Arming the South The Economics of Military Expenditure, Arms Production and Arms Trade in Developing Countries
Edited by
Jurgen Brauer and J. Paul Dunne
Arming the South
Also by furgen Brauer ECONOMIC ISSUES OF DISARMAMENT: Contributions from Peace Economics and Peace Science (with Manas Chatterji) PUBLIC ECONOMICS III: Public Choice, Political Economy, Peace & War. Economics Reading Lists, Course Outlines, Exams, Puzzles & Problems (with Ronald Friesen and Edward Tower) ECONOMICS OF CONFLICT AND PEACE (with William Gissy) THE ECONOMICS OF REGIONAL SECURITY: NATO, the Mediterranean, and Southern Africa (with Keith Hartley)
Also by J. Paul Dunne THE BRITISH ECONOMY AFTER OIL: Manufacturing or Services (with Terry Barker) QUANTITATIVE MARXISM STRUCTURAL CHANGE IN THE UK ECONOMY (with Ciaran Driver) THE PEACE DIVIDEND (with Nils Tetter Gleditsch, Adne Cappelen, Olav Bjerkholt and Ron Smith)
Arming the South The Economics of Military Expenditure, Arms Production and Arms Trade in Developing Countries Edited by
Jurgen Brauer College of Business Administration Augusta State University USA
and
J. Paul Dunne Middlesex University Business School London, UK
Editorial matter, Selection and Introduction © Jurgen Brauer and J. Paul Dunne 2002 Chapters 1-17 © Palgrave Publishers Ltd 2002 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T4LP. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2002 by PALGRAVE Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N.Y. 10010 Companies and representatives throughout the world PALGRAVE is the new global academic imprint of St. Martin's Press LLC Scholarly and Reference Division and Palgrave Publishers Ltd (formerly Macmillan Press Ltd). ISBN 0-333-75440-9 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Arming the South : the economics of military expenditure, arms production, and arms trade in developing countries / edited by Jurgen Brauer and J. Paul Dunne, p. cm. Includes bibliographical references and index. ISBN 0-333-75440-9 (cloth) 1. Developing countries—Armed Forces—Appropriations and expenditures. 2. Arms transfers—Economic aspects—Developing countries. 3. Defense industries—Developing countries. 4. Weapons industry—Developing countries. I. Brauer, Jurgen, 1957- II. Dunne, Paul. HC59.72.D4A75 2002 338.4'76234'091724—dc21 2001058216
10 9 8 7 6 5 4 3 2 1 11 10 09 08 07 06 05 04 03 02 Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire
With appreciation and affection to our activist friends in the worldwide peace movement
This page intentionally left blank
Contents
Figures and tables
xi
Contributors
xvi
Introduction Jurgen Brauer and J. Paul Dunne
1
PART I - MILITARY EXPENDITURE: OVERVIEWS
1
The Role of Demilitarization in Promoting Democracy and Prosperity in Africa Lloyd J. Dumas 2 Warlords and Logo Warriors: The Political Economy of Post-modern Conflict Neil Cooper 3 Military Expenditure and Development in Latin America Thomas Scheetz 4 Military Expenditure and Economic Development in Asia During the 1990s Geoff Harris
vn
15
35 51
71
PART II - ARMS PRODUCTION AND ARMS TRADE 5
6 7
The Arms Industry in Developing Nations: History and Post-Cold War Assessment 101 Jurgen Brauer Domestic Production as an Alternative to Importing Arms . . . . 129 Fotis Mouzakis Domestic Procurement, Subsidies, and the Arms Trade 161 Maria del Carmen Garcia-Alonso and Paul Levine
P A R T III - C O U N T R Y STUDIES
8 9
10
11
12
13
14
15
Saudi Arabia: Defense Offsets and Development Ron Matthews South Africa: An Econometric Analysis of Military Spending and Economic Growth Alvin Birdi and J. Paul Dunne Ludwig Erhard in Africa: War Finance and Post-War Reconstruction in Germany and Mozambique Tilman Briick Angola: Civil War and the Manufacturing Industry, 1975-1999 Manuel Ennes Ferreira Military Spending and Economic Development in Sub-Saharan Africa: A Supply-Side Analysis Oyinlola Olaniyi Greece: Military Expenditure, Economic Growth, and the Opportunity Cost of Defense Emmanuel Athanassiou, Christos Kollias, Eftychia Nikolaidou, and Stavros Zografakis A System Estimation of the Defense-Growth Relation in Turkey Jiilide Yddirim and Selami Sezgin The Military-Civilian Tradeoff in Guatemala: An Econometric Analysis Kanta Marwah, Lawrence R. Klein, and Thomas Scheetz vm
195
221
235
251
275
291
319
337
16 The Allocation of Resources to the Armed Forces in Chile: A Case of Limited Transparency Guillermo Pattillo
373
PART IV - THE PEACE MOVEMENT
17 Arms Sales and Development: The Role of the Peace Movement Tony Kempster
405
Index
417
IX
This page intentionally left blank
Figures and tables
Figure 3.1: Figure 4.1: Figure 6.1: Figure 6B.2: Figure 7.1: Figure 7B.2: Figure 7B.3: Figure 7B.4: Figure 7B.5: Figure 7B.6: Figure 7B.7: Figure 7B.8: Figure 7B.9: Figure 8.1: Figure 8.2: Figure 11.1:
Marginal cost and benefit of military expenditure . . . 61 Linkages in the military expenditure allocation process 84 Predicted and estimated demand curves for imported arms 142 Equilibrium when input/requires fixed costs 158 The arms market 173 Government purchases and exports against sellers' weight on consumption 188 Export price and subsidy against sellers' weight on consumption 188 Government purchases and exports against importers' variable benefit of defense 189 Export price and subsidy against importers' variable benefit of defense 189 Government purchases and exports against exporters' variable benefit of defense 190 Export price and subsidy against exporters' variable benefit of defense 190 Government purchases and exports against the weight of consumption 191 Export price and subsidy against the weight of consumption 191 Countertrade and offsets: a typology 199 The process of switch-trade 199 Angola: Index of manufacturing output by branch XI
Figure 11.2:
Figure 13.1: Figure 13.2: Figure 15.1: Figure 15.2: Figure 15.3:
Figure 15.4: Figure 15.5:
Figure 15.6: Figure 1 5.7: Figure 15.8: Figure 1 5.9: Figure 15.10: Figure 15.11:
(1975=100) and of military expenditures to government expenditures (1978=100), 1975-1991 257 Angola: Index of total manufacturing output (1975=100) and of military expenditures to government expenditures (1978=100), 1992-1998 . 264 Military expenditure as a share of GDP in Greece, NATO, and the EU 294 Military expenditure as a share of GDP and GDP growth rates 295 Guatemala: real military expenditure, 1969-1994 ( 4 000s of quetzales) 341 Ratio of real military expenditure to real GDP, 1969-1994 341 Average proportions of military expenditure (APMIL), consumption (APCON), gross fixed investment (APGFI), imports (APIM), exports (APEX), and government spending (APG) in GDP, 1969-1994 352 Actual and baseline simulated per capita consumption, 1984-1994 {quetzales per person) . . . 357 Actual and baseline simulated per capita national disposable income, 1984-1994 {quetzales per person) 357 Actual and baseline simulated GDP, 1984-1994 {quetzales per person) 357 Actual and baseline simulated gross fixed investment, 1984-1994 {quetzales per person) . . . . 357 Actual and baseline simulated imports, 1984-1994 (quetzales per person) 358 Actual and baseline simulated exports, 1984-1994 {quetzales per person) 358 Actual and baseline simulated price index, 1984-1994 (actual 1987=1.00) 358 Actual and baseline simulated exchange rate, 1984-1994 {quetzales per US$) 358 xn
Figure 15.12: Scenario 1: percentage effect, relative to baseline simulation, of yearly 30 million constant 1987 quetzales reduction in military expenditure on per capita GDP, price index, exchange rate, and per capita current account balance, 1984-1994 360 Figure 15.13: Baseline simulation and scenarios 1 and 2 effects on per capita consumption of annual 30 million constant quetzales reduction in military expenditures and annual 30 million constant quetzales reduction in arms imports, 1984-1994 . . . 361 Figure 15.14: Baseline simulation and scenarios 1 and 2 effects on non-military government expenditure, 1984-1994 (quetzales per person) 361 Figure 15.15: Baseline simulation and scenarios 1 and 2 effects on the price index, 1984-1994 361 Figure 15.16: Baseline simulation and scenarios 1 and 2 effects on the exchange rate (Q/US$), 1984-1994 361 Figure 16.1: Chile: the Ministry of Defense structure 378 Figure 16.2: Global fiscal income to the armed forces, 1989-1999 (as percentage of GDP) 381 Figure 16.3: Chile: the budget cycle 382 Figure 16.4: Chile: fiscal budget allocation to the armed forces (as percentage of GDP), 1989-1999 386 Figure 16.5: Chile: armed forces' income generated by the Copper Law (as percentage of GDP), 1989-1999 . . 389 Figure 16.6: Chile: The decision-making cycle of Copper Law funds 395
xm
Crowding-out of social and investment spending by military spending in Latin America, 1969-1994 . . . . 54 Table 3.2: The military burden in Latin America 55 Table 4.1: Asian military expenditure trends during the 1990s . 74 Table 5.1: Developing nations' arms producers/exporters, ca. 1985-1995 103 Table 5,2 The ladder of arms production 105 Table 6.1. Capital-labor ratio (arms imports per soldier) 133 Table 6.2: Estimated coefficients 145 Table 7.1 Model summary 171 Table 7A. 1: The model calibration 187 Table 8 1 Saudi Arabia: prime contracts and associated offset programs 204 Table 8.2: Saudi Arabian offset companies 208 Table 9.1: Aggregate estimation results, South Africa 227 Table 10.1: The nature of the wars in Germany and in Mozambique 241 Table 10.2: Economic indicators for Mozambique, 1987-1997 . 244 Table 11.1: Angola: Macroeconomic indicators, 1992-1999 . . . 262 Table 11A.1: Angola: Index of industrial output value, by branch and total, at 1987 prices (1975=100) and index of military/government expenditures (ME/GE), (1978=100) 274 Table 12A.1: Sample countries and sectoral GDP distribution . . . 289 Table 13.1: Estimation results, Greece, 1961-1996 301 Table 13.2: Greece: Effects on national accounting aggregates . 307 Table 14.1: Empirical studies on the Turkish defense-growth relationship 321 Table 14.2: Turkey: specification tests 324 Table 14.3: Residual correlations and lag-length statistics forVAR(2) 325 Table 14.4: Goodness of fit and diagnostic test results 325 Table 14.5: Cointegration analysis 326 Table 14.6: F M L model estimates 330 Table 14 A. 1: Unit root tests 335 Table 15.1: Guatemala: econometric model 345 Table 3.1:
xiv
Table Table Table Table
15.2: 15.3: 15.4: 15.5:
Table 15A.1: Table 15A.2: Table 16.1:
Estimated models, 1969-1994 Guatemala, scenario 1 Guatemala, scenario 2 Per capita real consumption baseline simulation and scenarios land 2 Guatemala military expenditure Data set for Guatemala study Chile: services' share in fiscal income in selected years
xv
347 359 359 362 367 368 385
Contributors
Emmanuel Athanassiou is a Research Fellow at the Centre of Planning and Economic Research in Athens, Greece. He also lectures on institutional economics and the economics of uncertainty at the Department of Economics, University of Athens. Alvin Birdi is Senior Lecturer in the School of Economics, Middlesex University Business School. His research interests include South African industry and military expenditure, and he has published in this area in Defence and Peace Economics. Jurgen Brauer is Professor of Economics at Augusta State University's College of Business Administration. He serves on the editorial board of Defence and Peace Economics and is a vice-chair of Economists Allied for Arms Reduction (ECAAR). Tilman Briick holds a PhD degree in economics from Oxford University and works at the Deutsches Institut fur Wirtschaftsforschung (DIW) in Berlin. He has conducted fieldwork in Mozambique, Tanzania, and the former Yugoslavia. Neil Cooper is a lecturer in politics and international relations at the University of Plymouth. He is author of The Business of Death; Britain 's Arms Trade At Home and Abroad and has published works on UK arms export policy, the arms trade and post-conflict demilitarization.
Lloyd J. Dumas is Professor of Political Economy and Economics at the University of Texas at Dallas. His most recent books is Lethal Arrogance: Human Fallibility and Dangerous Technologies (St. Martin's Press, 1999). J. Paul Dunne is Research Professor in the School of Economics at the Middlesex University Business School and chairs the UK affiliate of Economists Allied for Arms Reduction. He has co-edited a number of books, including The Peace Dividend (Amsterdam: North Holland, 1996). Manuel Ennes Ferreira is a professor in the Department of Economics at the Instituto Superior de Economia e Gestao at the Technical University of Lisbon. He is the author of A Industria em Tempo de Guerra: Angola, 1975-1991 (Lisbon: Edi<;ao Cosmos/IDN). Maria del Carmen Garcia-Alonso is Lecturer at the Department of Economics at the University of Kent at Canterbury, UK. Her work has been published in journals such as Defence and Peace Economics and the International Journal of Industrial Organization. Geoff Harris is Professor of Economics at the University of Natal, where he also teaches in the Conflict Resolution and Peace Studies program. He recently edited Recovery from Armed Conflict in Developing Countries (London: Routledge, 1999). Tony Kempster, DSc PhD MA, is a writer, singer, and campaigner. He is member of the Executive Committee of Campaign Against the Arms Trade (CAAT) with responsibility for strategic planning. He is also Secretary of the Anglican Pacifist Fellowship and a Director of the Peace Museum, Bradford. Lawrence R. Klein is Benjamin Franklin Professor Emeritus of Economics at the University of Pennsylvania, Philadelphia. A Nobel Laureate in Economic Science.
xvn
Christos Kollias is a professor of Economics at the Department of Business Administration of the Technological Education Institute of Larissa, Greece. His work has been published in journals such as Defence and Peace Economics, Journal of Peace Research, Labour, and Applied Economics. Paul Levine is Professor of Economics in the Department of Economics at the University of Surrey in Guildford, UK. He has co-authored or co-edited several books, including Rules, Reputation and Macroeconomic Policy Co-ordination (Cambridge University Press, 1993, with David Currie). He is also a CEPR fellow and a visiting Professor on the Regulation Initiative, London Business School. Kanta Marwah is Professor of Economics at Carleton University, Ottawa, Canada. She has published numerous articles in economics journals including Econometrica, American Economic Review, European Economic Review, Review of Economics and Statistics, and Economic Modelling. Ron Matthews is an economist at the Department of Defence Management and Security Analysis, Cranfield University, UK. Fotis Mouzakis holds the post of European Researcher in DTZ Research, where he analyzes and forecasts the European property market. He has a number of research publications to his credit, including two papers in Defence and Peace Economics, co-authored with Paul Levine and Ron Smith. Eftychia Nikolaidou is a professor in Thessaloniki, Greece, at City Liberal Studies, an institution affiliated with the University of Sheffield. Her applied work investigates the economic effects of defense spending in small industrialized countries as well as arms race issues and has been published in journals as well as collective volumes. Oyinlola Olaniyi is a senior lecturer and the head of the economics xvm
department at the University of Abuja, Nigeria. He has published journal articles in the Scandinavian Journal of Development Alternatives, the Nigerian Defence Academy's Journal of Defence Studies, and the Nigerian Journal of Economics and Social Sciences. Guillermo Pattillo is a professor at the Department of Economics of the University of Santiago and at the Institute of Political Science of the Catholic University of Chile. He is the editor of the monthly report on the Chilean economy published by the University of Santiago. In the field of military affairs he has published policy-oriented papers with a focus on the Chilean situation. Thomas Scheetz is Professor of Economics at Lincoln University College and teaches defense economics at the University of Buenos Aires. He has published widely in Spanish and English and co-authored Defensa No-Provocativa: Una Propuesta deReforma Militarpara la Argentina, and articles in the Journal of Peace Research and Defence and Peace Economics. Selami Sezgin is a professor in the public finance department at Pamukhale University in Denizli, Turkey. His interests include defense economics and economic growth in developing countries, especially Greece and Turkey. He has published in Defence and Peace Economics and other journals. Julide Yildirim obtained her Ph.D. degree in economics and econometrics at the University of Manchester, UK. She is a lecturer in Afyon Kocatepe University in Turkey. Her interests include monetary economics, econometrics, and defense economics. Stavros Zografakis is an advisor to the Greek Minister of Labor. He also lectures in the Department of Economics, University of Athens, on computable economic models and macroeconomics. He has published in the Journal of Population Economics and in a number of edited volumes.
