish demand from the pri- vate sector made them redundant and left the banks with considerable excess li- quidity. A marginal decline occurred in the share of commercial lending for private sector housing and consumption, and there was no noticeable increase in lending for industrial activities. The real money supply, particularly Ml, fell sharply in 1974 (table 4.1) be- cause the fiscal deficit was smaller and was financed mostly by nonexpansionary sources, and because extemal reserves declined. The reduction of the fiscal deficit would have been even more pronounced had it not been for large borrowings by public enterprises. Public enterprise borrowings were more restrained in 1975. Exchange Rate Policy The rupee depreciated substantially in nominal terms during this period because of periodic shifts in favor of the U.S. dollar and British pound in the aftermath of the breakdown of the Bretton Woods system (Hewavitharana 1975). By the end of 1975 the official exchange rate of the rupee had depreciated by about 15 percent relative to the SDR and 18 percent relative to the dollar from their 1970 levels, but this did not reflect a deliberate attempt at using exchange rate policy to respond to extemal sector disequilibrium. In fact, devaluation of the currency was almost completely absent from public policy discussions during this period. Both the govemment and the opposition appear to have agreed that devaluation was not an appropriate policy response. While it is known that discussions about the desir- ability of economic liberalization took place in 1973 in govemment circles and within the Central Bank (Lal and Rajapatirana 1989), there is no evidence that de- valuation figured prominently in those discussions. Undoubtedly, one reason for not implementing a nominal devaluation was that direct controls had a more immediate impact on the level of imports (Corden 1987). We drew attention earlier to the evidence which suggested that Sri Lanka could not have relied on substantial extemal funding to meet the looming pay- ments gap, at least without politically difficult changes in domestic policies. Un- der those circumstances, the resort to direct controls was probably unavoidable. 70 Macroeconomic Policies. Crises, and Growth in Sri Lanka. 1969-90 But it is nevertheless puzzling that a currency devaluation was not seriously con- sidered as part of the overall adjustment strategy. A part of the explanation for this may have been the widespread view that de- valuation of the currency was unlikely to enhance the volume of traditional ex- ports. There had been considerable discussion about the effects of currency devaluations in 1967, when the rupee had been devalued. The opposition at that time was based on the expected impact of devaluation on the cost of living, on the cost of intermediate goods imports, and on pervasive "export pessimism" (see Rasaputram 1972). That is, the low supply elasticities of the major tree crops were considered to make devaluation ineffective as a tool for enhancing exports. The long-term effects of altemative adjustment policies on the tree crop sector re- ceived little attention. The establishment of the FEECS multiple exchange rate sys- tem showed a recognition that a more favorable incentive structure could benefit exports, but this was not extended to the traditional tree crop sector. The same an- tidevaluation arguments seem to have influenced the government's policy stance in 1973-75. Though the nominal exchange rate depreciated because of exchange rate shifts, there was no significant real devaluation during this period. Indeed, the real rate may have even appreciated slightly (table 4.1). Certainly the rupee was gross- ly overvalued throughout this period, as attested to by the need for stringent ex- change and capital controls, and the high curb (black) market rate. Substantial evasion of exchange rate controls appears to have taken place throughout the pe- riod, though accurate quantification is not possible. There is much anecdotal evi- dence, as well as evidence presented in court when the alleged leader of a currency smuggling organization was brought to trial, that an active and large black market was in operation during the 1960s and early 1970s. According to tentative esti- mates reported by Fernando and Colombage (1976), the unrecorded outflow of foreign exchange during the 1969-74 period was about 10 percent of total export earnings during that period. Conservative Attitudes to Inflation Why did the UF regime decide in favor of an extremely unpopular austerity pro- gram rather than adopt an expansionary strategy that might well have been politi- cally less costly, at least in the short run? True, a conservative attitude to inflation had been inherited from the period of British rule and the Currency Board System. Moreover, the Monetary Act of 1950 had anti-inflationary principles built into it and required the Central Bank to act if, in a given month, the rate of inflation ex- ceeded 10 percent or the growth of the money supply (Ml) exceeded 15 percent of the rate for the corresponding month in the previous year (Gunasekara 1962). But the political pressures that came to bear on the government, particularly on the left-wing parties in the coalition, were unprecedented. Moreover, the left- wing parties exerted substantial influence on financial policymaking, particularly in the early years of the SF government. In the circumstances, it would not have Chapter Four 71 been surprising if the left had pushed for a more expansionary policy. But leaders on the left, including finance minister Perera, were as strongly opposed to an in- flationary strategy as the technocrats of the Central Bank and the conservative pol- iticians of the right. In fact, the left resisted pressure from SLFP back-benchers to reject austerity. The anti-inflation attitudes of the left reflected the views of their trade union constituency, whose real wages would have been rapidly eroded by inflation. In addition, they may have been influenced by their experience with inflation in Eu- rope during the interwar years. Most of the leftist leaders had spent their politically formative years in Europe during the late 1920s and 1930s, and saw a link between hyperinflation and the rise of fascism. The Pattern of Adjustment It could be argued that the effects on growth of the crisis management policies adopted during the 1973-75 period should be compared with trend rates of growth of the relevant variables in the immediately preceding period. However, there are problems in adopting this approach, since the years 1971 and 1972 were abnormal in many ways. The political disturbances of 1971 severely disrupted economic ac- tivity, and real GDP fell. Although the economy began to recover in 1972, recovery to more normal levels did not take place until 1973. The 1970-72 period was ab- normal in other ways as well. The economy was subjected to severe import and exchange controls, and land reform affected output in the plantation crop sector along with drought and pest attacks. Thus, the 1960-69 period is a better reference period for evaluating performance during 1973-75. Adjustment to the shocks of the early 1970s included major reductions in both consumption and investment. Private consumption had already fallen substantial- ly in 1972 because of the political disturbances of 1971 and remained depressed until 1975 (table 4.1). The reduction in private consumption was accompanied by a large reduction in government consumption. Even in 1975, when private con- sumption began to recover, the continued squeeze on public consumption damp- ened the recovery. It took more than a decade for government consumption to reach its pre-1973 level in per capita terms. Data on per capita food consumption indicate that food intake declined during this period. Per capita staple food availability declined from an average of 441 pounds of rice equivalents to 394 pounds between 1971 and 1974 (Gavan and Chandrasekara 1979). Real wages also declined markedly (table 4.1). To some ex- tent, private consumption expenditures were sustained because consumers were willing to reduce their savings. Private savings during 1973-75 declined to nine percent of GDP, compared with 16 percent during 1970-72 and 12.2 percent during 1960-69. Such behavior is consistent with the possibility that private agents 72 Macroeconomic Policies, Cr ises, and Growth in Sri Lanka, 1969-90 considered the effects of the shocks to be largely transient, with little effect on their permanent income. 17 This reduction of private savings was reflected in the slowdown in the growth rate of private financial savings held in the form of time and savings deposits in financial institutions (table 4.6). Time and savings deposits had grown rapidly dur- ing 1970-72, stimulated by the expansion of the banking infrastructure, the de- monetization of higher denomination currency deposits in 1970, and the raising of savings deposit interest rates. In 1973 the amount of private savings held in finan- cial institutions slumped sharply in real terms, and then grew sluggishly in 1974 and 1975. There was also a dramatic fall in investment. Private sector investment fell in both absolute and relative terms. During the 1973-75 crisis period, the generally unfavorable situation was aggravated by the import control policy that gave first priority in import allocation to public sector enterprises, particularly those produc- ing domestic intermediate goods. Analyses of this period have tended to portray this policy in a favorable light (see, for example, Kappagoda and Paine 1981). It should be noted, however, that many of these public sector enterprises were highly inefficient (Sirisena 1975; Wanigatunga 1974). Channelling intermediate goods through these enterprises raised costs to user industries downstream. Therefore it is probable that this method of producing intermediate goods imposed an addition- al burden on the economy. Similar comments apply to at least that part of the pub- lic sector investment which took the form of capital grants to state enterprises. Inflation and Real Wages Sri Lanka had very low inflation until the 1970s. The crisis years then witnessed a surge in inflation, according to all available price indexes. The GDPD showed the Table 4.6 Growth of Time and Savings Deposits, 1966-77 (annual percentage change) Year Nominal Real 1966-69 10.1 12.5 1970 17.3 15.0 1971 14.5 11.0 1972 14.4 10.1 1973 3.6 -13.8 1974 19.8 6.2 1975 9.7 2.4 1976 23.8 17.8 1977 41.8 23.1 Note: Real values are computed using the GDP deflator. Source: Central Bank of Ceylon. 4nnual Report (annual). Chapter Four 73 rate of inflation rising from 7 percent in 1972 to 16 percent in 1973 and 23 percent in 1974. In terms of the Bhalla-adjusted ccpI, the increase in prices was much higher, with the price level in 1974 being more than 75 percent above the 1972 lev- el. Supplyside factors were primarily responsible for this increase. Upward adjustment of the prices of consumer goods, the rationing of many commodities, and domestic supply shortfalls (mainly, the decline in paddy pro- duction) contributed to these price rises. Meanwhile, the restrictive stance of fiscal policy, the generally tight monetary policy, and the effects of declining extemal assets tended to dampen money supply growth. Because of the strong link be- tween money supply and prices in Sri Lanka (Ranaweera 1978; Athukorala and Huynh 1987; Jayamaha and Wilson 1989), inflationary pressure was contained by the nonaccommodating monetary policy. By 1975 inflation was under control. Real wages in all sectors declined considerably during this period without much resistance from the trade unions. The reductions in current real wages un- derstated the cuts in the real incomes of trade union members, due to one govern- ment action that received little attention at the time or subsequently. This was the use of employees' savings held in the Employees' Provident Fund to finance bud- get deficits. As the interest rates on these funds were artificially repressed, and as real interest rates were generally negative, this represented a tax on the future in- comes of employees, the bulk of whom were members of the trade unions. Trade union reaction to inflation during the period was strongly influenced by the link to the left-wing parties in the government. The leadership ensured that the unions did not express open opposition despite rising inflation and reduced subsidies. Thus the policy package could be pursued without major industrial conflict, which would have been unthinkable if any other government had been in power. National and Sectoral Growth Real GDP grew by 3.3 percent in 1974, 2.3 percent in 1975, and 3.3 percent in 1976 (table 4.1). Comparison with the 1960-69 period makes it clear that growth slowed down substantially, even though the economy did not contract in absolute terms. The average real GDP growth rate during 1973-75 was 2.7 percent, as against 4.0 percent during 1965-9. Real agricultural output during 1973-75 was only marginally higher than in 1965-69. This was mostly due to the poor performance of the three major export crops (table 4.1), and that of paddy in 1972 and 1975. At least part of the decline in output of plantation crops can be attributed to declining investment in replant- ings and to greatly reduced fertilizer use. Fertilizer imports during the 1971-74 period averaged 312,000 tons per year. They fell to 183,000 tons in 1975, and con- tinued to fall during the next two years. An even worse sectorwide performance was avoided thanks to an expansion in the subsidiary food sector, which respond- ed to the higher prices produced by import bans. The import restrictions, reduc- tions in food subsidies, and higher food prices tilted the sectoral terms of trade in favor of domestic food-production agriculture (table 4.1) and stimulated the culti- 74 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 vation of subsidiary food crops. Again, in any assessment of this growth, the effi- ciency costs of the distorted incentive structure must be recognized. The pattern of growth in this sector would have been considerably different in the absence of high effective protection. The import restrictions hurt the manufacturing sector as a whole because of its very high import dependence. The growth rate of manufacturing sector output slowed down in 1971 and turned negative in 1974. This slowdown was most se- vere in private sector manufacturing ventures. Overall capacity utilization levels remained quite low (table 4.2), and total manufacturing employment fell sharply in 1973 and 1974 (table 4.1). Reflecting the impact of increased selective incentives, nontraditional exports (notably gems and manufactured goods), showed a significant improvement from about 1970 and this trend continued during the period 1973-75. Recorded gem ex- ports grew from about SDR 0.7 million in 1970 to SDR 21 million in 1975 (table SA- 14). Manufactured exports during the same period grew from SDR four million to SDR 33 million. The combined share of nontraditional exports (excluding petroleum) in total nonpetroleum commodity exports increased from 10 percent to 17 percent between 1970 and 1975. However, given the low starting base, ex- ports remained small in absolute terms, and the expansion of these exports had only a marginal cushioning effect against the unsatisfactory performance of tradi- tional exports. In general, however, the steps taken to achieve structural changes in the econ- omy fell well short of what would have been necessary, particularly in view of an assessment of the terms of trade shock as a permanent one. The measures adopted to stimulate the tradables sectors were basically higher (and selective) incentives for nontraditional exports. Nothing in the adjustment strategy suggested that fun- damental changes in the economy were likely. Recoveryfrom the First Crisis Sri Lanka's external payments position had begun to improve by the end of 1975. The government's prompt response to the 1973-75 crisis by way of reductions in expenditures and realignment of priorities certainly played a role, but favorable external changes also helped. In 1976 the trade balance improved because the terms of trade became better thanks to a reduction in import prices and higher pric- es for exports. But the economy was slow to respond. The traumatic stresses of the 1973-75 period had sapped its vitality. In 1976 the government eased its deflationary fiscal stance, which was now attributed solely to Perera, who had been compelled to leave the government in 1975. A more expansionary budget was financed with a substantial increase in borrowing from the Central Bank. As a result, Ml rose by 34 percent, the highest rate of annual increase until then. Time and savings deposits rose by nearly 24 per- cent, helping to moderate the impact of the rise in Ml. Chapter Four 75 Unlike many developing countries, Sri Lanka emerged from the 1973-75 pe- riod with only a small external debt. But the UF government was to pay a heavy price for the deflationary approach. The import restrictions and the cuts in con- sumer subsidies were extremely unpopular. The government was thus unable to take advantage of the improvement brought about by the terms of trade changes. The import controls inhibited the effective use of rising external reserves to ame- liorate the consumer goods shortages and win back some of the lost popular sup- port before the election in 1977. The groundswell of unhappiness with the government paved the way for the UNP to come to power again. Chapter Five The Second Crisis: 1978-82 Two shocks struck the Sri Lankan economy in the 1978-82 period, the first do- mestic and the second external. The first was an enormous upsurge in investment, particularly public sector investment. We will call this the public sector invest- ment program (PSip) shock. The Psip shock was the result of a deliberate decision by the newly elected UNP government. The financing of this investment entailed a massive capital inflow, since domestic savings were an utterly inadequate source of funds. In addition to the capital flows directly related to the PSlp, a large increase in private inflow also took place. Inevitably, the inflows led to a big current ac- count deficit and a real appreciation. The second shock was a dramatic deteriora- tion of the terms of trade. The world price of tea fell, while import prices, including those of oil, fertilizer, and sugar, again rose strongly. The Background of the Crisis When the UNP govemment was elected in 1977 it inherited a considerably more favorable external payments situation than that which faced the UF government in 1970. By 1977 the world economy had recovered from the first oil price shock, and this was reflected in Sri Lanka's terms of trade. In addition to the general im- provement of the international economy, the country benefited from a sharp im- provement in tea prices that began in 1976. There was also a significant decline in import prices. Consequently, in 1977 Sri Lanka experienced its most favorable terms of trade since 1970, and (despite stagnation in export volume) the index of the import purchasing power of export earnings showed a 22 percent improvement over the 1976 level (table 5.1). These positive developments were not confined to the trade account. There had also been a small but significant improvement in the services account, owing to rising migrant-worker remittances and receipts from tourism. In 1977 these two 76 Chapter Five 77 Table 5.1 Movement of Major Economic Variables Before and During the 1978-82 Crisis Episode Economic variable 1976 1977 1978 1979 1980 1981 1982 Economic growth (annual percent) Real GDP 2.9 4.6 8.2 6.5 5.5 6.1 4.8 Per capita GDP 1.3 3.0 6.4 4.6 3.6 5.0 2.8 Sectoral growth Plantation agriculture -8.6 -0.8 4.1 4.1 -10.0 7.5 -1.5 Agriculture (other) 9.5 23.9 5.4 7.3 6.3 2.8 3.5 Industry 9.5 23.9 13.5 8.4 4.0 3.5 3.4 Manufacturing 5.7 1.5 10.8 4.8 6.1 4.2 9.1 Services 1.0 4.7 7.6 7.8 8.0 6.4 7.0 Employment growth (annual percent) Total employment 1.2 1.5 3.5 4.0 - - - Manufacturinga employment 4.0 5.7 14.8 7.4 5.6 6.4 3.7 Consumption, Salving and Investment Real per capita consumption (1970=100) 94 96 103 105 103 104 110 Real investment (1970=100) 84 64 116 160 217 218 272 Private 77 60 82 146 160 179 229 Public corporations 56 57 145 167 413 411 514 Government 124 82 165 192 205 192 182 Savings/GDP ratio (percent) 13.9 18.1 15.2 13.7 11.2 11.7 11.8 Private savings 14.2 17.8 16.5 13.3 14.8 13.4 13.3 Investment/GDP ratio (percent) 15.4 13.8 19.9 25.2 31.2 27.3 31.9 Private investment 8.7 7.2 8.9 13.1 13.7 13.0 14.7 Incremental capital/ output ratio 3.5 4.2 2.0 3.1 4.4 5.8 5.4 Prices and wages General price level (percent change) BACCPI 5.6 9.2 9.7 16.1 23.4 19.5 8.1 GDPD 6.3 20.2 7.6 8.3 21.1 16.0 17.1 wPt 8.6 20.9 15.5 9.5 33.7 17.0 5.5 CBCPI 3.8 15.0 8.9 18.6 37.1 23.8 11.1 Terms of trade Agriculture-manufacturing (1970= 1 oo)b 112 107 118 95 103 98 86 Plantation sector 100 127 136 127 131 126 117 Domestic agriculture 120 96 108 76 90 85 77 Real wage index (1978= 100) Private sectox5 69 77 100 100 94 98 83 Public sectord 105 97 100 98 82 75 88 Real interest rate (annual percent) NDo deposit rate -8.4 8.1 2.3 -5.1 2.3 0.3 -5.1 Commercial bank depositrate -11.1 7.7 1.1 -5.3 2.9 1.0 -9.3 (continued on next page) 78 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table 5.1 (continued) Economic variable 1976 1977 1978 1979 1980 1981 1982 Commercial bank lending rate -7.1 8.4 1.5 -3.2 2.6 1.6 -2.9 Exchange rates Nominal exchange rate (rupees/US dollars) Official rate 8.4 8.9 15.6 15.6 16.5 20.8 23.5 Official rate with tEEC premium' 14.5 14.81 - - - - - Curb-market rate 13.8 17.8 21.3 21.8 19.6 21.8 23.5 REER index (1980=100) Total exports 64 76 90 88 100 98 89 Traditional export 68 78 94 .91 100 95 86 Nontraditional exports 62 81 80 84 100 101 81 Trade and balance of payments (SDR millions) Trade balance -73 29 -144 -362 -758 -695 -890 Current account balance -5 117 -75 -117 -507 -381 -516 Basic balance -56 149 57 47 -201 -40 -39 Foreign reserves in import-month equivalent 3.1 10.6 5.9 5.1 2.4 3.1 3.3 Debt-service ratio 20.6 15.9 5.5 12.9 12.9 16.8 18.6 Ratio of outstanding debt to foreign reserves 4.5 2.4 2.3 2.0 4.5 4.9 5.0 Trade indices (1978=100) Export volume 102 94 100 101 99 102 112 Import volume 52 73 100 123 140 145 150 Terms of trade 80 96 100 72 58 46 38 Import purchasing power of export eamings 78 102 100 72 58 46 38 Gross producer margin of export production (percent) Tea 2.4 60.5 -2.1 -17.1 3.2 -5.4 2.5 Rubber 32.3 20.6 42.7 34.6 29.5 11.9 5.3 Fiscal operation (percent of GDP) Government expenditure 33.2 28.1 46.5 43.2 48.7 39.1 41.3 Government revenue 20.4 19.2 28.8 25.5 22.6 20.4 19.4 Budget deficit 12.7 8.8 17.7 17.6 26.1 18.7 21.0 Expansionary impactf 2.3 -2.3 0.4 1.2 10.3 4.6 5.8 Real money supply andfinancial deepening Ml (1980=100) 58 69 82 94 100 96 107 M2(1980=100) 44 54 71 87 100 109 131 M3 (1980=100) 49 55 73 89 100 108 133 M2/GDP 18 24 26 29 30 31 31 M3/GDP 25 32 34 38 38 40 40 a. Employment in organized manufacturing sector. b. Wages of workers in Wages Board trades. c. Wages of all central govemment employees. d. Ratio between national account deflators for agriculture and manufacturing. e. FEEC system was abolished in November 1977. f. Commercial bank borrowings and money creation for the purpose of financing the budget deficit. Source: Related tables in the text and appendix. Chapter Five 79 items contributed about 11 percent of current account receipts, as against three percent in 1975. Overall, the current account deficit had declined from SDR 102 million in 1975 to SDR SiX million in 1976. In 1977 there was a current account surplus for the first time since 1965. Reflecting the improved current account po- sition and the inflow of concessionary foreign capital in support of the market-ori- ented policy stance of the new government, foreign reserves increased from an average level of three months of import equivalents during 1970-76 to eleven months by the end of 1977. This was the largest figure recorded in a single year since 1951. By 1975 there had been no significant improvement in economic growth, even though the worst of the crisis was past. This probably increased unemploy- ment considerably, particularly in urban areas, but accurate assessment is not pos- sible. The 1973 Labor Force Survey of the Central Bank estimated unemployment at about 17.6 percent of the total labor force. A 1977 urban budget survey conduct- ed by the Census Department found an unemployment rate of 27.8 percent for the survey area and 29.4 percent for Colombo (table 3.7). Inflation had come down since the 1973-75 period. Although the acute shortages of that period had been alleviated, repressed demand for imports was certainly present for both consumer and investment goods. The implementation of the liberalization program was also the signal for the beginning of the investment surge. The Investment Boom and the Terms of Trade Decline The magnitude of the investment drive initiated in 1978 was striking. During the previous decade, total investment had been on average 15 percent of GDP (table 4.1). This figure began to rise so rapidly that by 1980 it exceeded 31 percent (table 5.1). This rise was primarily due to the PsIP, which raised the ratio of public capital expenditure to GDP fTom an average of about 6.5 percent in the period 1970-77 to about 18.5 percent by 1980.18 The most striking feature of the PSIP was the heavy emphasis placed on the Accelerated Mahaweli Development Project. During the period 1979-85, almost one-third of all government expenditure was devoted to the project (table 5.2), which also absorbed nearly 45 percent of the project funding committed by donors during this period. The second largest project of the program was a housing-cum- urban-development scheme (including the building of a new capital and a monu- mental parliament house), which accounted for about 10 percent of public invest- ment allocation in 1979-85. Thus, close to half of all public investment was channeled into only two projects. Work on the Mahaweli Project (see chapter 2) began in 1970, and the first trans-basin division was completed in 1976. The new UNP government announced its intention to accelerate the pace of the project in November 1977, four months after its accession to power. The new regime wanted to complete the entire project 80 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table 5.2 Allocations for the Accelerated Mahaweli Development Project, 1979-86 Source 1979 1980 1981 1982 1983 1984 1985 1986 Millions of rupees 1.603 2,991 3,991 5,469 7,302 7,195 5,541 34,092 As a percentage of: Total governrment expenditure 8 10 14 15 17 14 8 13 Goverrnent capital expenditure 21 25 34 34 44 37 23 32 GDP I 4 4 6 6 4 3 9 Aid-funded share percent of total investment 30 48 45 61 56 65 83 78 Source: Compiled from Govemment of Sri Lanka, Ministry of Finance and Planning, Public Invest- ment (annual). in six years, as against the original 30-year completion target, at an estimated cost of 11,000 million rupees (US$610 million). Yet past experience, both in Sri Lanka and elsewhere, gave ample reasons for doubting the project's technical feasibility as well as its economic benefits (kiyagolle 1978). In fact, an interim implementation report prepared in 1978 by a group of Dutch consultants recommended a slower rate of construction (Karunati- lake 1987). The report argued that any attempt to implement such a large project in such a short period of time was likely to place unmanageable strains on Sri Lan- ka's institutions and resources. The consultants' final report, released in 1979 and assuming considerable increases in the prices of oil and rice, predicted a rate of return of 11 percent for the entire project. That was not a very impressive figure, given the alternative investment opportunities offered by an economy whose cap- ital stock had been deteriorating for a long period. Although the economic ratio- nale for the project was weak, and the consultants themselves had expressed reservations about its value, the government decided to press ahead. The government's decision, which appeared to one analyst to be "nothing more than a substantial gamble" (Stern 1984), was based on the following ratio- nale. First, the govemment saw the project as an effective means of harnessing concessionary foreign funds. President Jayawardene felt quite sure that foreign aid could only be obtained for a relatively short period of time, and that the opportu- nity to lock up funding for large, long-term projects had to be grasped (Jayawar- dene 1979). Moreover, few project proposals other than the Mahaweli were ready to be presented to donors. In addition, the government wanted to cultivate a project that would capture the imagination of the people and thus persuade them to endure some economic hardship for the sake of future prosperity. The govern- ment's planned reduction of consumer subsidies and its substantial dismantling of import restrictions (which resulted in labor-shedding in such domestic industries as hand-loom textiles) were bound to arouse popular dissatisfaction. Hence, the vision of a mammoth irrigation scheme designed to achieve rice self-sufficiency Chapter Five 81 and thus rekindle the glories of ancient Sinhalese civilization was enormously ap- pealing to populist politicians (Levy 1985). As Stern (1984:27) pointed out, "no other development program could have appealed to the electorate in the same way as the Mahaweli program." Finally, and very importantly, a large public sector in- frastructure-cum-settlement project offered a mechanism for political patronage by enabling the govemment to provide employment on a large scale to youth and a promise of land to the rural poor. In short, economic considerations were pushed aside by political objectives. Yet donor response to the idea of accelerating the Mahaweli project was swift and positive. In many cases the donors announced their commitment well before the completion of feasibility studies and estimation of project costs (World Bank 1980). Before 1978 had ended the main donors had committed themselves to funding the three major upstream dams that were to be the cornerstones of the scheme. Moreover, the donors responded favorably to a suggestion by the World Bank that commitment to the project should be considered in addition to the amount of aid that would be forthcoming for Sri Lanka in the absence of the project. The World Bank itself became a major sponsor of the project.'9 The enthusiasm of the donors for the Mahaweli scheme spilled over to other projects as well, and total foreign commitments to Sri Lankan development projects increased from US$175 million for the entire 1970-77 period to US$390 million for the three years 1978-80. By 1980, however, it was clear that the investment program was generating strains in the economy. The estimated cost had shot up to about 18,000 million ru- pees (US$860 million) as a result of the project's acceleration and subsequent cost overruns. The most direct impact was on the fiscal deficit. Although a major share of the program costs was being funded by the donors, the local funds required for im- plementation were substantial (table 5.2), and curtailment of these local funds be- came virtually impossible once the program was set in motion. The government was forced to supplement the foreign funding with domestic borrowing to finance its large budget deficits. Subsequently, the government also resorted to foreign commercial borrowing. The donors' contribution was mostly intended to pay for the costs of imports needed for the project. However, they also paid a share of the project's domestic costs, producing a significant net inflow. This, together with remittances from mi- grant workers, created upward pressure on the foreign exchange rate. When the terms of trade shock occurred in 1979, the economy was already showing the strains imposed by the government's rush to implement its massive investment program. The Terms of Trade Shock The terms of trade decline started in late 1978 and gathered momentum in 1979, when it fell by 28.3 percent. This was followed by falls of 18.4 percent in 1980, 82 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 22.6 percent in 1981. and 16.7 percent in 1982. The cumulative decline between 1978 and 1982 was 62.3 percent (table 5.1). Overall, Sri Lanka was one of the de- veloping countries hit hardest by the second oil price hike and the associated world recession. Balassa and McCarthy (1984) estimated that the total negative ef- fect of exogenous shocks on Sri Lanka was equivalent to a drop of 25 percent of GDP during the 1979-81 period, compared with an average of five percent for de- veloping countries as a group. Total earnings (SDR 3,400 million) from commodity exports during the 1979- 82 period were almost 50 percent higher than for the preceding four years. How- ever, the escalation of import prices more than counterbalanced this increase. Con- sequently, the index of capacity to import (income terms of trade) declined by almost 30 percent between the two periods. The impact of the terms of trade change was aggravated by poor performance in the plantation crop sector. Despite much rhetoric, little was done to improve the management of state-owned plantations, and the decline in production levels that had started in the early 1970s continued. Average tea production during 1980-82 was similar to the 1959-61 average (table SA-6). Rubber production in 1982 (125 million tons) was the lowest in twenty years. Coconut production and exports showed some improvement during 1978-82, but this improvement was minuscule compared with the production loss in tea and rubber. Export earnings were also affected by the declines in export volume. In 1979, for instance, the decline in the world price of tea was accompanied by a decline of 3 percent in the volume of tea exports. Despite a five percent reduction in crude oil imports, the bill for imported oil rose very sharply. By the end of 1979, the bal- ance of trade showed a 23 percent rise in imports and stagnant aggregate exports. As a result, the trade deficit in 1979 rose by 150 percent over the 1978 figure. Jayatissa (1982) estimated that 80 percent of the negative effect on the trade bal- ance came from changes in the terms of trade and 20 percent from decline in ex- port volume. This deficit in the trade balance did not, however, place strains on the overall balance of payments in 1978 and 1979. The country had to run a deficit in the cur- rent account to absorb the inflow in foreign capital, and a larger trade deficit was a reflection of this absorptive process. About 20 percent of the imports were fi- nanced directly by foreign unrequited net transfers. Capital inflows (aid, conces- sional loans, migrant-worker remittances) increased so substantially that the balance of payments actually showed a surplus of SDR 35 million (table 5.1). The trade deficit of SDR 758 million in 1980 was a record, and double the amount of the previous year. Despite significant increases in the surpluses of ser- vices and transfers, due mainly to remittances, the current account deficit nearly tripled to SDR 507 million in 1980. External reserves had to be used to finance the current account deficit, since the capital flows were insufficient, and Sri Lanka ex- perienced a serious decline in external reserves for the first time since 1975 (Cen- tral Bank of Ceylon Annual Report-] 980). Chapter Five 83 Policy Response The terms of trade decline had no immediate influence on policymaking. The op- timism that characterized policy discussions in the 1978-79 period was barely af- fected by it. For instance, the Central Bank of Ceylon Annual Report-I 979:171 pointed out that: ... the high levels of trade deficit and the resultant current account deficits are also a reflection of the fact that the country has been uti- lizing the aid receipts, which otherwise would add to external assets. In the circumstances, the more useful indicator of the balance of pay- ments position of the country would be the overall balance which re- flects the net change in extemal assets. The rationale for this view rested in part on expectations of a J-curve adjust- ment-that is, the initial deterioration in the current account due to high levels of investment goods imports would be followed by current account surpluses once the investments generated higher exports. The devaluation of 1977 and the longer- term effects of trade liberalization, together with the future output from current in- vestments, were expected to generate a future expansion of tradables production that would be sufficient to meet financial obligations. The belief that resource constraints no longer mattered had become pervasive in Sri Lanka. For instance, the budget speech in November 1979, while noting the world recession and its adverse effects on Sri Lanka, outlined ambitious goals for 1980. There was no sign of a realization that Sri Lanka was in the first stage of its most dramatic terms of trade shock in decades. GDP growth rates of 8.2 percent in 1978 and 6.3 percent in 1979, and a seemingly never-ending flow of external re- sources, created an environment in which little attention was paid to the negative developments occurring in the larger world.20 The huge foreign aid flows encour- aged many within the government to engage in lavish expenditures; ministries were able to obtain additional funds whenever they requested them, even though the finance minister had begun to complain about a deterioration in financial dis- cipline within the government. Impact on the Fiscal Deficit As a share of GDP, government expenditures increased to 49 percent in 1980 from 43 percent in 1979 (table 5.1). This increase was mainly attributable to the mas- sive increase in capital expenditures, which rose from 11 percent of GDP to 22 per- cent. Revenues, on the other hand, declined from 25.5 percent to 22.6 percent of GDP. In other words, the substantial surplus in the current account of the budget generated through cuts in food subsidies disappeared. In part, revenues declined because of lower receipts from export duties. Im- mediately after the currency reforms the government had moved to siphon off a large part of the gains to exporters via progressive export taxes, but it soon became 84 Macroeconomic Policies, Crises. and Growth in Sri Lanka, 1969-90 clear that the tax burden on traditional exporters was destroying their viability. In 1979 there was a reduction in the export duty on bulk tea to assist the ailing tea industry. Export tax revenues on coconut declined sharply due to supply shortfalls and consequent lower exports. While tax revenues from rubber exports increased marginally, lower export volume reduced the potential gains. Overall revenues from export duties declined by 13 percent in current terms. On the other hand, im- port duty receipts increased substantially, so that the net effect was marginally higher revenues. Sri Lanka's fiscal deficit in 1980 was an unprecedented 26 percent of GDP. For- eign assistance funded a large proportion of the deficit, and the government was also able to engage in substantial noninflationary domestic borrowing, particularly from "captive funds." These, however, were inadequate to finance this deficit fully, and the government was compelled to borrow heavily from the Central Bank. This, not surprisingly, stimulated inflation, which rose to a historic high of 26 percent as measured by the Colombo cPi. Other indices indicated that the inflation rate might have exceeded 30 percent. In addition to the expansionary effect of the public investment program, rising international prices also had a larger impact on the domestic price level due to the increased size of the traded goods sector and more open economic policies. The impact of world price rises was magnified by the sharp depreciation of the nominal exchange rate, whose value declined by 10.8 percent against the SDR during 1980. Another contributory factor was the reduc- tion of consumer subsidies, leading to higher prices and subsequently to pressure to raise real wages (Stern 1984). Three factors helped to prevent inflation from rising even higher. First, the de- cline in external reserves brought about by the terms of trade shock tended to reduce the money supply. Second, high nominal interest rates had produced a very large buildup of time and savings deposits which provided noninflationary sources for public borrowing (table 5.1). These deposits constituted nearly 70 percent of M3 in 1982 and were equivalent to 5.9 percent of GDP that year, compared to 2.1 percent in 1981. Third, greater openness to imports helped to curtail increases in prices. It was only two years after the terms of trade deterioration began in late 1978 that measures were taken to adjust to external problems. These included a combi- nation of monetary and fiscal policy measures. The first response to the crisis came from the Central Bank in April 1980, and occurred chiefly because of concern about the rise in the inflation rate. The Central Bank raised the bank rate from 10 percent to 12 percent per year.21 As a further move to discourage commercial banks from resorting to Central Bank credit, the penalty rate was raised from a range of 15 to 25 percent to a range of 20 to 30 percent. Following these increases, the Treasury bill rate rose from 9 to 13 percent per year, and the rate on govern- ment securities went from 10 percent to 16 percent per year. The rates of interest on deposits at the National Savings Bank and at commercial banks were also raised, as were the lending rates of commercial banks and other lending institu- tions (table 3.9). Chapter Five 85 These measures were quite effective in stemming the rapid growth of the money supply. The annual growth rate of Ml dropped from 25 percent in April 1980 to 15 percent in September 1980. Commercial bank borrowing from the Central Bank fell by half. As a result, the interbank call money market rate rose from 12 percent earlier in 1980 to a peak of 25 percent toward the end of the year (Central Bank 1980). However, the slowdown in the growth of the money supply was short-lived as government borrowing again produced a sharp rise in the mon- ey supply toward the end of the year (table 5.1). When strong inflationary tendencies emerged in the first quarter of 1981 in as- sociation with rapid growth in the money supply and domestic credit demand, the Central Bank intervened directly. In May 1981 the Central Bank imposed credit ceilings on the commercial banks. The ceilings were removed in June, but other measures to restrain commercial bank credit were taken, such as raising the bank rate. In August 1981 the bank rate was raised to 14 percent, while the penalty rates were raised to a range of 21 to 35 percent. These measures were quite successful in reducing excess liquidity. By the end of 1981, money supply growth had slowed down (Ml rose by only six percent in 1981). Inflation, too, started to slow. Since a substantial part of the inflation during this period was due to the ex- pansion of base money, the rise in inflation can be viewed as an attempt by the government to finance its expenditures through the inflation tax. This was perhaps the most extensive use of the inflation tax by a government in Sri Lanka. Estimates by Edwards and Tabellini (1991) show that revenues accruing through the infla- tion tax between 1978-83 were 1.9 percent of GNP, compared with less than 0.7 percent during 1963-73 and 0.8 percent during 1973-78. Fiscal Policy To hold down the huge budget deficit, the government took steps in 1980 to raise gov- ermnent revenues, mainly through increases in import duties and other indirect taxes. However, the main thrust of the adjustment was expenditure restraint, which was to be achieved mainly through cuts in other capital expenditures while leaving the Ma- haweli and housing projects virtually untouched. Cutbacks in 1981 on these two projects amounted to only 8 percent, while other projects were cut by 32 percent. In 1982, the year of the first presidential election under the new constitution, fiscal stringency was relaxed. Government expenditures rose from 39.1 percent of GDP in 1981 to 41.3 percent in 1982. Both recurrent and capital expenditures went up. This loosening took place against a background of a further decline (16.7 percent) in the terms of trade, and widening deficits in the trade and current accounts. These deficits, measured in SDR terms, rose by 28 percent and 35 percent, respectively. Sluggish export growth in the face of rising import prices was behind the large trade deficit. Had it not been for higher earnings from tourism and an increase in worker remittances, the current account deficit would have been even bigger. A major component of fiscal restraint came from the cuts in consumer subsi- dies, in particular the cuts in food subsidies. By maintaining the nominal value of 86 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 food stamps unchanged in the face of high inflation, real expenditure on food sub- sidies was drastically cut. This contrasted with the large transfers to loss-making state economic enterprises like Air Lanka. Many such enterprises had become overstaffed through the hiring of government supporters. Such transfers substan- tially counteracted the net impact of the food subsidy cuts. The Pace of Trade and Exchange Rate Reforms Slows There was no systematic attempt during this period to reduce imports through di- rect controls or through higher tariffs, although a few import items were added to the licensing list from about 1979 on. While balance of payments considerations played a role in these moves, they were basically motivated by the vulnerability of many state enterprises to foreign competition. An across-the-board levy of 10 per- cent on all items dutied at over 50 percent was introduced in 1980. This was osten- sibly in order to gather funds to promote exports. In November 1982 some selective duty increments were applied to items already dutied at high rates. These were moderate, revenue-oriented duty changes applied to bring the trade balance into line. All in all, given the availability of very substantial external finance in the form of foreign aid and private transfers (mainly remittances), the government managed to avoid a retreat to import barriers as a way out of balance of payments pressures. Although fiscal and monetary policies were used with some success to re- spond to immediate pressures on the balance of payments in 1978-82, the funda- mental long-term issue remained that of stimulating growth of the tradables sector. Once again, as in 1973-75, Sri Lanka was unable to respond to the crisis by a rapid expansion of exports, even though export expansion was accorded high priority in the 1977 policy reform package. The UNP regime accepted the view that the exchange rate should be an impor- tant policy instrument in promoting exports, and the value of the rupee in relation to the major currencies was adjusted daily to reflect changes in demand and sup- ply. Central Bank intervention through the buying and selling of foreign curren- cies was limited at first, but then the Central Bank intensified its intervention and eventually abandoned (in November 1982) the practice of determining the ex- change rate daily. There had been a gradual appreciation of the real effective ex- change rate from 1979 onwards, except in 1981. when a marginal depreciation was recorded (table 5.1). When exchange rate policy ceased to provide a stimulus for exports, the al- ternative was to strengthen the various financial incentives offered by the EDB. However, this indirect approach tended to founder because of severe institutional and financial constraints (see Cuthbertson and Athukorala 1990). While export de- velopment was theoretically considered a high priority area, in practice it received only secondary support from annual budgetary allocations because of the priority given to the politically appealing "lead projects." This basic policy contradiction was discussed by the chairman of the EDB (Santiapillai 1981:1): Chapter Five 87 The pivotal role of exports in economic development ... is not easily seen by politicians or bureaucrats. Foreign trade, somehow, is generally viewed as something distinct that does not directly touch the day-to-day lives of peoples as would, for example, a multipurpose river diversion scheme concerned with human resettlement, agricultural expansion, en- ergy development, etc., all of which capture the attention of both the government and the people at large. This is at the root of the problems of mobilizing support for a concerted national effort to develop exports. Except for tea, the trend in incentives measured through the real effective ex- change rate (REER) over the period 1978-84 was clearly downward for broad ex- port categories (table SA-17). This downward trend became more evident after 1980. The incentive structure continued to show a bias against traditional exports, although the degree of discrimination had been reduced by the exchange rate re- forms of 1977. Within the broad category of nontraditional exports, manufactured exports were generally favored over nontraditional agricultural and mineral ex- ports. The REER for gem exports declined sharply after 1977, and in 1981 was even lower than in the mid-1960s. Currency depreciation under the new exchange rate regime did not fully compensate for the abolition of the CRA scheme. Moreover, gem exports were not eligible for any of the incentive schemes. According to one quantitative assessment of incentives to export-oriented and import-substitution manufacturing as of 1981, effective protection accorded to production for the domestic market was generally greater than that for exporting (Ratnayaka 1988). Overall, the favorable treatment accorded to manufacturing ex- ports by various incentive schemes was minor when compared with the domestic market bias generated by high tariff protection. Judging from the subsequent changes in tariffs, export incentives, and the exchange rate, the antiexport policy bias intensified after 1981. Thus the favorable effects of trade liberalization in the early years of the UNP regime were eroded over time. The basic problem with regard to export incentives arose from macroeconom- ic policy. For reasons associated with the public sector investment program, the maintenance of state enterprises, and the desire to restrain inflation, the goverm- ment refused to make the currency devaluation that was necessary to avoid over- valuation. A favorable exchange rate policy would have given nondiscriminatory incentives to exports without the red tape and opportunities for rent-seeking that were inherent in Sri Lanka's complicated scheme of direct export incentives. On the other hand, the capital flows tended to appreciate the real exchange rate and might have counteracted any attempt to make the exchange rate a more realistic one through depreciation of the currency. Employment and Wages Surveys during this period indicated a considerable expansion in employment leading to a significant fall in unemployment (table 5.3). The PSIP itself, as well as the services and construction sectors, were the sources of substantial employment 88 Macroeconomic Policies. Crises, and Growth in Sri Lanka, 1969-90 Table 5.3 Various Estimates of Unemployment Unemployment as a percentage of Year Source labor force 1953 Population Censusa 0.8 1959/60 Survey of Employmentb 10.5 1963 Population Census' 10.8 1964 Consumer Finance Surveyc 13.8 1968 Labour Force Surveyc 11.0 1969/70 Socio-Economic Survey' 14.0 1971 Population Census2 18.7 1973 Labour Force Participation Survey 18.1 1973 Consumer Finance Surveyc 24.0 1975 Land and Labor Utilization Surveya 19.8 1978/79 Consumer Finance Surveyc 14.7 1980/81 Socio-Economic Survey' 13.6 1981/82 Consumer Finance Surveyc 11.7 1985/86 Labour Force and Socio-Economic Survey' 14.1 1986/87 Consumer Finance Surveyc 15.5 a. Dcs-Department of Census and Statistics. b. IL-oInternational Labour Organization. C. CBC-Central Bank of Ceylon. Source: Authors' calculations. growth. This was supplemented by rapid expansion of the labor-intensive textile and garment industry. Despite rapid economic growth and higher aggregate ex- penditures. real wages in the organized sectors (both public and private) showed a declining trend (table 5.1). On the other hand, the limited data available on daily wages for workers outside the organized sector suggest a marginal improvement. For example, real wages for casual farm labor reported by Wickramasekara (1983) indicate a rise of over 20 percent between 1977 and 1982. The increase in employ- ment in the noninstitutional sector, combined with the substantial outmigration to the Middle East. is likely to have exerted some upward pressure on real wages, particularly in certain skilled worker categories. On the other hand, the decline in trade union power was probably a main reason for the poor wage performance in the institutional sectors. Capital Flows, Income and Expenditure Effects, and Inflation Much of the pressure on the price level and the real exchange rate during the 1978-82 period was a direct consequence of the higher domestic expenditures as- Chapter Five 89 sociated with the enhanced capital flows, which raised the demand for nontraded goods in the economy. A large literature has appeared since the late 1970s which analyzes the economic consequences of rises in expenditures brought about by fa- vorable exogenous shocks. These studies were initially stimulated by resource booms, and this body of literature has come to be known as the Dutch disease lit- erature (for an exposition and survey, see Corden 1984). We will draw on this lit- erature in our analysis of the effects of capital flows in Sri Lanka. These capital flows led to higher expenditures on domestic goods and services in a number of ways. Official capital inflows, which came as foreign aid and as- sistance, accrued directly to the government and were primarily meant to pay for the imports required by the PSIp. To the extent that they were spent on imports they did not have an impact on domestic prices. But there was a direct link between such official capital flows and the domestic expenditures of the government through the so-called counterpart funds, whereby the government agreed to spend matching funds on the domestic costs of the PSIP. Thus, every dollar of foreign capital inflow for the PSIP led to, in general, the addition of an equivalent amount in rupees to domestic government expenditures. Ultimately, such expenditures be- came the incomes of domestic private agents. The incomes of the private agents were directly supplemented by the remittances of migrant workers and the expen- ditures of tourists on domestic goods and services. Private agents can dispose of their higher incomes by spending on domestic (nontraded) goods and services and on imports, and by saving the remainder; the proportion saved depends on the ex- tent to which the additional incomes are thought of as an increase in their perma- nent income, since most transient income is saved. The evidence on the savings behavior of private agents during this period is consistent with the supposition that they thought that the increase in their incomes would be permanent. The private savings ratio had been 13.2 percent of GDP dur- ing 1970-76, but rose to 17.8 percent in 1977. In 1978 it fell to 16.5 percent and over the next five years it averaged 13.7 percent, only marginally higher than the 1970-76 figure. The 1977-78 figures probably reflect a lagged adjustment to the new situation. There are several reasons why private agents might have concluded that their income increase would be permanent. Although the very high initial level of in- vestment was not expected to be permanent, the investments themselves were widely expected to yield high future benefits, thus generating higher permanent incomes. Tourism appeared set to develop into a boom industry, and the country was enjoying the benefits of large remittances from overseas employment. In gen- eral, there was a mood of great optimism regarding the future of the economy. Thus, the capital flows led to higher expenditures by the government as well as private agents. Some of these expenditures were made for the acquisition of nontraded goods and services, and exerted upward pressure on their prices. In an economy with full employment, higher demand for nontraded goods can be met only by drawing resources from other sectors, producing a rise in the relative pric- es of the nontraded goods-that is, a real exchange rate appreciation.22 Such a re- 90 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 source reallocation gives rise to "Dutch disease" effects as higher domestic spending leads to a contraction of the tradable sectors. If the exchange rate is fixed, the domestic prices of tradables in a small country are exogenously given. Hence, the rise in the price of nontradables not only changes relative prices but also raises the general price level. Thus, an increase in the price level, as observed in Sri Lanka, was consistent with the normal consequences of an absorption of capital inflows. But there were also two factors in Sri Lanka that reduced such price pressures. First, as Lal (1985) has pointed out, trade liberalization that lowers the domestic price of imports will produce the required relative price change by means of a low- er increase in the general price level than would otherwise occur. The trade liber- alization of 1977-78 therefore dampened the inflationary pressure. Second, the existence of considerable unemployment (to which we have re- ferred earlier) would also have mitigated the price pressures, since the nontrad- ables sector could expand to some extent by drawing on unemployed resources without having to bid higher for resources already employed in the tradable sec- tors. The decline in open unemployment and the relatively muted real wage in- creases observed during these years suggest that a considerable proportion of those who found jobs in the expanding nontradables sector may have come from the ranks of the previously unemployed. Even allowing for these factors, however, an increase in the price level was probably unavoidable. The nominal depreciation of the rupee during this period contributed to a higher price level by raising the price of tradables. This meant that the required change in relative prices had to be brought about by an even higher increase in the nominal price of nontradables, thus raising the general price level more than would have been the case if the nom- inal exchange rate had been fixed. As would be expected from the above, the real exchange rate appreciated from 1979 onwards, and by the end of 1982 most of the real depreciation achieved by the 1977 devaluation had been lost (table 5.1). The resulting impact on domes- tic resource allocation was reflected in sectoral growth rates, which showed a rap- id expansion of the nontradable sectors. The share of the services sector in the economy rose by nearly 10 percent during 1978-83. Services was the fastest grow- ing sector of the economy in this period, followed by construction (table 3.4). What was the impact of the terms of trade deterioration on these develop- ments? As the figures on the magnitude of the terms of trade shock show, even if a part of the fall could be ascribed to normally expected volatility, the impact was large enough to be translated into a significant reduction in permanent income. The most influential analysis of capital flows into Sri Lanka during this period is that by La] (1985). The basic thrust of his analysis was that the large current ac- count deficits and the appreciation of the real exchange rate were necessary to ab- sorb the capital inflow and did not signify a current account problem that had to be corrected by expenditure switching and absorption reduction policies. Even the loan component of the aid flows was considered desirable, since the loans were made on concessional terms and were used to finance investments. According to Lal's empirical analysis, the magnitude of the observed rise in the price level and Chapter Five 91 the appreciation of the real exchange rate were consistent with the changes re- quired to absorb the capital flows. We will return to the issues relating to the investment program and its financ- ing in our discussion of long-term growth in the next chapter. Here, we simply note that the empirical analysis ignored the terms of trade shock, and that in any case the shock could not have been handled within the framework of his analytical model, which was based on the aggregation of importables and exportables into a single composite commodity. Given the magnitude of the change in terms of trade, such an aggregation was invalid for an empirical investigation of macroeconomic performance during the 1978-82 period.23 As described earlier, the government maintained its expenditure levels until 1981, when certain spending cuts were made. But even these were partially re- versed in 1982 in the period prior to the election. In any case, the 1981 reductions in spending were not sufficient to compensate for the real income losses flowing from the terms of trade fall, and the "autonomous" capital flows were inadequate to finance the resulting increase in the current account deficit. This was reflected in the greater resort to commercial borrowing and the related rise in the debt ser- vice ratio. Between 1980 and 1983 the debt service ratio rose from 12.9 to 21.8 percent. Thus, the real exchange rate and price level changes observed during this period were the direct consequences of a deliberate government policy to maintain the level of real expenditures. They were not merely the exchange rate and price level changes required to absorb the autonomous capital flows. A regression analysis of the movements of the real exchange rate during the 1970-87 period confirms that capital inflows and government expenditures tended to appreciate the real exchange rate (Athukorala and Rajapatirana 1990). Unfortu- nately. terms of trade effects could not be isolated because of econometric prob- lems in estimating the model.24 Bandara (1989). who conducted simulation experiments with a computable general equilibrium (CGE) model, decomposed the effects of the capital inflows and the terms of trade decline. The results confirm that the capital inflows and terms of trade decline had opposing effects on the real exchange rate (and on certain other key variables). Further analysis has shown that there would have been substantial declines in consumption, household incomes, and real wages, and a significant stimulus to the major export industries via a real exchange rate depreciation if the expenditure cuts had been large enough to offset the income effects of the terms of trade fall (Bandara and Jayasuriya 1991). The result of the decision not to cut spending was that the real exchange rate continued to appreciate rather than depreciate (except for a marginal depreciation in 1981). Hence, the Dutch disease effect on the tradable sector was allowed to persist. The poor performance of the export sector must be seen in the context of this policy of nonadjustment, or only partial adjustment, to the terms of trade de- cline. The traditional export crops, tea and rubber in particular, continued to stag- nate. Commodity specific real exchange rates appreciated, while producer margins remained depressed. There was no significant improvement in exports of other manufactures or of nontraditional agricultural products. In the case of tex- 92 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 tiles and garments, two factors unrelated to the domestic exchange rate and trade policy changes led to expansion of exports (Athukorala 1989). These were the protected world-market share ensured by the quota system under the Multi-Fibre Textile Agreement and the cost advantage achieved through lower wages. Regression analysis of short-run export performance using commodity spe- cific real exchange rates supports the view that trade and exchange rate policies during this period had adverse implications for the promotion of nontraditional ex- ports in general and manufactured goods in particular (see appendix 1). There were some limited attempts to counteract these effects, and the reduction in taxes on tree crop exports gave some relief. But these efforts fell short of what countries such as Malaysia and Indonesia did when they enjoyed temporary resource booms (Corden 1987; Barlow