xix
This page intentionally left blank
Introduction Jurgen Brauer and J. Paul Dunne
After marked reductions in military spending in the 1990s that followed the end of the cold war there now seems to be a leveling off and even the possibility of future increases in military budgets. This is not the result of any obvious significant change in the strategic environment, but would seem to be the result of countries' internal debate over their security and economic needs. The removal of the superpower confrontation had erased much of the pressure and support for the maintenance of high military burdens. This was particularly true in the developing world where the removal of superpower involvement in regional conflicts reduced tensions, military budgets and military-related aid, and the scale of conflicts. The general trend to reduced military burdens did hide more complex patterns. For example, some countries increased their military expenditure, mainly in response to local insecurities and local arms races, but also encouraged by the push for arms exports by developed countries. It is also important to recognize that the cuts in military spending that took place did not and do not necessarily imply a reduction in militarization. Weapons merely have become cheaper in the increasingly competitive world arms market, and the world remains a very dangerous place with many regional and civil conflicts. But the genuine cuts in military expenditure that have taken place since the end of the cold war are impressive. Military burden (military spending as a share of output) for the developing countries peaked in the mid-1970s, then declined and rose again peaking in 1982-83. It then declined to 1989, before rising briefly in 1989-90 and then continuing the downward trend. The overall picture for developing countries obscures regional differences. In South Asia, real military expenditure had continued on an increasing trend until the recent economic crisis, while
2 Arming the South
declining in Africa and South America. In the Middle East real military expenditure peaked in 1982-83, fell to 1989 and then rose again briefly. The rise in military expenditure, both in constant prices and as a share of GNP, in the developing world in the early 1990s was almost entirely the result of this increase in the Middle East (ACDA, 2000). In countries where the pressure to halt reductions in military spending have been strong, such as the US and UK, and South Africa in the developing world, the arguments have involved claims that military spending reductions carry economic costs as much as they include considered strategic threats. A particular claim has been made regarding the presumed benefits of offset trades and countertrade to arms importing countries, claims that are increasingly used to justify such imports. This has reignited the debate over the role of military spending in the domestic economy and its impact on development, challenging researchers to reconsider their previous work within the post-cold war economic and strategic environment. In this context it is therefore important that a new attempt is made to reconsider the economic impact of military expenditure, taking account of the changed political, strategic, and economic environment. This is particularly important for the developing world. As reductions in military expenditure can release resources that can be used for other purposes, the possibility exists of tackling some of the major threats to human survival, such as the problems of poverty, economic insecurity, and environmental degradation. For instance, the UNDP estimated that in developing countries the chances of dying from social neglect (e.g., malnutrition, preventable disease) were 33 times greater than from a war started by external aggression (UNDP, 1994, p. 50). If military spending is not shown to have a positive impact on development, then the cost in terms of foregone economic and social development can be huge and in most cases is unlikely to be justified by external threats. It is important to understand the dynamics of the military sector in the economy and society and its link to both military and economic security. It is equally important to develop policies to assist the developing world to move toward a future that is both peaceful and prosperous. This book makes a contribution to developing such an understanding. It offers a comprehensive coverage of issues and countries and brings
Introduction
3
together some of the top researchers in the field. It could be seen as a companion volume to the well-known and well-received Gleditsch et al. (1996) "peace dividend" survey but with its focus very clearly on the developing world. Our book is divided into four parts. Part I consists of four overview chapters which consider demilitarization and new forms of warfare in Africa as well as the relation between military spending and development in South America and East Asia. Part II provides three chapters to deal with the economics of arms production in developing economies and the arms trade, with Part III offering nine country/regional case studies in which various themes explored in earlier chapters come together. Finally, Part IV allows a research user a voice, supplying a contribution from an anti-arms trade campaigner. Considering Part I in more detail, the first chapter sees Lloyd J. Dumas considering a subject of vital importance to much of the developing world, but most significantly to Africa, namely the role of demilitarization in promoting democracy and prosperity. Dumas argues that militaries are unavoidably rigid, authoritarian, and hierarchical organizations meaning that societies dominated by the armed forces are not likely to be fertile ground for the growth of open and democratic political systems. Demilitarization in Dumas' view is therefore a prerequisite for democratization and for genuine economic development. Moreover, economic growth requires skill development which is expensive and difficult to achieve. In the face of the resources required to support extensive armed forces, this becomes virtually impossible. In addition to recognizing the role demilitarization can play in assisting development it is also necessary to recognize the changing nature of conflict. This is dealt with in chapter 2 where Neil Cooper spells out the links among post-modern peace in the developed world, post-modern weapons, and post-modern conflicts and takes a close look at the nature of those conflicts. He argues that the representation of contemporary conflicts as predominantly internal and driven by irrational ethnic hatred ignores the political economy of conflicts, where actors utilize control over local resources and links with global trading and other networks to accumulate the profits necessary to pay for arms, to reward supporters, and to accrue personal wealth. The chapter provides a contribution that is at the forefront of new thinking about the nature of
4 Arming the South
civil war in developing countries. The next two chapters deal with the relation between military expenditure and development in two other regions of the world that have seen high military burdens and military governments. Thomas Scheetz provides a stimulating and provocative analysis oi military expenditure and development in the Latin American context, in chapter 3. In an attempt to interpret numerous econometric studies showing a negative relation between military spending and economic development, Scheetz postulates that the negative impact of military expenditure on Latin American development (even after democratic transitions) is due to incorrect defense and military policies by the respective governments, and not to anything inherently negative about military spending in itself. He advocates military reform, namely adoption of non-offensive defense, arguing that this would stand the best chance of making the armed forces of Latin America positive factors in the development of their countries. In chapter 4, Geoff Harris examines trends in Asian military expenditure during the 1990s and supplies various interpretations of these trends. He then considers the main determinants of military expenditure, paying particular attention to the underlying motivating factors and to the relation of military expenditure to economic growth and development. To this Harris adds an interesting and extensive catalogue and discussion of a number of cost-effective alternatives to the military. Like Dumas and Scheetz, Harris updates our perspective on these issues and challenges us to study how the presumed objective of military forces - security - may be achieved, or even strengthened, with fewer resources. In Part II of the book the focus is on arms production in developing countries and on the arms trade. In chapter 5, Jurgen Brauer presents a history? and post-cold war assessment of the arms industry in developing nations. He reviews which developing nations are producing armaments, what their economic and non-economic motives are, and whether or not the promises of economic benefits from indigenous arms production are fulfilled Brauer finds that from the early 1980s to the late 1990s, a number of formerly developing nations have "graduated" from relatively low levels and sophistication of arms production to relatively high levels (e.g., South Korea, Taiwan, Singapore, Spain, Portugal, Israel). This coincides with the continued development of their civilian industrial
Introduction
5
capabilities. Strategic motivations apart, indigenous arms production efforts have been justified on economic grounds, that building up an indigenous arms industry will spur generalized industrialization by means of spill-over or spin-off effects, and that export sales will provide foreign-exchange earnings. After considering the evidence Brauer argues that, if anything, the development of indigenous arms industries in developing nations depends crucially on already established civilian capacities and that no one has ever presented a convincing case that arms exports provide net foreign-exchange earnings. Following on from Brauer's analysis an important question is what are the fundamental factors that can determine when countries are likely to start domestic arms production as an alternative to arms imports. This is considered by Fotis Mouzakis, in chapter 6, where he constructs a small country arms trade model that focuses on the properties of substitution between imports and domestic production by allowing the country an option of developing a domestic arms industry. Military capability is produced by labor, imported arms, and domestic arms, all of which are imperfect substitutes for each other. The transition of countries into arms production then represents a discontinuous change in the allocation of resources for defense. Solving the model Mouzakis finds that with the establishment of a domestic arms industry the demand for imported arms becomes less responsive to the military budget of the country and the effect of economic development on this demand changes from positive to negative. In contrast the impact of the price effect of arms imports remains negative. An econometric analysis tends to support the theoretical findings. A further analytical contribution is made, in chapter 7, by Maria del Carmen Garcia-Alonso and Paul Levine who produce a model to analyze the role of arms export markets in determining defense procurement policies in arms producing countries. Building on previous models, they consider arms producers and their governments as two distinct sets of decision-makers, with correspondingly different decision variables and objectives. Firms decide on the quantity of weapons that maximize their profits, while governments decide on the quantity of defense procurement that maximizes their utility. But there is of course a close relation
6 Arming the South
between firms and their governments: the firms are guaranteed positive profits by the choice of government's procurement price. Among other findings, the model shows that a decrease in the weight of security in the utility function of exporter countries decreases domestic defense procurement, increases the subsidy to the firms and the quantity of exports, and decreases the export price. However, a general decrease in the importance given to security in the utility function decreases both domestic procurement and exports. Part III of the book consists of nine country/regional case studies in which various themes explored in earlier chapters come together. The studied countries are Saudi Arabia, South Africa, Mozambique, Angola, sub-Saharan Africa, Greece, Turkey, Guatemala, and Chile. In chapter 8, Ron Matthews considers defense offsets and economic development in Saudi Arabia. This is an issue of growing importance to the arms trade with the number of countries in possession of formal offset policies now numbering about 150. Saudi Arabia's offset arrangements began in 1985. The relevant question is: do offsets actually work? Matthews argues that there is merit in the view that offsets are a mode of technology transfer that have the potential to facilitate local economic and technological development. However, whilst offsets may act as a stimulus to the promotion of indigenous production, by themselves they are not sufficient. It is important for there to be local capacity to absorb the technology transferred. Alvin Birdi and J. Paul Dunne consider the impact of military spending on economic growth in South Africa, in chapter 9. They argue that South Africa provides a particularly interesting case study because of the nature of its military industrial complex, the characteristics of the economy and the fact that it has undergone considerable change. In addition, the existence of relatively high quality data for a developing economy is unusual and valuable. They then provide a comprehensive review of the econometric analyses undertaken to date and present some new results which build upon previous work and use a cointegrating V AR modeling approach. At an aggregate level the results suggest a negative though statistically insignificant effect of military expenditure on growth. But when estimated at the level of manufacturing the results show a positive long-run relation between military spending and growth and a
Introduction
1
negative short-run effect. They argue that this is not surprising given the role the military played during the apartheid era. Overall the composite effect of the short-run coefficient on military spending and the errorcorrection term suggest that the short-run impact of cuts in military expenditure will at worst not be significant. In chapter 10 Tilman Briick contributes an intriguing comparative case study on Mozambique and Germany. He argues that both countries shared a number of crucial features following the end of their respective devastating wars, and that it might therefore be useful to reexamine how Ludwig Erhard engineered the German post-1945 "Wirtschaftswunder" in order to learn whether some of the lessons would have been applicable to Mozambique following the cessation of its war in 1992. The lessons learned suggest that wartime economic reforms in the areas of war debt, military expenditure, and economic institutions are crucial to maintain macroeconomic stability and household entitlements during the war. The lessons also are important as a foundation for equitable post-war economic growth, especially in economies with long-lasting or devastating conflicts. Briick concludes that Erhard's views of the secrecy surrounding war debt, of the negative effects of a large war debt, and of public expectations in the post-war period are also relevant to poor developing countries emerging from violent conflicts. Since 1945 exports and (fungible) aid have become more important in financing both wars and post-war reconstruction, thus requiring a modification of Erhard's analysis. In addition, Erhard's relative neglect of the social effect of war on households is not acceptable in the contemporary world. In chapter 11, Manuel Ennes Ferreira contributes a piece on Angola. In spite of forty years of nearly uninterrupted colonial and civil war, publications on the economics of Angola's war are rare. Ferreira examines Angola's manufacturing industry during the post-colonial period, 1975 to 1999, and finds that while the war and the associated military effort certainly made things difficult it was not nearly the only, and perhaps not the major, explanatory factor for the industry's abysmal performance. At least equally important in the catastrophic decline of Angolan industrial performance - a fact which he documents well with original source materials from Angola - were the introduction of a
8 Arming the South
misguided nationalization of industrial assets, mishandled conversion to a centrally planned, Soviet-style economy, and misdirected economic policies. Even following the brief respite from the civil war and the attempt to introduce economic and political reform - market economy and democracy- as from 1992, Angola's industrial fortunes continued to decline. But sheer ineptitude, lack of economic policy skills, consequent inadequate policies, continued political instability, political favoritism in privatizing previously nationalized assets, low transparency, and rampant corruption all kept investor uncertainty high. In addition, Angola's currency was overvalued, import trade restrictions severe, monetary expansion and inflation rampant, and world oil prices depressed. Clearly, the country's policy-makers had a way to go when the civil war resumed in 1994. Today, Angola continues in its economic and attendant social mires. Oyinlola Olaniyi examines, in chapter 12, the contribution of military spending to economic development in sub-Saharan Africa. As the foregoing African country case studies indicate, this is an important region to study. Military spending absorbs a large part of government budgets in this very poor region. For his sample of countries, for the early 1990s, Oyinlola's results suggest that military spending crowds out civilian investment in the economy, is less productive than civilian spending, and does not have any significant externality or spill-over effects on the civilian sector. Military spending would therefore appear to play no positive role in economic development in the region. Within NATO, Greece and Turkey have been involved in border disputes that have seen their military burdens remain the highest in Europe and to have been little influenced by the end of the cold war. In chapter 13 Emmanuel Athanassiou, Christos Kollias, Efi Nikolaidou, and Stavros Zografakis examine Greece's military expenditure, economic growth, and the opportunity cost of defense, using two distinctly different methods. The first estimates an augmented Feder-type model to assess the military expenditure - economic growth relation, finding that the defense sector in Greece has had no positive effect on growth. The second employs a computable general equilibrium (CGE) model to evaluate the likely effects if Greece had followed its fellow NATO members in reducing defense spending during the 1990s. The results suggest that a
Introduction
9
shift of expenditure from defense to non-defense public spending would have been beneficial for economic growth. In chapter 14, Jiilide Yildirim and Selami Sezgin examine the defense-growth relation in Turkey. Moving beyond the commonly used simultaneous equation models, they adopt a cointegrating vector autoregressive modeling (VAR) approach. A five equation VAR model is estimated, where the variables are GNP, military expenditure, savings, labor, and the balance of trade. The results indicate that there is some evidence of a positive relation between Turkish defense spending and economic growth in the long-run. While there is no direct effect in the short-run, they do find a positive indirect effect through the disequilibrium adjustment term. They thus show a positive effect on growth overall and suggest that this likely results from the sheer size of the Turkish army and its impact on aggregate demand. The final two case studies focus on Latin America. The first, in chapter 15, by Kanta Marwah, Lawrence R. Klein, and Thomas Scheetz, provides an econometric study of military-civilian tradeoffs in Guatemala. After discussing their unique database, the authors investigate the economic effects of military spending using a newly built econometric model for Guatemala. They find that contrary to much popular thinking along the lines of military Keynesianism, developing countries can prosper with less - rather than more - military spending, if governments were to move toward the Costa Rican model. Trimming the military and its demands on scarce resources can result in both short and long-run gains. In the long-run, the gains can be seen in the broadest economic measures, such as GDP, while both short and long-run gains are shown to occur in household consumption. Chapter 16, by Guillermo Pattillo, looks at the allocation of resources to the armed forces in Chile. His concern lies more with the transparency of budget allocation processes and procedures than with budgetary amounts, and he finds that the common denominator of resource allocation to armed forces in South America is their relative independence from effective social control. He finds that even though civil authorities formally possess the ability to make budget allocations, they ultimately occur to a great extent outside their scope of influence, thus lending the armed forces special status. This special status is argued
10 Arming the South
to reduce social welfare and to undermine the armed forces' legitimacy in society, thereby promoting conflict and distrust. It is important for researchers to listen to the consumers of their output and so the book concludes in Part IV with chapter 17, by Tony Kempster, an anti-arms trade campaigner in the UK. He examines the role of the peace movement and argues that peace campaigning is essentially about persuading those in positions of power to seek an end to existing conflict through justice and reconciliation, and to eliminate underlying reasons for potential future conflict. He then discusses the role of the peace movement in persuading decision-makers to ban, or at least severely restrict, arm sales to developing countries and discusses the information needs of campaigners and how researchers in the field can assist them. One of the most important ways to assist campaigners and policymakers is to produce volumes such as this one that bring together a comprehensive range of recent and up-to-date research output. This should ideally be based on an open event that allows debate and challenge within a constructive and supportive environment. Indeed, this describes the genesis of this book. A scholarly conference was held in 1998 at Middlesex University Business School in London, co-sponsored by the School, by the UK affiliate of Economists Allied for Arms Reductions (EC AAR), and by the Campaign Against the Arms Trade (C AAT). Of the papers presented, the editors made a selection and invited additional papers to round out the volume. All papers underwent a review and revision process. We thank our authors first and foremost for their contributions but also for their diligence and patience with the editors, and we thank our sponsors for hospitality and financial support. We were fortunate to have been able to draw on a diverse set of authors. They include a number of senior scholars, including one economics Nobel-Laureate, as well as relatively young scholars who had only recently completed, or were about to complete, their doctoral work. Four of the seventeen chapters are co-authored by women. The national origin or place of residence of our authors or their chosen country subjects also is diverse. These include, in alphabetical order, Angola, Argentina, Australia, Canada, Chile, Germany, Greece, Guatemala, Mozambique, Nigeria, Portugal, South Africa, Spain, Turkey, the UK,
Introduction
11
and the US. Brief biographies are provided for each author or co-author in the contributors section at the beginning of the book. The range and quality of the contributions to this collection make it a valuable resource for all those interested in the field and we hope will stimulate further research in an area of study that is vital for the prosperity of much of the world. References Gleditsch, N.P., O. Bjerkholt, A. Cappelen, R.P. Smith, and J.P. Dunne (eds.). The Peace Dividend. Amsterdam: Elsevier, 1996. United Nations Development Program (UNDP). Human Development Report 1994. Oxford: Oxford University Press, 1994. US Arms Control and Disarmament Agency (ACDA). World Military Expenditures and Arms Transfers 1998. Washington, DC: ACDA, April 2000.
This page intentionally left blank
PART I: MILITARY EXPENDITURE: OVERVIEWS
This page intentionally left blank
1 The Role of Demilitarization in Promoting Democracy and Prosperity in Africa Lloyd J. Dumas
Introduction In April 1994, the slaughter began. With guns, with knives, with clubs, the people of Rwanda butchered each other in an ethnic genocide carefully orchestrated by Rwanda's government, while the rest of the world watched in horror - and did nothing.1 By the time it was over, 500,000 to 1,000,000 people had died and several million more had fled their homes in terror. Squalid conditions in refugee camps took the lives of thousands who sought sanctuary in neighboring Zaire (a nation soon to be violently transformed into the Democratic Republic of Congo). Malnutrition and maltreatment weakened them, and cholera spread like wildfire (Villalon, 1998). Though the tragedy in Rwanda was monumental in scale, it was not all that different in kind from many other episodes of political and ethnic violence that have repeatedly plagued that troubled continent. The legacy of political and economic domination and exploitation of Africans by colonial powers laid the foundation for decades of political and economic domination and exploitation of Africans by other Africans. It is no longer the point who is to blame for the political oppression and depressed economic condition of the vast majority of the talented, culturally rich
16 Arming the South
African people. It has gone on far too long, and it is time for it to end. There are too many people with too many weapons in too many places in Africa. There is too much violence, too much poverty and desperation. The people of the continent that gave birth to all of humanity have the potential to do much better. Removing the obstacles that militarization has created is one key to unlocking that potential. Militarization as an impediment to development There is a substantial theoretical and empirical literature investigating the relationship between military spending and development. Some studies, such as Nabe (1982) and Willet (1997), have focused specifically on elements of this relationship in Africa. Others, such as Ball (1988) and the OECD (1997) have considered this connection in the developing world in general. Still other analyses have looked at these issues more generically, as in Dumas (1986) and Payne and Sahu (1993). For a variety of reasons, not all of this literature comes to consistent conclusions, including differing perspectives on the meaning and measurement of development itself. From my point of view, economic development is much more than growth in the money economy as measured by Gross Domestic Product (GDP), because economic activity is much more than the carrying out of money-valued transactions. The purpose of the economy as a social system is to provide for the material well-being of the population. Whether or not money is paid for it, any activity that results in the production and distribution of goods and services which add to material well-being contributes to that purpose. It is the quantity, quality, and availability of the products of such "contributive" activity that measures the true size and state of development of the economy. At the same time, activities that result in the production and distribution of goods and services that do not add to material well-being are "noncontributive," even if money is paid for them (Dumas, 1986; 1995). Military activity is one of the most important types of economically noncontributive activity in the modem world. Whatever else can be said for it, military activity does not grow food, it does not produce clothing, it does not build housing, and it does not keep people amused. Nor does
The role of demilitarization in promoting democracy and prosperity in Africa
17
it create the kind of machinery, equipment, and facilities that can be used to grow food, produce clothing, build housing, and the like. Military activity may have other kinds of value, but it has no economic value because it does not directly contribute to material well-being, to the material standard of living. But while military goods and services have no economic value, they do have considerable economic cost. Military expenditure cause labor, machinery, equipment, and other economically productive resources to be drawn into the service of the military sector. All of these resources could alternatively have been used to produce and distribute goods and services that do raise the standard of living. Their true cost is therefore their opportunity cost, the material well-being that has been sacrificed as a result of this noncontributive diversion of resources. To this cost must also be added the economic cost of the loss of human life, destruction of property, and economic activity foregone because of the turmoil and upheaval caused when military goods and services are put to the violent use for which they are designed. In the more developed countries, this cost has been a difficult burden to bear. For the weaker economies of Africa and much of the rest of the developing world, it has been nearly unbearable. Economic development cannot succeed without a great deal of economic investment. Education to raise the skill of labor and improved physical infrastructure are the two most important investments for generating strong, sustained development. Economies need substantial investment in the skill and education of their workforces and in their systems of roads, communications, water, and power supply in order to have any real prospect of freeing themselves from pure dependence on the export of primary products. Such investments will not make development happen by themselves, but they are a key part of the groundwork that must be laid before real development can be generated and sustained. There is no force for development more powerful than growth in the skills and capabilities of the labor force. People are both the reason for development and the most important means achieving it. But as important as it is, education is not magic. Education inherently raises people's expectations. Without critical investment in infrastructure and other
18 Arming the South
necessary capital, education alone will not generate enough good jobs for those who have acquired greater knowledge to put it to work in ways that are both economically productive and individually lucrative. This leads not to sustained development, but to growing frustration - frustration that is both personally cruel and socially dangerous. There is no way around the fact that large scale investments in education and infrastructure are very expensive. It is simply not possible for countries of limited means to make these investments on anything like the scale required as long as they insist on pouring large amounts of their limited resources into oversized military forces. Understanding this basic fact of economic life, the Costa Rican government eliminated its national military forces entirely in 1948 and directed the nation's limited resources to more economically contributive activities. For more than half a century now, in the absence of any national military, Costa Rica has managed to maintain its independence. More than that, it has been the most stable, democratic, and economically well off nation in a part of the world that has been plagued by economic trouble and wracked by terrible spasms of violence. It is an interesting and important example. Aside from draining resources and destroying people and property, a militaristic mindset replaces the economic urge to develop new resources and create needed goods and services with the military urge to take them from others by force. Quite apart from the obvious moral distinction, there is a profound difference between acquiring goods and services by confiscation and acquiring them by economic activity. Acquisition by force is at best a zero-sum game, a game of redistribution, where how much the winners gain depends on how much the losers have lost. More likely, it will be a negative-sum game, a game of net loss, because some goods and services will be destroyed in the battle. By contrast, economic activity is a positive-sum game, a game in which new wealth is created, a game in which the potential exists for everyone to win. The people of Africa have waited far too long for the dreams of economic prosperity that accompanied their political independence to become reality. Decades after decolonization, they find themselves still tied by bonds of economic dependence and falling farther behind. These bonds cannot be broken and replaced by more balanced, mutually beneficial trade relations until the nations of Africa find a way to
The role of demilitarization in promoting democracy and prosperity in Africa
19
accelerate their rate of economic development. Whether established by the government or disaffected rebel groups, oversized, overfed, overly influential military and paramilitary forces are not compatible with broadbased economic development. Substantial demilitarization is a necessary, though not sufficient, condition for success. Militarization as an impediment to democracy People do not ordinarily relish the idea of killing other people, nor do they look forward to putting themselves in the position of being killed or seriously injured. Yet stripped of the pomp and ceremony, of the uniforms and rituals, that is exactly what militaries are all about. Soldiers must be ready to kill or be killed or militaries cannot do what they have been designed to do. Military training is therefore most assuredly not simply a matter of teaching people to use weapons. It is very much a process of social and psychological conditioning, designed to take away their individuality and train them to do what they are told. There is no room for questioning authority, no place for free and open debate. In the midst of military action, votes cannot be taken on which tactics to use. It is difficult to see how militaries could be effective if they were not authoritarian organizations. They cannot be built around democratic principles. It is therefore very difficult for truly democratic political systems to develop and prosper in militarized societies. If democracy is incompatible with militarization, how is it that nearly all of the world's democratic nations have military forces and some (such as the UK, France, and the US) have very large, well-funded militaries? There is a difference between a society with a military and a militarized society. In the democratic nations, elected civilian officials control the governments. They are the only ones authorized to set the nation's military forces in action. The use of military forces within the borders of their own nations is severely circumscribed. There are also social norms, if not strictly enforced laws, that for the most part restrict the formation and limit the social impact and political influence of armed paramilitary forces. The long history of the Irish Republican Army and the Ulster Defense Forces in Northern Ireland, and the recent worrisome growth of private militias in the US make it clear that these restrictions and
20 Arming the South
protections are not completely effective. But even in Britain and the US, these paramilitary groups have had relatively limited impact on life in the wider societies. While the mere existence of national armed forces is not incompatible with democracy, real democracy is unlikely to take hold and fully develop in Africa (and elsewhere) unless there is a major shrinking of national and subnational armed forces in size, power, and influence. Even where democratic institutions are firmly embedded in law and custom, as in the US, overly powerful and influential military forces pose a real and present danger to democracy. One of the most successful military commanders of the twentieth century, Dwight Eisenhower, the 34th President of the United States, warned nearly forty years ago of the need to guard against the corrupting influence of what he called "the military-industrial complex" on the political life of the nation.2 The full flowering of democracy requires the establishment of democratically elected governments that truly reflect the interests and opinions of all of a nation's people. In that respect, even the wellestablished democracies still have some distance to go. Electing governments by popular vote is part of what is required, but it is not enough. The qualified electorate must be broad enough to express the opinions and interests of all the nation's people. There must be a sufficiently free flow of information in the society so that those who want to go to the polls can be well enough informed about the issues to cast a considered vote. People with a wide range of political viewpoints must be free to publicly argue their case. They must also have access to whatever it takes to seek political office and make themselves heard by the electorate. It is not particularly easy to establish these conditions. They are not fully established even in advanced democracies like the US where among other things the corrupting influence of money on politics has again emerged as a big public issue. But they are nevertheless the goals we must continue to pursue. The process of building democracy is not restricted to the establishment of the formal institutions of democratic government. It is also important to build the underlying political infrastructure of civil society from which democracy draws its strength and durability. Civil society is different from both government and the business sector. It is
The role of demilitarization in promoting democracy and prosperity in Africa
21
composed of the formal institutions and informal relationships and traditions that promote trust, a sense of belonging to and responsibility for the wider community, and an obligation to build a common future better than the past.3 Non-governmental nonprofit organizations, founded and energized by the ordinary men and women of a nation, are the formal institutions of civil society. They are brought into being for many different reasons, ranging from simply helping those who need help to raising public awareness of social inequities and political injustices. They act to garner support for social or political causes in which their members believe by providing information and encouraging peaceful civil action. A truly democratic society not only permits such organizations to exist and operate independently of control by business or government, it encourages and facilitates their formation. The informal relationships and traditions of civil society are also very important. Real democracy requires more than the ability to express some range of opinion. Civil discourse must be open enough and tolerant enough to allow the expression of opinions which may not only be critical of the government, but with which many of the people of the nation may disagree. Shouting down or shutting out those who express other than mainstream opinions - practices which I'm sad to say have become more common in the US in recent decades - have no place in civil society. But such "uncivil" behavior is inhibited more by culture and tradition than by government or formal civil institutions. And although simplistic slogans, satire, and ridicule do have their place in free political debate, they too tend to close the mind to alternative points of view. As it is said, "minds, like parachutes, always work best when they are open." The authoritarian structure and organizing principle of military forces leaves little opportunity for the formal institutions and informal traditions of civil society to operate. Militaries are unlikely to tolerate freestanding, independent organizations of soldiers that become alternative centers of power and influence. There is little room for wide-ranging, inclusive debate about policies, strategies, and tactics. The culture of obedience and discipline, the hierarchical command structure, and the military tradition Q{ rank and privilege do not encourage either open discourse or the independence and freedom of action that are so basic to civil democracy.
22 Arming the South
A society in which roving, armed bands of government, rebel, or criminal military and paramilitary forces are the order of the day is not a society in which open political discourse and free political activity can flourish. When force is the prevailing means of settling disputes and fear the prevailing means of enforcing order, elections and the other formal trappings of democracy become empty illusions. Between 1990 and 1995, some form of election was held in almost every African country. There were national elections in at least 39 countries in sub-Saharan Africa. Yet it is far from obvious that this flurry of electoral activity was in fact an indication of the spread of real democracy. According to Villalon (1998, pp. 15-16), This proliferation of elections ... tells us little about the extent of substantive change or the degree of democratization in fact occurring on the continent. ... [E]lections themselves may be a strategy for maintaining power, and many African ... elections in the 1990s have been clearly intended to forestall change, or even to strengthen the staUis quo.
The appearance of democracy is not democracy. Nevertheless, the pressures which led to these elections may have created a small opening to democratic processes. It is possible that this opening can be widened enough to make free and open elections a reality at some point in the future, at least in some places. The long-run value of democracy to development Many people seem to think that authoritarian governments, even if "nasty," really get things done. They can stimulate economic development because they can impose the discipline and organization necessary to make it happen. There may be something to this argument in the very short-run. But that kind of government tends to behave in ways that are destructive in the long-run. In the words of Olusegun Obasanjo (1993, p. 199), now in his second incarnation as President of Nigeria, ... one-man, one-party, non-pluralistic, or military regimes ... imposed with the excuse of having the potential to provide greater unity, or correcting the ills and
The role of demilitarization in promoting democracy and prosperity in Africa
23
abuses to which the operators of the different democratic experiments in Africa subjected their countries, have proven perhaps more divisive and prone to corruption than the regimes they ousted. In most cases, in fact, such dictatorial regimes have tended to exacerbate the ills against which they ostensibly forced their way into the seat of power.