and Jayasuriya 1987). Political Instability and Faltering Growth The Sri Lankan economy had not fully adjusted to the shocks of the 1978-82 pe- riod when the terms of trade improvement eased the immediate pressures. But po- litical conditions deteriorated sharply in 1983. This makes an overall assessment of the adjustment process difficult. However, the contrasts between this period and the 1973-75 crisis period are striking. While the decline in the terms of trade during the 1978-82 period was much larger than the decline during the 1973-75 period, the economy suffered only a relatively minor slowdown in overall growth performance. Indeed, in the popular consciousness the 1978-82 period is not seen as a period of stress, whereas the 1973-75 experience left strong negative memories. GDP growth was maintained at a respectable 5 percent level per annum throughout the 1978-82 period, and pri- vate consumption also grew throughout the period, except in 1980. Public con- sumption did fall somewhat during 1980-81 but grew thereafter. In absolute terms, total investment rose almost uninterruptedly. Was this relatively benign outcome a result of a far superior handling of economic shock through the use of more appropriate policies? Or, as some would argue, was it mainly a result of the easy availability of concessionary capital inflows? The rise in concessionary capital inflows and remittances did make a funda- mental difference. Despite the terms of trade deterioration, the country was able to sustain dramatically higher levels of imports and finance the investment surge without an inflationary crisis and with only a relatively minor buildup of high-cost external debt. The enormous importance of foreign finance can be illustrated by a few figures. Average import volume during 1980-84, after the terms of trade de- cline had already taken place, was 60 percent higher than during the 1978-79 pe- riod; and that, in turn, was 45 percent higher than the 1970-77 average. On the other hand, the average export volume index during 1980-84 was only 10 percent higher than that of 1978-79.25 To finance the average level of imports during 1980-84 at constant terms of trade, export volume would have had to be 50 per- cent higher than it actually was. Clearly, the surge in the GDP growth rate in the Chapter Five 93 immediate post- 1977 period was not brought about by an expansion of exports due to trade and exchange control liberalization. Instead, it had a distinctly Keynesian character, being brought about by the massive rise in govemment expenditure. The burst of inflation during the 1978-82 period, while not high by most de- veloping country standards, was nevertheless unprecedented in Sri Lankan histo- ry. But by 1982 there were signs that inflationary pressures were abating. All of the price indices showed inflation for the year falling to 10 percent or less. The Central Bank attributed this slowdown in large part to its restrictive credit policies. However, political uncertainty accompanying the 1982 presidential elec- tion probably also played an important role in reducing credit demand during the latter part of the year. Indeed, the targets established by the Central Bank in 1981 as part of the National Credit Plan were well above actual demand. A substantial buildup of excess liquidity in the commercial banking system then led the banks to reduce interest rates on time deposits. Also, the inflationary pressure generated by the investment program may have been moderated by the negative income ef- fects of the terms of trade deterioration. The emergence of the world economy from the 1980-82 recession brought an improvement in Sri Lanka's terms of trade. The impact of increases in world mar- ket prices of Sri Lanka's exports was considerable-export eamings in SDR terms rose by 8.2 percent in 1983, while the terms of trade rose by 16 percent (based on a 1978=100 index). However, the full benefits of the favorable world market situ- ation were lost because of continuing poor performance in the export crop sector. There were also negative developments in the services account. First, debt servicing soaked up an increased amount of export eamings, and the resort to short-term commercial borrowings (particularly after 1980) aggravated the debt service burden. Second, widespread violence against the Tamil minority in Co- lombo and other Sinhalese-dominated areas reversed the strong growth in the tourist sector. Despite these negative factors, the current account deficit declined in 1983. Although there was a 12 percent decline in the inflow of direct foreign investments and other short-term and long-term capital inflows, a marginal improvement in the overall balance of payments took place and net extemal reserves rose. There was a much greater improvement in the balance of payments in 1984, caused mainly by the tea price rise. Continuing growth in worker remittances from the Middle East also helped to improve the current account despite the decline in tourist num- bers. Given these favorable changes, the country no longer faced the balance of payments pressure of the 1979-82 period, and other issues took center stage. Chapter Six Macroeconomic Policies and Long-Term Growth In this chapter we turn to the central concern of this study by examining how pol- icy responses to the macroeconomic crises may have affected long-term growth. We proceed by analyzing changes in a number of broad, overlapping areas that are likely to influence long-term growth: the level, nature, and efficiency of invest- ment, the nature of domestic savings behavior, the functioning of economic insti- tutions, and the sociopolitical environment. The latter factors can be thought of as influencing growth through their impact on the level and efficiency of investment. Our interest here is not the quantification of the absolute or relative contribution of various factors. Rather, we seek to obtain some insight into how short-term cri- sis management policies influenced the behavior of key variables in a manner that affected long-term growth. Investment and Growth The selection of the above variables for our analysis is based on the generally ac- cepted view that investment level and its efficiency are key determinants of output growth. However, the underlying theoretical justification for such a choice has only recently been developed in a rigorous manner; the traditional neoclassical growth model implies no role for economic policy as a determinant of the long- run growth rate. We examine the role of investment and its efficiency using an equation based on Scott (1989), which hypothesizes that changes in long-run growth rates are due to changes in the rate and efficiency of investment. In this equation we regress the moving average of the rate of investment over three pre- vious years and a number of other variables that may have affected investment ef- ficiency on the rate of growth of real output in the nonresidential business sector. 94 Chapter Six 95 Note that the equation, while consistent with the Scott model, is consistent with several other models as well; changes in investment can have a positive impact on growth rates in the relatively short run even in the neoclassical model. The esti- mated function is presented in table 6.1 (for details, see appendix 1). Our results are consistent with the view that the rate of investment is a key determinant of the economy's growth rate. The coefficient of the trade liberaliza- tion index has the expected (positive) sign and is significantly different from zero, suggesting that a liberal trade policy stance promotes growth. This finding is di- rectly relevant for our discussion of the effects on growth of trade restrictions used as a macroeconomic policy tool. (A model was also estimated with the ratio of pri- vate sector investment to total investment as a separate variable, but it was highly correlated with the trade liberalization variable and its coefficient was statistically insignificant.) The investment scale variable (S2) has a statistically significant co- efficient with the expected (negative) sign. Again, this is consistent with the hy- pothesis that large increases in the rate of investment lead to reduced investment efficiency as relatively inefficient investment projects are implemented. Table 6.1 Determinants of the Rate of Growth of Real Output (G), 1950-83, Regression Results Rate of investment (S) 0.424* (6.72) Investment-scale variable (S2) -1.22* (4.90) Slope dummy of S for trade liberalization(T*S) 0.0244 (3.35) Rate of change of capacity utilization (U) 0.177** (2.30) Intercept dummy for output disruption (D71) -0.054 (4.17) Intercept dummy for 1950s (D50s) -0.046* (-5.42) Intercept dummy for civil war 1987-89 (D87-89) -0.029* (3.48) R2 0.68 F 13.13* DW 2.40 Note: Growth of real output (value added) in the nonresidential business sector (defined as total GDP minus the sum of the three national account categories, ownership of dwellings, public administration and defense, and other services). The t-ratios of regression coefficients are given within parentheses, with the level of significance de- noted as ** = significant at the one percent level and * = significant at the five percent level. For details on model specification and data sources, see the appendix. Source: Authors' calculations. 96 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 The implications of this regression equation for long-run growth must be in- terpreted with some caution. Growth rates are clearly influenced in the short run by investment, but no strong conclusions can be drawn regarding the long-run ef- fects. But to the extent that growth is stimulated by increases in investment, the importance of investment and its efficiency are emphasized in these results. Now we tum to an analysis of the behavior of domestic savings, a major determinant of investment in developing countries like Sri Lanka. Savings Behavior In principle, of course, investments can be financed with domestic or foreign sav- ings if a country has access to the world capital market. But domestic savings re- main a major detemninant of the capacity for investment because the cost of foreign capital rises with the scale of borrowing due to perceived lending risk. (In any case, domestic capital markets are insulated from the world capital market in many developing countries, and private agents do not have direct access to the lat- ter.) Hence, the level of domestic savings strongly influences the cost of capital. By lowering the domestic cost of capital, higher domestic saving reduces the level of foreign borrowings, and, given an upward-sloping foreign supply curve, lowers the cost of foreign borrowings as well. These considerations applied to Sri Lanka, even after the 1977-78 reformns. We estimated separate domestic savings func- tions for total (private plus public), private, and financial savings for the 1960-87 period to explore their behavior in a longer-term context. The estimated savings functions are presented in table 6.2. The rationale for the chosen specifications and the estimation procedures are discussed in appendix 1. Here we draw attention to some of the main points that emerge from the analysis. First, the results are clearly consistent with the hypothesis that foreign savings crowded out domestic savings, both total and private. The coefficient of foreign savings (capital inflow) is negative and significant in both the total and private do- mestic savings functions. This has particular relevance for analysis of the long-term implications of for- eign capital flows. However, a caveat is in order. It is theoretically possible that the causation goes the opposite way-that is, that the inadequacy of domestic sav- ings may have forced greater reliance on foreign savings. The results also suggest that domestic policies, particularly those associated with financial repression, sig- nificantly influenced domestic savings. Private savings (including private finan- cial savings) have been strongly influenced by the real interest rate. In the Sri Lankan context, the availability of foreign savings probably reduced the pressure to change financial repressionist policies that constrained domestic savings, there- by helping to maintain a low savings rate. On the positive side, the expansion of the banking network stimulated financial savings. Chapter Six 97 Table 6.2 Determinants of Savings: Regression Results Variable DS DPS FNS Constant 7.48** 2.38 -3.39** (3.83) (0.64) (2.78) PGDP 1.89** 1.22** 1.55** (7.00) (2.57) (4.59) GDPR 0.83 0.95 (1.62) (1.23) (I +RSTD) 1.68 1.98* 6.13* (1.37) (2.28) (2.23) D*(l + RSTD) -1.10 0.26 -1.99 (0.11) (0.19) (0.83) (I + PE) 5.84 (1.59) D*(I + PE) -0.45 (0.16) FS -0.05** -0.04* (4.81) (2.10) FIR 3.54** 3.37 (4.81) (1.69) RZ2 0.96 0.92 0.84 F 112.57** 51.35** 30.22** DW 2.16 1.94 1.60 Viariables: DS = Real total domestic savings; DPS = Real domestic private savings; FNS = Real Finan- cial Savings; PDGP = real per capita CDP; GDPR = Real GDP growth; RSTD = Weighted average real interest rate on savings and time deposits; PE = Expected inflation (proxied by the rate of inflation with a one- year lag); FS = Real foreign savings; FIR = Financial intermediation ratio (M3 as a percentage of nom- inal GDP); D = trade liberalization dummy (one for 1978-87 and zero for other years). All variables ex- cept growth rates are expressed in logarithms. Note: The t-ratios of regression coefficients are given in parentheses. The level of significance is de- noted as ** = significant at the one percent level and * = significant at the five percent level. For details on model specification and data sources. see appendix 1. Source: Authors' calculations. Overall, our savings functions suggest that the general policy of financial re- pression and the associated low interest rate policies hampered the mobilization of domestic savings in general. Given financial repression, inflation lowers ex- pected real interest rates, since the banks are unable to make upward adjustments in line with market forces. In these circumstances it can be argued that the general reluctance of the government to utilize inflationary policies probably helped avoid a poorer private savings performance. 98 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Private Investment In previous chapters we have discussed the political and policy parameters influ- encing public sector investment decisions. To gain more insight into private sector investment behavior, we estimated a private investment function for the period 1960-82. This is presented in table 6.3.26 Details about the theoretical rationale for the specification and the estimation procedure are given in appendix 1. Recall that credit to the institutional sector was usually rationed because of the interest rate and other regulations. It can be seen from the estimated function that a major influence on private investment was the availability of credit from the institutional sector. The short-run (one-year) elasticity of real private investment with respect to loanable funds is 0.63. This figure, when combined with the coefficient of the lagged dependent variable, yields a long-run elasticity estimate of 2.17. The results also support the hypotheses on the positive impact of income growth and the negative impact (the crowdingout effect) of public sector invest- ment on private investment. The regression results provide no evidence that high- er interest rates and the resulting high cost of investment inhibited private investment, as asserted by Jayawardana, Maasland, and Radhakrishnan (1987) and Roe (1982). Arguably, the role of financial repression was a key influence on private in- vestment, since it resulted in credit rationing to the private sector. Given that the Table 6.3 Determinants of Private Investment, 1960-82: Regression Results Constant (c) -0.33 (0.25) Institutional credit to the private sector (ICP) +0.63** (3.52) Lagged public-sector fixed investment (PBlv(-l)) -0.34* (2.26) GDP growth (I + YR) +0.05** (3.14) Index of rental cost of capital (RC) -0.07 (1.17) Lagged private investment (PRIV(-1)) +0.71 ** (4.73) -T2 0.92 F 49.35 DW 1.82 Note: All variables are measured at constant (1980) prices and expressed in logarithms. The t-ratios of regression coefficients are given in parentheses. The level of significance is denoted as ** = signif- icant at the one percent level and * = significant at the five percent level. For details on model specification and data sources, see the appendix. Source: Authors' calculations. Chapter Six 99 private sector had to rely entirely on domestic sources of credit because of foreign exchange regulations, such credit rationing probably constrained private invest- ment throughout the 1960-82 period, except during brief periods (such as in the 1973-74 period) when the constraints were nonbinding. As described earlier, the public sector conscripted a large chunk of institutional savings throughout the pe- riod, and it is plausible that this led to financial crowding-out, causing a decline in private investment. We draw on this analysis of the determinants of investment and savings to ex- amine the impact of crisis management policies on savings and investment pat- tems, and on the efficiency of investment. The 1973-75 Crisis: Long-Term Effects In chapter 4 we described how the policy response to the 1973-75 crisis involved a sharp reduction in the level of investment. Gross Domestic Fixed Capital (GDFC) declined from an average of 15.8 percent of GDP during the 1966-72 period to 13.3 percent during 1973-75. If cuts in education expenditure are included, the decline was even more severe. Total expenditure on education fell from an average of 4.4 percent of GDP during 1966-72 to 3.0 percent during 1973-75. However, we cannot conclude that this reduction in investment necessarily had a significant negative effect on long-term growth. Various factors suggest that the negative effects, if any, may not have been large. First, the fall in investment lasted a relatively short time. As economic condi- tions began to improve, investment recovered. And after the change in regime in 1977, it took a quantum leap to a level unmatched by any earlier period. Second, when the efficiency aspects of the investment pattems and their changes are considered the overall impact on growth may have been only margin- ally adverse, or even positive. While there was an overall reduction in both public and private sector investments, the sharpest cuts were in the public sector (table 5.1). Public corporations' investments declined from an average of 2.6 percent of GDP during 1966-72 to 1.5 percent during 1973-75. There is considerable evidence that many public sector enterprises (PSEs) were highly inefficient. These inefficiencies flowed from three main sources: (a) the protectionist import-substitution strategy resulted in the establishment of PSEs in industrial sectors where Sri Lanka had no comparative advantage; (b) such en- terprises typically were granted easy access to public funds via budgetary transfers and preferential access to bank credit; and (c) they often enjoyed a degree of mo- nopoly power in markets insulated from competition through protection or regu- lation. Under these conditions, the allocative inefficiencies arising from the trade regime were aggravated by various types of x-inefficiencies. Continued growth of such PSEs through additional investments could have en- larged the inefficiencies. with "immiserising" implications. There is evidence that 100 Macroeconomic Policies, Crises. and Growth in Sri Lanka. 1969-90 the "inefficiency spillover effects" of such industries were substantial, particularly since import compression policies were implemented with a distinct public sector bias in the allocation of imported intermediate goods. Since many PSEs were en- gaged in intermediate goods production (steel, tires, plywood, paper, and cement), the production costs of downstream user industries went up. In an inquiry into the operations of the steel corporation, which supplied about 50 percent of domestic rolled steel products in 1973, Sirisena (1975) found that domestic user costs, on average, were 41 percent higher for the steel corporation's metal than for compa- rable, or perhaps even better quality, imported steel. An additional longer-term consequence of such policies was to raise the import dependency of manufactur- ing industries. Analysis of input-output tables for 1965 and 1970 showed that 34 out of 41 manufacturing industries reported increases in total import content be- cause of greater use of locally processed imported inputs (Athukorala 1981). A similar exercise cannot be undertaken for the crisis period itself because of an ab- sence of data. But since the basic structure and mode of operations of PSEs re- mained virtually unchanged during the crisis years, it is safe to conclude that this would have applied also during that period. Therefore, to the extent that public sector investment in inefficient PSEs de- clined, there may have been little or no negative impact on long-term growth. In fact, we cannot even rule out the possibility that such cuts were beneficial. How- ever. where such cuts fell on infrastructure and efficiently utilized capital stock, such as trees on state-owned plantations, the effects would have been negative. In the absence of appropriately desegregated data it is not possible to quantify the ex- tent of such cuts, but anecdotal evidence suggests that the road system, for exam- ple, deteriorated badly during this period. In table 6.4 the incremental capital/output ratios (ICORs) for this period are presented. There is no discernible trend in that ratio until the end of the 1970s and no evidence of a marked deterioration in the immediate aftermath of the crisis. Of course, this measure is too crude to fully reflect the longer-term efficiency conse- quences of past investments. The generally depressed level of overall private investment is explained by the adverse economic and political climate, but the decline in total investment was minor. There is evidence that there were changes in the composition of private in- vestment during this period, as shown by the expansion of nontraditional exports and the development of new import-substitution sectors, particularly in agricul- ture. Again, the long-term growth consequences of these changes require careful analysis. These changes, particularly the growth of nontraditional exports, have been cited to support the view that, despite import controls and the absence of a nominal devaluation of the exchange rate, export dynamism was being generated through selective incentives (Kappagoda and Paine 1981; Hewavitharana 1975 and 1980). Was this expansion of nontraditional exports a reflection of structural changes in the economy toward dynamic long-term export growth based on the country's comparative advantage? We believe that the answer is no. Chapter Sir 101 Table 6.4 Gross Domestic Product (GDP), Gross Domestic Fixed Capital Formation (GDFC), and Incremental Capital/Output Ratio (ICOR), 1961-87 (1980 prices, millions of rupees) Year GDP GDFC"a lCORb 1961 29,076 3,330 5.7 1962 30,122 3,592 3.1 1963 30,744 4,158 5.7 1964 32,207 4,018 2.5 1965 33,218 3,604 3.9 1966 34,418 4,330 3.0 1967 35,849 4,883 3.0 1968 38,525 5,312 1.8 1969 40,371 6,975 2.8 1970 41,963 7,941 4.3 1971 41,937 7,158 _C 1972 45.222 7,828 2.1 1973 46,816 6,432 4.9 1974 48,667 7,654 3.4 1975 50,950 7,931 3.3 1976 53,211 8,620 3.5 1977 55,244 7,980 4.2 1978 59,098 11,848 2.0 1979 62,880 16,239 3.1 1980 66,527 22,465 4.4 1981 70,386 19,545 5.8 1982 73,975 22,747 5.4 1983 77,645 22,434 6.1 1984 80,800 20,866 7.1 1985 84,438 20,115 5.7 1986 88.194 20,866 5.3 1987 90,364 21,084 9.6 a. Current price data deflated by GDP (tradables) deflator. b. ICOR, = [GDP, - GDP, I]/GDFC,-X C. ICOR not defined due to negative GDP growth. Source: Based on data from the Central Bank of Ceylon. Review of the Economv (annual). First, there is much anecdotal evidence that export revenue from nontradi- tional exports was substantially inflated, though it is not possible to quantify pre- cisely its extent. Nontraditional export revenues could be partially utilized to finance imports under the CRA scheme. Under a highly restrictive trade regime, this attracted a scarcity premium to such exports. CRA credits to gem exporters, for examples, fetched a premium of 150 to 200 percent (Jayawardena 1982). This in- 102 Macroeconomic Policies. Crises, and Growth in Sri Lanka, 1969-90 duced widespread overinvoicing of exports. In the case of gem exports, which showed the largest growth, the higher figures were to a large extent a reflection of the channelling of previous illegal exports into legal (recorded) trade because of the enhanced profitability of legal exports. Second, some of the expansion came from the shifting of investment to indus- tries that did not enjoy any comparative advantage; the growth of many nontradi- tional agricultural commodities, for example, was at the expense of traditional plantation crops. Nor is there any evidence that the resulting diversification of the export structure made export eamings more stable, since the prices of many of the nontraditional agricultural exports were positively correlated with those of tradi- tional exports (Jayasuriya and Barlow 1986). Third, in the case of manufactured exports, about 85 percent of such exports in 1975 came from a handful of multinational subsidiaries which temporarily di- verted some of their domestic sales to the world market in order to become eligible for import entitlements under the cRA scheme. Most of these "new" exporters vir- tually disappeared after the 1977 trade liberalization, which raises a question about their intrinsic competitiveness (Athukorala 1984). In general, as is well known, selective promotion of exports involves resource reallocation to favored sectors. Without systematic analysis of such "backwash" effects, it is not possible to evaluate "export success" meaningfully. The nature of the trade and exchange rate regime suggests strongly that the growth of nontraditional exports and newer import-substitution activities often took place at the expense of sectors which clearly had a comparative advantage. Thus, the short-term crisis management policies designed to expand exports tended to enlarge the efficiency costs of the incentive structure and in many cases stifled the country's internationally competitive sectors. The cuts in investment, particularly public sector investment, encouraged moves by the govemment to increase domestic savings, but these moves were hampered by financial repressionist policies. Econometric analysis suggests that substantial increases in savings, particularly private sector savings, could have been achieved with more attractive interest rates on savings. Even the limited in- terest rate increases that were implemented and the expansion of the banking sys- tem facilitated a higher level of financial savings. But higher levels of financial savings within the institutional sector, under conditions of credit rationing and public sector bias, would not necessarily have helped private investment much. The econometric evidence on crowding out sug- gests that cuts in public sector investment would have at least mitigated the ad- verse conditions for private investment by reducing the public sector claims on savings. The policy response to the 1973-75 crisis-the tightening of already stringent controls on many markets and the extension of controls to new markets-certainly had severe short-term effects on economic activity. But, again, it is questionable whether they had any sustained effects on long-term growth. After all, they too lasted for only a relatively short period of time, being completely or substantially Chapter Six 103 removed by the policy reforms of 1977-78. Indeed, some measures, such as con- trols on domestic rice trade, were removed even earlier. Despite the brief lifespan of those measures, we would contend that they had a major impact on the longer-term evolution of the Sri Lankan economy. One rea- son is that Sri Lanka avoided creating a large external debt, whether by design or by force of circumstances. But probably the longer-term impact came primarily from the political consequences of the crisis management policies. They facilitat- ed the demoralization of the traditional left, the weakening of the trade unions, and the election of the UNP government in 1977 on a program that openly proclaimed a fundamental break with past economic policies. Without the first two conditions the implementation of the new policies would have been enormously difficult. The economic hardships suffered by the population during the 1973-75 peri- od, which did not entirely disappear even after the immediate crisis had passed, profoundly influenced the political environment. In the eyes of large sections of the population, "socialism" became identified with shortages, black markets, con- trols, rising unemployment, and large and sudden price increases which cut living standards overnight. Popular dissatisfaction with government policies led to inter- nal conflicts and splits within the constituent parties and weakened their capacity to withstand the challenge from the UNP. When the traditional defenders of con- sumer subsidies themselves discarded the principle that such subsidies were sac- rosanct, a fundamental change took place in the political arena. The attempts to justify sharp price increases and the rationing of many commodities focussed at- tention on the links between the domestic and the world economy. The illusion that the Sri Lankan economy could insulate itself from the world market and fol- low a path of self-reliant development was rudely shattered by the exposure of the vulnerability of the economy to external developments, even after decades of im- port-substitution policies. Greater popular awareness of international economic developments also highlighted the fact that Sri Lanka, which had one of the high- est standards of living in Asia during the 1950s, had fallen well behind nearby countries, such as Singapore, which had pursued liberal economic policies. In these circumstances, the program of liberal economic policies presented by the UNP appealed to large sections of the population. The widespread appeal of an "open economy" program produced a massive electoral victory for the UNP in 1977. It provided the new government with the political legitimacy and popular support essential for implementing fundamental policy changes. It is in this sense that the policies adopted during the 1973-75 crisis were of major long-term significance. The 1978-82 Crisis: Long-Term Effects In terms of GDP growth and investment, the post-1977 decade was sharply differ- ent from the preceding decades. The basis for these developments was laid during 104 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 the 1977-78 period, which started with important trade and exchange rate reforms and was followed by capital market reforms and the dismantling of certain con- sumer subsidies. While the effects of some of the policy reforms were modified by the countervailing consequences of other policies, some of the key elements of those early reforms were retained during the crisis years and survived the decade of the 1980s. In assessing the long-term impact of policies adopted to deal with the shocks inflicted by the investment drive and the terms of trade shock, we will look first at investments, with particular attention to their efficiency aspects. As described ear- lier, the investment boom was primarily, but not solely, a public sector phenome- non. Private sector investments rose to more than 13 percent of GDP after 1978, which was well above historical levels. Since the magnitude of the investments was so large, the key question in analyzing their long-term impact is, how efficient were they? The most commonly used indicator of investment efficiency on an economy- wide basis, the incremental capital/output ratio (ICOR), showed a steady decline in the efficiency of capital utilization throughout this period (table 6.4). Admittedly, the ICOR is a crude indicator of efficiency, particularly when the investments have long gestation periods. But it is striking that the figures show a consistent trend of increasing inefficiency. Indeed, many of the