Social order and a degree of discipline is important to economic progress. Disconnected, chaotic societies are not good prospects for development. But when authoritarian governments impose order by force and fear, an atmosphere is created that virtually guarantees any economic advantages will be transitory. For an economy to be successful in the long-run, an ongoing flow of people who possess a variety of managerial talents must be available. Key talents like organizational skill, innovativeness, entrepreneurial flair, and leadership ability are widely distributed among the human population. Centralized and authoritarian forms of social organization do not typically provide widespread enough opportunities for those who possess these talents to develop and employ them. They draw from too narrow an ingroup. When criteria such as kinship, social class, political loyalty, or ideology are used to determine who rises to positions of economic authority, some of the most capable people never get the chance to show what they can do. As time goes by, economic institutions come to be run by second or third rate managers. This is much less likely to happen in countries that are more politically and economically democratic, especially if careful attention is paid to breaking up concentrations of monopoly power which are also fertile ground for managerial inbreeding and sloppiness. Wisdom is perhaps even less highly concentrated in a specific ethnic or kinship group, social class, or adherents to a particular ideology than is managerial talent. Without the give and take of free and open debate, bad decisions made by the very small cadre of leaders at the top of the political pyramid are unlikely to be criticized before they have led to economic disaster, if then. As lengthy, irritating, and inefficient as it may be, the kind of public debate characteristic of democratic societies is more likely to explore alternative approaches and ultimately make better policy choices. At the very least, democratic societies are more likely to force the abandonment of wrong-headed policies before they have done
24 Arming the South
intolerable damage. That is not only true of government decisions, it is also true of decisions made in private economic organizations. Unfortunately, hierarchical authoritarian decision-making is often the rule in private business organizations, even within democratic societies with free market economies. This is one reason why less hierarchical worker decision- making schemes can be a real advantage. More often than not, the tendency to concentrate wealth within economic systems and power within authoritarian political regimes becomes self-serving and obsessive. The concentration of wealth and power is a breeding ground for corruption and other behaviors that rarely serve the long-term interests of the population at large. In the long-run, then, economically and politically democratic systems have great advantages in stimulating and sustaining real, broad-based economic development. Making demilitarization work: the process of effective demobilization In the United States, the former Soviet Union, and many of the more developed countries of the world, the end of the cold war raised the challenge of converting unneeded military bases and military industrial capacity to civilian use. Though it has been very poorly implemented in all too many places, this type of economic conversion has been wellstudied and is rather well understood. But the demobilization of armed forces has been given much less attention, largely because it has not been that much of a problem in countries with well-developed economies and integrated societies that can absorb soldiers returning to civilian life. In Africa, where few nations have significant military industry, demobilization is a first-order problem. In Ethiopia, for example, a series of wars of liberation wracked the nation beginning in the early 1960s (see Prendergast, 1997). By 1991, the Eritrean People's Liberation Front had pushed Ethiopian troops out of the territory that became the independent nation of Eritrea two years later. After Eritrea's war of independence ended, many Ethiopian soldiers were simply released from their military duties and told to go home. 4 No special attempt was made to integrate them into civilian life. A lot of them had been in the military so long that
The role of demilitarization
in promoting democracy and prosperity in Africa
25
they knew little, if anything, else. So when they were sent home, they took their guns with them. Used to a military life and without any real civilian skills, they turned into roving bandits who preyed on people in the countryside and further disrupted economic life. A great deal of damage was done before the government finally decided to undertake a variety of programs specifically aimed at retraining and otherwise reintegrating former soldiers into civilian life. These programs apparently met with considerable success. It was a painful but important lesson to learn. There is by now considerable experience with attempts at demobilization in Africa. Not all of it has been positive. In Somalia, several attempts have failed since 1992. In Angola, an agreement was reached in 1994 by both sides in the long civil war to demobilize about 73,000 ex-combatants. After some early progress, the demobilization ran into serious trouble when it appeared that some of the former fighters had not really been demobilized. In Sierra Leone, plans for demobilization were frustrated when the military seized power in May 1997. It remains to be seen whether that demobilization will now take place. On the other hand, Eritrea demobilized 55,000 former fighters since 1993, and Ethiopia about half a million since 1991. In Mozambique, some 70,000 government forces and 20,000 opposition forces were demobilized from 1992-1994. And in Namibia (1989), Uganda (1992-1995), and Liberia (1996-1997), a total of nearly 100,000 former soldiers have been returned to civilian life (Kingma, 1997a; 1997b). Each demobilization occurred for its own specific set of reasons within its own socio-economic and political context. But some common threads run through the process of demobilization and connect it to other types of economic conversion. As in military-industrial conversion, both retraining and reorientation of the personnel involved are crucial. Retraining means giving those in transition useful skills that they did not previously possess, skills that will help them take advantage of existing or newly created economic opportunities. Reorientation means getting them to leave behind the way of thinking they learned in the military world and replace it by a way of thinking that is more compatible with success in the very different civilian world. In military-industrial conversion, that means putting aside the way of
26 Alining the South
thinking appropriate to an environment in which maximum possible performance is crucial and cost is relatively unimportant, and adopting a way of thinking in which reasonable performance and lowest possible cost are the order of the day. In demobilization, it means shifting from a mindset in which unquestioning obedience and the use of extreme violence to achieve objectives is the norm to a way of thinking that is oriented to creation, taking the initiative, and achieving objectives through peaceful cooperation. In both cases, it is a major shift that is critical to success. And in both cases, it cannot be safely assumed that it will occur automatically. The process of demobilizing soldiers in Africa must contend with an especially difficult set of problems. According to Nicole Ball (1997, p. 86), of the Overseas Development Council in Washington, African ex-combatants constitute a specially disadvantaged group. The typical veteran is semi-literate at best, is unskilled, has few personal possessions, often has no housing or land, and frequently has many dependents. Some veterans are also physically and psychologically handicapped by wartime experiences. Many find it difficult to take independent initiatives and to cope with the ordinary demands of civilian life. Even when they possess a marketable skill, such as mechanic or driver, ex-combatants tend to have little or no experience in the labor market, having taken up arms at an early age.
Ball usefully divides the demobilization process into four phases (Ball, 1997): assembly, discharge, short-term reinsertion, and long-term reintegration. During assembly, primarily for security reasons, soldiers are brought to a particular area so they can be counted, registered, disarmed, and given identification cards. While in these areas, they need to be supplied with shelter, food, clothing, sanitary facilities, and medical care. All of this can be expensive. There are thus financial pressures as well as military and political reasons why they are often discharged as soon as possible. In fact, when demobilization is the result of negotiation rather than the defeat of one side by the other, a one year maximum time span for discharge is typically written into the agreement. For social and economic reasons, not the least of which is having the time and resources to properly carry out effective retraining and reorientation, a period of assembly and encampment stretching over a few
The role of demilitarization
in promoting democracy and prosperity in Africa
27
years might make more sense. This is especially true when large numbers of former combatants are being demobilized. Because they are gathered in one place, assembly areas may be the most cost effective locations to retrain and reorient them. Assembly is used to provide both information and tangible packages of cash and/or in-kind assistance, including "food, civilian clothing, household utensils, building material, seeds or agricultural implements/' In Uganda, for example, demobilized soldiers and their dependents were briefed about legal issues, family planning, and AIDS prevention as well as how to open bank accounts and start income generating activities. Their demobilization package also included payment of one year's school fees for their children (Kingma, 1997a). When they are ready for discharge, it is a good idea to provide former soldiers (and their dependents) with transportation to the areas where they intend to settle rather than leaving them to their own devices. In addition to being helpful to them, this helps assure that the ex-combatants do not all congregate in the same place and become a source of future trouble. Once they have been transported to their home areas, further reorientation sessions are useful to help them adjust to the specifics of their new surroundings and the roles they are now expected to play. These sessions are probably best conducted by, or at least in concert with, people from the local community. Other forms of transitory "reinsertion assistance," including providing additional food, household goods, and temporary shelter may also be a good idea. But it is important not to overdo it. It is not good for demobilizing soldiers to get the idea they will be permanently on the dole. Too much assistance is also likely to generate resentment on the part of their neighbors, resentment that can seriously interfere with their ability to successfully reintegrate into civilian society in the long-term. Even if all the other stages of demobilization are successful, there may still be real barriers to effective long-term reintegration of excombatants into civilian life. Reintegration is not just a matter of economic success. Former soldiers and their dependents must also be accepted by the communities in which they will now live. If they are veterans of a popular war or successful liberation struggle, they may be looked upon as returning heroes, and social acceptance may not be a
28 Arming the South
problem. But if they have been part of a government or guerrilla force that killed many of their countrymen and countrywomen, destroyed a great deal of property, and generally made the lives of people in the community miserable, long-term reintegration will be much harder to achieve. It will require careful attention to a lengthy, painful, and complicated process of national or regional reconciliation that may never fully succeed. This much is clear. Demilitarization is vital to the future political freedom and economic success of the African people. Even when those under arms are not killing people, destroying property, and disrupting life, they are draining the economy and holding back political progress toward democracy. Demilitarization is unlikely to succeed without effective programs of demobilization to help soldiers become economically productive civilians. The problem of security Reducing the size and influence of military forces is important to freeing critical resources for economic development and to encouraging the process of political democratization. Even so, the issue of internal and external security remains real and compelling. Neither democracy nor development can flourish in the midst of chaos. While militaries and police forces, controlled by democratic governments, do have a role to play in maintaining order, the inherent tension between democracy and economic development on the one hand and armed forces on the other makes it wise to try to implement reliable alternative security mechanisms. Interestingly enough, both democracy and economic development can themselves be sources of security. Because democracies rarely go to war with each other, the spread of democracy in Africa will likely lower the probability of interstate war (Russett, 1990). Furthermore, societies that allow free and open debate and embrace the principles of democracy have many more socially acceptable and politically viable pathways for expressing dissent and redressing grievances. Those who feel themselves to be marginalized or even unfairly disadvantaged, as well as those who have unpopular political views, are much less likely to engage in violence
The role of demilitarization
in promoting democracy and prosperity in Africa
29
to make themselves heard. For this reason, truly democratic states may also be more internally secure. When there is a well-developed and vibrant civil society to supplement the formal institutions of democracy, internal security is further enhanced. Africans know very well how destructive and violent exploitative economic relationships can be. But economic relationships that benefit all parties more or less equally increase economic well-being all around. Such balanced relationships are easier to build and sustain among countries that have achieved higher levels of development, because they have more to offer each other. Successful economic development will therefore facilitate the establishment of trade relationships that are mutually beneficial, rather than exploitative, among African nations and with the wider world. Mutually beneficial relationships also tend to reduce the threat of violence and war among the trading nations (Dumas, 1990). Similarly, more prosperous people are less likely to arm themselves and fight when conflicts arise within the nation. That idea that balanced economic relations can themselves be a source of security is not just a matter of theory or wishful thinking. There is powerful support for this proposition in the daily operation of the European Union (EU). Even a partial listing of the nations that are today part of the EU - Britain, France, Germany, Belgium, Italy, Spain, Portugal, the Netherlands - is striking. These countries were among the most aggressive in attacking and colonizing the people of Africa and many other parts of the world. They also fought countless wars with each other. Some of the same nations were key players in the two world wars that engulfed much of the globe and took tens of millions of lives. In fact, conflict, even serious conflict, still continues among these nations. And they still are ready to fight against outside states, as in the Persian Gulf war of in the early 1990s. But they are now a part of a web of economic relations that is so beneficial to all of them that they take great pains not to destroy it. They no longer think of settling their differences with each other by resorting to those mass spasms of organized violence we call war. Africans have experimented with establishing economic common markets in the past, one important example being the Economic Organization of West African States (ECOWAS). These experiments
30 Alining the South
have not been all that successful. But when the nations of Africa make the kinds of investments necessary to stimulate real economic development, they will begin to raise the living standards of their people enough to increase both the size of the market within their countries and the variety of goods and services they are able to exchange with their trading partners. That will greatly increase the viability and potential economic benefit of common market arrangements, which will in turn enhance their "peacekeeping" potential. Summary and conclusions Militaries are unavoidably rigid, authoritarian, and hierarchical organizations. Militarized societies dominated by the armed forces and their associated values are not likely to be fertile ground for the growth of open and democratic political systems. Reducing the size and influence of military and paramilitary organizations, and subordinating them to the control of governments put into office in free and fair popular elections, is critical to democratization in Africa and elsewhere. While the impact and influence of the military is shrinking, the formal institutions of democracy and the culture and traditions of civil society must be encouraged to grow until they are thoroughly and permanently woven into the political and social fabric. Demilitarization is not only prerequisite to democratization, it is also vital to stimulating and sustaining a process of real economic development in Africa. The African continent is rich in natural resources. But the resource most important to development is people. There is simply no substitute for large scale investment in their skills and education, investment that must be paralleled by investment in the capital with which they must work. This will be expensive. It is difficult enough for nations with limited means to afford it. In the face of the resources required to support extensive armed forces, it is virtually impossible. Militaries can only win if they are able to defeat their enemy. Military activity is at best a zero-sum game, a game in which only the winners gain, and they cannot gain any more than they are able to take from the losers. No new wealth is generated. Economic activity is very different. Because new wealth is generated, all participants can gain. Even better,
The role of demilitarization
in promoting democracy and prosperity in Africa
31
in the long-run each participant is actually likely to gain more when all of them share in the winnings. Despite the violence, corruption, and failed economic programs that continue to plague the continent, the dream of a more democratic, peaceful, and prosperous Africa remains alive. If that dream is to become a workable, practical vision, Africans must throw off the militarized, hierarchical paradigm that is a legacy of their colonial past and the authoritarian, exploitative way of thinking that goes with it. They must sharply reduce the flow of productive resources into noncontributive military and paramilitary activity, and redirect those resources to economically beneficial pursuits. None of this will be easy to achieve, and it will not happen quickly. But there is no doubt that it can be done. And when it is done, the Africa of tomorrow will indeed be a much more peaceful, democratic, and prosperous place than the Africa of today. Notes 1. In late February 1998, the Canadian general who had been commander of the UN peacekeeping troops in Rwanda testified before the international tribunal investigating war crimes in Rwanda that the 1994 massacre could have been stopped at an early stage if a small, well-armed force had been sent in with that mission. See James C. McKinley, "General Tells Rwanda Court Massacre Was Preventable," New York Times (28 February 1998). 2. This famous warning was delivered at the end of Dwight Eisenhower's second term as President of the United States in 1960, and was the focus of his farewell address to the nation. 3. See, for example, the Institute for Civil Society, One Bridge Street (Suite 101), Newton, MA 02158. 4. Unfortunately, the peace between Eritrea and Ethiopia remains fragile and subject to the periodic outbreak of new episodes of war, as in recent years.
32 Arming the South
References Ball, Nicole. Security and Economy in the Third World. Princeton, NJ: Princeton University Press, 1988. Ball, Nicole. "Demobilizing and Reintegrating Soldiers: Lessons from Africa" in Krishna Kumar (ed.). Rebuilding Societies After Civil War. Boulder, CO: Lynne Rienner Publishers, 1997. Dumas, Lloyd J. The Overburdened Economy: Uncovering the Causes of Chronic Unemployment, Inflation and National Decline. Berkeley, CA: University of California Press, 1986. Dumas, Lloyd J. "Economics and Alternative Security: Toward a Peacekeeping International Economy," in Weston Burns (ed.). Alternative Security: Living Without Nuclear Deterrence. Boulder, CO: Westview Press, 1990. Dumas, Lloyd J. "Finding the Future: The Role of Economic Conversion in Shaping the Twenty-First Century," in L.J. Dumas (ed.). The Socio-Economics of Conversion: From War to Peace. London, UK, and Armonk, NY: M.E. Sharpe, 1995. Kingma, Kees. "Post-war Demobilization and the Reintegration of ExCombatants into Civilian Life." USAID Conference "Promoting Democracy, Human Rights and Reintegration in Post-conflict Societies". October 30-31, 1997a. Kingma, Kees. "Demobilization of Combatants After Civil Wars in Africa and Their Reintegration into Civilian Life." Policy Sciences Vol. 30 (1997b), pp. 150-165. Nabe, Oumar. "Military Expenditures and Socio-Economic Development in Africa." Ph.D. Dissertation, Columbia University, 1982. Obasanjo, Olusegun. "Africa in the Twenty-First Century." Security Dialogue Vol. 24, Number 2 (June 1993). OECD (Organization for Economic Development and Cooperation) and the Government of Canada. "Final Report of the Ottawa Symposium on Military Expenditures in Developing Countries." Ottawa, 1997. Payne, James E. and Anandi P. Sahu (eds.). Defense Spending and Economic Growth. Boulder, CO: Westview Press, 1993. Prendergast, John. "Applying Concepts to Cases: Four African Case Studies," in John Paul Lederach (ed.). Building Peace: Sustainable
The role of demilitarization in promoting democracy and prosperity in Africa
33
Reconciliation in Divided Societies. Washington, DC: US Institute of Peace Press, 1997. Russett, Bruce. "Politics and Alternative Security: Toward A More Democratic, Therefore More Peaceful World," in Weston Burns (ed.). Alternative Security: Living Without Nuclear Deterrence. Boulder, CO: Westview Press, 1990. Villalon, Leonardo A. "The African State at the End of the Twentieth Century: Parameters of the Critical Juncture," in Leonardo A. Villalon and Phillip A. Huxtable (eds.). The African State at a Critical Juncture: Between Disintegration and Reconfiguration. Boulder, CO: Lynne Rienner Publishers, 1998. Willet, Susan. "Military Spending Trends and Developments in Southern Africa." Background paper for the Ottawa Symposium on Military Expenditures in Developing Countries, co-sponsored by the Canadian Government and the OECD Development Assistance Committee. July, 1997.
This page intentionally left blank
2 Warlords and Logo Warriors: The Political Economy of Postmodern Conflict Neil Cooper
Introduction This chapter has two aims. The first is to delineate the links between what has been termed post-modern peace in the developed world, post-modern weapons, and post-modem conflicts. There is a burgeoning literature on all three but little attempt has been made to consider the ways in which the first two have influenced the latter. The second aim is to examine the political economy of post-modern conflicts. I argue that the representation of contemporary conflicts as predominantly internal and as conflicts driven by irrational ethnic hatred ignores the political economy of conflicts where actors utilize control over local resources and links with global trading and other networks to accumulate the profits necessary to pay for arms, to reward supporters, and to accrue personal wealth. Post-modern peace The post-modern peace thesis is one that explains the absence of inter or intra-state war in the developed Western world as a function of the postmodern condition. This idea has been put forward most explicitly by Coker who has argued that post-modemity has made war an
36 Arming the South
"anachronism," a phenomenon that can be explained in particular by the end of ideology (1992). The post-modern peace thesis can however be usefully extended to encompass (a) the rise of what Luttwak has termed "post-heroic" societies in which wars fought for great purposes and involving mass societal sacrifice are unacceptable (1995; 1996), (b) the pacifying influence of interdependence in a globalized world, and (c) the spread of liberal democracy in the so-called "zones of peace." It might also encompass what Bradbury has described as "the advance of consumer democracy and the culture of glitz and shopping" (1995, p. 769) under which as Cooper (no relation) has noted, "individual consumption replaces collective glory as the dominant theme in national life" (1996, p. 36). Post-modern peace has influenced the dynamics of the global arms trade in a number of ways. For instance, it has brought large-scale reductions in the level of military aid and the armed forces. It has also increased the ability of the international community to deploy arms embargoes against actors in conflict. In consequence, whilst military aid from allies still remains an important source of arms for actors in conflict, it is also the case that commercial acquisitions of either arms or mercenaries from the legitimate or black market have become a more salient feature of the arms diffusion process in post-modern conflicts. Consequently, actors in conflict are increasingly forced to exploit global networks in order to raise funding, either through the sale of local resources on the international market, through remittances from the diaspora, or through the exploitation of international aid. Such activities have in turn been made easier by the conditions of post-modemity: for instance, the communications and IT revolutions and the globalization of markets. Moreover, post-modern peace has not, ironically, led to a more ethical approach to arms sales. Indeed, defense exporters have actually become less rigid in their attitude to potential buyers as cold war constraints on exporting to former ideological enemies have been removed. This has made it easier for recipients to play the arms traders off against each other in the search for higher technology at cheaper prices.
Warlords and logo warriors: the political economy of post-modern conflict
37
Post-modern weapons One of the ironies of the post-modem peace in the developed world is that, whilst it has brought large cuts in arms expenditure, one of its defining features has been the continued dominance of what has been termed an "armaments culture" based on the fetishism of the weapon or rather that of the advanced weapons system as distinct from the dominance in society of warlike values or militarism (Luckham, 1984, pp. 2-3). One of the consequences of this is that the armed forces of the major military powers are moving from manpower intensive militaries to an increased reliance on high tech weaponry - from militarism to an armament culture based on post-modern weapons. This is reflected most notably in the debate about the role of technology in the so-called revolution in military affairs (RMA). The extent to which state of the art military technology can be used by (as opposed to against) actors in sub-state conflicts is open to question. Many, for instance, will have neither the technological sophistication nor the economic resources to acquire much of the technology associated with the RMA. Consequently, many post-modern conflicts are, and will continue to be, dominated by relatively simple small arms and light weapons. Nevertheless, such actors are already drawing on both high and low tech civil and dual use equipment where relevant, and this trend is likely to grow. Thus, the US not only supplied the Contras with more conventional arms but also encryption machines as well as seismic, acoustic, magnetic, and infrared sensors; US troops in Somalia and Russian troops in Chechnya both found themselves confronted with warlords who used cell phones to warn each other of enemy movements; and insurgent and criminal groups have exploited the information revolution to launder digital money (from or for arms sales) through offshore banks which sometimes only exist in cyberspace (Rathmell, 1997, p. 41). Consequently, military capability is, and will increasingly be, a function not only of guns and bombs but also of less obviously lethal technologies. The growing salience of post-modem weapons has two implications. First, as post-modem armies draw on an increasing array of civil and dual-use technologies this is inevitably posing challenges to the
38 Arming the South
effectiveness of multilateral arms transfer regimes and national export control systems that have traditionally tended to focus on easily identifiable conventional weapons. Second, the technological fix does not come cheap. Indeed, despite the claim that the technological revolution will result in cheaper as well asfriendlierweapons, there is little evidence to suggest that the rise in equipment costs is abating. This has combined with the cut-backs in military expenditure that have flowed from postmodern peace to create a situation in which both companies and individual weapons development programs are only economically viable if they secure substantial export orders from abroad. In consequence, the contemporary arms market is now a buyers bazaar in which sellers are required to raise the technological sophistication of equipment they are prepared to sell, to lower prices, and to offer innovative approaches to the financing of arms acquisitions - in particular through the use of export credits and offsets. Post-modern conflict The power of post-modern states in the West has been eroded by the impact of globalization, the development of regional institutions such as the EU or NAFTA, and the privatization of state functions. They have thus become virtual states. Such states however still retain significant law and order and welfare functions and are able to command levels of legitimacy which are sometimes out of proportion to the scale of their autonomy. Many weak states in the developing world have also become virtual states, but in a different way. The autonomy of such states has been undermined by the influence of globalization, aid dependence, and structural adjustment programs which reduce both the authority and legitimacy of the state. At the same time however, local elites both inside and outside government have been able to manipulate control over local resources and their links to both the formal and informal global economy in order to extract wealth and generate political support via the distribution of rewards to supporters. At its worst this has meant, as in Zaire under Mobuto, that the apparatus of the state simply becomes a vehicle for personal acquisition. Furthermore, such approaches to profitmaking do not necessitate the development of inclusive state-building
Warlords and logo warriors: the political economy ofpost-modern conflict
39
strategies relying upon strong bureaucracies serving popular needs in exchange for popular support (Reno, 1997). Indeed, strong states can impede profits by developing regulations that restrict corruption, corporate vandalism, and the free flow of money out of the country and into Swiss bank accounts. Strong states can also provide alternative centers of power which can represent a potential threat to established elites. Consequently, as Reno has noted, the interests of weak state rulers and/or political elites in undercutting the power and authority of the state intersects with the agenda of external creditors advocating neo-liberal economic reform. This is in two ways. First, the process of state erosion can provide further opportunities for personal enrichment or to reward supporters. For instance, the privatization of nationalized industries often fails to introduce market efficiencies because old state-run enterprises are simply passed into the hands of cronies whose influence guarantees the continued distortion of market and tax mechanisms which preserves profits for owners whilst doing little to enhance productivity or to disperse the benefits of wealth throughout the general population (Reno, 1997, p. 169). Alternatively, to cite Reno at length (1997, p. 167), rulers of some weak states use creditor demands to privatise state agencies and liberalise markets as excuses to hire foreign firms that field mercenaries. These foreign soldiers serve the interests of foreign firms and weak state rulers to control resources and deny them to independent strongmen ... these new arrangements transform patron-client politics into trans-national alliances that generate profits in global markets ... paradoxically, the influence and interests of strong states, creditors and private firms permit weak state rulers to abjure the internally risky, but historically more effective strategy of maximising authority with bureaucracies capable of mobilising populations while imposing uniform central authority.