public sector investments, including the house-building program, were chosen and implemented with little or no attempt at rigorous ben- efit-cost analysis. Political imperatives were paramount in the choice of invest- ments, a process facilitated by the availability of generous donor funds. The capacity of the administrative system to undertake careful assessments of alterna- tive projects was limited, and the flood of foreign aid overwhelmed the system. In any case, it is not surprising that many projects of low (or even negative) benefit were funded. The massive leap in investment over such a short period of time was bound to lead to rapid exhaustion of higher return investments, given the con- straints imposed by an incomplete infrastructure, shortages of domestic skill, and weakness in other domestic inputs. Clearly, a more staggered pattern of invest- ment growth would have been more desirable from an efficiency viewpoint. But as discussed in chapter 5, there were political reasons for the PsIP. The important economic argument was that donor funds, which were either grants or loans on highly favorable terms, should be locked into long-term projects immediately to ensure they continued. Since these funds were a free, or nearly free, gift, the notion was that the country was likely to gain some advantage even if the projects were not very efficient. While the assertion that foreign aid flows would have dried up unless locked- in through long-term projects may have had some validity, it is important not to overlook the large domestic costs of the projects when assessing the net benefits. Although most imported inputs, and in certain cases even a part of domestic costs, were funded by foreign donors, the "local counterpart" funds allocated to these projects were extremely large and sometimes exceeded the amount of foreign Chapter Six 105 funding (table 5.2). This had three important effects. First, despite the availability of foreign funds, budget deficits that had to be financed from other sources rose. Second, the commitment of funds to these projects effectively precluded the fi- nancing of potentially more rewarding investments. Third, because the govern- ment as a whole as well as leading political figures had closely identified themselves with these projects, the government's ability to respond flexibly to macroeconomic imbalances was hampered. Such rigidities were aggravated by the interest that donor agencies had in maintaining projects to which they had al- ready committed funds. When expenditure cuts were made, they tended to fall more heavily on other, sometimes more socially desirable, projects. Further, there was a continuation of support for the expansion of many of the inefficient public sector enterprises that were mainly official conduits for the dispensation of polit- ical patronage, particularly in the form of jobs. As a consequence, they became even more inefficient and required substantial government transfers. Thus, these investment programs imposed substantial direct costs on the economy. There were important indirect costs as well. These were the Dutch disease ef- fects on the tradables sectors associated with capital inflows and higher domestic expenditures discussed in the previous chapter. By themselves, these would have tended to appreciate the real exchange rate, thereby reducing the profitability of the tradables sectors. But there were opposing effects on the real exchange rate due to the fall in national income brought about by the terms of trade deterioration. The decision of the government to maintain the growth of real expenditures, par- ticularly on the PSip, ensured the appreciation of the real exchange rate. Thus, the real exchange rate effects of the investment program tended to offset some of the incentives given to the tradable sectors by the trade and exchange rate reforms, whose goal had been to raise the profitability of tradable goods production. Those reformns, together with financial market reforms, did achieve that aim, as indicated by swift depreciation of the real exchange rate. But subsequent real appreciation eroded these incentives. This focus on the role of real exchange rate appreciation ignores the potential long-term benefits for the production of tradables arising from infrastructural in- vestments. The relevant question from the longer-term viewpoint is whether the adverse real exchange rate effects on the tradable sectors were outweighed by the expected future gains from the investments. Again, this depends on the efficiency of investment. Recall that maintenance of the investment program forced the government to obtain commercial loans, which came at market rates of interest. Hence, at the margin the investment program had to generate returns that exceeded market in- terest rates if they were to be justified on economic grounds. The nature of the pro- grams, with large indivisibilities built in, left little room for scaling down the more inefficient components; this further reduced their effective rate of return. Thus, the real exchange rate appreciation that imposed costs on other tradable sectors via Dutch disease effects appears to have had few, if any, significant offsetting bene- ficial effects. There is strong evidence that the major investments were inefficient. 106 Macroeconomic Policies, Crises. and Growth in Sri Lanka, 1969-90 Even the ex ante appraisal of the accelerated Mahaweli project in 1979 suggested that the overall rate of return would be no more than 11 percent, a rather low fig- ure.27 Evaluations of various components of the Mahaweli project suggest that they produced very low-and even negative-returns because of large cost over- runs and declining prices for rice and oil (World Bank 1988).28 There are no com- parable evaluations of the other major public sector investment programs, such as the housing program and construction of the new capital. In each of these cases the resulting macroeconomic effects reinforced the appreciation of the real ex- change rate. Hence, the timing of those programs is certainly debatable. The only agricultural sector that grew steadily after 1977 was the rice sector (table SA-6). But this was due mainly to a more favorable price environment re- sulting from the deregulation of rice marketing, reductions in consumer food sub- sidies, the adoption of high-yielding varieties, and positive discrimination for rice production provided by a range of implicit and explicit subsidies.29 The contribu- tion of the Mahaweli project to rice production has been relatively minor and was achieved at very high cost (World Bank 1988). The indirect costs of maintaining the large public sector investment program also included the crowding-out effects on private sector investment via the ration- ing of investment capital. The higher profitability of tradable sectors resulting from trade reform might have attracted larger private sector investment if crowd- ing out by the public sector had been absent. But, as indicated by the regression results on private investment behavior, such crowding out probably did occur and may have imposed additional long-term costs on the economy, given the ineffi- ciency of many public sector investments. An additional source of costs to the economy was the stimulus given to rent-seeking activities by the PSIP. Policy Responses, Income Distribution, and Political Stability We now turn to the relationship between macroeconomic policies and their impact on income distribution and political stability. This has become an issue of great concern in the Sri Lankan context, due to the social and ethnic conflicts that erupt- ed in the 1980s. Clearly, political stability is of central importance for long-term economic growth, and the political crisis of the late 1980s dealt a severe blow to the prospects for Sri Lanka's economic growth. The cost of defense expenditures alone has risen from less than one percent of GDP to nearly six percent.30 There- fore, it is natural to raise the question whether the policies pursued in the post- 1977 period contributed to the aggravation of social and ethnic conflicts through their distributional effects. A full answer to this question is beyond the scope of this study and is not attempted here. However, we do examine the available evi- dence on the distributional impact of the policy measures on income distribution as a contribution to a broader assessment of the issue. In Sri Lanka's case, the Chapter Six 107 overall changes in income distribution as well as the inter-ethnic distributional changes are of relevance. Any discussion of the distributional issues must be based on the clear recog- nition that fundamental changes in the economy had become unavoidable by the mid-1970s. By then, the economy had reached an impasse. Generous consumer subsidies, rising living standards, and employment growth were no longer sustain- able in the framework of the strategy pursued since independence. But all far- reaching changes in economic policy have distributional consequences. Indeed, often the very objective of such policy changes in a market economy is to produce distributional changes via changes in relative prices and intersectoral profitability. A failure to change course and lead the economy out of stagnation can itself push social and political tensions to the breaking point. If such changes produce general economic growth, absolute levels of income and welfare can rise even if relative inequalities worsen. By appropriate targeting of transfers, undesirable distribu- tional consequences can be minimized, even in the short run. On the other hand, a community will maintain social peace and stability in the midst of sharp changes in income distribution and immediate hardship only if the people can be persuaded of the long-term benefits of such changes. The credibility of the program and the policies becomes bound up with the trust and confidence that is inspired by political leadership. It is difficult to make generalizations about the impact of crisis management policies on the evolution of the political conflicts without an analysis of the specific policy package and the concrete conditions un- der which they were implemented. This is particularly so in our case, since the eth- nic and social conflicts that erupted in the 1980s have deep historical roots. In particular, the stagnation of the economy in the later postindependence period had created a fertile ground for the maturing of conflicts. Long-term social harmony could be achieved only in the context of a dynamic, growing economy. An exam- ination of the impact of post- 1977 economic policies on income distribution with- in and across communities can shed some limited light on whether those policies aggravated the already simmering tensions to the breaking point. Further, it is important to recognize that the observed changes were the result of the total package of policies and programs. In most discussions of the income distribution issue, there has been a tendency to focus on the trade and exchange rate reforms and the consumer subsidy cuts. These, however, were only a part of a whole range of policy measures, including the huge public sector expenditure surge. It is naive and misleading, for example, to attribute all income growth to policy reform alone or increases in poverty solely to subsidy cuts. The empirical evidence on changes in the level of poverty and inequality in the post- 1977 period has generated an extensive literature, but one which is main- ly confined to the analysis of distributional changes within the whole community. There is little or no rigorous analysis of the ethnic dimensions of those changes, partly due to quite serious data problems, and there are two main, and divergent, views on the changes. Bhalla and Glewwe have argued that the period after 1977 was marked not only by an acceleration of economic growth but also by a reduction 108 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 in income inequality (Bhalla and Glewwe 1986; Glewwe 1986, 1988a, 1988b). They attribute both of these results to the beneficial effects of liberalization. In our view, however, the methodological criticisms of their critics (Anand and Kanbur 1991; Divisekara and Flemingham 1986; Lakshman 1986; Ravallion and Jayasuriya 1988; Sahn 1985) are persuasive and indicate a deepening of ine- quality, with particularly adverse effects on the lowest income households. This view is supported by data from the Consumer Finance Survey of 1986-87, which became available only recently. According to Edirisinghe (1991 a), figures provid- ed by the Central Bank suggest a considerable worsening of poverty between 1978-79 and 1986-87, with the proportion of households living in absolute pov- erty increasing from 19 to 27 percent during this period. As we have argued earli- er, even the growth effects of the immediate post-1977 period owed more to the public investment surge than to the positive effects of liberalization. The growth of average income during this period appears to have masked pro- nounced distributional changes. Anand and Kanbur (1991:79), in a critique of the Bhalla/Glewwe analyses, point out that, between 1978-79 and 1981-82, with the rupees 60 (per capita monthly food expenditure in 1978-79 prices) poverty line, the incidence of poverty in Sri Lanka goes up from 12.9 percent to 13.3 percent. This increase is driven largely by an increase in the incidence of poverty in the rural sector from 12.8 percent to 13.6 percent. A small positive contribution is also made by the rise in estate sector incidence from 3.6 percent to 5.8 percent, but this is more than counteracted by a fall in urban poverty incidence from 14.3 percent to 12.4 percent. The direct cause of this increase in rural poverty is not obvious (particularly in view of the significant increase in rice production), but an important factor is likely to have been the reduction in food subsidies. It is probable that even the ru- ral poor were adversely affected by higher food prices induced by subsidy cuts; they were net consumers of food, and their incomes did not rise sufficiently to compensate for higher food expenditures. The cuts in food subsidies were drastic. Real government expenditure on them fell by 69 percent between 1979 and 1982, after rising in the 1977-79 period (Ravallion and Jayasuriya 1988). The negative impact of the 1977 subsidy cuts on the poor is also shown by data on daily calorie intake reported by Sahn (1985). The percentage of Sri Lankans with intake below 1,800 calories per day rose from 12.6 to 15.5 percent. The impact of the first real devaluation would have benefited those in the trad- able sectors, including rural producers of agricultural exportables and those who were able to obtain employment in the expanding tradable manufacturing indus- tries. But these positive effects were counteracted by the (exogenous) fall in the world prices of some key export crops. The erosion of the real depreciation rate over time would also have reduced the initial positive effects. The effects of trade liberalization and capital market reform on overall in- come distribution cannot be directly ascertained from the available data. But there Chapter Six 109 are no compelling reasons to expect that more complete data would show a gen- eral worsening of absolute poverty, even though the sectoral pattems of income distribution would have changed. In fact, the trade liberalization patterns that end- ed shortages of many consumer goods and reduced their effective prices were ex- tremely popular among all classes. This is not surprising, given that a large proportion of wage goods are imported. Even if the prices of certain imported goods rose, overall consumer welfare probably increased as a result of greater availability.31 Similarly, in principle, both net savers and net borrowers gained from capital market liberalization. Savers gained from earlier higher returns on savings, while borrowers gained from easier access to credit. Unlike the food subsidy reduction, however, trade liberalization could have had a differential impact on ethnic and regional patterns of income distribution. Some of the losers from trade liberalization were the farmers who had benefited from the import restrictions that had once raised the prices of certain food crops, such as chilies and onions. Since these crops are well-suited to the northern parts of the country, Tamil farmers in the Jaffna peninsula became major producers of these crops during the protectionist period and suffered from falling prices after trade liberalization occurred. This generalization does not apply to the entire Tamil community, however. Tamil consumers gained from lower prices and great- er availability of imports, and from the general revival of private sector commer- cial and trade activities, where members of the Tamil community had traditionally played an important role. Further, there was a large exodus of skilled Tamil work- ers (as part of the general emigration) to jobs in the Middle East in the late 1970s that was facilitated by some aspects of the broader liberalization package. Unfor- tunately, there exist no data on levels of income (or expenditure) in the immediate preliberalization period that could be compared with the data for the immediate postliberalization period. The overall effects of liberalization on the relative in- comes of the Tamil community are unclear, but there is no strong evidence to sug- gest that they were generally adverse. The more direct and important economic effects with a bearing on ethnic and regional income differences were probably related to the nature of the public sec- tor investment program and the decisions made on the distribution of transfers. The government's concentration on the Mahaweli project was perceived by the Tamil community as discriminatory. The Tamils saw the project as predominantly benefiting the Sinhalese community, and this perception was not entirely unfound- ed. During the planning of the Mahaweli project, the Tamil community had asked that the project's irrigation benefits be extended to traditional Tamil settlements, but these requests had been rejected. Ever since the 1930s, in fact, the Tamil community had objected to state- funded settlement schemes to resettle landless Sinhalese in the sparsely populated parts of the dry zone. These "colonization" schemes typically involved the reha- bilitation of ancient irrigation systems or construction of new ones, and were often inefficient. But the policymakers had other goals. The goal of rice self-sufficiency was one. There was also an overtly ethnic objective. The new settlers came mostly 110 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 from the southem part of the country and were ethnically Sinhalese. The Mahawe- li plan was proclaimed by leading Sinhalese politicians to be aimed at recreating the ancient glories of the Sinhalese kingdom in the dry zone. The large investment in the public sector also meant that a large proportion of new employment was being created in that sector, but Tamils encountered two ob- stacles in obtaining public sector jobs. One was the imposition of proficiency re- quirements in Sinhalese for public sector employment, and the other was the marginalization of the Tamil community from the mainstream political process. This made it difficult for Tamils to obtain the political patronage that had become increasingly more important for obtaining such employment. The imposition of racial quotas in higher education institutions compounded these problems. Tradi- tionally, Tamils had invested heavily in education and were particularly well-rep- resented in the technical and professional sectors. Racial quotas blocked these avenues of advancement for Tamil youth. The initial employment growth caused by the PSIP gradually slackened as projects neared completion. On the other hand, the real exchange rate effects of those programs tended to discourage the growth of labor-intensive export-oriented industries that could have provided a more stable long-term source of employment and income growth. The choices made in relation to the allocation of transfers made the distributional outcomes worse. We have already referred to the food subsidy cuts and their probable distribu- tional effects. The strongest economic rationale for this policy was the argument that the subsidies constituted a fiscal burden large enough to threaten macroeco- nomic stability. If the subsidy cuts were aimed at freeing up the financing needed for productive investments that would alleviate future poverty, then in principle the cuts could be considered a trade-off between present and future poverty alle- viation. But if the funds made available by the subsidy cuts were directly or indi- rectly used to finance programs that did not reduce future poverty, the trade-off was a different one. The govemment followed the latter path. Massive transfers were made to po- litically favored but loss-making public enterprises whose losses substantially off- set the impact of subsidy cuts on fiscal deficits. A particularly glaring example of these inequitable transfers involved the national airline, Air Lanka. During the 1980s, annual transfers to Air Lanka approached (and sometimes exceeded) the expenditures on food subsidies. The airline consistently obtained transfers larger than the combined total that went to the public bus and railway systems. Since the country was well-served by intemational airlines, there were no economic argu- ments to even remotely justify these subsidies. In a country as politically sensitive as Sri Lanka to consumer welfare changes, the political consequences of such choices would have been easy to predict. As the above discussion shows, the long-term growth consequences of the macroeconomic policies of the post-1977 period have to be related to the entire package of policies implemented during this period. The consequences of those Chapter Six 111 other policies often conflicted with the stated goals and objectives of the reform package and reduced their effectiveness. Initially, these conflicts appeared not to matter, but as macroeconomic imbal- ances emerged with the inflationary surge in 1980, the need for possibly painful adjustment measures became clear. However, there were political constraints. The initial popular euphoria gave way to disillusionment with the government, partic- ularly as inflation eroded the value of food stamps and exposed the reality of food subsidy cuts. The decline of popular support was seen in the election in 1982, which touched off widespread allegations of electoral fraud (Manor 1984). Under such clouded circumstances, a government in a democracy finds it difficult to im- plement painful adjustment measures. In these circumstances, there was pressure to resort to the traditional options of trade and exchange rate controls. This temptation was generally resisted, partly because of pressure from donor agencies but also because this approach would have been highly unpopular in the country itself. On the other hand, choices made about government transfers during this period did little to correct either macroeco- nomic imbalances or income distribution. To the extent that they aggravated social tensions, disrupted production and investment, and raised defense expenditures, the transfers did considerable damage to long-term growth. Chapter Seven Summary and Conclusions When Sri Lanka became independent in 1948 it appeared to have all the charac- teristics needed for rapid and sustained economic growth. It had what appeared to be a stable political system based on parliamentary democracy, and it had a high standard of living by developing country standards. In terms of income, literacy, and health care it was well ahead of most of its neighbors, and Sri Lankan society was not characterized by extremes of poverty and wealth. But violent social and political conflict in the 1980s shattered the country's outlook, and as the 1990s be- gan the prospect that the Sri Lankan economy would enter a sustained growth path in the near future appeared slim. Terms of Trade Fluctuations and Optimal Policy As a primary commodity exporter, Sri Lanka has periodically faced sharp swings in its terms of trade, and rational macroeconomic policymaking would have been rooted in this reality. If an intertemporal consumption-smoothing strategy was considered optimal, then it was necessary to build up external reserves during good times by running current account surpluses so that deficits could be financed in bad times. Otherwise, the periodic swings in the terms of trade would generate a pattern of boom and bust, and if domestic absorption was downward rigid, reg- ular and ever more serious balance of payments crises were inevitable. If both pri- vate agents and the government were well-informed and rational, their savings and investment decisions would have been based on this reality, and the current ac- count outcomes would have reflected it. However, as our analysis has shown, the actual history of Sri Lanka was quite different, and one of our objectives is to ex- plain why this was so. Certainly, neither the government nor private agents could have fully antici- pated the long secular decline in the real prices of Sri Lanka's traditional export 112 Chapter Seven 113 crops in the early postwar years. But this can be at most only part of the explana- tion. Even by the late 1950s there was a general recognition that the real prices of those commodities were unlikely to rise in the long run. Indeed, the rationale for pursuing industrialization and agricultural diversification was that primary com- modity prices would decline in the long run. It was also not the case that trends in the world economy were unfavorable. Opportunities for expansion of labor-intensive manufactures were at hand, and there were clear possibilities for export growth and efficient diversification. These might well have counteracted the declining trend in terms of trade implied by con- tinued reliance on traditional export crops alone. In other words, recurrent balance of payments crises due to periodic terms of trade swings or long-term backsliding due to adverse price movements were not inevitable. The reasons for the long- term problems have to be sought elsewhere, by examining the levels and patterns of savings and investment, the efficiency of investment, and how economic poli- cies influenced savings and investment. Welfare Expenditures: Not the Key Issue The issue of growth versus equity has figured prominently in recent debates on Sri Lanka's economic performance. This has been seen primarily in terms of a tradeoff between allocating resources to consumer subsidies or allocating them to investment. The achievements of Sri Lanka in the areas of health, education, and income have justly received wide acclaim and are considered to demonstrate the successes that can be achieved with targeted government intervention. However, high levels of expenditures on welfare are unsustainable without growth. If funds allocated to these programs are redirected to productive investment, growth and future incomes can be raised, and in that sense there is a clear tradeoff between current expenditures on welfare and the levels of future incomes. To the extent that economic growth is expected to reduce poverty, this raises issues of intertemporal and intergenerational tradeoffs.32 We have discussed the fiscal constraints imposed by welfare expenditures and their macroeconomic consequences, pointing out the link between fiscal deficits and balance of payments problems. Note, however, that the size of fiscal deficits could have been reduced without incurring major welfare losses. Given that con- sumer subsidies were provided for the whole population and not just the poor, there can be little doubt that welfare expenditures could have been better targeted and the alleviation of poverty achieved at lower cost. Further, other government expenditures, including certain capital expenditures, were quite inefficient. Note also the choice of instruments which were used to raise govemment rev- enues to finance welfare expenditures. In the main, revenues were raised by highly distortionary taxes on trade that exacerbated the adverse impact of overall govern- ment policies on traditional exports. 