Thus, weak state elites act as the intersection between the virtual states of the first and the developing world. In the former, the state (and the company) has contracted out extraction and production to other regions of the world. In the latter, weak state elites (both inside and outside of government) use the privatization of the state as an opportunity to maximize personal wealth and power by exploiting their control over local resources and their links with both the formal and informal global marketplace. Such strategies of course do little to develop either the
40 Arming the South
authority or legitimacy of the state and do even less to distribute economic rewards throughout society. This inevitably increases the likelihood that states will ultimately break up and that other elites within the state will attempt to follow similar strategies of personal enrichment, creating and also (in post-conflict societies) perpetuating, the conditions for conflict. When states do break up and descend, or redescend, into violent conflict such a process seems incomprehensible to outside observers they are neither colonial wars of national liberation nor cold war conflicts designed to save a country from the ideological extremes of capitalism or communism. Indeed, just as the post-modern era has seen the decline of ideology in the West so has it seen the decline of ideology as a motive or legitimizing force for conflict in the developing world. The only possible exception to this trend is for those regions of the world where Islamic fundamentalism has taken hold as an ideology legitimizing war-making. Even in these cases however, the role of economic factors has arguably become more important. For instance, the end of the cold war has seen a precipitous decline in the volume of military aid supplied to the Taliban from not only the US but also Muslim states such as Saudi Arabia. In consequence, Afghanistan has now become the world's largest producer of heroin as the warlords there have sought out new avenues from which to finance both the war effort and personal enrichment (The Guardian, 1998c). In regions where Islamic fundamentalism has not taken hold, the importance of economic as opposed to ideological considerations is even more apparent. Groups such as the Khmer Rouge and UNITA have moved from a strong ideological agenda to one dominated by economic aims (Keen, 1998, p.34). In conflicts begun after the end of the cold war, as in Liberia and Mexico, the absence of a revolutionary ideology and the importance of economic motives is even more pronounced. In the latter case, the refusal of the Zapatistas to articulate a specific revolutionary program and their relative disinterest in acquiring state power has been one of the factors that has led to them being described as the first postmodem revolutionaries (Gray, 1997, p. 6). In the former case, Ellis (1995, p. 185) has noted how the emergence of autonomous warlord systems produced a mosaic of militia zones of control, where civilians have some degree
Warlords and logo warriors: the political economy ofpost-modern conflict
41
of protection but must pay tribute in kind to the local warlord, constantly shifting frontier zones in which civilians are liable to raiding from all sides. The aim is control of people and acquisition of booty more than to control territory in the conventional military manner.
In this context, the state and state borders become increasingly meaningless. The state in effect is replaced by multiple centers of authority dominated by local elites or warlords who may not even have a particular interest in capturing the reigns of government, except perhaps as a means to extend their commercial activities. To describe conflicts characterized by such features as internal is misleading as the very phrase pre-supposes there is a "state" over which, and inside which, combatants are fighting. For Duffield, a more appropriate term is "post-modem conflict," a phrase which takes account of "the emergence of political projects in the South ... which no longer seek or even need to establish territorial, bureaucratic, or consent-based political authority in the traditional sense," yet at the same time need to establish international linkages (1998, p. 76). It might also be noted that supposedly internal conflicts often drift over what are effectively meaningless borders. For instance, conflicts can follow the flow of refugees into a neighboring state. Indeed, for Kaldor, one of the defining features of what she terms "new wars" is the fact that the main method of territorial control is based not on the popular support necessary for state-building but on population displacement - or, more precisely, the displacement of those labeled as "other" (1999, p. 98). Conflicts can also move over borders as actors search for new resources to plunder or because insurgent groups obtain arms from or establish military bases in neighboring states. As already noted, for outside observers such mosaics of conflict often appear meaningless and irrational, particularly when they can be no longer slotted into a larger cold war power game. At worst, such wars are characterized as conflicts between "erratic primitives of shifting allegiances, habituated to violence, with no stake in civil order" (Peters, 1994, p. 20) or between groups "motivated less by 'professionalism' than by fanatical, ideologically-based loyalties" (van Creveld, 1991, p. 196). At best, contemporary conflicts in the developing world are seen as conflicts which, albeit explicable in terms of ethnic and religious fundamentalism, poverty, and human rights abuse, are nevertheless
42 Arming the South
dysfunctional, in the sense that they interrupt a benevolent process of economic development. In this perspective then, conflicts can be resolved when the various parties are persuaded to see reason, to recognize the positive-sum advantages of peace for all, and when the state is reconstituted and the status quo ante resurrected. What this ignores however is that for many actors in conflict, war is not only rational but is rooted in the power relations existing in the status quo ante. War also, needless to say, provides further opportunities for elites to continue, or to begin, exploiting control of local resources and links with the global economy to generate economic benefits for themselves. Moreover, these same systems of profit and power are precisely the systems that are drawn on to finance arms acquisition. For instance, Charles Taylor effectively plundered Liberia to fund his military campaigns, exporting timber, diamonds, and rubber. Indeed, during the early 1990s Taylor was France's third largest supplier of tropical hardwoods. These activities were estimated to have netted him an income of around $400 to $450 million per year with which to finance both weapons acquisition and personal enrichment (Weissman, 1997, p. 104). Similarly, UNITA in Angola has compensated for the loss of military aid from the US and South Africa by selling diamonds. This and other commercial operations produced an estimated income of roughly $1.5 billion between 1992 and 1996 (Duffield, 1998, p. 22). The organization's continued access to such resources (one EU report has estimated it controls 6 percent of the world diamond market) was a major factor permitting it to resume fighting after the 1992 elections, and it continues to hamper prospects for a permanent peace in the country (The Guardian, 1998b; Vines, 1995, p. 40; UN, 2000). As already noted, such war economies are maintained through links with both the legitimate and illegitimate global economy. Perhaps the ultimate paradigm for this relationship however, is the new corporate mercenary company. On the one hand, such companies reflect the move to a virtual state in the developed world, representing the privatization of Western military force projection and foreign policy via companies such as the US firm MPRI (which claims to only operate in areas approved by the US State Department), the British firm Sandline, or the South African/British/Canadian firm Executive Outcomes (EO).
Warlords and logo warriors: the political economy ofpost-modern conflict
43
MPRI for instance was an instrument of US policy in the former Yugoslavia where the US government encouraged the Croatian forces to hire it to provide military training. For some commentators the professionalization this brought to the Croatian forces, coupled with covert arms shipments to the anti-Serb factions were "instrumental" in bringing about a change in the balance of forces on the ground and thus forcing the Serbs to the negotiating table (Glitman, 1996, pp. 74-75). Similarly, as the arms to Sierra Leone scandal in the UK demonstrated, at least one branch of the British government was prepared to use the mercenary company Sandline to achieve British foreign policy obj ectives. Needless to say, the use of such companies also chimes in with the concern of post-heroic societies to avoid military casualties at all costs. Privatizing foreign and defense policy in this way allows Western governments to disavow responsibility for the casualties of either side as well as the military excesses of private companies who implement government policy but at arms length. Furthermore, just as the corporate mercenary company reflects the move to a virtual state in the West so it does in the developing world. Security, supposedly the core function of the state, is privatized out to mercenary corporations which facilitate the prosecution of war and/or guarantee the protection of elites and key economic assets. These same companies also tap into grey and black arms markets to provide not only highly trained combatants but also the hardware of conflict, or in the case of Executive Outcomes, even come equipped with their own mini-air force (although EO has announced that it has now closed down its operations). For actors with the resources to hire such groups, they can represent a quick and efficient form of force multiplication. This is particularly the case in many developing countries where soldiers are often ill-equipped and ill-trained, and where sophisticated arms often remain unused for want of spares. In return, the mercenary companies not only receive cash payments but also diamond, oil, or other forms of economic concessions. In effect, then, such companies not only provide security but serve, at one and the same time, as the corporate partners of local elites and also the advance guard of a global economy through which the resources of a region can be translated not into state power but into personal enrichment. In Sierra Leone, for instance, Executive
44 A rn 11 ng th e So u th
Outcomes was not only paid in cash but one of its partner companies, Branch Energy, acquired concessions to the Kono diamond mines (Shearer, 1998, p. 52). Some commentators have credited EO with bringing about the defeat of the RUF and thus temporary peace in Sierra Leone. However, just six months after the signing of a peace agreement, the Sierra Leonian military and the RUF actually joined forces to overthrow the new government as peace threatened the economic interests of both parties. To quote Berdal and Keen (1997, p. 802): Unfortunately, the new democratic government of Tejan Kabbah was a threat to the system of exploitation that the RUF and elements of the military had conspired to create. Peace and democracy threatened the military's involvement in illicit diamond mining and both the military's and the rebels' practice of looting civilians and taxing their production.
Subsequently, Kabbah's government has been restored with the aid of outside powers and a peace agreement has been signed with the RUF. Notably, part of the UN-brokered peace deal gives Foday Sankoh, the RUF's leader, responsibility for oversight of the country's gold and diamond industry (The Guardian 1999; 2000). Notwithstanding this, however, the prospects for long-term peace in Sierra Leone are not good. Whilst the corporate mercenary companies may represent the ultimate expression of the relationship between local elites in weak states and the globalized economy, the same symbiotic relationship exists between highly legitimate civil companies and such elites. In Colombia, BP pays the government a $1.25/barrel war tax and has reportedly signed an agreement to provide an additional £39 million to establish anew military squad (The Observer, 1996; The Guardian, 1997); in Mozambique, the same company was obliged to pay what was effectively protection money to members of Frelimo, whilst Renamo obtained regular payments from a subsidiary of Lonhro, for protecting the Beira oil pipeline (Keen, 1998, p. 16); De Beers has bought up the diamonds traded by UNITA; and in the Congo Laurent Kabila reportedly received £30 million from a consortium including South African and Namibian businesses as well as the Namibian government to finance his military campaign against the rebels threatening to overthrow his new government (The Guardian, 1998a).
Warlords and logo warriors: the political economy of post-modern conflict
45
Other linkages to the global economy can also be exploited by actors in conflict as a means of obtaining funding for arms and the perpetuation of war. Remittances from ethnic or religious kin abroad can be an important source of revenue. Thus, remittances from Eritrean exiles abroad reportedly brought in $20 million a month for the EPLF (Clapham, 1996, p. 223); the LTTTE (the Tamil Tigers) have levied a revolutionary tax on its overseas diaspora; and the Kosovo Albanians have financed arms purchases and training from mercenaries through a combination of remittances from ethnic kin in Europe and the profits from a variety of black market activities (Judah, 2000). Needless to say, the transfer of such remittances has been made easier by the communications revolution that allows money to be moved from one continent to another at the touch of a button. Local elites or warlords, and of course their trading partners in the global economy, may well be the primary economic beneficiaries from post-modem conflict but economic considerations are also an important factor in determining the attitude of ordinary soldiers to the issue of war and peace. As Keen has noted, in both Liberia and Rwanda, genocide seems to have offered opportunities not only for looting but also for revenge against those perceived to enjoy unfair economic privileges, whilst in Kosovo a survey of demobilised KLA found that 36 percent were unemployed prior to joining the organization (International Organization for Migration, 1999). Similarly, recruitment to mercenary forces has been aided by the global downsizing of military forces that has accompanied the end of the cold war which has released a surplus of former soldiers onto the market. Once conflict has taken hold and war economies have become established, the economic incentives for becoming or remaining a soldier can be even more considerable. In economies where formal employment opportunities are limited, membership of an armed faction can provide regular pay as well as opportunities for extortion and banditry against the local population. As Berdal has noted, in such a context, "weapons always have an economic as well as a security value" (1996, p. 17). Indeed, economic motivations can be such that even trade with the enemy is not discounted. Government troops in Sierra Leone tended to avoid pitched battles with the RUF and instead preferred to sell them arms and
46 Arming the South
ammunition; during the Bosnian conflict, paramilitaries defending Mount Igman, overlooking Sarejevo, were prepared to "sell" their positions in order to control black market routes (Kaldor, 1999, p. 51); in Angola trade between both sides was so common that during a siege in Cuito, UNITA troops actually set up makeshift breweries to produce liquor to sell to government troops (Vines, 1995, p. 14). Moreover, as Kaldor notes, the symbols of global materialist culture - Rayban sunglasses, Adidas shoes, jogging suits, and caps are often adopted by paramilitaries in these new conflicts as a kind of quasiuniform. Ironically then, actors in post-modem conflicts also plug into a version of the culture of glitz and shopping which contributes to peace in the developed world, albeit a distorted one which underpins the status of these logo warriors. The economic benefits to be derived from ownership of a gun can sometimes manifest themselves in surprising ways. For instance, one of the few formal employment opportunities during the crises in Somalia was created by the aid agencies who used armed guards, often drawn from one or other of the competing clans, in order to protect food conveys. At the height of the crises the ICRC was reported to have had 15,000 to 20,000 armed guards on its staff, a phenomenon that created increased demand for weapons, ammunition, and soldiers (Natsios, 1996, p. 77). Indeed, as the latter example demonstrates, actors in conflict have become adept at manipulating the political economy of aid in order to gain economic rewards or to finance military campaigns. For example, refugee camps established by aid organizations can also become a source of funds, as they did in Zaire where the Hutu militia levied a 15 percent "war tax" on all rations received by those in the camps (de Waal, 1997, p. 205). It has also become a commonplace of aid politics that military factions will only allow emergency relief to the desperate and hungry on condition that a certain proportion is diverted for their own use. For instance, in Somalia, this relationship was formalized to the extent that relief organizations agreed to the diversion of food to the warlords as a quid pro quo for their own protection - the warlords then used the food to purchase more weapons (Natsios, 1996, p. 77).
Warlords and logo warriors: the political economy ofpost-modern conflict
47
Conclusion To conclude, the arms diffusion process in post-modem conflicts is influenced in various ways by the "triple whammy" effect of post-modem peace, post-modem weapons, and post-modem conflict. Post-modem peace has brought large-scale reductions in the level of military aid and global military forces. In consequence, whilst military aid from allies still remains an important source of arms for actors in conflict, it is also the case that commercial acquisitions of either arms or mercenaries from the legitimate or black market have become a more salient feature of the arms diffusion process in post-modem conflicts. Consequently, actors in conflict are increasingly exploiting global networks in order to raise funding, either through the sale of local resources on the international market, through remittances from the diaspora, or through the exploitation of international aid. Such activities have in turn been made easier by the conditions of post-modemity: for instance, the communications and IT revolutions, the globalization of markets, and the spread of transport links. At the same time, the ever rising costs of sustaining post-modem armies with high-tech weapons has combined with the cut-backs in military expenditure that have flowed from post-modem peace to create a buyers' market for arms which has lowered prices and encouraged sellers to offer innovative approaches to the financing of arms acquisitions, for instance via export credits or offsets. Moreover, postmodem peace has not, ironically, led to a more ethical approach to arms sales. Indeed, in many ways, arms exporters have, in general, become less rigid in their attitude to potential buyers as cold war constraints on exporting to former ideological enemies have been removed. Finally, post-modem conflicts increasingly occur in weak states where the agenda of local elites has conspired with that of external agencies to create a virtual state. One of the defining features of such states is, as noted above, the exploitation of local resources and global markets to fund personal enrichment and/or military campaigns. This has combined with the end of ideology and the decline of military aid to make the economic motives for conflict even more salient and certainly more transparent than in the past. What goes for local elites goes for the
48 Arming the South
ordinary soldier, too. In the war economies of conflict (and also postconflict) societies, where formal job opportunities are limited, ownership of a gun, the killing skills of the soldier, and links to the informal economy represent a form of comparative advantage rarely discussed by analysts at the IMF. The importance of this analysis is not just that it offers a better understanding of the dynamics of contemporary conflicts but that it also has important implications for the stmcturing of conflict resolution and post-conflict demilitarization strategies. As Brauer has noted, "an economic understanding of the perceived costs and benefits that drives individual actors can make worthwhile contributions to the analysis of force and perhaps help in the design of appropriate forms of ... intervention" (Brauer, 1999, p. 144). In particular, pre- and post-conflict intervention strategies need to recognize that many actors have a vested interest in either resurrecting conflict or in adapting the economy of war to post-conflict conditions in a manner that perpetuates lawlessness and violence. Failure to address these interests and to set in place an alternative system of carrots to reward peace and sticks to deter violence, is likely to make peace agreements both less likely and less durable. References Berdal, Mats R. "Disarmament and Demobilisation after Civil Wars." Adelphi Paper, No. 303. Oxford: Oxford University Press for the International Institute for Strategic Studies, 1996. Berdal, Mats and David Keen. "Violence and Economic Agendas in Civil Wars: Some Policy Implications." Millennium'. Journal of International Studies Vol. 26, No. 3 (1997), pp. 795-818. Brauer, Jurgen. "An Economic Perspective on Mercenaries, Military Companies, and the Privatisation of Force." Cambridge Review of International Affairs Vol. 13, No. 1 (Autumn/Winter 1999), pp. 130-146. Bradbury, Malcolm. "What Was Post-modernism? The Arts in and after the Cold War." International Affairs Vol. 71, No. 4 (1995), pp. 763774. Clapham, Christopher. Africa and the International System: The Politics
Warlords and logo warriors: the political economy ofpost-modern conflict
49
of State Survival. Cambridge: Cambridge University Press, 1996. Coker, Christopher. "Post-modemity and the End of the Cold War: Has War been Disinvented?" Review of International Studies Vol. 18, No. 3 (July 1992), pp. 189-198. Cooper, Robert. The Post-Modern State and the World Order. London: DEMOS, 1996. Duffield, Mark. "Post-Modem Conflict: Warlords, Post-Adjustment States and Private Protection." Journal of Civil Wars Vol. 1, No. 1 (April 1998), pp. 65-102. Ellis, Stephen. "Liberia 1989-1994: A Study of Ethnic and Spiritual Violence." African Affairs Vol. 94, No. 375 (1995), pp. 165-197. Glitman, Maynard. "US Policy in Bosnia: Rethinking a Flawed Approach." Survival Vol. 38, No. 4 (Winter 1996), pp. 66-82. Gray, Chris Hables. Postmodern War: The New Politics of Conflict. London: Routledge, 1997. The Guardian. Letter from Richard Brenner, Coalition Against BP, 9 May 1997. The Guardian. "Kabilla Rescue Squad Swells." 28 August 1998a. The Guardian. "Angola Rebels Provoke Crisis in Diamond Trade." 3 August 1998b. The Guardian. "Aid Work in the World's Poorest Country loses in America's War on the Terrorists." 1 September 1998c. The Guardian. "I will wait. Then I will kill him." 6 August 1999. The Guardian. "Rebels make the Rules in Ravaged Sierra Leone." 29 March 2000. International Organization for Migration. Socio-Economic and Demographic Profiles of Former KLA Combatants Registered by IOM. Geneva: IOM, 1999. Judah, Tim. Kosovo: War and Revenge. New Haven and London: Yale University Press, 2000. Kaldor, Mary. New and Old Wars: Organised Violence in a Global Era. Cambridge: Polity Press, 1999. Keen, David. "The Economic Functions of Violence in Civil Wars." Adelphi Paper, No. 320. Oxford: Oxford University Press for the International Institute for Strategic Studies, 1998. Luttwak, Edward N. "Toward Post-Heroic Warfare." Foreign Affairs
50 Arming the South
Vol. 74, No. 3 (1995), pp. 109-122. Luttwak, Edward N. "A Post-Heroic Military Policy." Foreign Affairs Vol. 75, No. 4(1996), pp. 33-44. Luckham, Robin. "Of Arms and Culture." Current Research on Peace and Violence Vol. 2, No. 1 (1984), pp. 1-64. The Observer. "BP does a U-tum on Rights Abuse." 10 November 1996. Peters, Ralph. "The New Warrior Class." Parameters Vol. 24, No. 2 (Summer 1994), pp.16-26. Natsios, Andrew S. "Humanitarian Relief Intervention in Somalia: The Economics of Chaos." International Peacekeeping Vol. 3, No. 1 (Spring 1996), pp. 68-91. Rathmell, Andrew. "Cyber-terrorism: The Shape of Future Conflict?" RUSI Journal Vol. 142, No. 5 (1997), pp. 40-45. Reno, William. "African Weak States and Commercial Alliances." African Affairs 96 (1997), pp. 165-185. Shearer, David. "Private Armies and Military Intervention." Adelphi Paper No. 316. Oxford: Oxford University Press for the International Institute for Strategic Studies, 1998. de Waal, Alex. Famine Crimes: Politics and the Disaster Relief Industry in Africa. Oxford, Bloomington, and Indianapolis: African Rights and The International African Institute in association with James Currey and Indiana University Press, 1997. Weissman, Fabrice. "Liberia: Can Relief Organisations Cope With the Warlords," in Medecins Sans Frontieres, World in Crises: The Politics Of Survival at the End of the Twentieth Century. London and New York: Routledge, 1997. van Creveld, Martin. The Transformation of War. New York, London: The Free Press, 1991. Vines, Alex. "Angola and Mozambique: The Aftermath of Conflict." Conflict Studies No. 280. London: Research Institute for the Study of Conflict and Terrorism, May/June 1995. United Nations. "Report of the Panel of Experts on Violations of Security Council Sanctions Against UNITA." New York: UN, 10 March 2000. See http://www.un.org/News/dh/latest/angolareport_eng.htm
3 Military Expenditure and Development in Latin America Thomas Scheetz
Introduction Often, there is an unwritten presupposition present in econometric studies dealing with the relation between a country's military expenditure and its economic growth. When military expenditure is found to have a negative effect on the economy, as I find in my experience in Latin America, it is taken as confirmation that the military per se are always deleterious to growth. Mutatis mutandis, the same could be said for researchers who have found a positive relation between military expenditure and the economy. In this chapter, I offer a nuanced interpretation of research findings, one that relates to the correctness of a government's policy. I suggest that if models are appropriately specified, military expenditure's positive (or negative) correlation with growth depends on policy-makers' "getting defense right" (or wrong) in terms of the military's mission and the resultant force deployment. These are what really cause a positive or negative effect of military spending on the economy. Using an inter-disciplinary approach (neoclassical economics, joined with political and military theory), this chapter postulates that the negative impact of military expenditure on Latin American development (even after democratic transitions) is due to incorrect defense and military policies by the respective governments, and not to anything inherently negative about military spending in itself1 The results from several of my own research projects dealing with the
52 Arming the South
fiscal and macroeconomic impact of military expenditure in six Latin America countries from 1969 to the present all point in the same general direction. The econometric evidence (e.g., Marwah, Klein, Scheetz, 2000; Scheetz, 1995; and Scheetz, 1991) indicates that military expenditure has a negative impact on GDP growth and on several other macroeconomic variables. Moreover, this negative impact on growth continued (although more study is required) even after the military left power. Fiscal analysis of crowd-out (e.g., Scheetz, 2000; Scheetz, 1992a; Scheetz, 1992b; Scheetz, 1992c; Scheetz, 1989; Scheetz, 1987) shows that while there occurred a displacement of social expenditures (or public sector investment) by military expenditure during the 1970s and part of the 1980s, that trade-off diminished notably following each country's transition to democratic rule. Furthermore, the defense burden (the ratio of military expenditure to GDP) has been declining over the past two decades. Since the onset of democratic transitions were accompanied by declining defense burdens 2 and an increasing fiscal share of social expenditure, one might have assumed that defense expenditures would now show a more benign effect on growth. Yet one does not find that (see, e.g., the chapter by Marwah, Klein, and Scheetz in this volume). In the next sections, I briefly summarize the results of research to date, followed by a discussion of what has changed since the military left power in Latin American countries. Then I interpret the negative impact of military expenditure on GDP in light of neoclassical economic theory. Finally, for medium-size states, like most of those in Latin America, I present a military-reform solution to the defense versus development dilemma. Research results to date Econometric
evidence
Using pooled or individual country time-series data, I studied the impact of military expenditure on growth (see papers listed in the introduction) for Guatemala (1968-1994), Argentina, Chile, Paraguay, and Pern (19691987), and Argentina and Chile (1969-1990). The results all show that military expenditure has had a significantly negative effect on GDP
Military expenditure and development in Latin America
53
growth and other variables, irrespective of the presence or absence of military governments. The most convincing is the model on Guatemala, published as chapter 15 in this volume (Marwah, Klein, Scheetz, 2001). It consists of nine equations and three identities and models the Guatemalan economy both from the demand and supply sides, long and short-run, for the entire period (1969-1994) and for the period of democratic transition. We find that even after Guatemala's transition to democratic mle in 1986, the economic impact of its military expenditure remains strongly negative. In Scheetz (1991) I employed a Deger-type model (Deger, 1986). Adapting Deger's four-equation model to a pooled data set (1969-1987) for Argentina, Chile, Paraguay, and Peru, I found a negative multiplier effect of the defense burden on real GDP growth rate (-0.57), on net domestic savings rate (-0.75), as well as on the current account balance as a share of GDP (-0.50). And in Scheetz (1995), I repeated an adaptation of the Deger model, but this time pooling only the two arms racing countries of Argentina and Chile (1969-1990). The results were similar to the earlier model, finding a negative multiplier effect of the defense burden on the real GDP growth rate (-0.43), on the gross national savings rate (-1.1), and on the current account balance as a share of GDP (-0.53). Fiscal data and military expenditure
crowd-out
The fiscal accounts listed in tables 3.1 and 3.2 are the product of local data gathering in each of these countries. 3 1 examined and rejected the use of international data sources as being inadequate for time series and cross country analysis in these countries (see Scheetz, 1991). In each country the entire central government executed budget was studied in situ in the agency responsible for government accounting, using the basic methodological definitions as found in the International Monetary Fund's Manual on Government Finance Statistics. Since I was concentrating on the cost of the defense function, I moved such expenditure as military retirement, health, and education programs to the defense function. Military expenditure definitions were held constant over time and across countries. 4 For my purposes crowd-out occurred when military spending
54 Arming the South
Table 3.1: Crowding-out of social and investment spending by military spending in Latin America, 1969-1994 Percentage of years with crowd-out during period listed Argentina
Bolivia
Chile Guatemala Paraguay
Peru
Total
Social expenditure, 1970-1979
50.0
30.0
40.0
30.0
30.0
10.0
31.7
Central government investment, 1970-1979
40.0
10.0
30.0
20.0
30.0
40.0
28.3
Social expenditure, 1980-
7.1
27.3
9.1
26.7
36.4
16.7
20.2
Central government investment, 1980-
21.4
27.3
18.2
33.3
36.4
46.2
28.0
0.0
25.0
0.0
22.2
0.0
18.2
14.0
10.0
37.5
0.0
11.1
0.0
41.7
22.7
Social expenditure postdemocratic transition Central government mvestmeni postdemocratic transition
Notes: Data coverage: Argentina, 1970-1993; Bolivia, 1970-1990; Chile, 1970-1991; Guatemala, 1969-1994; Paraguay, 1970-1991; Peru, 1969-1992. Only the period 19701979 is considered in the first two rows. 1980 through the end of data information is presented in rows three and four. Argentina is a federally organized country. Thus significant social and investment spending is done by the provinces, especially after 1977. Data taken from publications cited in note three, though these have been updated in case of Argentina, Chile, Paraguay, and Peru. Military expenditure is considered to crowd-out the listed expenditures if it grows as a percentage of government outlays, while other expenditure falls. This does not take into consideration the size of the crowd-out. Data for post-democratic transition: Argentina 1984-; Bolivia 1983-; Chile 1990-; Guatemala 1986-; Paraguay 1989-; Peru 1981-. Obviously earlier democratic governments existed in these countries. I use date for the latest wave of democratic ti'ansitions. Democratic transition actually passes through stages, with greater or lesser civilian control over military budgetary demands. The transition year used is that when incoming democratic government wrote the budget. The "total" column is the total number of country expenditure data points divided by the total number of country years. Sources: All data developed from official sources in each country listed.
Military expenditure and development in Latin America
55
Table 3.2: The military burden in Latin America (percentage of military expenditure in GDP)
Argentina Bolivia Chile Guatemala Paraguay Peru Simple averages
Average 1970-1979
Average 1980-
Average Post-transition
2.84 3.09 6.00 1.21 2.86 4.72 3.45
2.13 3.51 6.36 1.70 2.91 3.12 3.29
1.75 2.63 4.94 1.56 2.60 3.01 2.75
Notes: The table uses my definition of military expenditure (see text). All averages are simple, unweighted averages. Column 4 is the average for the most recent transition to civilian governments. Sources: All data developed from official sources in each country.
increased its percentage share of central government spending and the share of social outlays (education, health, and other social expenditures) or investment declined.5 Table 3.1 presents a condensed version of time-series results for military expenditure's crowding out, or displacement, of social and public investment expenditure for the six countries studied. The table groups the summarized data for the 1970s, 1980s (and 1990s where we have data), and for the period of the latest transition to democratic rule.6 This permits an examination of the overall country and regional evolution of the relation between military expenditure and social or public real investment over time. It also allows a rough comparison between military governments' and civilian transition governments' priorities in these three areas. Thus, table 3.1 shows that in both Argentina and Chile military expenditure displaced both social and investment outlays in declining fashion over time. Moreover, the transition to civilian rule brought with it even fewer occurrences of military expenditure displacing social or investment expenditures. The other countries in the survey (especially
56 Arming the South
Bolivia and Paraguay) suffer increases in displacement in the 1980s, but the overall tendency among the six countries is quite clear. Crowd-out of social and investment by military expenditure is declining over time and a result of the latest wave of democracy engulfing all of Latin America. The defense burden The evolution of the defense burden for the six countries is presented in table 3.2. Just as in table 3.1, the results show that while several countries spent more heavily on defense as a percentage of GDP during the 1980s, overall the defense burden fell during the 1980s and fell even further during the post-transitional period. While the defense burden says little about the "real unit of account" size of military expenditure, it is an indication of Latin American militaries' declining political power as a lobby, their fruitless search for significant missions or roles after the end of the cold war, and the priority given to other fiscal functions by civilian governments. Moreover, this change of priorities grew even stronger to the degree that these democracies began to break away from military tutelage. 7 The changed political context of the 1990s Viewed from the Latin American context, what has changed over the past decade with regard to the military expenditure - development debate? First, for the first time in history all countries of Latin America have passed to democratic regimes. Nonetheless, this is not to say that all Latin American governments exercise "objective control" 8 over their armed forces. Actually, only Argentina has evolved very far in significant civilian control over the armed forces, and even here it cannot be said that "objective control" over the military exists. While President Carlos Menem (1989-1999) maintained tranquility in the barracks, he never developed a defense policy that made the armed forces a useful tool for state policy. Although Argentine armed forces did participate in a significant way in recent international peacekeeping efforts, improving their professionalism in the process, the lack of serious military reform
Military expenditure and development in Latin America
57
nonetheless meant - ironically stated - that the soldiers did little more than collect "unemployment benefits" since they lack almost any operational capacity to defend the country. The incompleteness of civilian institutional control over the military is seen in the presence of diminished, though still extant, "crowd-out" during the 1980s. Second, Latin American countries have become economically far more open than they ever were. This has brought renewed growth up until the onset of the Asian financial crisis of 1997. It also implied that the markets' reaction to fiscal functional variables (such as defense) are now very definitely part of country risk analysis.9 Thus, military indebtedness for arms acquisitions, and fiscal subsidies to local defense industries, had to be controlled. This latter brought about the "conversion" (though better described as collapse) of local defense industries. These were largely sold off to the highest bidder, with immense losses over book and real value. Where possible, their assets were converted to civilian use (Scheetz, 1996).10 Third, the decline in Argentina's defense burden was triggered by a number of factors. First and foremost was its defeat in the Malvinas/Falkland War. The loss initiated the transition to democracy and human rights demands against the military. The military caste suffered the consequences, and part of this was a large cut in military expenditure. Second, by the time of the 1982 debt crisis the country (as was most of Latin America) was immersed in foreign debt, a very large share of which was due to arms imports. Debt refinancing and repayment meant that the principal fiscal functional expenditure was no longer defense, but interest payments. Consequently, all other fiscal functions were displaced, including defense. Third, with the transition to democracy there occurred an increase in civilian-oriented aspects of the budget, implying a decrease in the defense share. Fourth, with the end of the cold war the militaries' roles and missions simply disappeared, with no replacements to date that could be "sold" to voters and taxpayers. Finally, political actors were divided into three groups vis-a-vis the formulation of defense policy. There remained a few die-hard pro-militarists. And there were those who sought economic and political advantage against the weakened armed forces (these were often associated with corruption). Moreover, many supposedly progressive
58 Arming the South
civilian politicians took a position of benign neglect toward anything military, perhaps dreaming that the armed forces would simply wither away. Obviously this position is very naive. Non-attention to the problem of the armed forces will generate serious and unavoidable consequences. Argentina's military budget suffered severe declines for all of these reasons. And if indeed the description is made for Argentina, the factors apply to most of the countries of the region. What is surprising is that while the military burden declined, but we might have expected an improvement in the macro variables, none seems to have occurred. Moreover, military outlays have themselves become internally unbalanced. Salaries and retirement benefits have become a very large part of total military expenditure, with arms acquisitions and operational capacity sharply declining.11 The neoclassical vision and a hypothesis by way of explanation In light of the above description of a declining defense burden throughout the 1980s and 1990s, and given that for Argentina, Chile, and - most clearly - Guatemala the military expenditure variable continued in the 1980s and 1990s to show a negative sign in its relation to GDP growth, investment, current account balance, and other economic variables, what economic and military conclusions can be drawn from such evidence? Assuming that the results may be generalized, then one must either conclude that for these countries (or perhaps for developing countries in general) military expenditure and the presence of armed forces always have a negative effect on the economy, or one must conclude that these countries "simply haven't gotten defense right." That is to say, defense is not organized to form part of a general equilibrium system involving all other variables in the economy. These two opposing conclusions are the only ones logically viable. If one rules out a pacifist stance (associated with the first alternative), and assumes that at least some developing countries (I will explain later on just which ones) possess interests that could potentially benefit from the existence of armed forces, then the continued negative sign on the military expenditure coefficient implies that military reform is required in these countries.
Military expenditure and development in Latin America
59
Though many authors assume the contrary, defense should be a normal good. That is to say, military expenditure (and military expenditure/GDP) should fall with decreasing GDP and vice versa. Yet the consistently negative sign of the military expenditure coefficient suggests the opposite conclusion: as GDP falls, military spending rises. This would seem to indicate that defense is an inferior good. Presumably, the IMF (1986, pp. 183, 191) considers defense an unproductive service, since its capital stock (arms) is not considered an investment and, as such, has no internal rate of return. And Hartley (1991, pp. 12-13) counterposes defense spending to investment and growth, thus assuming that military expenditure always has a negative effect on GDP. He diagrams "the classic guns-versus-butter 'trade-off," modeling "the belief that defence is a burden on the economy." But other authors take a position opposed to this "defense burden as inferior good" view of military expenditure. Barro (1981, pp. 1090-91) writes that "in areas like the provision of a legal system and national defense, the public services are likely to enhance the marginal product of private factors." Moreover, from a public goods point of view he states that "even in the case of national defense, the benefits to individuals are likely to be relative to the total amount of property that is being defended, because the level of external threat would respond to the potential prize from conquest." He adds, "it is assumed that public services of this type have a positive marginal product" (p. 1092). Defense protects property rights against external risks, just as fire and police departments protect them against internal risks, thereby providing insurance to the investor so that the marginal product of capital might be obtained. This public good (not public burden) acts as a deterrent (Kennedy, 1983, p. 25), as an insurance policy, though of course not against all risk.12 Probably all these authors would agree with Devarajan et al. (1993, p. 10): "an increase in total government spending, since it has to be financed by taxes, will raise the steady-state growth rate only if the productivity of that government spending ... exceeds the taxes required to pay for it." From our point of view this is the crux of the issue. One must distinguish between what ideally could be the effect of defense spending versus that which de facto is its effect during a given period. Diamond (1989, p. 5) characterizes the problem from the
60 Arming the South view that centralized decision-making, a lack of profit motive, and the absence of competition - all of which typify government operations - imply that government production is always less efficient than private sector production. Assuming this lower productivity, any increase in government expenditure, by increasing the share of productive resources used by government, would slow economic growth in the economy as a whole ...
The theory of defense spending should therefore separate what should be military expenditure's productive role from what, in effect, has been historically its developmental role. We note the gap between reality and neoclassical theory, where all government spending (defense included) should follow the golden rule of Marshallian equilibrium (3.1)
MP def / pdet. = MPeduc / peduc = MPX / p x ,
where MP stands for marginal product, with simultaneous equilibria in factor markets and utilities. It is only when this equilibrium condition is met that defense spending can exercise its correct growth-supporting role in development. What the econometric analyses suggest is that military expenditure in many countries is a disequilibrium rather than a defense "burden"; they thus challenge our prejudices of defining the ratio of military expenditure to GDP as such a burden.13 Most frequently the marginal most (MC) of defense is greater than its marginal benefit (MB), that is to say, to the right of point E in the standard graph modeling the ideally correct level of public sector spending (see figure 3.1). As drawn, the figure states that marginal cost and marginal benefit are in equilibrium somewhere below the level where total protection against risk is provided. As Devarajan et al. stated, there exists an ideal relation between the distinct values to which each tax dollar is put. Neoclassical thought calls this the marginal utility of money (public or private). What then can be concluded if the empirical results of our research contradict the theory? Is defense no longer a public good, but rather a public bad? No, rather the results are telling us that defense policy is incorrectly formulated for the country, a not uncommon situation. This generally occurs because neoclassical rationality is not the criterion used for determining military expenditure. Nevertheless, as such it stands as
Military expenditure and development in Latin America
Price/ cost
0 percent protection against any risk
61
Price/
100 percent protection against all risk
Figure 3.1: Marginal cost and benefit of military expenditure
a severe critic of operative defense policy. Actual policy is based on institutional rationality, e.g., as described by Douglass North (1981, chapter 3) and Barbara Geddes (1994, chapter 1). According to Geddes, most Latin American countries like Argentina are de facto presidentialist, even if their constitutions present a system of checks and balances among the three powers of government. Such systems are based on clientelistic foundations and on political rentseeking. An essential element in the Argentine political system is the relation to the political boss or caudillo. For example, clientelism is clearly noted in the electoral lists of candidates for office. The list of candidates is designated internally by the party heads. Thus the elected legislators respond to party bosses and not to the interests of voters. Political parties capture state institutions as though they were booty of political war, with certain rights acquired by the electorally victorious party (including corrupt practices).14 And it is a tacit but well understood
62 Arming the South
rule that the caudillo has the right to place friends and allies without consideration of merit or ability, and only on the basis of political ties, in any administrative post under his control. Any person, or the critic, not well-integrated into the system is expelled. Voter interests are taken into consideration only when there are incentives favoring the caudillo. A civil service on the English style does not exist, nor does a concern for the efficient functioning of institutions. For example, in the last 17 years of continuous democratic government in Argentina, not a single minister of defense has been named who has had any prior knowledge or experience whatsoever with defense.15 Indeed, frequently enough, the provision of Argentine external security is not the primary interest of the Minister of Defense. Military reform as the solution for medium-sized
countries
Having contrasted neoclassical rationality, which maximizes some social welfare function, with the dominant, rent-seeking, institutionalist rationality, what could be done to improve defense in countries like Argentina? Or, responding to Benoit's admonition cited at the beginning of this chapter, what type of defense policy decisions are necessary that would contribute to, rather than hinder, economic growth and development 9 Any defense reform would have to simultaneously meet criteria in three areas: military, diplomatic, and economic. Militarily, it would have to prove itself genuinely capable of defending the country. Diplomatically, the defense posture should not be perceived as a menace by other countries of the region (in order to avoid an arms race and to promote regional economic cooperation). And economically, the cost of defense should foster growth and permit a "correct" balance of public sector expenditure in the Marshallian sense. A distinction should be made between countries and their choice of a defense posture. There are three types of countries. First are those countries which can benefit from power projection (thus possessing strategically offensive capabilities) or which become the "enforcer of last resort" in a region.16 The example/?c/r excellence is obviously the United States. Second, there are those small countries like Costa Rica or Uruguay
Military expenditure and development in Latin America
63
which are incapable of defending themselves militarily. They simply do not possess the GDP, manpower, or strategic depth that would enable them to assume even a strategically defensive posture. The best policy for them is to rely on a highly capable diplomatic corps and avoid spending scarce budget resources on a military force, but this does not rule out the use of border or coast guards. The third type, and the most numerous in Latin America, is the medium-sized country. The phrase "medium-sized countries" is used in Hill's sense (1990, p. 28), namely: [GJiven a certain minimum of basic resources, territory, population, education and industrial development, that a state be included in the category of medium-sized depends fundamentally on the perception which it has about itself ... What characterizes the medium-sized power is its desire to satisfy for itself the requirements for maintaining its existence as an entity. It must be capable of unleashing the required actions, although other states or organizations may come to its aid at some moment. It must assure itself control as often as possible whenever its interests are under attack. [My translation.]
Now a fundamental difficulty arises. Most medium-sized Latin American countries today vainly attempt to copy elements of strategic posture, and consequent deployment, from power-projecting countries, especially from the United States.17 By strategic posture I do not mean their publicly declared intentions, but rather their actual capabilities as implied by their deployment, and as interpreted by their neighbors. Often this deployment is quite chaotic but contains some menacing capabilities perceived by neighbors as permitting a strategic offensive. Obviously defense costs vary according to the strategic posture adopted, the offensive usually also being the most expensive one. In contrast, the maintenance of defensive capacity can be achieved at a significantly lower cost than that of an offensive army. Moreover, a strategically offensive posture for countries like Argentina is militarily impossible to sustain,18 economically ruinous, and diplomatically destructive of regional stability. This chapter posits that it is this military policy which is to be blamed for the negative sign on the military expenditure coefficients in our models. Often enough the irrationality of these countries' defense policy is
64 Arming the South
only too evident. But instead of changing their strategic posture to a more viable one, they simply begin a process of mission creep. Additional roles such as civil defense, anti-drug trafficking, ecology, or international peacekeeping are not essential roles for armed forces. These roles cannot justify the costs of maintaining armed forces19 and can usually be more cheaply accomplished by institutions specializing in that role.20 Only a defensive strategic posture such as non-provocative defense21 is capable of meeting the three requirements (economic, military, and diplomatic) simultaneously. Non-provocative defense is defined as A military posture in which the strategic and operational concepts, the deployment, organization, armaments, communications and command, logistics and training of the armed forces are such, that they are in their totality unambiguously capable of an adequate conventional defence, but as unambiguously incapable of a border crossing attack, be it an invasion or a destructive strike at the opponents territory (ter Borg and Smit, 1989, p. 1).
It is suggested that this type of military reform stands the best chance of making the armed forces of Latin America positive factors in the development of their countries. We also postulate that such a reform would make the military expenditure coefficient positive in our econometric models. Notes 1. In light of my own findings of military expenditure's negative impact on Latin American countries, there is a certain irony that Emile Benoit whose work is taken as the original thesis of the positive effects of militaty expenditure - made the same point almost three decades ago: 'The composition of defense programs maybe as important for economic growth as their size; thus a closer integration of economic and defense planning maybe decidedly helpful" (1972, p. 2). 2. For example, the Guatemalan defense burden from 1986 (initiating democratic transition) through 1994 fell to an average of 1.56 percent of GDP. During the five previous years it had averaged 2.03 percent (Scheetz, 2000, p. 118).
Military expenditure and development in Latin America
65
3. In Scheetz (1991,1992a, and 1994) I show how my data sets are superior to international military expenditure sources. 4. Scheetz (1994) contains a complete discussion of the definition of military expenditure and other methodological problems. A brief version in English is in Scheetz (1998). 5. Although a very important issue, no account is taken of the size of growth or decline. Though this is a somewhat crude measure of crowdout, it is the best definition available. 6. Virtually all of Latin America is currently under civilian, democratic regimes, although the degree of civilian control over the armed forces varies widely. Note, too, that civilian governments also held power sporadically during the 1970s. 7. This tutelage is still strong in Guatemala, Chile, and Paraguay even after ten years of democratic governments. It implies that civilian governments are less than free to decide on military policy and budgetary issues. 8. "Objective" versus "subjective" control is a significant distinction made by Samuel Huntington (1957). The concepts are further clarified in Feaver (1996). Huntington defines "subjective civilian control" as "maximizing the power of some particular civilian group ..." (p.80). And "civilian control in the objective sense is the maximizing of military professionalism" (p.82). 9. This became evident to me while working on the Peruvian external debt problem in the early 1980s. For instance, the minutes of the IMF Executive Board meeting of 18 November 1977 make very clear various board members' concerns over Peru's purchase of then-Soviet arms. This occurred again at a 1983 Paris Club meeting on Peruvian debt. 10. Some military industries survived, most notably in Chile.