114 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Thus, the focus on welfare expenditures as the sole, or even the main, con- straint to growth grossly oversimplifies the role of the broader economic policies followed by the govemment. The most striking feature of Sri Lankan macroeconomic history in the postin- dependence period was the reluctance to address basic macroeconomic imbalanc- es at their source and to adopt a long-term perspective in policymaking. A tendency to treat temporary favorable conditions as if they were permanent is a common thread of continuity through all the regime and policy changes. Perfect foresight was not needed, for example, to have realized that the commodity price boom during the Korean War, or the periodic tea price booms caused by weather- induced world supply shortages, would not be permanent. But time and again, driven by short-term political goals, Sri Lankan governments behaved as if these were indeed permanent. Current expenditures were raised during such times, but they could not be cut in bad times due to the downward rigidity of the social wage, a product of the political polarization of society and the strength of Sri Lankan trade unions. Locked into levels of expenditure that were unsustainable in the long run without rapid growth, and facing chronic fiscal and current account deficits, successive govemments resorted to increasingly tighter trade, exchange rate, and other capital market restrictions to achieve temporary external balance. Macroeconomic Imbalances, and Trade and Exchange Rate Controls The consequences of a long-term policy of ever-increasing reliance on trade and exchange controls to address (or, rather, to avoid addressing) macroeconomic im- balances are clear. Patterns of investment become distorted, bureaucratic regula- tions become pervasive, growth slows down, and, despite its quantitatively reduced dependence on imports, the economy finally ends up in a state of extreme vulnerability to external shocks. These problems are compounded by policies of financial repression that lead to low domestic savings and higher reliance on for- eign savings. Private agents are unable to generate stabilizing responses to chang- ing extemal conditions because price signals are heavily distorted and regulations severely constrain the range of available options. In such circumstances, private and social profitability diverge more and more, and the productivity of investment declines. The role of economic nationalist ideology in the Sri Lankan context was to rationalize and justify long-term reliance on trade and exchange controls. The ide- ology was based on a worldview that ignored the evolution of the Sri Lankan economy and thus propagated the myth that Sri Lanka could develop in isolation from the powerful forces molding the world economy. This view, untenable even for large economies, was completely at variance with reality for a small trade-de- pendent economy like Sri Lanka's. While the initial imposition of controls was dictated by pragmatic, short-term balance of payments considerations, vested in- Chapter Seven 115 terests gained from the continuation of controls and reinforced the ideology of economic nationalism. But adjustments to crisis situations that aggravated the in- ward orientation of the economy were doomed to lead to economic stagnation and failure. This lesson was driven home by the longer-term failure of the crisis man- agement measures adopted during the 1973-75 crisis, although these measures made it possible to avoid an immediate inflationary crisis or a large buildup of ex- ternal debt. Change and Continuity in the Post-1977 Period We have examined post-1977 Sri Lankan experience in light of the lessons of the preceding period of economic stagnation. The changes since 1977 have been widely accepted as representing a major break with previous policies, and certain- ly there have been important changes. The very acceptance of the idea that inward- oriented policies had reached an impasse was a major break with the past, and the trade and exchange rate reforms were extensive and far-reaching. However, the new regime's basic strategy was centered on the "big push" of a public sector in- vestment surge to achieve self-sufficiency in food. This strategy reflected the po- litical pressure for higher incomes and employment in a short period of time, and in the short run both could be achieved with a large public sector investment pro- gram. The rapid economic growth achieved in the early years created a euphoric optimism aroused by the massive capital flows accompanying the policy reforms. Initial reductions in welfare expenditures were made under conditions of rapid growth and a broad-based rise in income, and there was little effective resistance to the cuts. Inefficient enterprises in the public sector were made conduits for em- ployment generation and political patronage, a continuation of past practice, and they were maintained by large fiscal transfers and regulations that insulated them from the impact of trade liberalization. It seemed as if a painless transition to a rap- idly growing free market economy was becoming a reality. But this was a misleading picture. The rapid growth recorded in the early pe- riod primarily reflected the direct, short-term impact of the massive fiscal stimulus to the economy and was essentially a Keynesian-type demand-determined growth. Whether such growth could be sustained in the longer run depended on the pro- ductivity of the investments and the successes of the policy reforms, but the very nature of the surge undermnined the potential for enhancing investment efficiency. Large Projects, Macroeconomic Stability, and Growth Here, it is useful to summarize the lessons learned from the public sector invest- ment program built around the two lead projects, the Mahaweli scheme and the 116 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 housing program. The capital devoted to these two projects dwarfed all other in- vestments. Recall that at its peak, the Mahaweli project alone absorbed six percent of GDP, 17 percent of total government expenditure, and 44 percent of all govern- ment capital expenditure. Economic and political developments during the plan- ning and implementation of these projects make it possible to formulate the following generalizations: * Strategies that create boom conditions overnight appeal to politicians and donor agencies. * Implementation of large projects has immediate positive effects on over- all growth indicators, irrespective of their longer-term impact. (In much the same way, pyramid-building would have dramatically raised the growth rate in the construction and services sectors and therefore of over- all GDP in ancient Egypt!) * Impressive growth indicators suggest a vibrant economy and help attract, for a time, greater economic and political support both at home and abroad. * The success or failure of such projects comes to be seen as an indicator of the success or failure of the regime (not to mention donors). Objective appraisals are ignored. and changes become difficult to implement. The temptation to throw good money after bad becomes almost irresistible, due to political imperatives. * Such investments crowd out other, more socially beneficial but less glamorous investments. Further, by their overall impact on the wider economy they distort the pattern of private investment. * To the extent that benefits from these projects are not widely diffused, they may aggravate internal social tensions. * Such projects reduce the government's ability to respond flexibly to shocks by making it impossible to undertake expenditure reductions. D Other sectors then bear the brunt of the necessary cuts. * Such projects can have large and potentially adverse impacts on macro- economic stability and sectoral performance that run counter to the long- er-term interests of the economy. Indeed, in the Sri Lankan case, the public investment program, by appreciating the real exchange rate, coun- teracted the structural adjustment goals of trade and exchange rate liber- alization, which required a real depreciation. Sri Lanka's Experience in a Comparative Setting The history of Sri Lanka during the latter half of the 1980s suggests that the prob- lems that have plagued the economy since independence remain unresolved. Ex- port growth has remained sluggish, domestic savings remain low, and balance of payments pressures have not ceased. Given that many other small developing Chapter Seven 117 economies also were confronted by major shocks during this period, it is useful to briefly compare Sri Lanka's performance with that of similar countries to place it in perspective. The adoption of trade restrictions in response to balance of payments crises was not unique to Sri Lanka; it was a common practice among developing coun- tries until the 1970s. And as in Sri Lanka, these restrictions have often tended to perpetuate themselves. The consequences, too, have been similar. What distin- guishes Sri Lanka from other countries was the pervasiveness and stringency of the restrictions, particularly during the early 1970s. Perhaps the only comparable Asian country (outside the centrally planned countries) was Burma under the Ne Win regime. But in response to the 1973-75 crisis, Sri Lanka's actions differed from those of most developing countries in that its fiscal policy was strongly deflationary and its crisis-induced rate of inflation was moderate by most developing country stan- dards. Moreover, and unlike many developing countries, Sri Lanka passed through the 1973-75 crisis without incurring a large external debt. All of this, however, was achieved at a high cost in terms of investment and growth. Sri Lanka's post-1977 development had many similarities with public sector spending booms during the mid-1970s in a number of developing countries, such as Indonesia, Nigeria, Pakistan, and Turkey (see Corden 1991 for a detailed dis- cussion). In most such cases, however, the surge in spending was related to an ex- port boom of some sort (oil in Indonesia and Nigeria) or to the availability of migrant worker remittances and cheap credit (Pakistan and Turkey). Although Sri Lanka also benefited from migrant worker remittances, the main source of exter- nal funding was aid. In terms of the magnitude of the rise in public spending, Sri Lanka towered above all others. No other country in recent history raised public sector spending so massively in such a short period of time. But its boom ended the same way as many of the others: faltering growth after an initial jump, accompanied by balance of payments and debt problems. Perhaps the worst aspect of such booms (and one shared by Sri Lanka) was that they made it possible to postpone necessary adjust- ments at the cost of even more severe problems later. Conclusions Up until the ethnic upheaval of the 1980s, Sri Lanka's extensive array of welfare subsidies facilitated social peace and contributed to political stability. But despite their effectiveness, such policies were unsustainable without economic growth, which required policies based on a long-term perspective and devoted to exploit- ing the trade opportunities offered by the international economy. Thus, Sri Lanka missed the chance for greater growth offered by the widespread postwar boom. 118 Macroeconomic Policies, Crises. and Growth in Sri Lanka, 1969-90 A comparable era of sustained worldwide economic growth is not on the ho- rizon. The l990s promise to be a period of great volatility in the world economy. The decade started badly for Sri Lanka as the Gulf War added to the burdens of the ongoing civil war by destroying two major sources of external finance: Iraq was a large buyer of Sri Lankan tea, and the Middle East was the largest source of employment for Sri Lanka's migrant workers. While political stability appeared to be a possibility after the crushing of the JVP uprising in 1989-90, and some impor- tant policy initiatives were undertaken, tensions continued to simmer beneath the surface. Thus, the country's economic prospects will depend to a large degree on po- litical developments that are difficult to predict. But the depths of Sri Lanka's long crises can only be removed by fundamental change. Long-term growth must be based on policies that stimulate savings and raise investment efficiency, allow flexible responses to domestic and external circumstances, and recognize that the Sri Lankan economy is closely tied to the world economy. Only such policies will achieve economic growth and poverty alleviation over the long term and thus per- manently resolve the country's ingrained social and ethnic conflicts. Developing such policies is the challenge facing Sri Lanka today. Appendix I In the orthodox neoclassical growth model (Solow 1956; Swan 1956; Meade 1961; Uzawa 1961) the long-run growth rate depends on the growth rate of the la- bor force and the exogenously given rate of technological change. Changes in the rate of investment, according to the model, have no long-run effects on growth. Technological change is assumed to be independent of the rate of investment, which implies that any change in the latter can have only transitory effects. Vari- ations on this model, such as the vintage model in which technical change is em- bodied in investment, do not alter this basic conclusion (for an exposition and comparison of these models, see Hamberg 1971). Development economists have been uncomfortable with these models. While prescribing all manner of policy packages that were supposed to make a real difference in the long run, the models attach no importance to policy. Note, however, that, even in the traditional model, investment does change the growth rate, except in the long run. Attempts to develop theoretical founda- tions for models in which policy affects long-run steady-state growth have led to a large and growing body of theoretical literature (see, for example, Barro 1991; Grossman and Helpman 1989; King and Rebelo 1990; Lucas 1988; Rebelo 1991; Romer 1986). A major contribution to this literature has been that of Scott (1989), who has produced a fundamental critique of the traditional neoclassical growth model and the methodology of the empirical studies used to support that model. He argues that the level and quality of investment are crucial determinants of the long-run growth rate, and that investment is better regarded as the cost, in terms of consumption foregone, of changing economic circumstances than as physical increments to the capital stock. Technical progress, so the argument goes, is not independent of investment, and new capital seldom reduplicates old capital. (Some of the best-known empirical studies, such as Denison (1967), treat invest- ment as if new capital merely reduplicates existing assets.) The rate of invention is "endogenized" in the rate of gross investment rather than treated as exogenous. Technical change is not unimportant; rather, it cannot be separated from measures of new investment. In the Scott model the growth rate is explained by two main variables, the growth of employment and the rate of investment. The capital stock is not of cen- tral interest, and there is no independent rate of technological progress variable. Thus, the basic Scott growth equation is: 119 120 Macroeconomic Policies. Crises. and Growth in Sri Lanka. 1969-90 G = aQS + bL (1) where, G = rate of growth of output Q = investment efficiency S = rate of investment L = rate of growth of employment and a and b are country specific parameters. The variable Q allows for such factors as diminishing returns to the rate of in- vestment. Thus, as S increases, investment efficiency can decline, since fewer and fewer productive investment projects will be included in the total investment pro- gram. Other relevant variables are those that may influence the quality or efficiency of investment, such as the tax system, capital market efficiency, and the pressure of competition. (In principle, variables that affect the quality of labor also enter into the equation.) We can express these relationships affecting investment efficiency in terms of the following equation: Q=Q(S.T)=(I +a'S+aal) (2) The first tern in equation 2 can be allowed for by adding the term S2 to the right-hand side of equation 1. The expected sign for the coefficient of this variable is negative. The term T is an index of the degree of trade liberalization (see Cuth- bertson and Athukorala 1990). We hypothesize that trade liberalization is likely to improve allocative efficiency in the economy and thereby impact on investment efficiency. T is entered as an interactive term for S (that is, T*S) whose coefficient is expected to be positive. We adapted this hypothesis in examining the Sri Lankan growth pattern and estimated the following equation: G=aS+bS2+c(T*S)+L+U+DI+D2 (3) Following Scott, the G variable refers to the rate of growth of real output in the nonresidential business sector. Gross investment is conventionally defined as the measure of material investment net of maintenance. The rate of investment is simply the gross investment to GDP ratio, estimated using both series (in current prices). To allow for the gestation of investment, we use the moving averages of the investment rate over three previous years. Rate of growth of employment (L) should, in principle, be measured using "quality adjusted" employment data. In the absence of this ideal measure, we measured employment growth in terms of gross employment. It should be noted that our employment series is of doubtful quality even as an indicator of change in gross employment. When equation I is estimated using time series data, rather than fairly long period averages, it is im- portant to allow for cyclical fluctuations in growth. We use U, an index of capacity utilization, as an additional regressor to allow for that. In the absence of a better AppendLi I 121 alternative, the series for U was constructed as the ratio of actual output to trend output obtained by fitting an exponential trend line. We also include three dummy variables to incorporate the effects of the 1971 events, major changes in the economic and political environment after the 1950s, and the impact of the civil war during the 1987-89 period (the variables having a value of I during the relevant period and 0 otherwise). The model was first estimated with data for 1950-79, since the data series for L was available only for this period. L was subsequently dropped and the equation was re-estimated, since its coefficient was consistently statistically insignificant and had an unexpected (negative) sign. The lack of significance of this variable may be due to the very poor quality of the employment data series (see the statis- tical appendix) and in any case does not fully capture the changes in underemploy- ment that are important in labor-surplus economies like Sri Lanka's. Further, note that we could not adjust it for quality due to lack of a suitable measure. The dele- tion of this variable enabled us to use a longer time series, and the final equation was estimated for the period 1950 to 1983. The sample period could not be extend- ed beyond 1983 because of an inexplicable inconsistency in the subsequently available investment series. The equation was estimated using TSLS; there was ev- idence, in terms of the Wu-Hausman test, that S is asymptotically correlated with the contemporaneous disturbance term. The results are reported in table 6.1, and the data series are given in table SA-19. The equation explains nearly 70 percent of the variation in the annual rate of growth of real GDP over the sample period. This is quite a satisfactory fit given the absence of the employment variable, the quality of the data, and (in particular) the absence of a direct measure of capacity utilization to satisfactorily account for cy- clical influences on output.1 Determinants of Domestic Savings The following regression models were tested, using annual data for the period 1960-87. Total domestic savings (DS) and domestic private savings (DPS): DS or DPS =f(YR, PGDP, RSTD, D*RSTD, FS, FIR, D) Domestic financial savings (FNS): FNS =f(PGDP, RSTD, D*RSTD, PE, D*PE, D) The explanatory variables (with expected signs of the regression coefficient in pa- rentheses) are: YR (+) = growth rate of GDP (1980 prices) measured as logarithmic differences 122 Mac roeconomic Policies. Crises. and Growth in Sri Lanka, 1969-90 PGDP (+) = per capita GDP (1980 price) in logs RSTD (+) = weighted average real interest rate on savings and time deposits PE (-1) = expected rate of inflation proxied by the rate of inflation with a one-year lag FS (-) = foreign savings (1980 price) in million rupees FIR (+) = financial intermediation ratio, measured as the ratio of M3 to GDP D = a dummy variable which takes value I for years 1978 through 1987 (post-trade liberalization period) and 0 for other years Domestic saving (DS) is estimated by subtracting foreign saving (Fs) from gross domestic investment. Foreign saving is taken as equivalent to the balance of payments deficit on the current account. Private saving (PS) is estimated by sub- tracting budgetary saving (GB) from DS. Budgetary saving is defined as the excess of govemment revenue over govemment current expenditure plus capital outlays for military hardware. Gross (rather than net) saving is used because of the un- availability of data on depreciation allowances. A major limitation in estimating domestic saving by subtracting foreign sav- ing from GDI is that inaccuracies in either series may lead to measurement errors. However, if the biases are consistent over time and errors are random, the use of this series need not necessarily yield misleading results (Fry 1988). In studies of savings behavior in developing countries, private and public sav- ings have usually been treated as a lump-sum variable. Aggregating these two components, however, can bias the coefficient of the savings function because the underlying determinants of the two tend to be different. In particular, public sav- ings usually exhibit a higher degree of volatility than private savings. To allow for these differences, we estimated a function for domestic private savings in addition to the one for total domestic savings. A separate function for financial savings is estimated as a way of shedding light on the implications of interest rate policy for the mobilization of private sav- ings. Financial savings are proxied here by annual changes in time and savings de- posits in commercial banks and other noncontractual saving institutions (that is, the change in M3 - M I). Dummy variables (slope dummies for the interest rate and inflationary expec- tation variables, and the intercept dummy, D) are included to capture the impact on observed relationships of changes in broad government policies and possible shifts in private sector factors brought about by the financial market and trade lib- eralization policy package introduced in late 1977. The other variables are those commonly used to capture the essence of the major hypotheses advanced to explain savings in developing countries.2 The growth rate variable (YR) is justified on the ground that rapid growth leads to in- creased transitory income in relation to permanent income, and the former influ- ences saving more than the latter. The per capita GDP variable (PGDP) reflects the Appendix 1 123 state of development in the economy, which has a direct bearing on the level of savings. Foreign savings (FS) is included to test the hypothesis that such inflows may "crowd out" domestic savings by allowing the government and residents to consume more at a given rate of capital accumulation (Weisskopf 1972; Newlyn 1977). This variable is included linearly (in million rupees) rather than in log form because it assumes negative values in some years in the sample period. The finan- cial intermediation ratio (FIR) is included to capture the impact of changes in the degree of financial deepening on domestic savings mobilization. Financial deep- ening helps raise the domestic savings rate by raising the level of financial savings and by improving the efficiency with which loanable funds are allocated.3 The results of the regression exercise are reported in Table 6.1 in the text. Note that the intercept dummy, D, was excluded from the final equations. In ex- perimental runs the high multicollinearity between it, PGDP, and the slope dum- mies resulted in distorted coefficient estimates. Preliminary estimates using OLS indicated the presence of significant autocorrelation among residuals in the total domestic savings and domestic private savings functions. Therefore, an iterative maximum likelihood correction procedure was adopted. Determinants of Private Investment The model estimated to explain private investment behavior is the following: PRIV = f(ICP, YR, RC, PBIV(-1), PRIV(-I)) where PRIV is real private sector fixed capital formation. The explanatory variables (with the expected signs of the regression coefficients in parentheses) are: 1cP (+) = total real institutional credit to the private sector in log YR (+) = growth rate of real GDP estimated in terms of log differences PBIV' (><0) = real public sector fixed investment in log, lagged year RC = an index (1980 = 1) of rental cost of capital (to be explained below). All variables are measured at 1980 prices. The index of rental cost of capital (RC) is constructed as: RC = PIM * (I + NLR)/P) where PIM price of investment goods proxied by the import price index of investment goods (1980=100) adjusted for the 124 Macroeconomic Policies. Crises. and Growth in Sri Lanka, 1969-90 dual exchange rate (FEEC) premium for the period 1968-77 NLR = nominal one-year (maximum) lending rate of commercial banks P = implicit GDP deflator (1980 = 100) for nonservice activities. The cost of capital depreciation was ignored because of lack of data. The in- clusion of GDP growth (YR) as an explanatory variable implies an accelerator-type relationship between the level of domestic economic activity and capital forma- tion. A view that has gained considerable currency is that one of the principal con- straints on investment in developing countries is the quantity of available financial resources rather than their cost. The rudimentary nature of the capital market in these countries limits the financing of private investment to retained profits and bank credits. Of these, the flow of bank credit is perhaps the most important. We therefore include the annual change in real credit (ICP) as an explanatory variable. PBIV was chosen as an explanatory variable in the light of the debate on the crowd- ing out effect (Blejer and Khan 1984). PRIV(-l) is included as a regressor on grounds that the adjustment of actual investment to the desired level is not com- pleted within a single period. The model was estimated by OLS using annual data covering the period 1960- 82.4 The results are reported in table 6.2. Short-Run Supply Response of Exports In this section we report on an analysis of the response of Sri Lankan commodity exports to economic incentives. Given the heterogeneity of exports and the selec- tivity with which export-incentive (disincentive) policies were implemented, re- sults obtained for total exports (or even for the two conventionally identified categories: traditional and nontraditional exports) might be of little policy rele- vance. In this exercise we therefore treated exports under eight subcategories. The sample period was 1965-85. Methodology The export supply equation was of the form: RX = f (REER, LPRD or TIME, LBD) RX = real exports REER = real effective exchange rates LPRD = production lagged by one year TIME = time trend (1965 = 0) Appendix 1 125 LBD = a dummy variable set equal to I for the years 1978-85 and 0 otherwise. All regression coefficients were expected to be positive. For tea, rubber, and coconut kernel products (henceforth coconut for short), the Central Bank's export volume indexes were employed as the dependent variable (RX). For other com- modity categories other than gems, the RX series was constructed by deflating ex- port earnings by the respective export price (unit value) indexes. Due to the lack of an appropriate price index, the dependent variable for gems is measured in units of a million dollars each. REER, the key explanatory variable in the model, measures changes over time in the relative profitability of exporting and selling domestically. It brings together changes in the nominal exchange rate, the effective value of financial incentives, and domestic and world market (foreign currency) prices. It is defined as: REER= IOER (1 - TX + SX)1 PW PD where OER = official exchange rate (Sri Lanka rupees per one U.S. dollar) X = export tax rate sx = rate of total effective subsidy PD = index of domestic prices PW = index of foreign (export) prices. The numerator of the above formula is simply the index of the local currency price of exports (LCPX). REER, which is measured as the ratio of LCPX to domestic price index (PD), is therefore a composite indicator of relative export profitability. The U.S. dollar is the principal invoicing currency in foreign trade for Sri Lanka, accounting for over 60 percent of total export receipts. Ideally, the rate should have been deflated by an index of composite value of all currencies used in export transactions to take into account fluctuations in exchange value of these currencies compared with the dollar. This was not possible because of lack of sufficient data. Normally, the composite index is compiled using the share of each partner country in total exports as the proxy weight. We did not adopt this procedure because it would have resulted in a misleading indication of the average exchange rate change, since trade with any particular country is not conducted exclusively in the currency of that country. The subsidy rate (sx) incorporates import duty rebates, rebates on the tum- over tax, the subsidy equivalent of export credit and outright cash grants, and the scarcity premium of credit to convertible rupee accounts (for the period 1973-77). Tax concessions at the income tax level were not taken into account because of identification difficulties. For all commodity categories, the foreign-price index (PW) used was the ex- port unit value index adjusted for exchange rate changes. For tea, rubber, and co- 126 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 conut, the cost of production index was used as the domestic price variable. Domestic price indexes used for the other commodity categories were the implicit national product deflator for nonplantation agriculture (all minor agricultural ex- ports and spices), the implicit national product deflator of manufacturing (for total manufactures and other manufactures), and the GDP deflator (for coconut by-prod- ucts and gems). The REER variable is expected to capture only the changes in real exports re- sulting from movement along the production possibility curve triggered by rela- tive price changes. In practice, even if REER remains unchanged, exports would still change depending on shifts in the country's production capacity. The net ef- fect of changes in the REER on exports can therefore be meaningfully delineated only if the latter changes are taken into account. On these grounds, we used two alternative proxies to account for shifts in the production possibility curve: pro- duction lagged one year (LPRD) in the three manufactured goods equations, and time trend (TIME) in the other equations. The third explanatory variable, LBD, is included to capture the distinction (if any) between the pre- and post-trade liberalization periods with regard to export performance. The existing literature suggests that the favorable impact of trade liberalization on export performance is not limited to relative price aspects (which are captured in the REER). Various institutional changes embodied in a liberaliza- tion package (such as easy access to imported inputs, increased flexibility in the exchange control mechanism, and a more conducive atmosphere for private sector activity) may also contribute to improved export performance (Bhagwati 1978; Krueger 1978). An important assumption implicit in the model is that foreign demand for Sri Lankan exports is perfectly price-elastic, which implies that export prices are de- termined exogenously (the "small-country" assumption). Judging from market share data, this assumption seems justified for all commodity categories except tea. Yet even for tea, Sri Lanka's market share has been declining (27 percent in 1980-83 as against 32 percent in 1972-74). A short-run export supply model of this nature has limited relevance for pe- rennial crops, which are characterized by a high degree of inflexibility in the short run.5 However, output of these products can be varied by varying the harvesting intensity (that is, intensive tapping in the case of rubber, intensive plucking in the case of tea) and to a certain extent by varying the application of inputs, such as fertilizer and other yield stimulants. Moreover, in cases where a significant portion of output is absorbed locally, export volume can be varied through changes in the rate of local absorption. Regression Results The regression equation estimated for the selected commodity categories using or- dinary least squares was reported in table SA- I. All the equations have been cor- rected for autocorrelation using the maximum likelihood iterative technique Appendix 1 127 (AR I). The model was estimated in both linear and log-linear forms. While broadly similar results were obtained with both forms, the log linear form generally pro- vided a better fit. The coefficient of REER in each regression can therefore be di- rectly interpreted as an estimate of the export supply elasticity. Note that in all regressions REER has been included with a one-year lag. This specification gener- ally gave better results than when current period REER was included. Alternative specifications with higher-order lags, as well as a partial adjustment specification (that is, the inclusion of lagged dependent variables as an explicator), failed to yield better results. The coefficient for REER in the equation for rubber is statistically insignifi- cant, but is statistically significant for tea and coconut. The estimated export sup- ply elasticities for these two commodities are 0.3 and 0.7, respectively. The results generally support the view that Sri Lanka's traditional exports are price inelastic in the short-run. However, the weak year-to-year link between REER and real ex- ports does not mean much with regard to the implications of the incentive struc- ture for long-term export performance. The relatively higher elasticity estimate for coconut is not surprising because, in contrast to tea and rubber (of which less than 10 percent is absorbed locally), nearly 75 percent of total coconut production is consumed locally. Therefore, the level of domestic demand is a very important determinant of export levels. Throughout the period under study, coconut output expanded at a slower rate than the rate of expansion in local demand. This helped generate a highly lucrative do- mestic market, thereby reducing the exportable surplus. As regards minor agricultural products, the coefficient of REER is not signifi- cantly different from zero in the equations for the aggregate as well as for coconut by-products and spices. The coefficient of REER in the equation for the residual mi- nor agricultural commodity category (2.3) is significant and has the expected sign. The estimate suggests that a one percent change in relative price is reflected in a 1.2 percent change in real exports with a one-year lag. This commodity group con- tains a variety of products, ranging from crops with a long gestation period (such as cocoa and coffee) to crops with a shorter gestation period, such as cut-flowers and vegetables. The high elasticity coefficient reported above is probably strongly influenced by the export behavior of products in the latter category, which have shown considerable dynamism since the early 1 970s. Gem exports (category 3) appear to be as highly elastic (2.6) with respect to changes in incentives. A well-known feature of the Sri Lankan gem trade is that a significant portion of total exports occurs through illicit channels. Therefore, ex- porters can respond to changes in export incentives simply by shifting exports from illicit channels to official channels. This makes official exports quite sensi- tive to incentive changes. The coefficient of REER in the equation for total manufactured exports is not statistically significant. The estimates for the two manufacturing subcategories suggest that this result for total manufactures was determined by the export pat- terns of textiles and garments. Since the mid-1970s, export growth of manufac- 128 Macroeconomit Policies, Crises, and Growth in Sri Lanka, 1969-90 tures has primarily reflected growth in this category (Rajapatirana 1988). Its share in total manufactured exports increased from 30 percent in 1970-72 to 75 percent in 1983-85. While factors exogenous to the domestic exchange rate and trade pol- icy scene were very important, the trade liberalization in 1977 and the receptive approach of the government toward foreign investors undoubtedly provided the groundwork for exploiting those opportunities. This is borne out by the highly sig- nificant coefficient for LBD in the equation for textiles and garments. The coeffi- cient of REER for other manufactures is highly significant and suggests a 1.5 percent response to a given percentage point change in lagged relative prices. The interpretation is that even when allowance is made for the special factors affecting exports of garments, exchange rate and incentive policies were important influ- ences on the promotion of manufactured exports. In sum, our estimates generally support the view that Sri Lankan exports in the aggregate are price inelastic in the short run. The weighted-average elasticity coefficient obtained from the desegregated regressions (treating all insignificant coefficients as zero) is 0.42. At the same time, the results suggest that even in the short run an active exchange rate and export incentive policy could have played an important role in promoting new export lines. Appendix II Notes on the Statistics Sri Lankan statistics on the economy suffer from a number of serious limitations. even though Snodgrass has argued that "the body of statistics available for Ceylon (Sri Lanka) must certainly rank as one of the most complete and accurate in the world for a country of such low per capita income" (Snodgrass 1966:300; see also Pyatt and Roe 1977:35, for a similar observation). In this section we point out some of the major limitations.X Sources A large number of Sri Lankan agencies are engaged in data collection, and their published figures often show discrepancies. For example, the Central Bank and the Department of Census and Statistics (DCS) compile data on many aggregate economic variables. They produce independent sets of national income accounts that are often at variance with each other. Hence, different sets of global estimates of gross domestic product, gross national product. and their components are avail- able, both at current and constant prices. Such discrepancies between sources are found in many other statistics, including those relating to the domestic price level, foreign trade, and employment. In this study we generally rely on published Central Bank statistics and sup- plement them from other sources where it is felt to be desirable. One important reason for our choice is that more recent data are published by the Central Bank, whereas the DCS data are normally available only after a considerable delay. How- ever, there are problems of consistency in the Central Bank data. Data series found in the documents of international financial institutions that give the Central Bank as their primary source are often at variance with data given in Central Bank pub- lications. The conflicts in the mid-1980s then raised problems in the collection of certain data series, and even certain key statistics (such as GDP) need to be treated with caution. 