66 Arming the South
11. Many countries, such as Chile, have been planning to renew outdated arms stocks. If the economic situation improves, this may occur, thus running the risk of a renewed arms race in Latin America. 12. Knight, et al. (1996) also take a position similar to Barro's. Anderton and Anderton (1997, pp. 56, 58) employ an Edgeworth-box "model of conflict, production and of exchange [in which] ... the contract curve is the set of points 'along which the pleasure-forces of the contractors are mutually antagonistic' (Edgeworth, 1881, p. 29). The resolution of the mutual antagonism occurs through bargaining. The bargaining is a form of 'war' because the parties' interests are antagonistic." 13. In the same sense one should speak of the "health burden," the "education burden," the "police burden," and so on. 14. Among many examples, we might cite the illegal sale of arms by the Menem administration and the Argentine army to Croatia during the 1990s, arms which Argentine peacekeeping forces might have had to confront. While widely known within the military and the government, no one lifted his voice questioning the legality of the arms sales. 15. There have been ten defense ministers in those 17 years. This is not to deny the existence of some highly professional specialists, both civilian and military. The problem is that intra-party conflict (the interna in local jargon) tends to expel these professionals at the moment of their first criticism. In this context even suitable professionals must come to accept "certain realities" in order to keep their post. If they do not learn, sooner or later they resign. 16. No examples of such a country are in Latin America. 17. The developed countries sell both arms and symbolic constructs defining what genuine armies, navies, and air forces are really to look like. 18. Unterseher and Conetta (1994, p. 9) write that for incomplete strategically offensive deployments, such as exist in all Latin American
Military expenditure and development in Latin America
67
countries, "...when low force levels prevail, the offensive emphasis of threat-based defense postures produces the 'revolving door phenomenon,' in which cross-border aggression is met with cross-border retaliation." 19. Argentina has a US$4 billion budget for defense, of which only US$30 million are for peacekeeping. Clearly, taxpayers cannot be convinced of the centrality of the peacekeeping role, however valuable it may be. Some other mission (external defense) must be put forward, or else the scarce resources should be spent elsewhere. 20. Police forces (trained in the use of minimum force) are generally cheaper and more efficient than the military (who should be trained in the use of maximum force). A national Red Cross should be more efficient and cheaper in coordinating all segments of society in moments of disaster. 21. Or any of its various synonyms: non-offensive defense, alternative defense, defensive defense, defensive security, confidence-building defense, etc.(see M0ller, 1995, pp. 242-243; Unterseher and Conetta, 1994, pp. 5-6). References Anderton, Charles H. and Roxane A. Anderton. "The Economics of Conflict, Production, and Exchange," pp. 54-82 in Jurgen Brauer and William G. Gissy (eds.). Economics of Conflict and Peace. Aldershot, UK: Avebury Press, 1997. Barro, Robert J. "Output Effects of Government Purchases." Journal of Political Economy Vol. 89, No. 6 (1981), pp.1086-1121. Benoit, Emile. "Growth Effects of Defense in Developing Countries." International Development Review No\. 14, No. 1 (1972), pp. 2-10. Deger, Saadet. Military Expenditure in Third World Countries: The Economic Effects. London: Routledge & Kegan Paul, 1986. Devarajan, Shantayanan, Vinaya Swaroop, and Heng-fu Zou. "What do Governments Buy 9 The Composition of Public Spending and Economic Performance." Working Papers, Country Economics Department, World Bank, WPS 1082. Washington, DC: The World
68 Arming the South
Bank, February 1993. Diamond, J. "Government Expenditure and Economic Growth: An Empirical Investigation." IMF Working Paper, WP/89/45. Washington, DC: International Monetary Fund, 15 May 1989. Feaver, Peter D. "The Civil-Military Problematique: Huntington, Janowitz, and the Question of Civilian Control." Armed Forces and Society Vol. 23, No. 2 (Winter 1996), pp.144-178. Geddes, Barbara. Politician 's Dilemma: Building State Capacity in Latin America. Berkeley, CA: University of California Press, 1994. Hartley, Keith. The Economics of Defence Policy. London: Brassey's, 1991. Hill, J.R. Estrategia maritima para potencias medianas. Buenos Aires: Centro Naval, Instituto de Publicaciones Navales, 1990. [Originally published as Maritime Strategy for Medium Powers. Annapolis: Naval Institute Press.] Huntington, Samuel. The Soldier and the State: The Theory and Politics of Civil-Military Relations. Cambridge, MA: Harvard University Press, 1957. International Monetary Fund. Manual on Government Finance Statistics. Washington, DC, 1986. Kennedy, Gavin. Defense Economics. London: Duckworth & Co., 1983. Knight, Malcolm, Norman Loayza, and Delano Villanueva. "The Peace Dividend: Military Spending Cuts and Economic Growth." IMF Staff Papers Vol. 43, No. 1 (March 1996), pp. 1-37. Marwah, Kanta, Lawrence R. Klein, and Thomas Scheetz. "The MilitaryCivilian Trade-Off in Guatemala: An Econometric Analysis," chapter 15 in Jurgen Brauer and J. Paul Dunne (eds.). Arming the South: The Economics of Military Expenditure, Arms Production, and Arms Trade in Developing Countries. London: Palgrave, 2002. Moller, Bjorn. Dictionary of Alternative Defense. Boulder, CO: Lynne Rienner Publishers, 1995. North, Douglass. Structure and Change in Economic History. New York: W.W.Norton, 1981. Scheetz, Thomas. "Public Sector Expenditures and Financial Crisis in Chile." World Development Vol. 15, No. 8 (August 1987), pp. 10531075.
Military expenditure and development in Latin America
69
Scheetz, Thomas. "Costos de seguridad y gastos sociales dentro del sector publico paraguayo, 1969-1988." Working Document, BASEIS, Asuncion, Paraguay, October 1989. Scheetz, Thomas. "The Macroeconomic Impact of Defence Expenditures: Some Econometric Evidence for Argentina, Chile, Paraguay and Peru." Defence Economics Vol. 3, No. 1 (1991), pp. 65-81. Scheetz, Thomas. "The Evolution of Public Sector Expenditures: Changing Political Priorities in Argentina, Chile, Paraguay and Peru." Journal of Peace Research Vol. 29, No. 2 (1992a), pp. 175-190. Scheetz, Thomas. "Los gastos militares y sociales en Bolivia: 19691990." Working Document. La Paz, Bolivia, 1992b. Scheetz, Thomas. "Una vision fiscal y regional del gasto militar chileno: Analisis de los presupuestos ejecutados 1969-1991." Document prepared for the Chilean Congress under the auspices of the Programa de Asesoria Legislativa, Santiago, 1992c. Scheetz, Thomas. "Gastos militares en America del Sur," pp. 195-220 in Proliferacion de armamentos y medidas defomento de la confianza y la seguridad en America Latina. Centro Regional de las Naciones Unidas para la Paz, el Desarme y el Desarrollo en America Latina y el Caribe, Lima, Peru, 1994. Scheetz, Thomas. "Los costos economicos de la defensa en Argentina y Chile y el esbozo de una solucion." Revista de Ciencias Sociales [Universidad Nacional de Quilmes, Argentina] No. 3 (November 1995), pp. 157-174. Scheetz, Thomas. "A Peace Dividend in South America? Defense Conversion in Argentina and Chile," pp. 403-423 in Nils P. Gleditsch etal. (eds.). The Peace Dividend. Amsterdam: North-Holland, 1996. Scheetz, Thomas. "Transparency of Military Expenditures in Developing Countries." Editorial Archive on webpage: http://www.unibonn.de/milex, 1998. Scheetz, Thomas. "Military Expenditures and Guatemala's Peace Process," pp. 113-127 in Jorn Brommelhorster (ed.). Demystifying the Peace Dividend. Bonn International Center for Conversion. BadenBaden: Nomos Verlagsgesellschaft, 2000. ter Borg, Marlies and Wim Smit. "Non-provocative Defence, Conventional Stability and Reasonable Sufficiency," pp. 1-13 in
70 Arming the South
Marlies ter Borg and Wim A. Smit (eds.). Non-provocative defence as a principle of arms reduction, And its implications for assessing defence technologies. Amsterdam: Free University Press, 1989. Unterseher, Lutz, and Carl Conetta. Confidence-Building Defense: A Comprehensive Approach to Security and Stability in the New Era. Study group on Alternative Security Policy (Bonn) and Project on Defense Alternatives (Cambridge, MA), 1994.
4 Military Expenditure and Economic Development in Asia During the 1990s Geoff Harris
Introduction There is a striking diversity amongst the countries of Asia, more so than in any other region. Compare the wealth of the newly industrialized countries - Singapore, Taiwan, Hong Kong, and South Korea - and of Japan with the poverty of Laos and Nepal or compare the immensity of China and India with all other countries. In writing this regional survey, therefore, I have worked largely at the subregional level (South, East, and Southeast Asia), with occasional reference to individual countries. Several studies (Denoon, 1987; Harris, 1988; Soesastro, 1994) have reviewed the issues surrounding Asian and Asia-Pacific military expenditure up to the mid-1980s, and from the mid-1980s to the early 1990s, respectively. Their main findings maybe summarized as follows: 1
In the 1980s, Asia's military expenditure grew much more rapidly than that of the world as a whole. The principal reason for this was the rapid growth of most of the region's economies, which enabled increased military expenditure in real terms even if the share of GDP devoted to military expenditure was roughly constant.
72 Arming the South
2
Whilst war, or the perceived likelihood of war, was the main determinant of military expenditure, the size of a country's GNP what the country could afford - normally determined its broad magnitude. Domestic economic conditions influenced its level in any year and the size of government in the economy (central government expenditure, CGE, as a proportion of GNP) was positively related to military expenditure.
3
The general conclusion concerning the net effect of military expenditure on a country's growth rates - that it reduced economic growth via a strong negative impact on savings and investment - was found to apply to Asia. Whilst this was of no great importance to the newly industrialized countries (NICs), whose GNP per capita were very high, it was of great significance to low-income countries where the opportunity costs of military expenditure were much more obvious.
To this summary may be added information concerning the arms trade. Asian countries accounted for a modest proportion of world arms imports at the start of the 1980s, but these were growing rapidly, both in real terms and as a proportion of the (declining) global market. Whilst world arms imports fell by 33 percent in real terms between 1982 and 1990, Asian arms imports rose by 40 percent (SIPRI, 1992, p. 308). Asian countries took 33 percent of all arms imports in 1990 compared with 15.5 percent in 1982 (SIPRI, 1993, p. 476), and six Asian countries were included in the top 15 arms importers between 1988 and 1992, with India being the largest importer worldwide (SIPRI, 1993, p. 445). This chapter begins with an examination of trends in Asia's military expenditure during the 1990s and differing interpretations of these trends. It then considers the main determinants of military expenditure, paying particular attention to the underlying motivating factors and to the relation of military expenditure to economic growth and development. Finally, in light of the changed natured of security, a number of costeffective alternatives to the military are discussed. Attention is drawn to important areas for future research.
Military expenditure and economic development in Asia
73
Trends in the 1990s In the decade following the end of the cold war, i.e., 1989-1998, and due to severe economic constraints, world military expenditure fell by about one-third in real terms to $745 billion at current prices or $696 billion in 1995 prices (SIPRI, 1999, p. 270). Most of this was due to falls in Russia's military expenditure, which in 1998 amounted, in real terms, to less than 10 percent of the USSR figure for 1989. Falling military expenditure was, nonetheless, a worldwide phenomena, except for the Middle East, North Africa (almost entirely due to Algeria), and Asia. In real terms, Asian military expenditure rose from $106 billion in 1990 to $130 billion in 1998, a 22.6 percent increase overall. South Asia's military expenditure grew by 27.3 percent over this period, whilst East and Southeast Asia increased by 22.1 percent. Against a background of falling world military expenditure, Asia's share of world military expenditure rose from 10.6 percent in 1990 to 18.7 percent in 1998. The most logical explanation is that Asia's relatively rapid economic growth allowed this continued expansion. Between 1975 and 1995, when GNP per capita for all developing countries grew at an annual rate of 2.3 percent, the figure for East Asia was 7.3 percent, for Southeast Asia and the Pacific 4.4 percent, and South Asia 1.4 percent (UNDP, 1999, p. 183). Military expenditure growth in South Asia was sustained by the IndianPakistani arms race based on the Kashmir conflict, exacerbated by nuclear testing by both countries in 1998. Table 4.1 presents data on military expenditure, by country, for the 1990s. As regards imports of conventional weapons, SIPRIs imputed trendindicator values show that Asia again went against world trends, rising by 8.6 percent in real terms between 1989-1992 and 1995-1998 whilst world arms imports fell by 26.1 percent (SIPRI, 1999, p. 450). Asia took 28.5 percent of world arms imports in the former period but this rose to 41.9 percent in the latter period. Between these periods, there were big decreases in imports by India (66 percent), Japan (55 percent), and Pakistan (25 percent), and a 496 percent increase by Taiwan which resulted in it becoming the world's largest arms importer (SIPRI, 1999, p. 428). For the latter period, eight of the 15 largest arms importers were
74 Arming the South
Table 4.1: Asian military expenditure trends during the 1990s Milex average 1990-1992 (US$mnl995)
Milex/ GDP 1990
Milex/ CGE 1990
Milex average 1996-1998 (US$mnl995)
Milex/ GDP 1997
Milex/ CGE 1997
South Asia o Bangladesh o India o Nepal o Pakistan o Sri Lanka
400 7,200 32 3,347 278
1.4 2.8 0.8 6.8 2.1
n/a 16.6 4.5 n/a 7.4
533 9,035 38 3,442 607
1.5 2.4 0.7 5.3 4.2
n/a 14.9 5.2 n/a 17.3
East Asia o China o Japan o South Korea o Mongolia o Taiwan
12,000 47,823 12,488 44 9,853
2.7 1.0 3.7 5.7 4.9
14.8 4.6 23.0 11.6 n/a
15,167 51,233 15,362 18 10,418
1.8 1.0 3.1 3.2 3.5
12.9 4.4 17.2 11.3 n/a
Southeast Asia o Brunei o Indonesia o Malaysia oMyanmar o Philippines o Singapore o Thailand
338 2,230 1,859 2,880 884 2,732 2,713
n/a 1.6 2.6 3.4 1.4 4.8 4.9
n/a 6.7 8.7 22.3 10.7 23.9 17.3
29 3,057 2,224 4,199 1,367 4,449 3,429
5.6 1.5 2.2 3.5 1.6 4.6 2.0
n/a 7.1 11.6 36.5 7.8 27.5 12.8
Source: Derived from SIPRI (1999) and IMF (1998).
from Asia, with 18 percent going to South Asia and 82 percent to East and Southeast Asia. The financial crisis, which began in the second half of 1997, affected a number of countries - particularly Indonesia, South Korea, Malaysia, Thailand and, to a lesser extent, the Philippines, with side effects on Japan, Singapore, and Taiwan. Some military budgets adopted for 1997 were cut. SIPRI (1998, p. 194) reports a 30 percent reduction in Thailand's proposed military expenditure by the end of 1997, and small reductions by Malaysia and Thailand for 1998. As regards arms imports, the depreciation of Asian currencies led to a reduction in the purchasing
Military expenditure and economic development in Asia
75
power of defense budgets and, in the case of South Korea, postponement of almost all arms procurement. SIPRI (1999, appendix table 7 A) reports cuts in military expenditure between 1997 and 1998 by seven Asian countries totaling $2,154 billion but an increase of some $2 billion by China. IMF data (Gupta et al., 1999) presents a much more extreme picture, suggesting that the economic downturn was associated with a cut in military expenditure by NICs of $6.4 billion between 1997 and 1998 and $1.9 billion by the developing countries of Asia. Together, these represented more than half the world reduction of $15.3 billion. There is mounting evidence that Asia has made what has been called an "astonishing bounce-back" in economic terms. SIPRIs most recent assessment (1999, pp. 285-288) is that the effect of the financial crisis on military expenditure has been relatively mild and not as strong as the impact on employment, income distribution, and poverty (1999, p. 287). It seems that the constraints of the financial crisis have had only a modest and temporary impact on military expenditure, though rather a larger one on amis imports, and that its expansionary path has been, or very shortly will be, resumed. The reductions which did occur were the result of government revenue shortfalls and in no sense represented a peace dividend. The determinants of military expenditure are discussed later on. Suffice to say here that the big or most rapidly increasing spending on the military is undertaken by those countries facing a perceived threat (India/ Pakistan, China/Taiwan, the Koreas, Sri Lanka), or those whose economic strength enables large expenditure (e.g., Japan, Singapore), or military regimes intent on retaining power (Myanmar). Within Southeast Asia, leaving aside Brunei and Singapore, the rate of growth of military expenditure has been broadly similar. Given the long history of nonconflict among countries in the region, similar growth in military expenditure may stem from a desire to keep up with the neighbors in terms of the sophistication of weaponry rather than being an arms race in the conventional sense. The costs, widely defined, of this military expenditure will be the same, however, irrespective of the precise motives which underpin it. Finally, the domination of Asian military expenditure by Japan should be noted. Although the country famously limits its
76 Arming the South
military expenditure to one percent of its GNP, the size of the country's GNP is such that it makes up around 45 per cent of the total Asian military expenditure during the 1990s. Interpreting Asia's military expenditure trends Two recent articles illustrate the breadth of difference in interpreting the trends outlined in the previous section. In an article titled "Asia's insecurity," Paul Dibb et al. (1999) argue that in the years since mid1997, Asia's strategic outlook has changed dramatically as a result of three events: the Asian economic crisis, the demise of the Suharto regime in Indonesia, and the escalation of nuclear readiness by India and Pakistan. They refer to the traditional areas of tension - the Korean peninsula, between China and Taiwan, and between India and Pakistan - and identify three other matters of concern: Japan's weak regional role, increasing tension between the US and China, and the prospect of a reduced US military presence. The uncertain strategic outlook, the weakness of the region's multilateral bodies, and the limits to US involvement in the region, lead Dibb et al. to suggest three basic security aims, in the conventional sense, for Asian countries. First, they need military forces capable of deterring a major power. Second, they should enter bilateral security arrangements with their neighbors. Third, they should enhance their security ties with a powerful ally (i.e., the US). This analysis can be questioned at several levels. First, it is very short term in emphasis. It is clear that only some countries were significantly affected by the economic crisis and it is generally acknowledged that almost all countries are making a remarkable recovery. Indonesia, despite the loss of East Timor, has hardly disintegrated. Second, whilst their recommendations are relevant and have been adopted for decades by Asian countries facing a large external threat, they are far less if at all relevant to most countries. The last armed conflict between any of the ASEAN nations ended in the mid-1960s and the three major traditional areas of tension have been generally quiet, in part due to the balancing effects of military power. Third, it is based on possibilities rather than probabilities and gives too much emphasis to uncertainty. The authors acknowledge that no country in Asia has the military capability needed
Military expenditure and economic development in Asia
11
for a major ground assault. Invasion of one country by another, then, seems highly unlikely. Border disputes are few and, with the exception of Kashmir, are of small significance. Fourth, the dismissal of the role ASEAN and the ASEAN Regional Forum (ARF) fails to take into account these organizations' long-term achievements, particularly given their strongly-held credo of non-interference in domestic affairs by outsiders, including other ASEAN members. Finally, there is no consideration of the opportunity cost of military expenditure which is extremely high for many low-income Asian countries. Bjorn Moller (1998), in common with researchers such as Simon (1996), takes a much longer view of Asian security and draws more optimistic conclusions. He attempts to answer a series of assertions about Asia, of which two are relevant to this chapter: that Asian countries are fundamentally aggressive and inclined to use force, and that the region lacks international organizations and therefore tends to act as a "raw international anarchy." On the first, concerning aggression, Moller cites Weisburd (1997) to demonstrate that Asia is rather below the world average concerning wars between states (and Weisburd includes, but I omit, Iran, Iraq, and Kuwait as part of Asia). Of the 16 Asian wars he identifies, ten occurred prior to 1970 and only one (India/Sri Lanka) occurred in the 1990s. Of particular interest, given Dibb's first suggested security aim, is that only one of Weisburd's ten "classic invasions" which have occurred since 1945 was in Asia (Vietnam/Kampuchea 1979) despite a wide range of territorial disputes and other potential causes of war. Moller points to the "centuries-long history of peaceful relations with neighbours" by countries such as India and China, and to the high degree of self-restraint shown in wars such as those between India and Pakistan (except for the first in 1947), India and China, and China and Vietnam. This brings to mind Matthew Melko's (1993) contention that whereas war is interesting and exciting, times of peace are regarded largely as the exception and a mere background for the real events of history. In reality, as he convincingly demonstrates, peace is far more prevalent than war: it is "ubiquitous, incessant, normal." Simon (1996, p. 388) describes the five to ten year external Asian security environment as being "remarkably benign."
78 Arming the South
As regards the lack of international organizations, Moller agrees there are fewer in Asia than elsewhere but points out that this does not preclude effective practices which promote order. That is, there maybe "functional substitutes" for such organizations. He (1998, pp. 13-16) and several contributors to his book have a much more positive opinion concerning ASEAN than have Dibb and associates. ASEAN has dealt with possible external threats by embracing states such as Myanmar as new members - in Simon's words, it has "promoted security with rather than against states" (1996, p. 383); it has been declared a nuclear free zone; it has made much progress toward economic integration and free trade and it has institutionalized significant conflict prevention and management mechanisms. The ASEAN Regional Forum has, in its short life, successfully encouraged confidence-building measures and transparency agreements, including the publication of defense white papers by its member states and China. The weaknesses which Dibb et al. identify in ASEAN in fact seem to relate to its principle of non-interference in members' domestic affairs. Indeed, Moller (1998, p. 19) envisages that Asia in general might go the "ASEAN way" of peaceful coexistence and respect for national sovereignty, in which case the general level of armaments, he suggests, might fall. An Asian arms race? Can the rapidly increasing military expenditure in Southeast Asia since the mid- to late-1980s, including purchases of major conventional systems, be appropriately considered an arms race? Drawing on Simon (1996), Moller (1998), and others, I suggest six possible interpretations: 1
There is in fact no arms race, but sustained economic growth has allowed countries to continually increase their military expenditure. Moller (1998, p. 5) seems to support this explanation, noting that while military expenditure has risen in real terms, it has not risen as a proportion of GDP (see table 4.1). If his interpretation is correct, then true arms races are not occurring in Asia. Indeed, between 1989 and 1997, the share of military expenditure in GDP fell in eleven Asian countries, remained fairly constant in five, and rose in only two
Military expenditure and economic development in Asia
79
- Sri Lanka and Myanmar (SIPRI, 1999, table 7A4). As regards military expenditure as a proportion of CGE, for the 12 countries for which Government Finance Statistics Yearbook data were available (IMF, 1998), three (Myanmar, Sri Lanka, and Singapore) increased their share between 1990-1992 and 1995-1997, three decreased it, and in five it remained roughly unchanged. 2
Southeast Asian military expenditure is driven principally by concern about China, whose military expenditure is thought to be increasing although, as Moller notes, from an unknown level and at an unknown rate. Greater certainty about China's military expenditure and ambitions could reduce these fears but China's leadership seems obsessed with secrecy. The concerns are encouraged by territorial ambitions of China, focusing on the Spratlys, and that China is the only Asian nation possibly bent on power projection beyond its maritime economic exclusion zone (Simon 1996, p. 391). An argument against this explanation is the behavior of Vietnam which, itself a claimant to the Spratlys, concluded a treaty with China following their armed conflict of 1979. Vietnam has since cut its military expenditure massively: SIPRI data (1999, p. 312) indicates a 54 percent reduction in real terms between 1989 and 1994. Another argument against is the lack of military cohesion among Southeast Asian countries (Joon, 1998, p. 86), whose defense forces are overwhelmingly national in focus. China's military expenditure between 1989 and 1998 has been carefully studied by Wang (1999, pp. 334-349), who draws the following conclusions: (a) China's military expenditure has consistently been about 1.7 - 1.8 times its official defense budget; (b) in real terms, China's military expenditure has increased about 75 percent since 1979; and (c) given its rapid economic growth, the share of military expenditure in China's GDP has steadily declined over the period and is lower than those of all the major powers and its neighboring countries, except Japan. Simon's (1996) assessment is that China poses not much of a threat in the region and that even if it became so inclined, such a threat would take 15 years to materialize (i.e., circa 2010), while
80 Arming the South
Conetta and Knight (1998, pp. 34-35) suggest it would take until around 2025. 3
The countries may be in a special sort of arms race where weapons acquisitions are essentially status symbols to impress neighboring countries. Moller (1998, p. 14) suggests that such an arms race might not be a serious one, but "one of posturing and make-believe." Joon's (1998, pp. 81-83, 88) description of ASEAN nations' weapons purchases as being demonstrations of regime legitimacy, national sovereignty, and virility supports such an explanation.
4
The countries are indeed in a traditional arms race with each other, perceiving each other as a threat or at least as a potential adversary if disputes over land and sea boundaries were to escalate. If correct, this explanation is based on a desire to maintain a balance of power rather than hegemony (Simon 1996, pp. 390-391). The lack of armed conflict over many decades represents an argument against this explanation or, at least, is testimony to the conflict management skills of ASEAN and national governments. There is some evidence (Tan, 1999) that a more traditional arms race may be beginning between Singapore and Malaysia. Singapore has the best armed and equipped forces in Southeast Asia and a wellpublicized doctrine of forward offensive defense, i.e., a pre-emptive military strategy. To date, Tan argues, the "unrelenting nature of Singapore's military acquisitions ... has surprisingly not alarmed its neighbours and sparked an interactive arms build-up" (1999, p. 465). However, he points to Malaysia's major modernization program, including the construction of two major bases in Johor, as evidence that an arms race may be beginning, although the program has been largely deferred owing to the financial crisis. This may destroy the very security that Singapore's military buildup was meant to achieve and Tan recommends a much closer link between defense and foreign policy.
5
With the ending of internal conflicts in a number of countries, the military has been reverting to its original role of defense against
Militaiy expenditure and economic development in Asia
81
external threat (Findlay, 1996). This seems to be a rather dated explanation but a more contemporary case is that of the Philippines which was, prior to the closure of major US bases, in the mid-1990s able to leave its external defense in the hands of the US. Obligations to patrol expanded (200 nautical mile) economic exclusion zones, EEZs, may explain why the arms build-up has a strong maritime emphasis (Joon, 1998, p. 85), as might the increase in illegal refugees, smuggling, piracy, and poaching of marine resources. 6
A motive possibly related to both 4 and 5 above are that arms purchases are simply part of the continual modernizing process and replacement of obsolete equipment. There is frequently a concentration of re-equipment over a few years, as is about to commence in South Korea and the Philippines (SEPRI, 1999, p. 288), and this may encourage neighboring countries to do likewise.
Even if the build-ups were originally for innocuous reasons, states may nonetheless come to regard a neighbor's military strength as a threat and a true arms race might begin, as Tan suggests is occurring with Malaysia's response to Singapore's long-term build-up. And as they grow in military strength, states might develop "ambitions which go beyond their present one of national defence" (Moller 1998, pp. 15-19). Determinants of military expenditure Various possible motives for Southeast Asian countries' involvement in a prolonged arms build-up were discussed. It was noted that the nature and extent of military expenditure, in Southeast Asia in particular, was not closely linked to threats. What, then, determines military expenditure allocations? Five broad explanations of military expenditure have been identified (West, 1992). 1
The past. Countries inherited a military establishment at independence and through the annual budget setting process have tended to at least maintain the share going to military expenditure. That is, the best predictor of military expenditure in any year is that
82 Arming the South
of the previous year. 2
Geo-strategic considerations. A country's strategic situation: involvement in armed conflict or the perceived likelihood of such involvement, or involvement in an arms race, is a major determinant of military expenditure.
3
The politics of the budgetary decision-making process. The relative strength and ability of the military pressure group will have a strong impact on military expenditure allocations.
4
Pressure from arms suppliers. The arms industry, from within or from outside a country, may in various ways press a government to allocate more to military expenditure than it otherwise would.
5
Financial and economic factors. This will determine the ability of a society to allocate resources.
If a country is invaded, or the ruling regime's survival is threatened from within, then very large allocations to military expenditure can be expected in an attempt to ensure survival. This clearly explains high military expenditure allocations by countries involved in tense interfaces. In the absence of war, military pressure groups may invent threats in order to justify greater military expenditure. Writing in the US context, Conetta and Knight (1998) criticize the tendency for simulation exercises, e.g., a scenario of war with China and North Korea in 2015, devised as an aid to planning, to be transformed into arguments in future decision-making. Simulations can aid planning but Conetta and Knight caution that they can be misused. "Exploring 'wild cards' to identify warning signs or to define limits is one thing. Using them to establish force structure or modernization requirements is quite another" (1998, p. 34). They also question whether uncertainty about the future - especially futures 15, 20 or more years into the future - should justify large military expenditure allocations in the present. It is clear that the above list does not get at the underlying motives, some of which (e.g., the quest for national prestige and status) have been
Military expenditure and economic development in Asia
83
mentioned earlier in the chapter. In addition, two beliefs - almost universally accepted but open to question - underlie the military expenditure allocation process, namely, that a strong military is the best means of achieving security and that a strong military is a necessary part of achieving a level of national pride and status. These beliefs underpin the efforts of the military pressure group, which includes the military, related politicians and public servants, and arms suppliers (domestic and/or foreign). The military pressure group basically acts so as to maximize its share of government resource allocations in a geopolitical context and with an economic constraint. In times of war or high perceived threat, the former dominates. In more normal times, the size of a country's GNP, and the share of GNP which a government can capture, act as general limits to military expenditure. Government revenues, and the purchasing power of such revenues in the face of exchange rate depreciations, may fluctuate between years. During economic recession, there may be some deferral of equipment purchases but a permanent reduction in the share of military expenditure in CGE is unlikely. Figure 4.1 presents a possible way in which military expenditure allocation decisions can be understood. Leaving aside the past, which is itself subject to a range of earlier influences, determinants number 2, 3, and 4 may be called motivating factors (M2-4) whilst number 5 is an enabling factor (El). Three other variables underlie the motivating factors: the self-interest of the military pressure group (Ul), the belief held by the military pressure group, but also by the public, that security requires a strong military (U2), and the widely-held belief that national prestige and status can be cost-effectively achieved through military expenditure (U3). An important issue is the strength of Ul if the beliefs U2 and U3 were to be proven false. As regards U2, there is evidence that expansion of military power may erode rather than enhance security, particularly when wider definitions of security are considered (e.g., Johansen, 1991; Felice, 1998). There is little doubt that the military contributes to national pride and status at least among some segments of the population, but the real issue is its cost-effectiveness. The military is costly and unproductive in a national accounting sense. There may be less costly ways of building national pride and prestige (e.g., through comprehensive public education
84 Arming the South
/ *
M4 Arms suppliers
----->
M2 Geo-strategic considerations
({
\ Ul Military pressure group
/ / / / / /
i i • i 11 11 11
u % % \ \ \
U2 Belief that milex results in security U3 Belief that milex results in prestige
-* ^w W
Milex allocation
M3 Budgetary politics i
/ ^ / /
El Financial and economic factors
w >
Main causal paths Secondary causal paths
Figure 4.1: Linkages in the military expenditure allocation process
and health systems) which also have positive economic outcomes. A number of research questions arise from this model of military expenditure determination: how is security to be understood? is the military a cost effective means of achieving security? are there alternative ways which could increase security and how cost effective are these likely to be? how is national pride and status to be understood? is the military a cost effective means of enhancing national pride and status? are there alternative ways of enhancing national pride and status and how effective are these likely to be? what are the dynamics of the military pressure group in any country? what is the appropriate response to the understandable tendency for the military pressure group to exaggerate perceived threats? The multiple regression analysis approach favored by economists may assist in determining the motivating factors but carefully planned interviews with decision-makers are more likely to bring out the underlying beliefs and values. The study of leadership perceptions in the area of security by Ayoob and Samudavanija (1989) is an important example of this latter type of research and may help explain arms race-
Military expenditure and economic development in Asia
85
type behavior in benign security environments. Military expenditure, growth, and development Three issues are discussed in this section: the relation between military expenditure and economic growth, the relation between military expenditure and development, and the possibility of capturing a peace dividend. Studies of the relation between military expenditure and economic growth in developing countries began with Benoit's (1973, 1978) straightforward correlations, moved to simultaneous equations (e.g., Deger and Smith, 1983) and Granger-causality tests (e.g., Joerding, 1986) in the 1980s, and then on to various forms of neoclassical (e.g., Biswas, 1993) and Keynesian models (e.g., Dunne and Mohammed, 1995). This is not the place to evaluate this research. Dunne's comprehensive survey concludes that military expenditure has at best no effect on growth but is likely to have a negative impact. Certainly there is no evidence of a positive effect... Overall, these results suggest that disarmament need not be costly, and can indeed provide an opportunity for improved economic performance (1996, p. 459).
An inspection of Dunne's list of studies (1996, pp. 460-464) suggests that for Asian countries with high levels of GNP per capita and high rates of growth, the effect of even high military burdens on growth is likely to be non-significant. On the other hand, negative net effects on growth are more readily found amongst low-income Asian countries. It is unlikely that further research in this area will result in any change to these wellestablished conclusions. The relationship between military expenditure and poverty is closely related to the previous conclusion. Economic growth is accepted as a necessary but not sufficient condition for significant reductions in poverty, so factors which retard growth lead to the maintenance of large numbers and/or proportions of the population in poverty. Whilst poverty in the rapidly growing Asian economies has fallen over the past 20 years, some 860 million people in Asia fall beneath their national poverty lines, almost 70 percent of all those in poverty in developing countries. The
86 Arming the South
Human Poverty Index (UNDP, 1999, pp. 146-148), which measures poverty along four dimensions - a long and healthy life, knowledge, economic provisioning, and social inclusion - reveals that 36.6 percent of South Asia's population were in poverty. The respective figures from East Asia and Southeast Asia (and the Pacific) were 19 and 25 percent. One specific aspect of the link between military expenditure and poverty concerns trade-offs between military expenditure and other CGE expenditure categories. It is clear enough that if governments allocate their financial resources according to social rates of return, then society's welfare will be maximized. Since national accounting criteria classify military expenditure as a government consumption item, it yields no rate of return; any reallocation toward government investment activity where social rates of return are positive, then, must increase society's welfare. That said, most studies have not been able to demonstrate a negative trade-off between military expenditure and other CGE categories. An exception is a study of Sri Lanka between 1981 and 1992 (Harris, 1996a) where trade-offs between military expenditure and "Economic Affairs and Services," and with capital expenditure in general, were identified. For the 13 Asian countries for which IMF Government Finance Statistics Yearbook data were available for the 1990s there was no evidence of trade-offs, except again with "Economic Affairs and Services" in the case of Sri Lanka. The long-term effect of diverting government expenditure away from productive activities and particularly capital expenditure, is lower economic growth. Harris (1996a, p. 145) estimated a reduction of half a percentage point in Sri Lanka's annual rate of growth between 1983 and 1992 as a result of military expenditure, and the country's military burden has increased considerably since the early 1990s. Finally, the central governments of six Asian countries (India, Pakistan, China, Mongolia, Myanmar, and Singapore) spent more on their military than on education and health combined in the late 1990s. There is clearly opportunity for reallocations to take place. At the same time, it should be noted that there is no clear inverse relation between military expenditure and human development measures. After a careful examination of the data, Brauer (1996, p. 120) concludes that "the supposedly clearly negative relation between military expenditures and human development (human freedom) for all nations is
Military expenditure and economic development in Asia
87
in fact driven - statistically - solely by the data for 'Western' countries and does not appear to hold for 'non-Western' countries." His interpretation of this result, which parallels that of a number of trade-off studies, deserves to be carefully read. A crucial question is why, given the benign long-term security environment and the potential for saved expenditure to make inroads into poverty, Asian military expenditure continues to rise in real terms? Mohammed Ayoob (1991, p. 279) argues that whilst research demonstrating that military expenditure retards growth and development is valid, it misses the "essential motivation" behind military expenditure. Governments, according to Ayoob, have no great interest in spending on development, which ranks a poor third priority after keeping their regimes in power and meeting threats from the regional environment. Economists are not oblivious to this problem. Paul Dunne concludes his survey of military expenditure and growth by noting that "the potential reward [of disarmament in developing countries] is an unprecedented contribution to human survival and welfare - but we must ask if the political will really does exist" (1996, pp. 459-460). Greater scrutiny of the determinants of military expenditure, as suggested here, seems an important research task, as does ways in which aid flows from developed countries might be linked to reduced military expenditure by the recipients, though this is relevant to only some (low income) Asian countries. The next section also suggests alternative and less costly ways of achieving security. Much has been written on the potential of a peace dividend which can have positive effects on growth and development. The simplest understanding of the peace dividend is that if governments were to cut their military expenditure there would be greater financial resources to allocate to more socially desirable and/or more productive expenditure categories. This thinking has been strongly influenced by the events in Europe between 1991 and 1993 which led to the end of the cold war and by the continuing struggle of many low-income developing countries to save sufficient resources and then to allocate these to best effect. A more sophisticated version of the peace dividend embraces the concept of the conversion of an economy, or parts of it, from a military orientation to a civilian orientation. Clearly, this approach includes the conversion
88 Arming the South
experiences of countries after, say, the first and second world wars when major reallocations of human capital and physical capital - indeed entire productive systems - took place. A peace dividend, it should be noted, can occur at any time but it is often easiest to cut military expenditure when a major armed conflict comes to an end. Building on the work of Chan (1995), Intriligator (1996) emphasizes that such adjustments involve considerable short-term costs, in the form of unemployment of resources and various conversion costs, before benefits are realized in the medium and long-term from the new allocation of resources. Disarmament, then, can be usefully considered as a type of investment for which rates of return can be calculated. There is no certain way of knowing what happened to any dollar not spent on the military. This said, there are three broad possibilities for a peace dividend: 1 2 3
Military expenditure is reallocated to other CGE categories, i.e., CGE does not fall as military expenditure falls. The government's budget deficit is cut, i.e., CGE does fall as military expenditure falls. Taxes are reduced, i.e., central government revenue (CGR) falls.
Given the current economic emphasis on small government and fiscal responsibility, it would not be surprising to find that a classic peace dividend rarely occurs, i.e., that there is no reallocation to other CGE categories. Normally, reduced military expenditure allows budget deficits to be reduced, thus leaving more financial resources available for investment by the private sector. Given that interest rates are also likely to fall as a result of reduced borrowing from domestic and foreign sources by the public sector, private investment is likely to rise, ceteris paribus, and thereby economic growth. For Asia, this discussion remains an abstract one since only two Asian countries reduced their military expenditure, in real terms, during the 1990s. To summarize this section, it seems that there is limited scope for further insights into the effects of military expenditure on growth and development and peace dividends can only be reaped if military expenditure allocations are cut. Ayoob's contention that governments
Military expenditure and economic development in Asia
89
give low priority to development needs to be borne in mind. The next section suggests some alternative ways of achieving security which may allow such cuts to take place. New concepts of security The meaning of security and the nature of armed conflict has changed significantly in the past two decades. A stream of literature (e.g., Azar and Moon, 1983/84,1988;Buzan, 1983; UNDP, 1994; Felice, 1998) has emphasized that the traditional definition of security is of limited relevance to developing countries. This definition focused on the external military threats to nation states and military power to deter or resist external attack was therefore seen as the prerequisite of national security. Azar and Moon (1983/84) also noted the assumption that it is super and middle powers which set the agendas for armed conflict in the developing world. Developing countries were assumed to have little role in such conflicts, other than forming the backdrop for the competition of such powers. Proponents of redefinition point to the changed nature of warfare from old to new wars, as Mary Kaldor (1999) puts it. It is also well known that almost all armed conflicts now occur within countries rather than between them. Typically, there is fighting between government forces and groups wishing to secede or take over government. In 1998, there were 27 major armed conflicts, defined as "prolonged use of armed force ... incurring the battle-related deaths of at least 1000 people during the entire conflict and in which the compatibility concerns government and/or territory" (Sollenberg et al. 1999, p. 15). Only two of these 27 were interstate wars, and one - between India and Pakistan over Kashmir - was in Asia. Of the remaining 25 internal conflicts, eight were in Asia. This is not to deny regions of major military tension in Asia but, as already noted, these are noteworthy for their military restraint. Invasions are very rare events. Another approach to a wider definition of security comes from peace theory, which extended the concept of violence from direct or physical violence to structural violence. The term "structural violence" was coined by Johan Galtung as a result of fieldwork in colonial Rhodesia where he
90 Arming the South
became increasingly aware of the limitations of defining peace as the absence of violence. He noted that while there was little direct violence by the colonial authorities against the native population, there were structures which had significant negative effects on African people: In a certain sense, there was harmony, cooperation, integration. But was this peace? With the blatant exploitation, with blacks being denied most opportunities for development given to whites, with flagrant inequality whereby whites were making about twenty times as much for exactly the same job as blacks? Not to mention the basic fact that this was still a white colony (Galtung, 1985, p. 145).
These structures, procedures, and policies, it should be noted, were not intended to cause harm but nonetheless did so. To Galtung, they represented a "quiet process, working slowly in the way misery in general, and hunger in particular, erode and finally kill human beings" (1985, p. 145). Structural violence, then, describes those systems which maintain the dominance of one central group over another group, often the majority, at the periphery. At a practical level for those at the periphery, structural violence can mean low wages, landlessness, illiteracy, poor health, limited or non-existent political representation or legal rights and, in general, limited control over much of their lives. If those who suffer structural violence resist or try to change it, they are likely to be met with direct violence. "Negative peace" involves cessation of physical violence whilst positive peace occurs when the underlying causes of the conflict - exploitation, inequality, and the like - are dealt with. The proponents of a revised concept of security suggest that the type and source of threats now facing developing countries are multiple and complex. Azar and Moon (1983/84) pointed out three major dimensions of security, other than security from external military threat: economic security, which focuses on freedom from deprivation, ecological scarcity, and ethnic tensions which may involve "protracted social conflicts." More recently, in a post-cold war context, the UNDP (1994, chapter 2) argues for a redefinition which focuses on human security, which it explains as ... the child who did not die, a disease that did not spread, a job that was not cut, an
Military expenditure and economic development in Asia
91
ethnic tension that did not explode in violence, a dissident who was not silenced. Human security is not a concern with weapons - it is a concern with human life and dignity (UNDP, 1994, p. 22).
The UNDP identifies seven aspects of human security: economic, food, health, environmental, personal, community, and political. It points out (1994, p. 50) that a person in a developing country is 33 times more likely to die as a result of social neglect (preventable disease and malnutrition) than as a result of an inter-country war. Finally, as Felice (1998) argues, there may well be an inverse relation between militarism and human rights which is most clearly seen in three areas, physical security, economic well-being, and self-determination and freedom. Despite these well-known arguments, the structure and force projections of the military establishments of most Asian countries have little relation to current strategic realities. They appear to be based on possible strategic events rather than probable ones. Of course, it is always possible that the government of some currently passive country may become aggressive and attempt an invasion of one or more neighbors. The questions which need to be asked, obviously, are how likely this possibility might be and whether it would occur unexpectedly or only after some observable preparations. In this focus on possibilities, Asian military decision-makers are acting in a similar way to the US military whose principle is that it needs the capacity to be able to fight two nearly simultaneous major wars. In fact, it has been involved in only three major wars since the end of the second world war, and these have occurred decades apart (Conetta and Knight, 1998, p. 35). Alternative ways of achieving security We have seen that Asian military expenditure continued to rise in the 1990s whilst that of most other regions fell, and this despite the absence of major insecurities in the traditional sense. We have suggested the importance of investigating the determinants of this military expenditure, particularly the underlying factors. We have also considered the ways in which security has been traditionally envisaged and pointed to alternative and wider definitions. This section builds on this theme and suggests
92 Arming the South
alternative ways by which security may be enhanced. A major task of economists is to determine the most cost effective way of achieving a policy objective. With security, however, there has been an extreme emphasis on a strong military as the appropriate means and little consideration of alternatives. These alternatives are certainly less costly than the military. Like the military, their potential effectiveness is uncertain, but the military maybe preferred because of its greater apparent tangibility. This section briefly examines seven alternatives to the military and excludes some other alternatives which remain essentially military in character, e.g., moving to a defense-only military capability and engaging in military confidence-building measures (e.g., transparency concerning the nature and extent of military capabilities). These could be applied along with the non-military options and many of these alternatives could be applied at the same time. They are presented roughly in descending order of cost and ease of introduction. 1
First, a government can unilaterally demilitarize, irrespective of what its neighbors or its internal opponents may choose. An impressive example in Costa Rica, which chose to demilitarize in 1948 and which, in 1983, also announced its "permanent, active and unarmed neutrality in respect of armed conflicts elsewhere in Central America." Despite its location - in a region with a long history of armed conflict, oppression, and human rights abuses - Costa Rica has a flourishing democracy and social indicators far ahead of countries in the region and of countries elsewhere with similar levels of GDP per capita. The explanation for the latter appears to be the direct application of resources not spent on the military to social expenditure (Harris, 1996b). In many developing countries, the message which would be sent by demilitarization is that future disputes will be settled at the elections. In Costa Rica, for example, governments have been voted in and out of office regularly.
2
Second, a government may negotiate non-aggression treaties, possibly in conjunction with bilateral reductions in military capabilities, with one or more neighbors. Indeed, UN members are under an obligation
Military expenditure and economic development in Asia
93
to settle their disputes peacefully using one or more of negotiations, enquiry, mediation, conciliation, arbitration, judicial settlement, resort to regional agencies or arrangements, or other peaceful means of their own choice (UN Charter, Chapter VI, Article 33). Many countries have non-formalized but powerful understandings with their neighbors that they will not settle disputes by force. A functioning democracy can deal with internal differences of opinion by debate and through the electoral system. 3
Third, forums may be established where the concerns, positions, and interests of disputants can be aired and clarified. These may allow for negotiation with a view to securing agreements and may defuse disputes before they reach a heightened level. Clearly, ASEAN and the ARF have made much progress in this respect and such bodies could be established to allow internal disputes to be aired.
4
Fourth, governments and NGOs may establish institutions to arbitrate and adjudicate in disputes between countries and, insofar as existing legal systems do not meet such needs, within countries. The International Court of Justice, an arm of the UN based in The Hague, acts in this capacity.
5
Fifth, governments and/or NGOs can establish institutions to help in the resolution of conflicts, both between and within countries. The difference between these institutions and those in the previous suggestion relate to the three approaches to tackling a disputes (Ury et al., 1988) - power (physical, economic), rights, and reconciling interests. The previous suggestion is based upon establishing which party is "right" according to some legal code or assessment of natural justice. Leaving aside that those with most economic power are more likely to succeed in rights-based contests, these result in a winner and a loser. Therefore, the dispute may not be settled in the long-term and the relation between the disputing parties may be permanently strained. In contrast, conflict resolution seeks a win-win outcome which is satisfying to both sides. On criteria such as cost, durability of the
94 Arming the South
outcome, and the quality of the relation between the disputants, attempts to reconcile the interests of disputants are superior to the alternatives. A classic example is the five year mediation carried out by the Papacy between 1979 and 1985 concerning the dispute between Argentina and Chile over three islands in the Beagle Channel. The peaceful resolution of this conflict stands in stark contrast to the UK-Argentina war over the Falklands, despite many similarities in the two disputes. 6
Sixth, governments and NGOs can train and educate a population in nonviolent conflict resolution and conflict management. The philosophy behind this suggestion is that people are educated, in the widest sense of the word, to deal with disputes in certain ways. Often these emphasize power and rights, and recourse to violence is a distinct possibility. This suggestion involves educating people in an alternative way of dealing with conflict and has similar advantages to those mentioned under the fifth suggestion.