129 130 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 National Accounts The Central Bank started compiling its national accounts in 1959 and has revised the method of compilation a number of times, the latest significant revisions being in 1978 and in 1983.2 The new series is available from 1982 in the Central Bank's annual report, and Savundranayagam (1983) has unified and revised the figures for the previous years back to 1950 in line with 1978 revisions. The Central Bank series (defined to include both the "Savundranayagam" data and "official" Central Bank data) conforms more closely with the UN system of National Accounts. However, problems remain: A substantial proportion of the variables (comprising one third of GDP in 1970) are estimated indirectly, and some variables (about 15 percent) are assumed to grow at a steady trend rate of growth. The operation of a dual exchange rate system in 1968-77 created problems in imputing correct prices to tradable goods. In the latest revisions to Central Bank series for this period, all external transactions have been valued at the higher FEEC- inclusive rate on the assumption that it would be a reasonable proxy for the rate that would have prevailed in the absence of controls. However, the considerable differences observable between the FEEC-inclusive rate and the curb market rate, and year-to-year changes in these differences (table SA-15), suggest that this may not be entirely appropriate. There are some differences between estimates arrived at by the "product method" and those arrived at by the "expenditure" method, and since about 1982 there has been a considerable enlarging of these differences. There are reasons to suspect that the former may have overstated the level of economic activity (Jayawardena, Maasland, and Radhakrishnan). For this reason we used expendi- ture account data in constructing the activity variable to be used in our regression analysis of savings and investment. However, in analyzing growth performance on a sectoral basis, there is no alternative to product-account data. In constructing constant price series, the current figures for the services sector are usually deflated by the Colombo Consumer Price Index (cCPI) and the wage rate indices for workers.3 The CCPI, as discussed below, is a very poor index for such use. Likewise, since the Wages Boards report wage rates only in the organized sector, the wage rate index does not properly reflect wage changes in the large unorganized sector. These facts should be borne in mind when using the implicit GDP deflator rather than the CCPI for inflation adjustments; the two series are not entirely independent. Employment Data Employment data are perhaps the weakest link in the Sri Lankan national data sys- tem. There are no data on overall employment which are directly compiled and Appendix 1! 131 published on an annual basis. Snodgrass (1966) and Hallett (1981) provide total employment series constructed using data from periodic surveys, supplemented with International Labor Organization (ILO) population projections for the periods 1950-59 and 1960-79, respectively. The various surveys of employment available to analyze changes in employ- ment and unemployment over time (as listed in table 3.7) have often used different definitions, making strict comparisons impossible. Annual figures on employment are available only for the "organized" sector, which accounts for only 10 to 15 per- cent of total employment. In general, unemployment figures should be treated with great caution. Trade Data Trade data are produced by the Central Bank as well as the DCS. Further, trade data for particular commodities are also available from other sources, such as the Food Commissioner's Department, the Petroleum Corporation, and the Gem Corpora- tion. The Central Bank adjusts the figures obtained from customs records, using data obtained from these various sources. Substantial discrepancies are present be- tween the trade data from different sources. Sri Lanka's import data are presented only on a CiF basis. Naturally, these trade figures do not capture the effects of ex- tensive underinvoicing of imports and overinvoicing of exports. Note also that the import and export price indices published by the Central Bank are really unit value indices rather than true price indices. Aggregate export unit value indices become quite misleading indicators of the underlying price movements when the compo- sition of the product categories changes substantially. This has certainly occurred in Sri Lanka since the late 1970s, with the notable shift in the export structure from primary commodities to manufactured exports. Thus, the terms of trade move- ments indicated by the available series should be used with caution. Fiscal and Monetary Data These are probably the best data available at a disaggregated level. Wages Available data going back to 1950 refer only to the minimum wages in the orga- nized sector that, in many cases, represents only a minority of wage earners. Fur- ther, the ratios between minimum wage and total wage earnings vary between 132 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 years and sectors. Informal sector wage data are available (from Central Bank sources). in most cases, from the mid- or late-1970s only. Price Data The two longest time series available are the CCPI and the implicit GDP deflator.4 Bhalla (1987) has presented a revised version of the CCPI for the period 1952- 86). A WPI and an (unpublished) Central Bank Consumer Price Index (CBcPI) have been in existence since 1974. A Construction Cost Price Index (ccl) has been available since 1977. The ccPi, though the most widely used price index, suffers from major prob- lems. It is based on a set of weights based on a 1949-50 family budget survey cov- ering 455 working-class households in Colombo. The small sample was unrepresentative of the wider community even at that time. Colombo workers' families constituted only three to five percent of the total population, and there were major differences in the expenditure patterns between different groups. The set of weights and the basket of goods used in 1952 have continued in use despite large changes in expenditure patterns. Furthermore, weighted average prices of controlled goods have been calculated on the basis of a constant set of weights be- tween open and controlled market prices. These weights have little correspon- dence to the relative household expenditures on the two categories. Typically, black market prices are not captured, which makes the index a particularly poor guide to price movements during the early 1970s, when rationing and widespread scarcities resulted in thriving black markets for many consumer goods. Generally, the ccPi underestimates actual price changes. For instance, using survey data pro- vided by the DCS, Bhalla (1987) found that food grain prices rose by over 380 per- cent during the 1970s, as against 166 percent indicated by the CCPi. Furthermore, the subindex for expenditure on house rent shows practically no increase over 30 years. The adjustment of the CCPi by Bhalla (1987) involved the replacement of con- trolled prices for the food and rent components of the index with open market pric- es. In this study we generally use the Bhalla-adjusted CCPI (BACCPI) rather than the original CCPi as the measure of consumer price movements. As noted above, in national accounts the constant price estimates of value added in services sectors are derived using the CCPi and wage rate indexes. There- fore, the overall GDP deflator suffers from the above-mentioned limitations of these base indices. But the implicit deflator estimated separately for the tradable sectors can be considered a more reliable indicator of price changes at the produc- er level than the overall GDP deflator. This is because constant price estimates for these sectors are directly obtained using volume indices, which are considered to be fairly reliable (Ranaweera 1978). Appendix /1 133 The CBCPI has been developed to overcome some of the major weaknesses of the ccpi and is considered to give a much higher weight to tradable goods and open-market prices. For reasons that have not been made public, this index is not publicly available. 134 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table SA-1 Gross Domestic Product (GDP), Mid-year Population (POP), Per Capita GDP (PGDP), and their Growth Rates, 1950-89 (1970 prices) GDP' (millions POP PGDpa Growth rates (percent) Year of rupees) (thousands) (rupees) GDP POP PGDP 1950 6,669 7,678 868 - - - 1951 7,084 7,876 899 6.2 2.6 3.6 1952 7,409 8,074 917 4.5 2.5 2.1 1953 7,553 8,290 911 1.9 2.6 -0.7 1954 7,758 8,520 910 2.7 2.8 -0.1 1955 8,212 8,723 941 5.8 2.4 3.4 1956 8,266 8,929 925 0.7 2.4 -1.7 1957 8,387 9,165 915 1.5 2.6 -1.2 1958 8,631 9.388 919 2.9 2.4 0.5 1959 8,752 9,620 910 1.4 2.5 -1.1 1960 9,074 9.900 917 3.7 2.9 0.8 1961 9,260 10,140 913 2.0 2.4 -0.4 1962 9,593 10,380 924 3.6 2.4 1.2 1963 9,791 10,650 919 2.1 2.6 -0.5 1964 10,257 10,900 941 4.8 2.3 2.4 1965 10,579 11,160 948 3.1 2.4 0.8 1966 10,961 11,440 958 3.6 2.5 1.1 1967 11,417 11,700 975 4.2 2.3 1.9 1968 12,269 11.990 1,023 7.5 2.5 5.0 1969 12,857 12,250 1,049 4.8 2.2 2.6 1970 13,364 12,520 1,067 3.9 2.2 1.7 1971 13,251 12,690 1,050 -0.8 1.4 -2.2 1972 13,530 12,860 1,052 2.1 1.3 0.8 1973 14,078 13,091 1,075 4.1 1.8 2.2 1974 14,536 13,284 1,094 3.3 1.5 1.8 1975 14,873 13,496 1,101 2.3 1.6 0.7 1976 15,312 13,720 1,116 2.9 1.7 1.3 1977 16,021 13,942 1,149 4.6 1.6 3.0 1978 17,329 14,190 1,221 8.2 1.8 6.4 1979 18,461 14,470 1,275 6.5 2.0 4.6 1980 19,468 14,740 1,320 5.5 1.8 3.6 1981 20,651 14,897 1,377 6.1 1.1 5.0. 1982 22,482 15,189 1,550 428b 2.0 2.8b 1983 23,604 15,420 1,574 4.9 1.5 3.5 1984 24,796 15,599 1,574 5.1 1.2 3.9 1985 26,026 15,840 1,627 4.9 1.5 3.4 1986 27,140 16,117 1,668 4.3 1.7 2.5 1987 27,534 16,361 1,667 1.5 1.5 -0.1 1988 28,277 16,586 1,689 2.7 1.4 1.3 1989 28,914 16,808 1,704 2.3 1.3 0.9 a. The new (1982-89) GDP series has been linked to the old (1950-81) series by deflating the joined current price series by the GDP deflator (1970 = 100). Figures for the two subperiods are not strictly comparable because of differences in the method of compilation. b. Growth rates based on the old series. Source: Savundranayagam (1983), for the 1950-81 period; Central Bank of Ceylon, Annual Report (annual). Stazistical Appendix 135 Table SA-2 Gross Domestic Fixed Capital Formation, 1966-88 As a percentage of GDP' At constant (1970) pricesb Public Public Govern- corpo- Govern- corpo- Year ment rations Private Total ment rations Private Total 1966 4.6 2.7 6.9 14.2 486 275 772 1,483 1967 4.9 2.1 8.1 15.1 559 237 993 1,729 1968 4.3 2.6 7.7 14.6 513 307 905 1,725 1969 4.8 2.9 11.8 19.5 841 344 1,420 2,605 1970 4.1 3.3 9.8 17.2 570 451 1,338 2,359 1971 3.5 2.6 8.9 15.2 465 346 1,208 2,019 1972 3.3 2.0 9.2 14.5 462 285 1,223 1,970 1973 3.3 1.6 8.6 13.5 433 224 1,175 1,832 1974 3.4 1.4 7.7 12.5 412 174 923 1,509 1975 4.1 1.6 8.2 13.9 491 191 975 1,657 1976 5.4 1.9 7.9 15.2 706 255 1,029 1,990 1977 4.2 2.3 7.3 13.8 468 261 798 1,527 1978 7.3 4.8 7.9 19.9 943 663 1,093 2,749 1979 7.2 5.0 13.0 25.2 1,095 753 1,959 3,807 1980 7.0 11.3 12.9 31.2 1,171 1,879 2,135 5,185 1981 4.8 9.8 12.7 13.2 923 1,870 2,415 5,208 1982 5.1 11.6 15.2 31.9 1,040 2,339 3,091 6,470 1983 5.2 25.8 31.0 1,096 5,400 6,496 1984 5.1 25.2 28.3 1,116 5,123 6,239 1985 5.2 20.7 25.9 1,231 4,863 6,095 1986 5.8 20.0 25.8 1,449 4,916 6,365 1987 5.7 17.6 23.3 1988 5.8 16.7 22.5 a. From 1983 onwards, public corporations and private sector investment figures are not available in disaggregated form. b. Estimated using GDP (nontradable) deflator. Source: Compiled from data series obtained from Central Bank of Ceylon, Review of the Economy (annual) and Savundranayagam (1983). 136 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 Table SA-3 Gross Domestic Saving, 1959-87 (percentage of GDP) Year Total saving Private saving 1959 11.2 10.3 1960 10.6 12.2 1961 12.1 12.6 1962 12.2 11.9 1963 13.2 11.8 1964 11.5 12.2 1965 12.3 11.8 1966 10.3 10.7 1967 11.8 11.6 1968 11.6 11.6 1969 9.5 9.0 1970 15.6 16.5 1971 15.1 17.0 1972 15.7 17.0 1973 12.5 11.9 1974 8.2 7.2 1975 8.1 8.6 1976 13.9 14.3 1977 18.1 17.8 1978 15.3 16.6 1979 13.8 13.4 1980 11.2 14.8 1981 11.7 13.5 1982 11.9 13.3 1983 13.8 13.7 1984 19.9 16.1 1985 11.9 8.9 1986 12.0 8.1 1987 12.8 8.7 Note: Ratios have been estimated using current price data. Source: Central Bank of Ceylon, Annual Report (annual) and Review of the Economy (annual). Table SA-4 Sectoral Distribution of Investment, 1966-88 (percentage of gross domestic capitalformation) Sector 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 19871988 Gross domestic fixed capital formation 99 99 94 101 91 89 84 99 80 89 94 96 100 98 93 99 99 101 100 99 100 100 99 Private sector and public corporations 67 68 66 76 69 69 64 75 58 63 61 66 64 70 72 81 83 84 82 73 77 75 73 Planting, replanting and land development 6 6 5 4 4 5 4 4 3 3 2 2 2 1 1 1 1 1 1 2 3 2 2 Building and other construction 33 38 41 36 38 41 35 41 36 31 32 28 26 28 33 29 29 34 35 35 41 41 41 Plant and machinery 14 11 10 17 12 12 9 15 5 13 14 17 18 17 17 18 21 24 22 19 20 21 20 X Transportequipment 9 6 4 14 9 7 11 10 7 9 6 10 11 15 20 21 23 15 11 9 5 4 4 Other capital goods 5 7 5 4 5 4 5 5 7 8 7 9 7 8 5 8 9 9 13 9 9 7 6 Government and public enterprises 33 32 28 25 22 21 20 23 22 26 33 29 36 28 21 17 16 17 18 27 23 24 25 Changes in stocks: I 1 6 -1 9 11 16 1 20 11 6 4 1 2 7 1 1 -1 I 1 0 0 1 Private sector and public corporations 4 -2 2 7 1 5 -I 3 1 -I -2 5 3 3 2 1 0 1 0 0 0 1 - Govemment and public enterprises -3 -4 -3 1 10 12 3 18 10 7 6 -4 0 4 -I 0 0 0 0 0 0 0 - Gross domestic capital formation 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Private sector 48 54 48 61 57 33 55 55 50 52 50 50 45 51 40 47 48 83 82 73 77 76 74 Public corporations 22 14 16 15 20 36 14 19 10 11 10 15 23 21 34 36 36 0 0 0 0 0 0 Government and public enterprises 30 32 27 25 23 31 31 26 39 36 40 35 32 28 25 17 16 17 18 27 23 24 26 Note: Figures may not add up to 100 due to rounding; zero means less than 0.5 percent. Source: Central Bank of Ceylon, Annual Report (annual). 138 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table SA-5 Index of Industrial Production, 1970-89 (1977=100) Public Year sector'a Totalb 1970 81 67 1971 82 68 1972 101 71 1973 100 71 1974 101 60 1975 102 73 1976 103 101 1977 100 100 1978 125 100 1979 111 113 1980 120 117 1981 118 120 1982 125 125 1983 115 126 1984 107 142 1985 118 149 1986 116 165 1987 116 176 1988 114 188 1989 90 198 a. Central Bank's Public Sector Major Industry Output Index. b. Compiled using data relating to real manufacturing value added (excluding processing of planta- tion products) obtained from national income accounts. Source: Based on Central Bank of Ceylon, Review of the Economy (annual). Statistical Appendix 139 Table SA-6 Production of Major Agricultural Crops, 1960-89 Tea Rubber Paddy (millions of (millions of Coconut (thousands of Year kilograms) kilograms) (millions of nuts) metric tons) 1960 197 99 2,183 901 1961 207 98 2,601 903 1962 212 104 2,811 1,005 1963 220 105 2,549 1,028 1964 219 117 2,991 1,055 1965 228 118 2,676 759 1966 222 131 2,461 955 1967 221 143 2,416 1,152 1968 225 149 2,601 1,350 1969 220 151 2,440 1,377 1970 212 159 2,510 1,616 1971 218 142 2,610 1,396 1972 214 140 2,963 1,312 1973 211 155 1,935 1,312 1974 204 132 2,031 1,602 1975 214 149 2,398 1,154 1976 196 152 2,330 1,253 1977 209 146 1,821 1,677 1978 199 156 2,207 1,891 1979 206 153 2,393 1,917 1980 191 133 2,026 2,133 1981 210 124 2,258 2,230 1982 188 125 2,521 2,156 1983 179 140 2,398 2,479 1984 208 142 1,983 2.420 1985 214 138 2,958 2,661 1986 211 138 3,059 2,588 1987 213 122 2,487 2,128 1988 227 122 1,996 2,477 1989 207 111 2,517 2,063 Source: Central Bank of Ceylon, Annual Report (annual). 140 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 Table SA-7 Basic Indicators of the General Price Level, 1950-89 Year CCPI GDPD WPI CBCPI BACCPI 1950 - 60.47 - - - 1951 62.12 - - - - 1952 100.00 58.48 - - 98.3 1953 100.30 61.32 - - 99.9 1954 101.40 63.86 - - 96.3 1955 102.00 63.13 - - 98.0 1956 102.60 64.17 - - 100.0 1957 102.80 67.29 - - 102.0 1958 105.00 69.93 - - 106.3 1959 105.20 71.49 - - 107.1 1960 103.20 76.69 - - 107.0 1961 104.80 75.53 - - 108.8 1962 106.30 75.24 - - 110.4 1963 108.80 76.20 - - 112.9 1964 112.20 77.04 - - 117.4 1965 112.50 75.32 - - 117.5 1966 112.30 77.94 - - 116.1 1967 114.80 80.19 - - 120.7 1968 121.50 91.96 - - 126.0 1969 130.50 100.21 - - 138.2 1970 138.20 100.00 - - 142.9 1971 141.90 103.45 - - 148.9 1972 150.80 110.85 - - 171.0 1973 165.80 128.20 - - 216.4 1974 185.80 15.00 10.00 100.00 289.5 1975 198.30 17.00 10.00 103.00 300.5 1976 200.70 18.00 11.00 107.00 316.2 1977 203.20 21.00 13.00 123.00 348.4 1978 227.30 23.00 15.00 134.00 385.7 1979 252.20 25.00 17.00 159.00 462.3 1980 318.20 30.00 22.00 218.00 603.4 1981 375.40 35.00 26.00 270.00 749.2 1982 416.10 42.00 28.00 300.00 815.6 1983 474.20 48.00 35.00 334.00 932.2 1984 553.10 56.00 444.70 1,086.00 - 1985 561.20 56.00 377.10 1,106.70 - 1986 606.00 60.00 366.00 1,201.90 - 1987 652.80 64.00 414.90 - - 1988 744.10 71.00 488.70 1989 830.20 78.00 532.90 - - Note: ccpi = Colombo Consumer Price Index (1952 = 100). GDPD = Gross Domestic Product Defla- tor (1970 = 100). wpi = Wholesale Price Index (1974 = 100). CBCPI = Central Bank Consumer Price Index (1974 = 100) (unpublished). BACCPI = Bhalla-adjusted ccpi (ccpi as adjusted by Bhalla (1986) by replacing controlled prices of food and rent components by open-market prices. ) Source: ccpi and wPI: Central Bank of Ceylon, Annual Report (annual). GDPD: Compiled from sourc- es listed in table SA-1. CBCPI: World Bank data files. BACCPI: Bhalla (1986). Statistical Appendix 141 Table SA-8 Wage Rate Indices, 1965-89 (1978 = 100) Workers in wages board tradea Government employeesb Year Nominal Real' Nominal RealL 1965 26.5 86.9 45.2 148.2 1966 26.7 88.7 46.0 152.8 1967 27.8 89.4 47.7 153.3 1968 31.9 97.6 55.5 169.7 1969 31.9 89.1 57.1 159.5 1970 32.3 87.2 62.1 167.8 1971 32.8 85.0 62.1 160.8 1972 35.5 80.1 62.3 140.6 1973 38.4 68.4 65.5 116.8 1974 48.1 64.1 73.7 98.3 1975 55.3 71.0 81.5 104.6 1976 56.7 69.1 86.1 105.0 1977 69.8 77.2 87.2 96.6 1978 100.0 100.0 100.0 100.0 1979 119.6 99.8 117.2 97.8 1980 147.3 94.2 129.1 82.5 1981 152.2 98.4 146.1 75.3 1982 175.8 83.2 187.8 88.4 1983 188.8 78.1 215.7 89.3 1984 228.8 81.3 246.4 87.7 1985 247.9 86.5 297.3 99.1 1986 261.3 83.6 297.4 95.2 1987 277.7 82.7 297.4 88.5 1988 335.8 87.6 390.0 101.5 1989 388.1 90.8 421.8 99.1 a. Refers to minimum wages of tea growing and manufacturing, rubber growing and manufacturing, coconut growing and manufactures, engineering, printing, transport, match manufacturing, building, dock, tea, rubber and cinnamon trade. b. Refers to wages of all central govemment employees. c. Nominal wage rate index deflated by the Bhalla-adjusted Consumer Price Index up to 1986, and by the Colombo Consumer Price Index. Source: Nominal wage rate indices: Central Bank of Ceylon, Annual Report (annual)-the old series (base 1952) has been spliced to the base of the new series (1978). Consumer Price Indexes: as in table SA-6. 142 Macroeconromtic Policies, Crises, and Growthr in Sri Lanka, 1969-90 Table SA-9 Terms of Trade Between Agriculture and Industry, 1960-85 (1970 = 100) Implicit deflatorfor agriculture Implicit deflatorfor terms of trade Year Plantatio&s' Domestic Total Manufacturing' (agriculture/manufacturing) 1960 87.4 76.2 83.0 79.8 109.5 95.5 104.0 1961 81.5 73.0 78.2 81.5 100.0 93.4 96.0 1962 80.9 70.6 76.7 82.7 97.8 85.4 92.7 1963 73.1 71.2 72.3 82.4 88.7 86.4 87.7 1964 80.4 72.2 77.0 84.7 94.9 85.2 90.9 1965 85.7 69.7 80.1 91.8 93.4 75.9 87.3 1966 84.7 71.6 79.5 85.3 99.3 83.9 93.2 1967 75.8 86.1 80.3 89.2 85.0 96.5 90.0 1968 94.6 101.9 98.0 94.7 99.9 107.6 103.5 1969 104.6 100.0 102.3 97.8 107.0 102.2 104.6 1q70 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1971 101.8 100.0 108.2 90.9 94.1 98.4 90.9 1971 98.4 101.8 100.0 108.2 90.9 94.1 92.5 1973 125.8 155.9 142.7 138.2 91.0 112.8 103.3 1974 189.0 247.3 224.3 174.8 108.1 141.5 128.3 1975 169.6 275.2 227.7 231.5 73.3 118.9 98.4 1976 204.7 248.9 231.0 205.8 99.5 120.9 112.2 1977 315.3 237.8 265.0 248.0 127.1 95.9 106.9 1978 353.6 280.4 306.0 259.8 136.1 107.9 117.8 1979 412.3 248.4 308.3 325.0 126.9 76.4 94.9 1980 501.8 343.8 394.5 382.6 131.2 90.0 103.1 1981 569.7 383.2 443.1 452.3 126.0 84.7 98.0 1982 585.4 384.3 423.1 500.5 117.0 76.8 86.3 1983 807.9 455.4 535.8 548.0 147.4 83.1 97.7 1984 1,107.0 558.0 678.8 596.6 185.6 93.5 113.7 1985 769.2 581.4 627.8 625.1 123.1 93.0 100.4 a. Tea, rubber and coconut production and processing for exports. b. Manufacturing excluding processing of tea, rubber, and coconuts. Source: Estimated using data from Savundranayagam (1983) (for the period 1960-81) and Central Bank of Ceylon, Review of the Economy (1983), and Annual Report (1985). Table SA-10 Summary of Government Fiscal Operations, 1966-88 (millions of rupees) Items 1966 /968 1970 1972 1974 1976 1978 19813 1982 1984 1986 1988, Current revenue 1,233 2,156 2,815 4,034 5,084 5,739 11,688 14,068 17,809 37,731 41,644 45,675 Expenditure 2,515 3,006 3.886 4,647 6,386 9,314 18,853 30,343 37,900 53,592 69,715 88,916 Recurrent expenditure 1,860 2,186 2,659 3,386 4,506 5,534 10,408 13,249 20,110 28,926 34,722 46,613 Capital expenditure 644 849 1,026 1,207 1,841 3,448 6,614 13,845 18,669 21,750 35.112 44,620 Recurrent expenditure 1,860 -29 201 54 39 312 1,831 3,240 -879 2,916 -170 2,732 Budgetdeficit 682 850 1,150 1,366 1,599 3,576 7,165 16,274 20,091 15,861 28,071 36,176 Debt repayment 116 135 214 330 564 662 1,165 1,810 2,612 2.229 7,525 8,441 Net cash deficit 566 716 936 1,306 1,035 2,914 6,000 14,772 17,479 13,632 20,546 34,800 Financing of budget deficit Domestic non-market borrowing 112 10 16 199 -54 485 453 -216 1,713 904 -660 1,685 Domestic market borrowing 498 630 815 803 964 1,799 2,200 9,826 9,899 3,041 12,137 16,451 Foreign finance 154 247 345 478 650 1,326 4,454 6,735 8,794 11,251 15,834 18,925 Decline in cash balance -82 -36 -26 -115 39 33 58 -72 -315 666 760 -885 *2 Financing of net cash deficit Domestic non-market borrowing 112 10 16 199 -54 485 280 -487 1,664 807 -667 1,673 Domestic market borrowing 418 552 720 668 671 1,505 1,710 9,195 8,010 2,374 7,639 20,296 Foreign finance 188 190 226 284 378 957 3,953 6,136 8,112 9,785 12,814 13,716 Decline in cash balance -82 -36 -26 115 39 -33 58 -72 -315 666 760 -885 Expansionary impact of fiscal operations 122 267 425 112 24 605 173 7,029 3,765 -2,095 3,047 10,153 Public debt outstanding 4,268 5,689 7,237 7,558 11,027 15,621 27,746 46,779 71,250 95,741 149,690 220,003 As percentage of GDP Revenue 23.6 21.7 20.7 22.2 20.5 20.4 28.8 22.6 19.4 26.9 23.1 20.6 Aggregate expenditure 32.4 30.2 29.4 31.5 27.4 33.2 46.5 48.7 41.3 38.3 38.7 40.0 Recurrent expenditure 24.0 22.0 20.1 23.0 19.3 19.8 25.7 21.2 21.9 20.7 19.3 21.0 Government savings -0.3 -0.3 0.5 0.7 1.2 0.6 3.1 0.3 -2.5 6.3 3.8 -0.4 Budget deficit 8.8 8.5 8.7 9.2 6.8 12.7 17.7 26.1 21.0 11.3 15.6 15.7 Expansionary impact of fiscal operations 1.5 2.5 3.2 0.8 0.1 2.3 0.4 10.3 5.8 1.5 1.7 4.6 a. Provisional. Source: Central Bank of Ceylon, Annual Report (annual) and Review of the Economy (annual). 144 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 Table SA-ll Budgetary Transfers to Public Enterprises, 1960-89 (millions of rupees) Total transfers As percent As Current of percent of Transfers transfers as government budget Yeara Current Capital Total percent of total expenditure deficit 1960 - 70 70 - 3.9 12.5 1961 - 73 73 - 3.7 13.5 1962 - 101 101 - 4.9 15.2 1963 - 115 115 - 5.8 14.6 1964 - 87 87 - 3.9 20.1 1965 - 147 147 - 6.0 19.3 1966 3 152 155 1.6 5.9 22.5 1967 1 199 200 0.5 7.1 27.7 1968 7 210 217 3.0 6.9 25.5 1969 12 286 298 4.2 8.3 31.5 1970 12 283 295 4.2 7.5 25.6 1971 24 281 305 7.8 7.7 23.0 1972 28 333 361 7.8 6.7 24.4 1973 166 385 551 30.1 10.9 38.7 1974 164 378 542 30.3 9.3 33.9 1975 134 770 904 14.8 12.6 33.5 1976 129 965 1,094 11.8 12.6 30.5 1977 146 868 1,014 14.4 11.5 33.0 1978 1,081 2,070 3,151 34.3 17.8 44.0 1979 920 3,112 4,032 22.8 19.8 46.0 1980 1,583 6,086 7,669 20.7 26.9 47.1 1981 1,350 7,174 8,524 15.8 28.9 57.1 1982 845 10,591 11,436 8.0 30.2 56.9 1983 1,275 10,123 11,398 11.2 24.3 52.8 1984 1,082 12,192 13,274 8.2 24.8 83.7 1985 751 11,592 12,343 6.1 19.1 48.1 1986 1,936 14,174 16,110 12.0 23.1 57.4 1987 1,264 10,721 11,985 10.5 16.6 43.8 1988b 1,725 9,683 11,408 15.1 12.8 31.5 1960-69c - - 2.3 5.6 17.1 - 1970-75c - _ 15.3 9.6 29.9 - 1977-89c - - 14.9 21.3 39.1 - a. Data for the period 1960-71 are for the fiscal year, for other years on a calendar-year basis. b. Provisional. c. Annual averages; figures for 1960-69 are for 1966-69 period only. Source: Compiled from data from Central Bank of Ceylon, Review of the Economy (annual). Statistical Appendix 145 Table SA-12 End-of-year Foreign Reserves, 1950-89 Year Import-month SDR millions equivalenta 1950 227.5 11.1 1951 244.3 9.0 1952 171.2 5.7 1953 121.0 4.2 1954 188.4 7.9 1955 242.4 9.4 1956 230.8 9.1 1957 204.4 6.6 1958 176.1 5.9 1959 137.6 4.0 1960 96.2 2.7 1961 92.7 3.0 1962 85.5 2.5 1963 75.4 2.3 1964 64.1 1.9 1965 85.7 2.6 1966 59.2 1.7 1967 85.5 2.5 1968 70.1 2.1 1969 55.1 1.5 1970 61.7 1.9 1971 74.2 2.5 1972 100.3 3.7 1973 119.0 3.6 1974 92.6 2.2 1975 92.4 1.9 1976 136.9 3.1 1977 294.8 10.6 1978 370.1 5.9 1979 475.5 5.1 1980 295.5 2.4 1981 387.2 3.1 1982 477.7 3.3 1983 498.5 3.5 1984 735.1 4.0 1985 610.9 3.5 1986 491.7 3.5 1987 423.3 3.2 1988 428.1 3.1 1989 446.8 3.1 a. Estimated by dividing end-of-year reserves by the average monthly import requirement during the year. Source: Central Bank of Ceylon, Review of the Economy (annual) and Annual Report (annual). 146 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table SA-13 Import Structure, 1950-89 (percentage shares) Consumer goods Intermediate goods Investment Investment Year Total Food Other Total Oil goods goods 1950 72.2 45.0 23.2 15.2 5.2 12.6 27.8 1955 66.6 45.7 20.5 18.4 8.7 15.0 33.4 1959 60.0 39.9 20.1 19.8 6.7 19.4 40.0 1960 61.0 38.3 22.6 20.1 6.3 18.1 39.0 1961 57.1 39.5 17.6 22.5 7.5 20.0 42.9 1962 54.6 38.0 16.6 24.6 7.6 20.3 45.4 1963 52.4 42.2 10.2 25.0 8.0 21.7 47.6 1964 64.0 51.4 12.6 20.0 4.7 15.4 36.0 1965 52.9 41.0 11.9 28.1 7.4 17.7 47.1 1966 57.3 47.1 10.1 23.2 6.4 17.8 42.7 1967 53.6 45.1 8.5 25.4 6.4 19.0 46.4 1968 52.8 45.5 7.2 29.0 9.0 17.6 47.2 1969 48.0 38.4 9.5 23.3 6.1 27.5 52.0 1970 56.0 46.3 9.7 20.0 2.5 23.6 44.0 1971 57.2 46.3 10.9 20.7 1.3 21.1 42.8 1972 51.8 44.4 7.4 24.3 1.8 21.2 48.2 1973 52.5 46.0 6.5 30.0 10.9 16.7 47.5 1974 47.0 42.9 4.1 42.2 19.9 10.4 53.0 1975 50.5 48.0 2.5 36.0 16.6 12.4 49.5 1976 36.4 32.2 4.2 48.6 25.1 13.8 63.6 1977 42.2 36.3 5.9 44.1 24.0 12.4 57.8 1978 38.3 28.1 10.1 38.1 16.4 22.9 61.7 1979 34.7 21.3 13.4 40.5 17.4 24.2 65.3 1980 29.9 18.9 11.0 47.7 23.8 24.0 70.1 1981 25.2 13.4 11.8 52.7 27.2 21.7 74.8 1982 20.5 8.5 12.1 51.6 29.3 27.6 79.5 1983 25.6 11.8 13.8 47.7 24.2 26.5 74.4 1984 17.0 10.5 6.5 56.2 22.5 25.6 81.8 1985 19.4 10.9 8.4 54.2 20.3 19.2 73.4 1986 22.5 12.4 110.1 52.4 11.5 19.4 71.8 1987 22.8 12.3 10.5 57.2 14.4 18.7 75.9 1988 24.5 14.4 10.2 56.8 11.0 17.0 73.8 1989 26.1 16.4 9.7 56.4 10.4 15.0 71.4 Note: Textile and clothing imports previously classified as consumer goods are classified as inter- mediate goods from 1984 onwards. Source: Central Bank of Ceylon, Annual Report (1950-85), and Review of the Economy (1974-88). Table SA-14 Commodity Exports, 1966-86 Commodity Note 1966 1967 1968 1969 197) 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Major agricultural products (a) 328.0317.6306.6288.6 302.6271.1 250.0261.8310.A352.3345.8457.5489.6472.3441.9 456.4420-5499.4790.4613.7409.8 (b)-8.7 -3.2 -3.4 -6.0 4.9-10.5 -7.7 4.3 Its - - 32.3 7.1 -3.5 -6.4 -3.3-178 18.7 53.2-32.4-33.3 (c)- - ---9 9.3--- - - 83.0.- - - - 65.7 Tea (a) 2JS.6 222.7 195.1 178.3188A 195.5 171.4 165.2169,3 228.9215.8 336.1 3369284.3286.7284.3275.9329.8604.5434.5381.2 (b) -15.2 3.2-12.4 -7.7 5.7 2.2-11.0 -3.5 2.6 35.2 -5.7 55.1 -2.U-13.5 0.9 0.9 -2.9 19.5 83.3-28.1-35.3 (c) - - - - 6.2- - - - 54.0- - 42.7 -- 29 Rubber (a) 70.9 59.3 55.5 72.4 73.9 31.6 39.1 77.6 92.0 77.5 913 89.3103.1 123.6120.2127.5101.0 114.5126A 92.9 79.7 (b) 11.1-16.4 -6.5 30.4 2.1-51.3 23.7 98.9 1B.5 15.7 18.1 2.3 153 19.9 2.7 5.8-20.5 12.3 11.4-26.7-142 (c)- - - - 22.0 - -- - 18.2- - - - 17.8 -8.3 Coconut kemel products (a) 41-, 35.6 56.0 37.9 40.3 47.0 39.5 19.0 49.1 45.9 33.5 32.1 49.6 64.4 35.0 44.6 43f6 56.1 59,5 86.3 48.9 (b) -18.3-14.3 57.3-32.3 6.3 16.6 16.0-51.1 159.4 -6.5-12.2 -16.6 54.5 29.8-45.7 27.4 -2.2 28.6 6.0-45.0-43.3 (c)- - - 12. - - --10.8 ---- 5.2 - - - 5.1 Minor agricultural products (a) 18.9 18.8 19.2 20.5 21.6 21.5 20.4 26.1 38.6 29.6 36.6 48.0 51.8 63.4 63.1 80.5 86 79.4 75.0 75.1 69.3 (b)-10A -0.5 1.1 0.3 5.3 -0.5 -5.1 27.9 47.8-23.3 23.6 31.1 7.9 22.4 40.5 27.6 7!5 8.3 -5iS 0.1 -7.6 (c) - - - 6.4 - - - - 7.0 --- 9.4 - - - - 7.1 Minerals (a) 3.0 2.4 2.1 2.2 2.9 2.9 4.2 21.0 16i 24.4 30.0 31.4 32.1 31.5 37.4 34.9 37.4 45.0 32.0 31.3 35.9 (b) - - --- - 31.8 - 44.8400.0-293 46.1 23.0 4.7 2.2 -1.9 18.7 -5.5 7.1 20.3-28.8 -2.1 14.6 (c) - - - - 0.9- - - - 5.8- - - - 5.6- - - - - 3.7 Gems (a) 0.8 0.7 0.3 0.1 0.7 0.6 1.8 18.5 16.5 21.3 26.8 28.6 27.1 24.6 30.9 27.9 29.8 37.4 23.6 20.3 231 (b) - - - - - -14.3 2.8108.1184.4 2.7 430 52.9 12.1 54.3 42.0 38.7 24.7 12.4 51.4 2.0 91) (c) - - - - 0.2 - - - - 5.0- - - - 4.6 - - - - - 2.4 Industrial products (a) 1.2 1.3 18.0 2.1 4.2 3.6 :3.7 7.7 21.9 22.5 333 50.9 57.1 88.1125.1 173.52172244.0369.3377.1411.0 (b) - - - - - -14.3 2.8108.1 184.4 2.7 48. 52.9 12.1 54.3 42.0 38.7 24.7 12.4 51.4 2.0 9.0 (c) - - - - 0.1 - - - - 5.3- - - -8.6 - - - - - 4.3 Textilesandgannents (a) - - 0.7 0.6 1.3 1.1 1.0 2.6 42 3.8 8.6 15.1 26.5 55.5 84.9133.3t1524183.3290.1288.1292.7 (b) - - - - - -15.4 -9.0160.0 61.5 9.6126.3 75.5 75.5109.4 53.0 57.0 14.3 23.6 54.1 -0.7 1.5 (Table continues on following page.) Table SA-14 (continued) (c)- - - - 0.4- --- 0.9- - - - 12.6- ---- 30.0 Petroleum (a) - - - - - - 2.7 17.5 425 41.8 51.3 56.3 48.2 95.8145.1148.9142.8106.6126.1140.3 71.7 (b) - - - - - - - 548.0142.8 -1.6 22.7 9.7-14.3987.7 51.5 2.6 -1,6-24.4 19.0 11.3-49.0 Total commodity 1, 11, 1, exports (a) 352.1 342.6332.0315.2335.3 324.2 2940340.2429.6466.249..0658.7674.2 759.5817.71928.3933.7 997.7 432.1 310.5 035.7 including petroleum (b) -12.5 -2.7 3.1 -5.0 6.3 -3.4 -9.4 15.6 26.4 8.6 6.2 33.1 2.4 12.7 7.3 13.5 0.6 6.9 43.5 -8.5-29.9 X Total commodity exports (a) 352.1 342.7332.0315.23353324.2291.3322.7 37.1424.4443.7 602.4620663.7612.6 779.47909891.11,3061,170.2964 excluding petroleum (b) -12.5 -2.7 -3.0 -5.2 6.4 -3.4-10.1 10.8 20.0 9.5 4.5 35.9 4.0 6.0 1.3 16.0 1.5 12.6 46.5-10.4-17.7 a. Value in million SDR (current). b. Annual percentage growth. c. Percentage share in total commodity exports (excluding petroleum). Source: Central Bank of Ceylon, Review of the Economy (annual). Statistical Appendix 149 Table SA-15 Nominal Exchange Rates of the Sri Lankan Rupee, 1950-89 (rupees per unit offoreign currency, period average) Formal offcial rate US dollars Curb- normal FEECS market rate Year SDR rate rate (US dollars) 1950-66 4.78 4.78 - 13.00f 1967a 4.86 4.86 - 14.15 1968 5.95 5.95 7.26b 11.35 1969 5.95 5.95 8.91c 12.00 1970 5.95 5.95 9.23 15.20 1971 5.95 5.93 9.23 18.50 1972 6.48 5,95 9,96d 14.40 1973 7.63 6.40 10.56 11.25 1974 8.00 6.65 10.97 12.10 1975 8.50 7.00 11.58 14.80 1976 9.71 8.41 14.57 13.80 1977 10.35 8.87 14.81e 17.75 1978 19.54 15.61 - 21.25 1979 20.12 15.57 - 21.75 1980 21.52 16.53 - 19.60 1981 22.69 19.24 - 21.75 1982 22.97 20.81 - 23.55 1983 25.15 23.52 - 34.60 1984 26.07 25.43 - 34.75 1985 27.58 27.16 - 33.00 1986 32.87 28.02 - 29.25 1987 43.64 29.44 - - 1988 44.45 31.81 - - 1989 52.57 36.05 - - a. Devaluation of November 1967. b. Established on May 6 based on a 44 percent rate on normal official exchange rate. c. Rate raised to 55 percent on June 18. d. Rate raised to 65 percent on November 11. e. Abolished on November 15. f. Figure for 1966. The rate increased from the average level of 6.35 during 1956-59 to 6.8 in 1960 and 11.5 in 1965. Source: Official exchange rates: IMF, International Financial Statistics (annual); Curb-market rate: World Currency Yearbook. 150 Macroeconomic Policies, Crises. and Growth in Sri Lanka, 1969-90 Table SA-16 Trade-Weighted Exchange Rate Indexesa and Related Data, 1970-86 (1980 = 100) Year NERX WPX NERM WPM MERT WPT WPDb RERX RERM RERT 1970 43.3 37.3 40.6 40.6 41.9 39.1 32.4 49.8 51.0 50.5 1971 47.1 39.0 42.8 44.0 44.0 41.1 33.3 55.1 53.4 54.3 1972 44.4 41.2 44.3 42.9 42.9 42.9 35.4 51.6 52.1 51.9 1973 46.2 48.7 51.6 45.9 45.9 50.2 38.8 58.1 60.6 59.4 1974 47.2 60.1 65.4 46.4 46.4 63.1 43.6 65.1 68.8 67.2 1975 50.3 66.0 69.0 49.5 49.5 67.8 45.0 73.9 75.1 74.7 1976 55.1 70.5 72.0 55.3 55.3 71.2 48.7 79.8 81.9 80.9 1977 58.9 76.9 76.3 59.2 59.2 76.6 59.0 76.8 77.1 76.9 1978 89.8 80.3 79.3 90.3 90.3 79.7 68.3 105.6 105.2 105.3 1979 92.8 88.0 86.1 93.2 93.2 86.8 74.8 109.2 107.5 108.1 1980 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1981 108.5 108.8 107.8 109.2 109.2 108.2 117.0 101.0 101.0 101.0 1982 109.3 108.9 107.5 109.1 109.1 108.0 123.4 96.5 94.9 95.5 1983 121.8 114.8 111.7 121.9 121.9 112.7 154.3 90.6 88.3 89.0 1984 127.5 120.2 116.8 128.1 128.1 118.1 193.9 79.1 77.5 78.1 1985 136.6 118.9 118.2 132.7 132.7 118.5 164.3 98.9 93.5 95.7 1986 155.0 117.4 115.3 157.9 157.9 116.2 159.4 114.1 115.7 115.1 Note: NERX = Export-weighted nominal exchange rate. NERM = Import-weighted nominal exchange rate. MERT = Trade-weighted nominal exchange rate. wpx = Trade-weighted wholesale price index. wPM = Import-weighted wholesale price index. wpT = Trade-weighted wholesale price index. WPD = Domestic (Sri Lankan) wholesale price index. RERX = Export-weighted real exchange rate, NERX x (WPX/WPD) X 100. RERM = Import-weighted real exchange rate, NERM X (WPM/WPD) X 100. RERT = Trade- weighted real exchange rage. NERT X (WPT/WPD) x 100. a. The estimates are based on bilateral exchange rates (expressed in rupees) for Sri Lanka's 10 major trading partners (accounted for 92 percent of Sn Lanka's foreign trade in 1980) and wholesale price indices (foreign currency) used as weights. For the period 1970-77 nominal exchange rates are adjust- ed for the "effective" FEECS premium (FEEc rate multiplied by the share of trade covered by FEEc scheme). CRA premium is not taken into account. b. The Sri Lankan wholesale price index dates from 1974. It was extrapolated to 1970 on the basis of annual changes in the implicit GDP deflator for tradable production. Source: Compiled using data from IMF, International Financial SJatistics (annual), and Directions of Trade (annual). Table SA-17 Real Effective Exchange Rate (REER) Indices for Exports, 1965-86 (1980=100) 1965 1968 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Traditional exports 76 67 64 63 61 58 60 57 68 78 94 91 100 95 86 98 110 94 73 Nontraditional exports 28 30 36 41 47 53 54 59 62 81 80 84 100 101 81 78 80 83 84 minor agricultural products 68 76 74 79 76 72 75 69 05 14 16 13 100 97 90 81 71 69 61 gems 87 30 48 41 85 01 80 74 00 92 28 20 100 100 03 90 91 89 93 manufactures 21 32 31 51 51 56 66 74 84 95 11 89 100 105 04 96 85 93 94 Total exports 50 54 51 52 48 55 57 54 64 76 90 88 100 98 89 70 96 91 80 Note: The REER index shows the change in real return to exporters in local currency from exporting goods worth one unit of foreign currency and is estimated as: REER = |TWER (I - Tx + SB)] Px/Pd where TWER = Trade weighted nominal official exchange rate index (1980=100) Tr~ = Export tax rate SB = the rate of total effective subsidies P. = Export price index (export unit value index except for gems, for which consumer price index of Japan. the major importing country of Sri Lankan gems, is used). Source: Cuthbertson and Athukorala (1990) Appendix B, supplemented with IMF. International Financial Statistics (for the Japanese cpi). Table SA-18 Export Indices, Import Indices, and Terms of Trade, 1950-89 (1978 = 100) Export volume Export unit value Import volume Import unit value Year S 4 2 t| 1950 71 88 167 79 17 23 15 17 III 81 40 62 11 7 7 8 208 1955 85 74 245 91 22 24 9 19 125 100 60 74 12 8 1( 9 201 1960 97 74 164 92 18 25 12 19 174 128 85 101 It 8 12 9 185 1961 101 65 231 95 17 20 9 16 130 124 74 81 11 7 9 9 172 1962 107 74 259 102 17 20 9 16 122 138 84 82 11 7 9 9 178 1963 108 69 205 98 16 18 10 16 101 131 67 70 12 7 10 10 161 1964 107 84 273 107 17 17 11 16 156 130 49 86 14 7 19 12 135 1965 117 88 214 III 16 17 14 16 101 124 42 65 13 7 19 It 142 1966 103 99 169 101 15 17 12 15 157 141 53 89 13 7 19 11 117 1967 112 96 156 105 15 14 11 14 123 135 47 76 13 7 21 11 120 1968 108 109 184 108 17 15 19 17 119 151 46 77 17 8 24 14 117 1969 104 104 152 103 16 20 16 17 115 141 81 82 18 9 26 15 110 1970 108 117 147 107 16 19 17 17 131 116 54 77 18 10 31 16 106 1971 108 94 170 104 17 16 17 17 101 130 50 68 19 11 30 17 99 1972 98 94 211 102 18 14 13 17 135 193 41 67 22 11 25 18 94 1973 107 117 720 114 64 39 60 28 16 29 24 82 1974 91 93 78 89 23 39 50 31 77 62 23 42 52 27 43 42 72 1975 110 116 141 107 2 28 33 29 83 76 45 52 66 28 43 49 58 1976 103 100 131 102 31 44 34 34 85 92 53 57 48 34 52 44 78 1977 95 98 48 94 57 47 56 55 123 86 62 73 56 48 53 54 102 1978 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1979 97 93 98 101 92 132 153 109 139 95 144 123 114 196 145 152 72 1980 97 88 44 99 99 143 163 126 116 145 172 140 177 296 160 217 58 1981 97 97 69 102 104 148 149 129 59 143 293 145 255 394 149 282 46 1982 96 95 94 112 103 120 109 119 75 229 251 150 227 448 228 309 38 1983 86 90 93 109 152 146 159 165 116 190 289 180 230 562 298 375 44 1984 109 89 58 127 226 173 289 207 106 218 283 185 239 571 414 415 50 1985 108 88 114 122 174 144 189 183 87 150 264 158 258 504 198 347 53 1986 113 81 118 131 128 160 133 163 92 184 275 180 273 386 258 321 51 1987 110 78 85 133 152 185 173 188 85 194 290 184 321 437 279 364 52 1988 119 73 52 137 161 250 243 214 96 216 231 189 380 496 395 443 48 1989 111 63 107 148 192 243 204 237 90 196 214 250 487 611 431 525 45 Note: Indices from 1978 are based on 1978 weights. The figures priorto 1978 have been spliced to the new index (1978=100). Indices from 1985-89, based on 1981 weights, have been spliced to 1978 = 100. Textile and clothing imports previously classified as consumer goods are classified as intermediate goods in figures from 1985. Export unit value index/import unit value index 100. Source: Central Bank of Ceylon, Annual Report (annual). Table SA-19 Export Agriculture: Prices, Costs, Export Duties, and Gross Profit Margins, 1950-86 Tea Rubber Year BP$ BP EDR NBPa PPRa COP GPMb BPS BP EDR NBRa ppRa COP GPMb 1950 1.16 5.56 17.50 4.59 4.65 3.05 52.5 0.70 3.37 4.50 3.21 3.42 - - 1951 1.20 5.78 20.54 4.59 4.23 3.06 39.8 1.16 5.58 3.10 4.85 4.74 - - 1952 1.05 5.07 15.84 4.27 3.86 3.07 25.7 0.81 3.88 2.10 3.41 3.94 1.85 112.5 1953 1.13 5.42 16.82 4.51 4.23 2.73 13.4 0.71 3.40 0.00 3.06 2.98 1.88 58.5 1954 1.43 6.86 19.00 5.56 5.76 3.16 82.3 0.63 3.00 11.20 2.66 2.45 1.96 25.0 1955 1.52 7.28 25.10 5.45 4.85 3.25 49.2 0.72 3.44 10.50 3.08 2.82 2.02 39.6 1956 1.37 6.61 21.11 5.24 4.83 3.23 49.5 0.69 3.31 18.00 2.71 3.20 2.01 59.2 1957 1.27 6.13 23.45 4.69 4.10 3.20 28.1 0.66 3.15 20.00 2.52 2.56 1.82 40.1 1958 1.26 6.06 24.85 4.55 3.81 3.15 21.0 0.57 2.73 18.00 2.21 2.05 1.64 25.0 1959 1.25 6.00 18.40 4.90 4.08 3.21 26.4 0.66 3.18 15.20 2.70 2.76 1.65 67.2 1960 1.16 5.88 12.68 5.13 4.14 3.21 28.9 0.74 3.55 13.80 3.06 2.73 1.65 63.4 1961 1.20 5.78 13.21 5.02 4.03 3.10 30.0 0.60 2.88 10.28 2.58 2.22 1.53 45.1 1962 1.17 5.60 14.15 4.81 3.85 2.98 29.1 0.59 2.84 8.92 2.59 2.16 1.47 47.0 1963 1.16 5.57 16.25 4.66 3.70 2.96 25.0 0.56 2.68 6.32 2.51 2.05 1.55 32.2 1964 1.15 5.53 18.32 4.52 3.70 3.04 21.7 0.53 2.53 5.45 2.39 1.98 1.47 34.6 1965 1.13 5.40 17.95 4.43 3.85 3.25 18.5 0.52 2.51 7.24 2.33 2.00 1.61 24.2 1966 1.07 5.14 19.76 4.12 3.45 3.15 9.5 0.52 2.49 4.20 2.39 1.96 1.55 26.5 1967 1.02 4.90 20.79 3.88 3.30 3.26 1.2 0.44 2.13 0.57 2.12 1.74 1.48 17.5 1968 0.94 5.56 19.81 4.46 3.85 3.25 18.5 0.38 2.22 5.44 2.10 1.94 1.64 18.2 1969 0.89 5.27 20.04 4.21 3.37 3.35 0.5 0.51 3.01 15.57 2.54 2.29 1.57 45.8 1970 0.90 5.38 18.99 4.36 3.59 3.35 7.2 0.46 2.73 12.24 2.40 2.00 1.52 31.6 1971 0.93 5.57 17.53 4.59 3.99 3.45 15.7 0.40 2.38 3.20 2.30 1.74 1.68 3.6 1972 0.98 6.10 17.46 5.03 4.18 3.52 18.8 0.33 2.05 3.77 1.97 1.78 1.65 7.8 1973 0.96 6.12 16.36 5.12 4.23 4.47 -5.3 0.58 3.68 23.91 2.80 2.57 1.87 37.7 1974 1.16 7.76 22.07 6.05 5.88 5.55 5.9 0.86 5.75 38.50 3.54 2.82 2.46 14.6 1975 1.30 9.08 16.90 7.55 6.15 6.70 -8.2 0.58 4.05 21.35 3.19 2.49 2.75 -9.5 1976 1.25 0.49 19.78 8.42 7.80 4.42 2.4 0.77 6.50 22.19 5.06 4.34 3.28 32.3 1977 2.12 8.86 22.39 4.60 3.15 8.19 60.6 0.77 6.85 28.00 4.93 4.51 3.74 20.5 1978 2.13 3.22 49.84 6.66 1.55 1.80 -2.1 0.96 4.90 49.53 7.52 6.92 4.85 42.6 1979 1.96 0.57 47.11 6.17 1.14 3.43 17.1 1.25 9.44 49.69 0.03 9.15 6.80 34.5 1980 2.02 3.41 32.60 2.52 7.73 7.17 3.3 1.30 1.42 53.05 0.06 0.62 8.20 29.5 1981 1.83 5.14 30.35 4.47 7.71 8.73 -5.4 1.14 1.80 48.94 1.13 0.04 8.97 11.9 1982 1.68 5.04 24.89 6.31 2.52 1.97 2.5 0.85 7.68 31.62 2.09 0.18 9.66 5.8 1983 2.23 2.52 25.61 9.06 6.96 6.37 40.1 0.97 2.77 29.72 6.00 3.95 9.90 40.9 1984 3.03 7.20 31.67 2.75 6.45 4.00 36.6 1.03 6.16 30.57 8.17 4.94 2.06 23.8 1985 2.23 0.62 18.43 9.45 5.39 5.00 1.1 0.73 1.33 9.89 9.22 6.17 3.67 18.2 1986 1.56 3.95 10.76 9.22 0.28 6.00 15.0 0.85 3.83 11.70 1.04 7.72 3.70 29.3 Note: BPS = border price (per kilogram) in US dollars. BP = border price (per kilogram). EDR = export, duty rate (percent). NBP = after-duty border price in rupees. PPR = producer price in rupees. cop = cost of production in rupees. GPM = gross profit margin (percent). a. The difference between NBP and PPR represents trade and transport margin. b. GPM = |(PPR - COP)/COP1 X 100. Source. Compiled using data from Central Bank of Ceylon, Annual Report (annual) and Review of the Economy (annual); Govemment of Sri Lanka, Department of Census and Statistics. Statistical Abstract (annual). 156 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Table SA-20 Data Used for the Estimation of the Scott Growth Equation, 1950-89 Year G S L 1950 - 0.099 - 1951 0.063 0.119 0.019 1952 0.041 0.142 0.024 1953 0.008 0.118 0.018 1954 0.014 0.097 0.213 1955 0.055 0.113 0.027 1956 -0.002 0.115 0.023 1957 0.011 0.119 0.010 1958 0.026 0.113 0.026 1959 0.009 0.159 0.047 1960 0.039 0.141 0.070 1961 0.014 0.137 0.030 1962 0.038 0.142 0.200 1963 0.016 0.147 -0.004 1964 0.046 0.142 -0.012 1965 0.024 0.131 0.089 1966 0.037 0.139 0.040 1967 0.044 0.149 0.014 1968 0.071 0.139 -0.011 1969 0.052 0.177 0.005 1970 0.036 0.177 0.034 1971 -0.018 0.156 0.040 1972 0.018 0.147 0.033 1973 0.030 0.138 -0.017 1974 0.023 0.130 0.056 1975 0.021 0.146 0.028 1976 0.027 0.166 0.012 1977 0.046 0.173 0.016 1978 0.082 0.210 0.035 1979 0.060 0.281 0.040 1980 0.055 0.346 - 1981 0.064 0.312 - 1982 0.041 0.320 - 1983 0.042 0.310 - 1984 0.053 0.282 - 1985 0.064 0.259 - 1986 0.041 0.259 - 1987 0.014 0.257 - 1988 0.027 0.245 - 1989 0.022 0.238 - Note: G = GDP growth (growth of nonresidential business, defined as total GDP less value added in following GDP categories: (a) ownership of dwelling, (b) public administration and defense, and (c) other services). S = rate of investment (gross domestic fixed capital formation as a ratio of GDP). L = growth of employment. Source: L series was constructed from data in Snodgrass (1966), for the period 1950-60, and Hallett (1981): G and S series are based on data from Central Bank of Ceylon. Annual Report (annual). G and L are logarithmic differences of the relevant series. S has been estimated using current-price data. Table SA-21 Money Supply, 1960-85 (millions of rupees) Time and savings Time and savings Time and savings Demand deposits at deposits at other deposits as a Currency deposits M(M) commercial banks M(2) savings institutes M(3) percentage of M3 Year (1) (2) (3)=(I)+(2) (5) (5)=(3)+(4) (6) (7)=(5)+(4) (6)+(5)=(7) 1960 595.3 613.6 1,208.9 363.6 1,259.7 451.1 1,710.8 29 1961 692.2 596.4 1,288.6 354.4 1,643.0 459.9 2,102.9 39 1962 716.6 630.0 1,342.7 404.6 1,747.3 470.3 2,217.6 39 1963 828.4 677.7 1,506.0 468.0 1,974.0 487.1 2,461.1 39 1964 853.0 768.8 1,621.8 520.2 2,142.0 521.3 2,663.3 39 1965 901.4 814.3 1,715.7 567.1 2,282.8 555.5 2,838.3 40 1966 882.5 776.4 1,658.9 584.5 2,243.4 586.1 2,829.5 41 1967 979.9 827.7 1,807.6 686.6 2,494.2 622.5 3,116.7 42 1968 1,066.2 847.0 1,913.2 811.2 2,724.4 664.9 3,389.3 44 1969 1,038.9 799.7 1,883.1 969.3 2,852.4 674.8 3,527.2 47 , 1970 935.1 1,031.5 1,966.6 1,148.2 3,114.8 780.7 3,895.6 50 1971 115.3 1,033.8 2,149.1 1,235.8 3,434.9 992.5 4,357.4 51 1972 1,202.3 1,278.8 2,481.1 1,493.1 3,974.2 1,032.8 5,007.0 50 1973 1,406.7 1,341.0 2,777.7 1,376.2 4,153.9 1,241.4 5,395.3 52 1974 1,539.3 1,406.3 2,945.6 1,622.5 4,568.1 1,512.6 6,080.7 52 1975 1,609.8 1,478.4 3,088.1 1,668.9 4,757.0 1,769.7 6,526.7 53 1976 2,080.5 2,085.1 4,165.6 2,155.3 6,320.9 2,101.4 8,422.3 51 1977 2,791.7 2,574.2 5,365.8 3,441.5 8,807.3 2,959.5 11,402.8 53 1978 3,015.5 2,920.8 5,936.4 5,025.5 10,961.9 3,165.9 14,127.8 58 1979 3,774.2 3,895.1 7,669.3 7,388.3 15,057.6 4,286.6 19,344.2 60 1980 4,180.8 5,247.3 9,428.2 10,432.0 19,860.2 4,953.1 24,813.3 62 1981 4,822.9 5,201.5 10,024.4 14,422.4 24,446.8 5,439.3 29,886.1 66 1982 5,987.7 5,772.1 11,759.8 18,750.1 30,509.0 7,673.4 38,183.3 69 1983 7,200.1 7,547.8 14,747.9 22,509.0 37,256.9 9,472.3 46,729.2 68 1984 8,560.9 8,262.9 16,823.8 26,603.6 43,427.4 11,642.5 55,069.9 69 1985 9,815.5 8,945.5 18,761.0 29,647.9 48,408.9 13,267.6 61,676.5 70 Source: Central Bank of Ceylon, Annual Report (annual). Notes Chapter Notes 1. For a concise history of Sri Lanka, see de Silva (1981). 2. The standard reference on the plantation industries is Snodgrass (1966). 3. For more details on the FEECS and related trade policy reform measures, see Cuthbertson and Athukorala (1990). 4. The history of the Mahaweli project goes back to the early 1950s. It is said that the project was initially conceived by Don Steven Senanayaka as a means of rejuvenating the North-Central province, the heartland of the ancient irrigation-based civilization of Sri Lanka. The feasibility of the project was first considered by a joint Sri Lanka/United States mission in 1958 and ex- plored further by the World Bank in 1961 (Levy 1985). A master plan, which detailed a three- phase, thirty-year program of implementation was prepared by a joint UNDP/FAO team between 1965 and 1968. This program was targeted at irrigating 900,000 acres of land for over half a mil- lion people and also at developing fifteen multipurpose projects, four transbasin diversion canals, and several power stations with a total capacity of 500 megawatts. The cost of the project (at 1968 market prices) was estimated at 6,700 million rupees (Mendis 1973). 5. These policy reforms have been the subject of substantial research. See Athukorala (1986), Ra- japatirana (1988), Lal and Rajapatirana (1989), Athukorala and Rajapatirana (1990), and Cuth- bertson and Athukorala (1990). 6. Relations between Sri Lanka and India had already been strained by Sri Lanka's alleged decision to provide limited naval facilities for the United States at Trincomalee, a large natural harbor on the east coast, the establishment of close links with China and Pakistan, and the permission grant- ed to Israel to establish a quasi-diplomatic presence in Colombo. 7. Data used in this paragraph are from Central Bank of Ceylon Annual Report-1990. 8. This increase is even sharper (from 42-48 percent to 65 percent) when shares are estimated using current price data. 9. For more detailed discussions of govemment intervention in the rice market, see Edirisinghe (1975), Gavan and Chandrasekara (1979), and Gunawardana (1989). 10. For details on this topic, see Gunasekara (1962). Karunatilake (1986), and Hettiarachchi (1986). 11. By statute, the NSB is required to invest at least 60 percent of its deposits in govemment securities. Since its inception it has always surpassed this norm. usually investing about 95-97 percent of those deposits. Under the EPF charter, the Monetary Board of the Central Bank decides which in- vestment practices can be used by the EPF. In reality, over 95 percent of the accumulated funds are invested in government securities. 12. Khatkhate (1982) provides an anatomy of financial retardation in the Sri Lankan economy during this period. 13. The nature and reliability of these indices are discussed in appendix 11. 14. For a useful account of the growth of the tourist industry, see Central Bank of Ceylon Annual Report- 1979. 158 Notes 159 15. Given the high degree of import-intensity of garments and other newly emerging manufactured exports, the use of export shares estimated in gross terms for analysis of the changing export pat- tem exaggerates the reduction in the importance of primary exports in the export structure and overall export growth. An analysis of the growth and structure of exports based on net export se- ries (derived using input-output tables) shows that the combined share of traditional agricultural exports in total net export eamings declined between 1977 and 1985 by only 14 percent (94 per- cent to 81 percent) as against a 33 percent decline (86 percent to 59 percent) observed in gross terms. The average annual growth rates of exports during 1977-85 in gross and net terms were II percent and 7 percent, respectively (Athukorala and Bandara 1989). 16. Imported rice and wheat flour contributed to more than half the domestic cereal supply in these years (Kappagoda and Paine 1981). 17. Altemative explanations of this savings behavior include Ricardian equivalence (private agents respond to govemment austerity with lower savings in the expectation of lower taxes in the fu- ture) or the simple Keynesian response (lower savings in response to lower current income). 18. For a detailed account of the PSin, see Karunatilake (1987). 19. The World Bank did not finance any of the big dams, but it financed the construction of the orig- inal diversion barrage (a barrage is a type of dam) across the Mahaweli River. It also financed the power distribution networks and the feeder canals and civil works needed to make use of the out- put of the dams. 20. When the terms of trade change is taken into account, real growth in 1979 was only 3.1 percent. 21. The rate at which the Central Bank grants advances to commercial banks for temporary liquidity purposes. 22. The real exchange rate. in the presence of factor specificity and labor market distortions, can be- have in a manner that does not conform to the predictions of simpler models, and these results should be interpreted with care (Edwards 1988; Hazari, Jayasuriya, and Sgro, forthcoming). 23. Lal (1985) invokes the composite good theorem to justify his aggregation of the importables and exportables into a single composite good, tradables. Such an aggregation is valid only if the rel- ative prices of the goods being aggregated remain constant. As Lal's analytical model focussed on the price and real exchange rate adjustments to capital flows. it was a useful simplification. However, its use in empirical analysis was not justified given the large change in the relative price of importables and exportables expressed in the terms of trade change. 24. The TroT was strongly (negatively) correlated with net capital inflows as well as government ex- penditure, which was another explanatory variable. 25. Comparisons of export volumes in this period with previous periods should be made with con- siderable caution because of the substantial changes in composition due, in particular, to the growth of manufactured exports. 26. The latter function could not be extended for the post-1982 period because the Central Bank ceased to publish a separate private investment series in 1983. 27. Quoted in Karunatilake (1987:405). 28. Analysis of the agriculture components of the projects indicates very low retums. For example, the cost of production of rice in the project areas has been well above world prices (Ekanayake 1987; World Bank 1986). A comprehensive evaluation of the Mahaweli scheme remains to be done. as available studies have focussed on the irrigation and power components separately. This has led to underestimation of some of the major infrastructural costs, since each assessment has imputed different (and higher) proportions of total costs to the other component. These suggest that, even if considerably higher values are placed on the outputs. the rates of return obtained in the available studies-low as they are-should be considered as being on the high side. 29. While most important agricultural products have been discriminated against by govemment pol- icies, the rice sector has received a significantly high level of protection. Effective protection for rice has been estimated to be around 15 percent (Edirisinghe 1991 b). 30. Some estimates put the extent of this damage closer to two-thirds of GDP (World Bank 1991). 31. For a formal demonstration of this argument. see Lipton and Sachs (1990). 160 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 32. In the context of the intertemporal tradeoff involved, recall that a significant pan of the budget deficit was financed by an implicit tax (via low interest rates) on the savings of wage earners held in the Employees' Provident Fund. Appendix I Notes I. On the basis of his experience from attempting to estimate his growth equation using time-series data for the United Kingdom and the United States, Scott notes that cyclical influences tend to reduce the degree of the overall fit of estimates based on arnual observation much more than those based on period averages. 2. For useful surveys of this literature, see Mikesell and Zinser (1973); Fry (1988). 3. Of the explanatory variables discussed where, FIR and FS are not included in the financial saving function for obvious reasons. 4. The Central Bank data series on private investment stops in 1983. 5. Sophisticated econometric analysis of supply response of rubber and tea which addresses some of the complexities of modelling perennial crop supply responses is reported by Hartey, Nerlove, and Peters (1987) and Akiyama and Trivedi (1987). The focus of their analyses are different from ours and, due to time and data constraints, no attempt was made to adapt their approaches in our study. But their supply price elasticity estimates confirm our general conclusions. Further, it should be noted that they obtain quite low long-run elasticities; this may be partly due to the acute uncertainties that faced tea and rubber producers for a very long time. Appendix II Notes I . For further details on the socioeconomic data structure of Sri Lanka, see Karunatilake (1980 and 1984). Shourie (1974), Pyatt and Row (1977), and Hallett (1981). 2. See Savundranayagam (1983), and Central Bank of Ceylon, Review of the Economy (1978:22- 24, and 1983:11-12) for details. 3. See Shourie (1974) for a revealing discussion on the deflator problem of Sri Lankan national accounts. 4. For a detailed discussion of these indices, see Khatkhate (1978). Bibliography The word "processed" describes informally reproduced works that may not be commonly available through libraries. Akiyama, Takamasa, and Pravin K. Trivedi. 1987. "A New Global Tea Model: Specification, Estimation, and Simulation." World Bank Staff Commodity Working Papers. Washington, D.C. Alailima, Pat. 1985. "Evolution of Government Policies and Expenditure on So- cial Welfare in Sri Lanka during the 20th Century." Ministry of Finance and Planning, Colombo. Processed. Amarasinghe, Y. R. 1979. "Emergence and Growth of Public Enterprises in Sri Lanka." In W. D. Lakshman, ed., Public Enterprises in the Economic Develop- ment of Sri Lanka. Colombo: National Institute of Business Management. Amerasinghe, C. F. 1964. "The Ceylon Oil Expropriations." American Journal of International Law 58(3):445-50. Anand, Sudhir, and S. M. Ravi Kanbur. 1991. "Public Policy and Basic Needs Pro- visions: Intervention and Achievement in Sri Lanka." In Jean Dreze and Amartya Sen, eds., The Political Economy of Hunger. Vol. 3. Oxford: Clarendon Press. Arasaratnam, S. 1964. Ceylon. Englewood Cliffs, NJ: Prentice-Hall. Athukorala, Premachandra. 1981. "Import Substitution, Structural Transforma- tion and Trade Dependence: A Case Study of Sri Lanka." The Developing Economies 24(2): 119-42. . 1984. "Direct Foreign Investment and Manufactured Export Expansion: The Case of Sri Lanka." Vidyodaya Journal of Arts Science and Letters 12:21-41. 1986. "The Impact of 1977 Policy Reforms on Domestic Industry." Up- anathi I (1). . 1989. "Export-oriented Direct Foreign Investment in New Exporting Countries: Patterns and Determinants with Evidence from Sri Lanka." World Competition 13(1):61-81. . 1990. "International Contract Migration and Reintegration of Return Mi- grants: The Experience of Sri Lanka." International Migration Review 24(1):323-46. . and Jayatileke S. Bandara. 1989. "Growth of Manufactured Exports, Pri- mary Commodity Dependence, and Net Export Earnings: Sri Lanka." World Development 17(8):897-903. 161 162 Macroeconomic Polic ies, Crises, and Growth in Sri Lanka, 1969-90 and Frank C. H. Huynh. 1987. Export Instability and Growth: Problems and Prospects for the Developing Countries. London: Croom Helm. , and Sisira K. Jayasuriya. 1988. "Parentage and Factor Proportion: A Comparative Study of Third-World Multinationals in Sri Lankan Manufactur- ing." Offord Bulletin of Economics and Statistics 50(4):409-24. and Sarath Rajapatirana. 1990. "The Domestic Financial Market and the Trade Liberalization Outcome: The Evidence from Sri Lanka." World Bank Policy, Research and External Affairs Working Papers WPS590. World Bank, Washington, D.C. Balassa, Bela, and Desmond McCarthy. 1984. "Adjustment Policies in Develop- ing Countries." World Bank, Washington, D.C. Processed. Bandara, Jayatileke S. 1989. "A Multisectoral General Equilibrium Model of the Sri Lankan Economy with Applications to the Analysis of the Effects of Exter- nal Shocks." Unpublished Ph.D. dissertation, LaTrobe University, Bundoora, Australia. Bandara, Jayatileke S., and Sisira K. Jayasuriya. 1991. "Capital Flows, Terms of Trade Shocks, and Economic Adjustment: Sri Lanka 1978-82." LaTrobe Uni- versity, Bundoora, Australia. Processed. Bandaranayake, R. Dias. 1984. "Tea Production in Sri Lanka: Future Outlook and Mechanisms for Enhancing Sectoral Performance." Occasional Paper Number 7. Central Bank, Colombo. Barlow, C., and Sisira K. Jayasuriya. 1987. "Structural Change and Its Impact on Traditional Agricultural Sectors of Developing Countries." Agricultural Eco- nomics 1(2): 159-74. Barro, Robert J. 1991. "Economic Growth in a Cross Section of Countries." Quar- terly Journal of Economics 95(May):407-43. Bhagwati, Jagdish. 1978. Anatomy and Consequences of Exchange Control Re- gimes. Cambridge, Mass.: Ballinger Press. Bhalla, Surjit S. 1987. "Political Economy of Agricultural Policies: A Case Study of Sri Lanka." World Bank, Washington, D.C. Processed. , and Paul Glewwe. 1986. "Growth and Equity in Developing Countries: A Reinterpretation of the Sri Lankan Experience." World Bank Economic Re- view 1(l):35-64. Blejer, Mario and Moshin Khan. 1984. "Private Investment in Developing Coun- tries." Finance and Development 21(2):26-29. Central Bank of Ceylon. Annual Report (annual). Colombo. . Review of the Economy (annual until 1988). Colombo. Ceylon Brokers' Association. Rupee Company Yearbook, various years. Colombo. Corden, W. Max. 1984. "Booming Sector and Dutch Disease: Survey and Consol- idation." Oxford Economic Papers 36(3):359-80. . 1987. "Protection and Liberalization: A Review of Analytical Issues." IMF Occasional Paper 54. Washington, D.C. . 1989. "Exchange Rate Policy in Developing Countries." Policy Research and External Affairs Working Paper 412. World Bank, Washington, D.C. Bibliography 163 1991. "Macroeconomic Policies and Growth: Some Lessons of Experi- ence." In Proceedings of the World Bank Annual Conference On Development Economics 1990. World Bank, Washington, D.C. Corea, Gamani. 1965. "Ceylon." In Cranky Onslow, ed., Asian Economic Devel- opment. London: Weidenfeld and Nicolson. . 1971 a. "Aid and the Economy." Marga Quarterly Journal I(1): 19-54. . 197 lb. "Ceylon in the Sixties." Marga Quarterly Journal 1(2):1-29. 1973. "Economic Planning, the Green Revolution and the Food Drive in Ceylon." In W.L. David, ed., Public Finance, Planning and Economic Devel- opment: Essays in Honour of Ursula Hicks. London: Macmillan. Cuthbertson, Andrew G., and Premachandra Athukorala. 1990. "Sri Lanka." In D. Papageorgiou, M. Michaely, and A. M. Choksi, eds., Liberalizing Foreign Trade: Indonesia. Pakistan and Sri Lanka. Oxford: Basil Blackwell. , and M. Z. Khan. 1981. "Effective Protection to Manufacturing Industry in Sri Lanka." Ministry of Finance and Planning, Colombo. Processed. Dahanayaka, P. A. S. 1977. Economic Policies and their Implications for the For- eign Exchange Resource Availability in Sri Lanka 1956-72. Central Bank, Colombo. Das Gupta, B. B. 1949. A Short Economic Survey of Ceylon. Associated News Pa- pers of Ceylon Ltd., Colombo. Denison, Edward F. 1967. Why Growth Rates Differ: Postwar Experience in Nine Western Countries. Washington, D.C.: The Brookings Institution. De Silva, K. M. 1981. A History of Sri Lanka. Delhi: Oxford University Press, Delhi. De Silva, H. W. Sumith. 1979. "The Coconut Industry In Sri Lanka: An Analysis of Government Intervention Measures." Unpublished Masters' Thesis, Aus- tralian National University, Canberra. De Silva, S. B. D. 1982. The Political Economy of Underdevelopment. London: Routledge and Kegan Paul. Divisekara, Sarath, and B. S. Felmingham. 1986. "Economic Growth, Ethnicity and Personal Income Distribution in Sri Lanka." Economics Discussion Paper 86-5, University of Tasmania, Hobart, Australia. Duesenberry, James S., John Sheahan, and Susan Hickok. 1981. "Stabilization Policy for Sri Lanka." Central Bank, Colombo. Processed. Dutta, Amita. 1973. International Migration, Trade and Real Income: A Case Study of Ceylon, 1920-1938. Calcutta: World Press. Economic and Social Commission for Asia and the Pacific. 1983. "Sri Lanka." In The Integration of Tax Policy into Development Planning in the ESCAP Re- gion. Development Papers No. 3. Bangkok: Economic and Social Commission for Asia and the Pacific (ESCAP). Edirisinghe, Neville. 1975. "Implications of Government Intervention in the Rice Economy of Sri Lanka." Unpublished Masters' Thesis, Cornell University, Ith- aca, NY. Processed. 164 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 . 1991 a. "Poverty in Sri Lanka: Its Extent, Distribution and Characteristics of the Poor." Paper submitted to the World Bank for the Sri Lanka Poverty Al- leviation Project. Processed. . 1991b. "Comparative Advantage and Policy Incentives in Rice Produc- tion in Sri Lanka." Paper prepared for the joint International Food Policy Re- search Institute Agrarian Research Training Institute project, Washington, D.C. Processed. , and T. T. Poleman. 1977. "Rice Economy of Sri Lanka: Consumption Characteristics and Production Trends." Marga Quarterly Journal 4(3): 1-59. Edwards, Sebastian. 1988. "Terms of Trade, Tariffs, and Labor Market Adjust- ment in Developing Countries." World Bank Economic Review 2(2): 165-85. , and Guido Tabellini. 1991. "The Political Economy of Fiscal Policy and Inflation in Developing Countries: An Empirical Analysis." World Bank PPR Working Papers No. 703, Washington, D.C. Ekanayake, S. A. B. 1987. "Micro-Level Planning Economic Performance and Hu- man Capital: A Study of the Settler Farmers in the Mahaweli Project in Sri Lan- ka." Unpublished Ph.D. dissertation, Australian National University, Canberra. Farmer, B. H. 1957. Pioneer Peasant Colonization in Ceylon. London: Oxford University Press. . 1967. "Nine Years of Political and Economic Changes in Ceylon." World Today 1965:189-98. Fernando, L. E. N. 1972. "Multinational Enterprise in Manufacturing Industry with Special Reference to Ceylon." Staff Studies 2(2):25-48. Central Bank, Colombo. Fernando, M. R., and S. Colombage. 1976. "The Performance of the Convertible Rupee Account Scheme." Staff Studies 6(l):47-77. Central Bank, Colombo. Femando, W. Nimal. 1978. "Land Reform in Plantation Agriculture: An Analysis of the Case of Sri Lanka with Special Reference to Tea Plantations." Research Paper of the Land Tenure Center, University of Wisconsin, Madison. . 1980a. "Continuity and Change in Plantation Agriculture: A Study of Sri Lanka's Land Reform Program on Tea Plantations." Unpublished Ph.D. disser- tation, University of Wisconsin, Madison. . 1980b. "A Study of the Impact of Land Reform Laws of 1972 and 1975 on Tea Production in Sri Lanka." Staff Studies 10(l):54-85. Central Bank, Colombo. Fields, Gary S. 1986. "Public Policy and the Evolution of the Labor Market in Sri Lanka." World Bank, Washington, D.C. Processed. Fitter, Jomn. 1973. "Ceylon on the Road to Long-term Decline." Intereconomics 1:25-27. Fry, Maxwell. 1988. Money, Interest, and Banking in Economic Development. Baltimore, Md.: Johns Hopkins University Press. Gavan, J. D., and 1. S. Chandrasekara. 1979. "The Impact on Public Foodgrain Distribution on Food Consumption and Welfare in Sri Lanka." Research Re- port 13, IFPRI, Washington, D.C. Bibliography 165 Glewwe, Paul. 1986. "The Distribution of income in Sri Lanka in 1969-70 and 1980-81: A Decomposition Analysis." Journal of Development Economics 24:255-74. . 1988a. "Economic Liberalization and Income Inequality: Further Evi- dence on the Sri Lankan Experience." Journal of Development Economics 28:233-46. . 1988b. "Response to Ravallion and Jayasuriya." Journal of Development Economics 28:257-60. Gooneratna, Wilbert and Dewpriya Esumperuma. 1984. Plantation Agriculture in Sri Lanka. International Labor Organization, Bangkok. Government of Sri Lanka. 1971. Ministry of Finance and Planning. The Economy of Ceylon Trends and Prospects. Colombo. . 1974. Report of the Commission of Inquiry on Agency Houses and Brok- ering Firms. Sessional Paper No. XII of 1974, Colombo. . 1985. Ministry of Plan Implementation. Foreign Employment: Sri Lanka Experience. Colombo. . Annual. Statistical Abstract. Department of Census and Statistics, Colombo. -. Annual. Statistical Pocket Book. Department of Census and Statistics, Colombo. . 1982. Population, Sri Lanka. Department of Census and Statistics, Colombo. . 1983. Socio-Economic Indicators of Sri Lanka. Department of Census and Statistics, Colombo. . 1972. Five-Year Plan. Ministry of Planning and Employment, Colombo. Grossman, Gene, and Elahan Helpman. 1989. "Growth and Welfare in a Small Open Economy." National Bureau of Economic Research Working Paper No. 2970, Cambridge, MA. Gunasekara, H. A. De S. 1962. From Dependent Currency to Central Banking in Ceylon. London: G. Bell & Sons. 1. 963. "Ceylon." In W. F. Crick, ed., Commonwealth Banking System. Oxford: Clarendon Press. Gunawardana, P. J. 1989. "Rice Pricing and Marketing Policies in Sri Lanka, 1952-84: A Welfare Evaluation." Unpublished Ph.D. dissertation, LaTrobe University, Bundoora, Australia. Hallett, A. J. Hughes. 1981. "Data Analysis as a Sufficient Condition for Econom- ic Planning: The Case of Sri Lanka's Development Plans." Review of Public Data Use 19(l):283-300. Hamberg, Daniel. 1971. Models of Economic Growth. New York: Harper and Row. Hartley, M. J., Marc Nerlove, and R. Kyle Peters, Jr. 1987. "An Analysis of Rub- ber Supply in Sri Lanka." American Journal of Agricultural Economics 69(4):755-61. Hazari B. R., S. K. Jayasuriya, and P. Sgro. "Tariffs, Terms of Trade, Unemploy- ment. and the Real Exchange Rate." Southern Economic Journal 58(3): 721- 73. 166 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Hettiarachchi, Wimal. 1978. "Some Reflections on the August-September 77 In- terest Rate Reform in Sri Lanka." Staff Studies 7(2):71-91. Central Bank, Colombo. . 1986. "Growth and Expansion of Banking." In W. Rasaputra, W. M. Tilakaratna, S. T. G. Femando and L. E. N. Fernando, eds., Facets of Develop- ment in Independent Sri Lanka. Ministry of Finance and Planning, Colombo. Hewavitharana, Budhadasa. 1975. "Recent Trends in the Management of External and Internal Finance in Sri Lanka." Marga Quarterley Journal 2(4):1-37. . 1980. "New Patterns of Developments for Sri Lanka." Economic Bulletin for Asia and the Pacific 31(1):20-44. International Labor Organization (ILO). 1971. Matching Employment Opportuni- ties and Expectations. A Program of Action for Ceylon. International Labor Office, Geneva. Intemational Monetary Fund (LMF). Annual. Directions of Trade. Washington, D.C. Annual. International Financial Statistics. Washington, D.C. Annual. International Financial Statistics Yearbook. Washington, D.C. Iriyagolle, Gamini. 1978. The Truth of Mahaweli. Published in Colombo by the author. Isenman, Paul. 1980. "Basic Needs: The Case of Sri Lanka." World Development 8:237-58. Jayamaha, Rani, and P. R. D. Wilson. 1989. "The Transmission of Monetary Pol- icy in Sri Lanka, 1977 to 1985: A Dynamic Simulation Approach." Interna- tional Journal of Development Banking 7(2):3-48. Jayasena, W. G., and H. M. Gamini Herath. 1982. Socio-Economic Conditions of Rubber Smallholders in Sri Lanka. ARTI, Colombo. Jayasuriya, Sisira K. 1976. The Dynamic Replacement Problem in the Rubber In- dustry of Sri Lanka. Canberra: Australian National University Press. , and Colin Barlow. 1986. "Effects of Industrial Protection and Exchange Rate Policies on a Traditional Export Crop: A Study of the Sri Lankan Rubber Industry, 1950-1980." Journal of Natural Rubber Research 1(3): 145-54. , C. Barlow, and R. T. Shand. 1981. "Farmers' Long Term Investment De- cision: A Case Study of Rubber Smallholders in Sri Lanka." Journal of Devel- opment Studies 18(l):47-67. Jayatissa, R. A. 1982. "Balance of Payments Adjustment to Exogenous Shocks During 1970-81: The Case of Sri Lanka." Staff Studies 12(1):41-92. Central Bank, Colombo. Jayawardena, Lal. 1970. "Recent Approaches to Planning in Ceylon." In E. A. G. Robinson and M. Kidrow, eds., Economic Development in South Asia. New York: St. Martin's Press. , A. Maasland, and P. N. Radhakrishnan. 1987. Sri Lanka. Country Study 15, World Institute of Development Economics Research (WIDER) Series on Stabilization and Adjustment Policies and Programs, Helsinki. Jayawardena, N. U. 1982. "Exports of Blue Sapphires." Economic Review 8(5): 17-18. Bibliography 167 1985. "Exchange Rate: Inevitability of Devaluation of the Sri Lankan Ru- pee." Paper presented at the seminar on The Role of Banks in Export Develop- ment organized by the Federation of Chambers of Commerce and Industry of Sri Lanka, March 2, 1985, Colombo. Processed. Jayawardena, V. K. 1972. The Rise of the Labour Movement in Ceylon. Durham, NC: Duke University Press. Jayawardene, J. R. 1979. A New Path. Ministry of Information and Broadcasting, Colombo. Jennings, W. Ivor. 1950. The Economy of Ceylon. Oxford: Oxford University Press. Jiggins, Janice. 1976. "Dismantling Welfarism in Sri Lanka." Overseas Develop- ment Institute Review 2. Jones, Gavin W., and S. Selvaratnam. 1971. "Some Problems of Employment Creation in Ceylon." Marga Quarterly Journal 1(l):38-54. .1972. Population Growth and Economic Development in Ceylon. Colom- bo: Marga Institute. Kappagoda, Nihal, M. 1967. "Foreign Exchange Budgeting-Ceylon's Experi- ence." Monthly Bulletin of the Central Bank of Ceylon. September. , and Susan Paine. 1981. The Balance of Payments Adjustment Process: The Experience of Sri Lanka. Colombo: Marga Institute. Karunatilake, H. Nevil S. 1974. Central Banking and Monetary Policy in Sri Lanka. Colombo: Lake House Investments Ltd. . 1980. "National Primary Socio-Economic Data Structure, Sri Lanka." International Journal of Social Science 32(3):536-53. . 1984. "Main Socio-Economic Statistical Series of South Asian Coun- tries: A Critical Survey." International Journal of Social Science 36(2):369- 402. -. 1986. The Banking and Financial System of Sri Lanka. Center for Demo- graphic and Socio-Economic Studies, Colombo. 1987. The Economy of Sri Lanka. Center for Demographic and Socio- Economic Studies, Colombo. Khatkhate, Deena M. 1980. "Anatomy of Financial Retardation in a Less-Devel- oped Country: The Case of Sri Lanka 1951-1976." World Development 10(9):829-40. -, W. L. Coats, and K. W. Riechel. 1978. "The Sri Lanka Financial System and A Framework for Monetary Policy." Washington, D.C.: International Monetary Fund. Processed. King, Robert G., and Sergio Rebelo. 1990. "Public Policy and Economic Growth: Developing Neoclassical Implications." National Bureau of Economic Re- search Working Paper No. 3338, Washington, D.C. Korale, R. B. M. 1987. Employment and the Labour Market in Sri Lanka: A Re- view. Department of Census and Statistics, Ministry of Plan Implementation, Colombo. 168 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 Kravis, Irving B., Alan Hetson and Robert Summers. 1982. World Production and Income: International Comparisons of Real Gross Product. Baltimore, MD: Johns Hopkins University Press. Krueger, Anne. 1978. Liberalization Attempts and Consequences. Cambridge, Mass.: Ballinger. Lakshman, Weligamage D. 1973. "Terms of Trade, Public Policy and Economic Development of Ceylon, 1948-1958." Ph.D. dissertation, Oxford University. . ed., 1979. Public Enterprises in the Economic Development of Sri Lanka. National Institute of Business Management, Colombo. . 1981. "Some Thoughts on the Political Economy of the Growth of Public Enterprises in Sri Lanka." Sri Lanka Journal of Management Studies 2(1):67-90. . 1986. "State Policy in Sri Lanka and Its Economic Impact." Upanathi l(1):5-36. . and Premachandra Athukorala. 1985a. "Analysis of Trading Channels for the Foreign Trade of Sri Lanka." In Transnational Trading Corporations in Se- lected Asian and Pacific Countries. Bangkok: United Nations. and Premmchandra Athulcorala. 1985b. "Transnational Trading Corpora- tions and the Structure of Trading Channels in Sri Lankan Foreign Trade." Bangkok: ESCAP/UNCTC. Processed. . and N. A. Obadage. 1986. Public Enterprises in Sri Lanka. Manila: Asian Development Bank. Processed. Lal, Deepak. 1985. "The Real Exchange Rate, Capital Inflows and Inflation: Sri Lanka 1970-1982." Weltwirtschaftliches Archiv 121(4):682-701. . and Sarath Rajapatirana. 1989. "Impediments to Trade Liberalization in Sri Lanka." Thames Essays No. 51. Aldershot: Gower. Lall, Sanjaya, and Senake Bibile. 1977. "The Political Economy of Controlling Transnationals: The Pharmaceutical Industry in Sri Lanka (1972-76)." World Development 5(8):677-97. Levy, Brian. 1985. "Economic Aid in the Making of Economic Policy in Sri Lanka." World Bank, Washington, D.C. Processed. Lipton, David, and Jeffrey Sachs. 1990. "Creating a Market Economy in Eastern Europe: The Case of Poland." Brookings Papers on Economic Activity 1990(l):75-133. Lucas, Robert. 1988. "On the Mechanics of Economic Development." Journal of Monetary Economics 22 (July):3-42. Manor, James. 1984. Sri Lanka in Change and Crisis. London: Croom Helm. Marga Institute. 1974. Welfare and Growth in Sri Lanka. Marga Research Studies No. 2, Colombo. . 1978. "Technology Transfer and Reverse Flows." Special Issue of Marga Quarterly Journal 5(2). McDonald, Donogh C. 1985. "Trade Data Discrepancies and the Incentive to Smuggle." IMF Staff Papers 32(4):668-92. Meade, J. E. 1961. A Neo Classical Theory of Economic Growth. London: George Allen and Unwin. Bibliography 169 Mendis, M. W. J. G. 1973. The Planning Implications of the Mahaweli Develop- ment Project in Sri Lanka. Colombo: Lake House Investments Ltd. Mikesell, Raymond and James Zinser. 1973. "The Nature of the Saving Function in Developing Countries: A Survey of the Theoretical and Empirical Litera- ture. " Journal of Economic Literature 11 (1): 1-26. Moore, Mick. 1990. "Economic Liberalism Versus Political Pluralism in Sri Lanka?" Modern Asian Studies 24(2):341-383. Motha, J. G. Guy. 1971. "The Impact of Import Policies on the Economy of Cey- lon." Unpublished Ph.D. dissertation, Stanford University. Stanford. Calif. Myint, Hla. 1985. "Growth Policies and Income Distribution." Development Pol- icy Issues Series-Discussion Paper VPERS3. World Bank, Washington, D.C. Newlyn, W.T. 1977. The Financing of Economic Development. Oxford: Claren- don Press. Oliver, H. M. 1956. "Industrialization of Ceylon's Opinions and Policies." Ceylon Economist 3(3): 178-225. -. 1957. Economic Opinion and Policy in Ceylon. Durham, NC: Duke Uni- versity Press. Olson, Richard S. 1977. "Expropriation and International Economic Coercion: Ceylon and the West 1961-65." Journal of Developing Areas 11 (2):205-26. Peiris, Gerald H. 1977. "Plantation Agriculture." In K. M. de Silva, ed., Sri Lanka: A Survey. London: C. Hurst and Co. . 1978. "Land Reform and Agrarian Change in Sri Lanka." Modern Asian Studies 12(4):611-28. Perera, N.M. 1974. Budget Speech for 1975. Colombo: Government Publications Bureau. Ponnambalam, Satchi. 1980. Dependent Capitalism in Crisis: The Sri Lankan Economy. London: Zed Press. Pyatt, F. Graham. and Allan Roe. 1977. SocialAccountingforDevelopment Plan- ning with Special Reference to Sri Lanka. Cambridge: Cambridge University Press. Rajapatirana, Sarath. 1988. "Foreign Trade and Economic Development: The Case of Sri Lanka." World Development 16(10):1 143-57. Ramachandran, N. 1963. Foreign Plantation Investment in Ceylon, 1889-1958. Central Bank of Ceylon, Colombo. Ramanayake, Dennis. 1984. "Sri Lanka: The Katunayake Investment Promotion Zone." In Eddy Lee, ed., Export Processing Zones and Industrial Employment in Asia. International Labor Organization/Asian Team for Employment Pro- motion, Bangkok. Ranaweera, T. W. Y. 1978. "Money Supply, Inflation and Balance of Payments in Sri Lanka 1959-74): A Comment." Journal of Development Studies 15(l): 106-11. Rasaputram, Wamasena. 1972. "Exchange Rate Experience and Policy in Sri Lan- ka since World War IU." In H. G. Grubel, and T. Morgan, eds., Exchange Rate Policy in Southeast Asia. Lexington, MA: Lexington Books. 170 Macroeconomic Policies, Crises, and Growth in Sri Lanka. 1969-90 Ratnayaka, Ravindra. 1988. "Trade Policy and the Performance of the Manufac- turing Sector: Sri Lanka." Developing Economies 26(1):12-33. Ravallion, Martin, and Sisira K. Jayasuriya. 1988. "Liberalization and Inequality in Sri Lanka: A Comment."Journal of Development Economics 28(2j:247-55. Rebelo, Sergio. 1991. "Long Run Analysis and Long Run Growth." Journal of Political Economy 99(3):500-21. Richards, L. J. 1971. Employment and Unemployment in Ceylon. Paris: Organiza- tion for Economic Cooperation and Development. Rodrigo, Chandra. 1983. A National Wage Policy for Sri Lanka. Sri Lanka Insti- tute of Development Administration, Colombo. .1988. "Employment, Wages and Income Distribution in Sri Lanka: 1960- 1985." ARTEP Working Paper, International Labor Organization, New Delhi. Roe, Alan. 1982. "High Interest Rates: A New Conventional Wisdom for Devel- opment Policy? Some Conclusions from Sri Lankan Experience." World De- velopment 10(3):211-22. Romer, Paul. 1986. "Increasing Returns and Long Run Growth." Journal of Po- litical Economy 94(Oct): 1002-37. Sahn, David E. 1985. "Food Consumption Patterns and Parameters in Sri Lanka: The Causes and Control of Malnutrition." International Food Policy Research Institute, Washington, D.C. Samarasinghe, S. W. R. de A. 1975. "Ceilings on Bank Credit: The Experience of Sri Lanka." The Ceylon Journal of Historical and Social Studies 142:1-15. Santiapillai, Victor E. 1981. "Export Expansion and Economic Development." Progress 1(3):1-5. Savundranayagam, Terance. 1983. "Estimates of Sri Lanka's Gross National Product from 1950 to 1981." Staff Studies 13(142). Central Bank, Colombo. Scott, Maurice F. 1989. A New View of Economic Growth. Oxford: Clarendon Press. Sen, Amartya. 1981. "Public Action and Quality of Life in Developing Coun- tries." Oxford Bulletin of Economics and Statistics. Sen, Amartya. 1982. "The Food Problem: Theory and Policy." Third World Quar- terly 4(3):447-59. Shourie, Aruin. 1974. "Sri Lanka's National Accounts-Notes for Regression En- thusiast." Marga Quarterly Journal 2(3):21-42. Singh, Manmoha. 1971. "Development Policy in Ceylon: 1966-1971." Economic Bulletin for Asia and the Far East 23:19-36. Sirisena, P.T. 1975. "An Evaluation of the Efficiency, Foreign Exchange Savings and the Welfare Impact of the Steel Industry." Staff Studies.' Central Bank of Ceylon 5(2): 15-32. Snodgrass, Donald R. 1966. Ceylon. An Export Economy in Transition. Home- wood, IL: Richard D. Irwin. . 1974. "Sri Lanka's Economic Development During Twenty-Five Years of Development." Ceylon Journal of Historical and Social Studies 4:119-25. Solow, R. M. 1956. "A Contribution to the Theory of Economic Growth." Quar- terly Journal of Economics 70(Feb):65-94. Bibliography 171 Staley, C. E. 1959. "Export Taxes in Ceylon, 1948-1952." Public Finance 314:249-68. Stem, Joseph J. 1984. "Liberalization in Sri Lanka: A Preliminary Assessment." Ministry of Finance and Planning, Colombo. Processed. Swan, T. W. 1956. "Economic Growth and Capital Accumulation." Economic Record 32(Nov):334-61. Tambiah, S. J. 1963. "Ceylon." In R. D. Lambert and B. F. Haselitz, eds., The Role of Savings and Wealth in Southern Asia and the West. Paris: UNESCO. Thorbecke, Erick, and J. Svejnar. 1984. "Effects of Macroeconomic Policies on Agriculture Performance in Sri Lanka 1960-92." Organization for Economic Cooperation and Development, Development Center. Processed. Tilakaratna, W. M. 1963. Agricultural Credit in a Development Country-Ceylon. Colombo: Central Bank of Ceylon. . 1970. "How the FEEC Scheme Came to be Introduced." Research and /ndustry. Colombo Industrial Development Board. Uzawa, H. 1961. "Neutral Inventions and the Stability of Growth Equilibrium." Review of Economic Studies 28(Feb): 117-24. Wanigatunga, R. C. 1987. "Direct Private Overseas Investment in Export-Orient- ed Ventures: Recent Developments in Sri Lanka." In V. Cable and B. Persaud, eds., Developing with Foreign Investment. London: Croom Helm. 1974. "Tyre Industry." Staff Studies: Central Bank of Ceylon. 4(3):33-52. Weisskopf, T.E. 1972. "The Impact of Foreign Capital Inflow on Domestic Sav- ings in Underdeveloped Countries." Journal of International Economics 2(1):25-38. Wickramasekara, Piyasiri. 1982. "Transnational Corporations and the Tea Export Industry of Sri Lanka." Bangkok: Economic and Social Commission for Asia and the Pacific. Processed. -. 1983. Expansion of Employment and Income through Local Resource Mobilization: A Study of Two Sri Lankan Villages. ARTEP-ILO, Bangkok. Wickramasinghe, Wimal. 1976. "Sri Lanka's Dilemma of Exchange Rate Man- agement: Origin and Evolution of the Sri Lanka Rupee." Colombo: Central Bank of Ceylon. Processed. - . 1985. A Theory of Multiple Exchange Rates and Exchange Rate Manage- ment in Sri Lanka. Colombo: Central Bank of Ceylon. Wickrameratne, L. Ananda. 1977. "Peasant Agriculture." In K. M. de Silva, ed., Sri Lanka: A Survey. London: C. Hurst and Co. Wijewardana, D. 1973. "The Problem of a Multiple Currency System in a Devel- oping Economy with Special Reference to Sri Lanka." Staff Studies 3(l):1-44. Central Bank of Ceylon, Colombo. Wilson, Jeyaratnam A. 1979. Politics in Sri Lanka: 1947-1979. London: MacMillan. . 1980. The Gaullist System in Asia. London: Macmillan. World Bank. 1953. The Economic Development of Ceylon. Baltimore: Johns Hop- kins University Press. 172 Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90 . 1977. "Sri Lanka: Country Economic Memorandum, Report No. 1425- CE." Washington, D.C. . 1980. "Sri Lanka: Key Development Issues in the 1980s." Report No. 3466 South Asia Regional Office, Washington, D.C. . 1981. "Sri Lanka: Policies and Prospects for Economic Adjustment." Washington, D.C. Processed. .1982. "Economic Adjustment in Sri Lanka: Issues and Prospects." Wash- ington, D.C. Processed. . 1984. "Sri Lanka: Recent Economic Developments, Prospects and Poli- cies, Report No. 5038-CE." Washington, D.C. Processed. . 1986. World Development Report 1986. New York: Oxford University Press. . 1988. "Sri Lanka-A Break with The Past: The 1987-90 Program of Economic Reforms and Adjustment, Report No. 7220-CE." Washington, D.C. Processed. . 1991. World Development Report 1991. New York: Oxford University Press. . 1989. "Recent Macroeconomic Developments and Adjustment Poli- cies-Sri Lanka." Report No. 8193. Washington, D.C. Processed. World Currency Yearbook. Various Years. International Currency Analysis Inc. Brooklyn, N.Y. Wriggins, W. H. 1960. Ceylon: Dilemmas of New Nations. Princeton: Princeton University Press. Index Accelerated Mahaweli Development Program(AMDP). See Mahaweli project agriculture (see also plantation crops, tea, rubber, coconut): diversification of, 15, 113; exports, 29-30; investment in, 10; labor force, 32-33, 59; low growth in, 56, 106; public sector investment in, 9, 15; sector, 27, 73-74; share of GDP, 30 Air Lanka, 86, 110 balance of payments crisis, 64-65 Bank of Ceylon, 37-38, 46 banking sector, 37-39; expansion of, 38; govemment involvement in, 46 banks: foreign, 39 bonus voucher system (for exports), 14 business tumover tax, 50 Business Undertakings Acquisition Act, 18, 46, 57 capital formation, 31 capital inflows, 76,82,92; analysis of 90-91; Dutch disease and, 105; govemment expenditure and, 89 Central bank, 8; founding of, 9; terms of trade shock and, 84 Ceylon Federation of Labor (CFL), 34 Ceylon Federation of Trade Unions (CFru), 34 Ceylon Mercantile Union (cMU), 34 coconut (see also plantation crops): industry: 36-37; export ban, 64; exports, 64, 84, 127; production, 82 collective bargaining, 34 Colombo consumer price index (ccpi), 44-46 commodity: price boom, 64 (see also Korean war commodity price boom); mar- kets, 35, 39 Communist party (CP): 6; trade unions and, 34 computable general equilibrium model (CGE), 91 consumer subsidies (see also food subsidies and welfare services), as impediment to growth, 16; trade unions and, 34 convertible rupee account (cRA),19, 101-02; abolition of, 87 counterpart funds, 89 173 174 Macroeconomic Policies, Crises. and Growth in Sri Lanka. 1969-90 credit controls, 69. 85; investment and, 98-99; on nonessential goods, 11 Currency Board System , 9 currency black market, 14, 70 debt service ratio. 54 defense expenditure, 18, 106 Democratic Workers Congress (DWC), 34 devaluation. See exchange rate policy Development Finance Corporation (DFCC), 38, 39 development planning, 10-Il domestic savings ratio, 31 Dutch disease, 89- 91, 105; capital inflows and, 105 economic nationalism, 114-15 Employee's Provident Fund (EPF), 38, 68, 73, 159n ethnic: (see also Tamil) composition of work force, 33; groups, 4-5, 25; poverty, decline of, 107-08 ethnic conflict, 6, 11, 23, 93; effect on growth, 28, 56, 71 Exchange Control Act, 19 exchange rate: controls, 14-15, 37; 69-70, 86-87, 114-15; fixed, 9; multiple, 14; policy, 53-54, 86; public investment and, 105-06, 110; restrictions, 7-8, 9; unification of, 21 export: boom, 7-8; bias against, 15; dependence, 29-30; diversification of, 15; ex- pansion of 16. 21-22; performance, 52; pessimism, 70; promotion, 50, 86- 87, 102, 124; taxes, 84 Export Development Board (EDB), 22, 86 Export Development Fund (EDF), 22 finance act, 21 Federal Party (FP), 13 food imports, 60, 64-65; effect on subsidies, 68 food subsidies, 7, 11, 20-21, 48; abolition of, 48; cuts in, 67, 68, 80, 85-86, 108, 110; govemment commitment to, 8, 12,62,103; replacement by food stamp scheme, 21 foreign currency banking units (FCBUs), 39 foreign aid: cut off of, 12; increase in 13, 20, 67, 79, 81; deterioration in financial discipline and 83 Foreign Exchange Entitlement Certificate Scheme (FEECS), 14-16, 50, 70 foreign savings: reliance on, 54 GDP growth, 56, 73 garment industry, 30, 47, 52, 88, 158n gem: exports, 74, 87, 101-02; industry, 19, 52, 127 Government-Owned Business Undertakings (GOBUs), 19 Index 175 government: budget deficit, 8, 10; commercial borrowing by, 23; commitment to food subsidies, 8, 12, 62, 103; control over foreign trade, 12; disengagement from manufacturing sector, 9; investment policies, 37; view of commodity booms, 114 Greater Colombo Economic Commission (GCEC), 22 gulf war, 118 Hartal, 86 housing sector program, 22, 79, 115-16 import compression, 60 import controls, 10- 13 import restrictions (see also QRs, tariffs), 9, 22, 54, 61-62, 67-68, 74, 86, 114-15; dismantling of, 7-8, 80; manufacturing sector and, 30 import-substitution, 10, 12, 52; agriculture and, 15; industries, 16; strategy, 61,99 income distribution, 106-11 incremental capital/output ratio (ICOR), 100, 104 India: PPp comparison, 28; relations with, 23, 158n industrialization scheme, 11 inflation, 23,44,45-46,70-71,72-73,84,93; attitudes toward,46; oil price shock and, 44; remittances and, 45; savings and, 97; trade unions and, 73 informal credit, 37, 44 Insurance Corporation of Ceylon (icc), 39 insurance sector. 38-39; government involvement in, 46 interest rate: government manipulation of, 21, 37, 69,98 International Monetary Fund (IMF), 13, 14, 66; drawings, 62, 67 investment: as percentage of GDP, 59; efficiency of, 94-96; foreign direct, 19; government policies for, 79-81, 104; private, 72; six-year program for, 10 investment capital: rationing of, 106 Investment Promotion Zones (IPZs), 22 Janatha Vimukthi Peramuna (JvP), 17-18, 23; armed conflict and, 24; destruction of, 20 joint stock companies, 9 Korean war commodity boom, 7. 44, 54 labor: force, 32-35; strife, 11, 13, 47 land reform, 18, 56 Land Reform Act of 1972, 18 Lanka Sama Samaja Party (LSsP), 6; trade unions and, 34 leftist opposition: 6-13, disappearance of, 13 long-term lending, 44 luxury goods ban, 11 176 Macroeconomic Policies. Crises, and Growth in Sri Lanka, I969-90 Mahaweli project, 15, 22,79-81, 115-16, 158n, 159n; cut backs in, 85; rate of re- turn on, 80,106; rice production and, 106; Tamil view of, 109; World Bank support of, 81, 159n manufactured exports, 74, 87 migration, 27 Multi-Fibre Textile Agreement, 92 National Credit Plan, 93 National Savings Bank (NSB), 38, 42, 68, 84, 158n nationalization: 10; of Agency houses, 36; of banks, 12, 37; of oil companies, 11- 12; of plantations, 18-19, 47 oil price shock, 54, 62-64, 82; and inflation, 44 open general licensing system, 15 outmigration, 27, 32, 109 Paddy Marketing Board (PMB), 18, 36, paddy purchasing scheme, 35 People's bank, 46 Pettah money market, 44 plantation crops (see also agriculture, coconut, tea, and rubber), 15, 29, 52; diver- sification of, 15; production decline in, 82 plantation sector, 30, 32, 37; labor force, 32-33, 59; origin of, 5-6 population structure, 26-27 poverty: alleviation of, 24; increase in 107-08; rural, 105 public sector enterprises, 18, 72, 86; as source of political patronage, 105, 110, 115 ; government capital transfers to, 98; government transfer to private sec- tor, 9; inefficiencies in, 99-100; investment in, 45, 47, 100; labor force, 47; share of GDP, 46 public sector investment program (PsiP), 54, 79, 87, 89, 104-05, 115 quantitative restrictions (QRs), 18, 21 remittances, 22,52, 76, 89, 92; current accounts balance and, 93; inflation and, 45 repatriation of capital, 19, 21 repatriation of profits: moratorium on, 12, 14 rice: production and the Mahaweli project, 106; self sufficiency, 15, 28, 109, 115; subsidy, 8, 11, 19, 34-35, 159n rubber industry (see also plantation crops), 7, 27, 127 savings rate, 72, 96, estimation of, 122; inflation and, 97; ratio, 89 services sector, 30 Scott, Maurice, 94, 119-20 share brokers association, 39 Index 177 Six-Year Development plan, 10 Sri Lanka Freedom Party (SLFP), 10, 66 social wage, 34; downward rigidity of, 114 State Industrial Corporation Act, 11 steel corporation, 100 stock exchange, 39-40 sugar ration, 19 Tamils, 23, 25. 33; effects of trade liberalization upon, 109; emigration of, 109; in public sector employment, 110; view of Mahaweli project, 109 Tamil United Liberation Front (TULF), 20 tariffs, 15, 87 taxation: as percentage of government revenue, 49; collection of, 50; of exports, 36, 50, 83-84; indirect, 49-50, 159n; welfare expenditures and, 113 tax holidays, 22 tea (see also plantation crops): boom 22, 44, 54, 64; exports, 82, 127; industry, 27, 52, 84 terms of trade: decline in, 62, 64-65, 89-91, 93 tourism, 51-52, 76, 89; ethnic conflict and, 93 trade dependence, 29-30, 59-61 trade dependence ratio (TDR), 51 trade restrictions: comparison with other countries, 117; effects on growth 95; re- moval of, 9 trade unions, 34; inflation and, 73; decline of, 88 Treasury bill market, 41 unemployment, 24, 33, 59, 79; decline in, 87-88; inflation and, 90 United Front (UF), 17, 70, 75 United Left Front (ULF), 13 United National Party (UNP), 6-7, 8,9, 21,75,76, 103; agricultural development and, 9; Mahaweli project and, 79; open economy program and, 103 wage employment, 32, 88 Wages boards, 34-35 wages: minimum, 35; increases, 13, 14 welfare services: government deficit and, 10, 47-48, 113; reduction in, 20, 115 welfare system: growth and 113-14; origin of, 7; taxation for, 113; use to control social conflict, 7, 117 World Bank: conditionality, 62; support for Mahaweli project, 81, 159n I SE6TT$ 968ZTOOOOOO: 11 111 1j1 11 1101 [ l 1 1~11 S31DI1Od ZIWONOH30JVW 11-LGZZ-ETZS 80 ASC L6£ . _U