7
Finally, populations can be trained in social defense - in nonviolent resistance and non-cooperation with an invader or internal usurper of power - so as to make their takeover extremely costly and perhaps untenable. Writers such as Martin (1984), Sharp (1987), and Summy (1994) have pointed to the success of informal nonviolent resistance by, for example, many independence movements and the Norwegian and Danish populations to Nazi occupation during the second world war. Stephen Zunes (1994) has documented the increase, since the early 1980s, of unarmed insurrections against authoritarian regimes in developing countries.
This does not exhaust the list of nonviolent, non-military alternatives which promote security from both internal and external aggression. Preventive diplomacy is, of course, widely practiced, educational and cultural exchanges between countries could be multiplied many times, and development and democratization can reduce the intensity of internal sources of insecurity. The central point of this section is that there are many relatively cheap ways of working for security and that the current
Military expenditure and economic development in Asia
95
focus on military-based security may well be economically unjustifiable. The cost effectiveness of such alternatives deserves close study. For the successful implementation of a demilitarization program, at least two necessary conditions must be met. First, a government and its people need to be convinced that security, however they conceive it, will be adequately met; education concerning alternatives such as those listed above will therefore be necessary. Second, a government and its people need to be able to see a net advantage to changed security arrangements. One way of demonstrating this is to draw up a range of specific programs and projects which would be undertaken using the net resources saved from military expenditure cuts. That is, the financial resources saved should not go simply to reduce the size of the budgetary deficit, despite the possible benefits of such an action, because these would be insufficiently direct and tangible to convince a people to accept extensive military expenditure cuts. Members of the armed forces can be encouraged to cooperate with demilitarization by a well-planned demobilization program which gives them significant economic benefits. References Ayoob, M. and C.A. Samudavanija. "Leadership and Security in Southeast Asia: Eexploring General Propositions," pp. 256-277 in M. Ayoob and C.A. Samudavanija (eds.). Leadership Perceptions and National Security: The Southeast Asian Experience. Singapore: Institute of Southeast Asian Studies, 1989. Ayoob, M. "The Security Problematic of the Third World." World Politics Vol. 43, No. 1 (1991), pp. 257-283. Azar, E. and C-I Moon. National Security in the Third World. The Management of Internal and External Threats. Aldershot: Edward Elgar, 1988. Azar, E. and C-I Moon. "Third World National Security: Toward a New Conceptual Framework." International Interactions Vol. 11, No. 2 (1983-84), pp. 103-135. Benoit, E. Defense and Economic Growth in Developing Countries. Lexington: Heath, 1973. Benoit, E. "Growth and Defense in Developing Countries." Economic
96 Arming the South
Development and Cultural Change Vol. 26, No. 2 (1978), pp. 271280. Biswas, B. "Defense Spending and Eeconomic Growth in Developing Countries," pp. 223-235 in J. Payne and A. Sahu (eds.). Defense Spending and Economic Growth. Boulder, CO: Westview, 1993. Brauer, J. "Military Expenditures and Human Development Measures." Public Budgeting and Financial Management Vol. 8, No. 1 (1996), pp. 106-124. Buzan, B. People, States and Fear. Hemel Hempstead: Wheatsheaf, 1983. Chan, S. "Grasping the Peace Dividend: Some Propositions on the Conversion of Swords into Plowshares." Mershon International Studies Review Vol. 39, No. 1 (1995), pp. 53-96. Conetta, C. and C. Knight. "Inventing Threats." Bulletin of the Atomic Scientists Vol. 54, No. 2 (1998), pp. 32-38. Deger, S. and R. Smith. "Military Expenditure and Growth in Less Developed Countries." Journal of Conflict Resolution Vol. 27, No. 2 (1983), pp. 335-353. Denoon, D. "Defence Spending in ASEAN: An Overview," pp. 48-74 in Chin Kin Wah (ed.). Defence Spending in Southeast Asia. Singapore: Institute of Southeast Asian Studies, 1987. Dibb, P., D. Hale, and P. Prince. "Asia's Insecurity." Survival Vol. 41, No. 3(1999), pp. 5-20. Dunne, J.P. "Economic Effects of Military Expenditure in Developing Countries: A Survey," pp. 439-464 in N.P. Gleditsch, O. Bjerkholt, A. Cappelen, R.P. Smith, and J.P. Dunne (eds.). The Peace Dividend. Amsterdam: Elsevier, 1996. Dunne, J.P. and N. Mohammed. "Military Spending in Sub-Saharan Africa: Some Evidence for 1967-85." Journal ofPeace ResearchVol. 32, No. 3 (1995), pp. 331-343. Felice, W. "Militarism and Human Rights." International Affairs Vol. 74, No. 1 (1998), pp. 25-40. Findlay, T. "Turning the Corner in Southeast Asia," pp. 141-172 in M.E. Brown (ed.). The International Dimensions of Internal Conflict. Cambridge, MA: MIT Press, 1996. Galtung, J. "Twenty-Five Years of Peace Research: Ten Challenges and
Military expenditure and economic development in Asia
97
Some Responses." Journal of Peace Research Vol. 22, No. 2 (1985), pp. 141-158. Gupta, S., C. McDonald, L. DeMello, and R. Sab. "Military Expenditure Continues to Stabilize: Some Countries Increase Social Spending." IMF Survey (7 June 1999), pp. 186-188. Harris, G.T. "Economic Aspects of Military Expenditure in Developing Countries: A Survey Article." Contemporary Southeast Asia Vol. 10, No. 1 (1988), pp. 82-102. Harris, G.T. "Financing the Sri Lankan Armed Conflict, 1983-92." Public Budgeting and Financial Management Vol. 9, No. 1 (1996a), pp. 125-149. Harris, G.T. "Military Expenditure and Social Development in Costa Rica: A Model for Small Countries?" Pacifica Review Vol. 8, No. 1 (1996b), pp. 93-100. International Monetary Fund. Government Finance Statistics. Washington, DC: IMF, 1998. Intrihgator, M. "The Peace Dividend: Myth or Reality?," pp. 1-16 in N. Gleditsch, O. Bjerkholt, A. Cappelen, R. Smith, and J.P. Dunne (eds.). The Peace Dividend. Amsterdam: Elsevier, 1996. Joerding, W. "Economic Growth and Defense Spending: Granger Causality." Journal of Development Economics Vol. 21, No. 1 (1986), pp. 35-40. Johansen, R. "Do Preparations for War Increase or Decrease International Security?," pp. 224-244 inC. Kegley(ed.). The Long Postwar Peace: Contending Explanations and Projections. New York: Harper Collins, 1991. Joon, N.M. "International Cooperation in Regional Security: 'Noninterference' and ASEAN Arms Modernization," pp. 77-92 in B. Moller (ed.). Security, Arms Control and Defence Restructuring in East Asia. Aldershot: Ashgate, 1998. Kaldor, M. New and Old Wars. Stanford: Stanford University Press, 1999. Martin, B. Uprooting War. London: Freedom Press, 1984. Melko, M. "The Delineation of Peaceful Societies," pp. 242-246 in D. Ringler (ed.). Dilemmas of War and Peace. Madison, WI: University of Wisconsin Madison Press, 1993.
98 Arming the South
Moller, B. "Introduction: Defence Restructuring in Asia," pp. 1-35 in B. Moller (ed.). Security, Arms Control and Defence Restructuring in East Asia. Aldershot: Ashgate, 1998. Sharp, G. "Deterrence and Defence by Nonviolent Sanctions." Social Alternatives Vol. 6, No. 2 (1987), pp. 9-18. Simon, S.W. "Alternative Visions of Security in the Asia Pacific." Pacific Affairs Vol. 69, No. 3 (1996), pp. 381-396. Soesastro, H. "Military Expenditure and the Arms Trade in the Asia Pacific Region." Asian-Pacific Economic Literature Vol. 8, No. 1 (1994), pp. 27-47. Sollenberg, M., P. Wallensteen, and A. Jato. "Security and Conflicts, 1998," pp. 15-75 in SIPRI Yearbook 1999. Oxford: Oxford University Press, 1999. Stockholm International Peace Research Institute (SEPRI). SIPRI Yearbook. Oxford: Oxford University Press, annually. Summy, R. "Nonviolence and the Case of the Extremely Ruthless Opponent." Pacifica Review Vol. 6, No. 1 (1994), pp. 1-30. Tan, A. "Singapore's Defence: Capabilities, Trends and Implications." Contemporary Southeast Asia Vol. 21, No. 3 (1999), pp. 451-474. United Nations Development Programme (UNDP). Human Development Report. New York: Oxford University Press, annually. Ury, W., J. Brett, and S. Goldberg. Getting Disputes Resolved. San Francisco: Jossey-Bass, 1988. Wang, S. "The Military Expenditure of China, 1989-98," pp. 334-350 in SIPRI Yearbook 1999. Oxford: Oxford University Press, 1999. Weisburd, A. Use of Force. The Practice of States since World War II. University Park, PA: Pennsylvania State University Press, 1997. West, R. "Determinants of Military Expenditure in Developing Countries: Review of Academic Research," pp. 113-145 in G. Lamb and V. Kallas (eds.). Military Expenditure and Economic Development. World Bank Discussion Papers, No. 185. Washington, DC: The World Bank, 1992. Zunes, S. "Unarmed Insurrections Against Authoritarian Governments in the Third World." Third World Quarterly Vol. 15, No. 3 (1994), pp. 403-426.
PART II: ARMS PRODUCTION AND ARMS TRADE
This page intentionally left blank
5 The Arms Industry in Developing Nations: History and Post-Cold War Assessment Jurgen Brauer
Introduction Even before the end of the cold war, the study of the arms industry in developing nations was ill-attended to. It was, and still is, more fashionable to study military expenditure and their impact on economic development and growth in general rather than to study developing nations' arms industry in particular. Part of the explanation is that some, although often dubious, data on military expenditure and economic growth is readily available and can be subjected to statistical analysis with relative ease. In contrast, the detailed case study of an arms industry requires field work - in an industry that for obvious reasons always has much to hide. Nonetheless, a number of studies have emerged - on Brazil, South Korea and Taiwan, on Israel, on India, to name a few examples. Many of them, at least early on, came out of or were otherwise connected to the work of the Stockholm International Peace Research Institute (SIPRI). Partly in response to these writings, a set of justifications arose that posited not only the existence of good political and military reasons for the establishment and maintenance of an indigenous arms industry, but hypothesized the existence of good economic reasons as well. In the
102 A) in ing th e Sou th
present chapter, I evaluate some of these arguments. The issue of arms industry and arms exports in and by developing and so-called emerging countries has been at the forefront of two relatively recent remarkable political revolutions: that of the "velvet" revolution of Czechoslovakia in 1989 and that of the democratic election, in 1994, in South Africa. In both cases, leaders as revered as Vaclav Havel and Nelson Mandela quickly articulated odd-sounding reasons to promote the continuation of their countries' indigenous arms production capacities and even to promote arms exports to a stable of strange and fearsome clients. This suggests that the issues and questions surrounding arms production capabilities, including and especially of technologically relatively unsophisticated weaponry ("small arms"), are highly important matters that are likely to reemerge from time to time. Clearly, an assessment is warranted. Arms production in developing nations The production of arms in developing nations ranges from relatively simple to very sophisticated weaponry. This section has two objectives: (a) to list the names of developing nations known to produce and/or export weapons and (b) to list and assess the supposed non-economic and economic motives for indigenous arms production; the bulk of attention is paid to economic motives. Which developing nations are arms producers? Table 5.1 draws on three sources, ACDA (1997), Brzoska (1995), and Rana (1995). The US Arms Control and Disarmament Agency (ACDA) lists countries' arms exports, by year, between the years 1985 and 1995. Arms exports do not necessarily imply arms production: in a small number of cases, e.g., countries such as Nicaragua, previously imported weapons have been re-exported. In addition, some countries might produce but not export arms and would therefore not be captured in ACDAs listing (e.g., Bangladesh). Michael Brzoska (1995) lists African, Asian, Latin American, and Middle Eastern producers of conventional arms as of the "early 1990s," mostly drawn from SIPRI sources, and
The arms industry in developing nations
103
Table 5.1: Developing nations' arms producers/exporters, ca. 19851995 ACDA (1997)*
Brzoska (1995)*
Rana (1995)*
Afghanistan Argentina Brazil Cape Verde Chile China, PRC Cuba Egypt Ethiopia Greece
1994 all years all years 1985 all years all years 1985, 1988, 1989 all except 1991, 1995 1990 all years
Algeria Argentina Bangladesh Bolivia Brazil Burkina-Faso Burma Cameroon Chile China, PRC
Argentina Bangladesh Brazil Chile China, PRC Cuba Dominican Republic Egypt India Iraq
India Indonesia Iraq Iran Jordan North Korea Kuwait Libya Malaysia Mali
all except 1988, 1989, 1992 all except 1986, 1987 1985-1990 1991-1995 1986-1989, 1994 all years 1988, 1992 1985-1992 1994, 1995 1989
Columbia Dominican Republic Egypt India Indonesia Iraq Iran Ivory Coast Libya Malaysia
Iran Libya Malaysia Mexico Namibia Nigeria North Korea Pakistan Peru Philippines
Mexico Nicaragua Nigeria Oman Pakistan Panama Philippines Saudi Arabia South Africa Sudan
all years 1992,1995 1986,1989 1986 all years 1992,1993 1985,1986 all except 1990, 1991, 1993 all years 1986
Mexico Morocco Nigeria North Korea Pakistan Peru Philippines Saudi Arabia South Africa Sri Lanka
South Africa Saudi Arabia Turkey Venezuela Yugoslavia
Syria Thailand Turkey Venezuela Vietnam Yugoslavia — Slovenia Zimbabwe
1986,1992 1988,1989 all except 1986 1990 1985,1987, 1988, 1992 1985-1991 1992, 1994, 1995 1994
Sudan Syria Thailand Venezuela
Nole: * For ACDA, "all years," refers to all years from 1985 to 1995; see text for details and sources. ACDA lists arms exporters; Brzoska lists only African, Asian, Latin American, and Middle Eastern nation, counting Turkey as "European"; Rana only lists producers of "small arms". Countries listed in bold typeface appear on all three lists.
104 Arming the South
Swadesh Rana (1995, p. 29) lists more than fifty producers of "small arms," half of whom might be considered developing countries and whose names are reproduced in the table. For simplicity, I have chosen to designate countries as "developing" when their per capita GNP is estimated as below US$ 9,700 in 1995, i.e., those countries not classified as "high-income economies" by the World Bank (World Bank, 1997). One may of course quibble with that approach, not least because some important arms producers once considered "developing" thus do not appear in my table (e.g., Israel, South Korea, Taiwan, Singapore, Spain, and Portugal). Similarly, I have left out East-Central and Southeast-Central European nations, i.e., erstwhile satellites and republics of the former Soviet Union that nowadays would appropriately be designated as "developing." By my ordering, ACDA lists 37 countries (counting Yugoslavia and Slovenia but once), Michael Brzoska lists 34, Swadesh Rana lists 25 nations. Eighteen countries appear on all three lists. What exactly do we mean by arms production? One must be clear from the outset that arms production is not a selfevident term. In reviewing the literature, Keith Krause (1992a, p. 171) finds eleven ways by which one may conceive of an activity as constituting arms production, ranging from simple maintenance tasks on imported arms to completely independent R&D and production capabilities (see table 5.2). But on account of two reasons this "laddering" of stages of arms production, as culled from the literature, is inadequate. First, the laddering suggests that any country wishing to produce arms starts at stage 1 and works its way up until the highest stage is reached. Put differently, there is no notion that countries may choose an entry stage other than stage 1 and a goal stage other than the highest stage. For example, Singapore early on appeared to follow a deliberate strategy of servicing naval vessels, and Greece followed a strategy of servicing NATO aircraft, as a means of amassing knowledge and experience valuable for potential entry into more sophisticated stages of arms production later on. At least some countries are able to choose the entry point, and the entry point may differ across different arms classes.
The arms industry in developing nations
105
Table 5.2: The ladder of arms production 1 2 3 4 5 6 7 8 9 10 11
Capability of performing simple maintenance Overhaul, refurbishment and rudimentary modification capabilities Assembly of imported components, simple licensed production Local production of components or raw materials Final assembly of less sophisticated weapons; some local component production Co-production or complete licensed production of less sophisticated weapons Limited R&D improvements to local license-produced arms Limited independent production of less sophisticated weapons; limited production of more advanced weapons Independent R&D and production of less sophisticated weapons Independent R&D and production of advanced arms with foreign components Completely independent R&D and production
Source: Krause (1992a, p. 171).
Conversely, some countries appear to have chosen a goal stage that cannot be achieved given their capacities (e.g., Israel, South Africa, Egypt, India), and yet other countries have chosen to produce weaponry at a stage below their likely capabilities (e.g., Mexico). Second, it is increasingly apparent that there should be at least two, or even three, tables of stages of arms production, one table referring to weapon platforms (call it "Krause-table 1"), the other to weapons and associated control-units and sub-systems ("Krause-table 2"). Since control-units and sub-systems are often electronic modules, one might even distinguish between complete weapon systems production and weapon module production ("Krause-table 2a" and "Krause-table 2b"). For example, a country may be able to produce a coast-guard patrol craft but still need to import the actual weapon to be mounted onto the platform. Or, of the weapon to be mounted, it may be able to produce some, but not all, needed components or modules. Thus, any given country may be relatively advanced with respect to the production of platforms, but rank low with respect to producing weapons or weapons components. A few years ago, for instance, Indonesia bought naval vessels from the former East Germany, stripped the vessels of everything but the hull, and imported new weaponry and guidance systems to be
106 Arming the South
mounted onto the old platform. Once the conceptual separation between weapons and platforms is made, one immediately realizes that much of the activity involved in platform production is generic work with dual-use features applicable to both military and civilian purposes. For instance, the famous export success of Brazil's arms industry in the 1980s relied in good measure on the success of its civilian commuter and trainer aircraft (Franko-Jones, 1992), a point to which I return later on. Non-economic motives Observers are united in their opinion that the initial motivation for indigenous arms production in developing nations almost always is strategic. Chief among the strategic reasons are weapon embargoes or other threats to an existing arms import supply line. For example, in the 1950s both China and Egypt used to rely on Soviet weaponry before the supply line became uncertain and unreliable, and both went on to develop indigenous arms industries. Taiwan and South Africa suffered generalized arms supply embargoes, and both went on to set up indigenous arms industries. Turkey and Brazil suffered specific arms supply embargoes from the administration of US President Carter, and both went on to build up indigenous arms industries (see, e.g., Pearson, 1994, pp. 19-20, generally; on Turkey, see Giinluk-§enesen, 1993; on Brazil see Franko-Jones, 1992). Noting this pattern, some analysts suggest that the increasing threat, in the post-cold war era, to impose arms embargoes of various sorts might drive even more countries into efforts to produce indigenous weaponry and to achieve some degree of selfsufficiency in at least some arms category (e.g., Baek et al., 1989; Brzoska, 1995, p. 28). This suggestion may not hold for the new century, however, since unlike the 1960s, there are so many more alternative weapon supply lines available that an embargoed country could draw upon. A relatively new version of the acute strategic motive is the preemptive strategic motive. It runs as follows: even if no actual conflict involving the country is thinkable, a country may nonetheless wish to produce arms indigenouslyy't/s/ in case a conflict emerges. Moreover, the
The arms industiy in developing nations
107
arms in question will have to be suffused with high-technology so as not to place oneself at a military disadvantage in case an armed conflict actually arises. Pearson (1994, p. 48) spells out this argument for Brazil, but it is illustrated to perfection by Joe Modise, a former South African Minister of Defense. Defending continued South African arms production, he is quoted as follows (in Batchelor, 1995, p. 2): "The immediate danger lies in the instability around us. We have to face up to it and cannot safely assume that it will not spill over into South Africa or otherwise affect our interests." Several authors speak of political motives for arms production as separate from strategic ones. For the most part, these involve considerations of foreign policy and the potential influence and leverage that one may bring to bear on the recipients of one's arms production and arms exports (e.g., Krause, 1992a; Smith, Humm, Fontanel, 1985). Exceptions notwithstanding, this "arms-for-good-behavior" policy was predominantly the domain of the so-called first and second-tier producers, i.e., of the US, Russia, France, and Britain in particular. This kind of influence peddling is on the decline since the number of arms export desperate second-tier producers has increased rapidly in recent years and now includes a number of former developing nations such as the aforementioned South Korea, Taiwan, Singapore, Israel, Spain, and Portugal, i.e., former third-tier producers that have "graduated" to secondtier rank in some respects. This means that desperate arms importers are now offered a broad choice of suppliers - the arms market is a buyers' market - thus reducing the effectiveness of influence peddling, putting purely political considerations on the back-burner and bringing commercial considerations into the foreground. This situation is likely to persist. Economic motives The effective limiting factor to indigenous arms production is industrial capability, including the human-capital constraint of sufficiently welltrained production personnel, and scientists and engineers. That is the overriding expert opinion (e.g., Wulf, 1987; Ball, 1988, p. 375; Brauer, 1991a; Brauer, 2000; Krause, 1992b, p. 141). But perhaps, or so the
108 Arm ing the Sou th
reasoning once went, one can simultaneously kill two birds with one stone: engage in capacity building by investing in an indigenous arms industry. The specific form of this hope comes in two major clusters of arguments. They are (a) arms production as industrial policy and (b) export-promotion industrialization and foreign-exchange earnings. For any one specific country, these two arguments might be used in combination rather than in the isolation in which I present them. (An additional "economic motive" argument - I refer to it as the distress argument ~ is dealt with toward the end of this section.) A R M S PRODUCTION AS INDUSTRIAL POLICY. Among the earliest arguments, in line with the state of thinking in the field of development economics at the time, was that indigenous arms production could be viewed as a form of import-substitution industrialization. The logic of the argument roughly runs as follows. We will install an arms industry. Not only will this be useful for strategic reasons, but there will be specific economic benefits. An indigenous arms industry will prevent or mitigate "brain drain," i.e., it will keep our best scientists and engineers in the country. It will tell us what other industries we need to build up in order to build arms. Therefore, upon an arms industrial base, generalized industrialization and capacity building will follow. The arms industry will be our leading sector, promoting backward linkages to support industries, especially in heavy manufacturing and the chemical, electrical, and electronic industries. It will also provide industrial spin-offs useful for civilian industry. Our arms industry will be a pod from which seeds of generalized industrialization will grow up; it will be a development pole around which other economically useful industrial activities will cluster. Moreover, over time as the first and second-tier arms exporters compete for markets, we gain increasing leverage - in a buyers' market - to ask for and receive co-production agreements, licensed production, and various offset and barter deals, even if unrelated to arms (see, e.g., Hartung, 1994, p. 249 on South Korea), in which our agreement to buy some arms (or licenses) from others allows us not only to produce or co-produce our own arms, but obligates others in return to buy something else from us, thereby spurring on that unrelated market as well. This line of reasoning has been advanced, in one form or another, for countries such as Turkey and Brazil, South Korea and Taiwan, Israel,
The arms industry in developing nations
109
Spain, and South Africa (on the latter, see Batchelor, 1995, p. 13). For instance, Gunluk-$enesen (1993, p. 260) paraphrases a 1991 news interview with the head of Turkey's Under-Secretariat for Defense Industry as follows: It has been officially expected that domestic spill-over effects from the modern arms industry would include more diverse industrial production, more efficient production, product quality improvement, foreign exchange savings, acceleration of economic growth, increased value added, less unemployment, increases in the overall technology level, and improvement of the quality of the labour force and university education, especially engineering.
What is the evidence? After the Cyprus war, Turkey came under an arms embargo by the US, the first-tier producer that attempted to use its arms exports to Turkey as a policy tool to influence Turkey and events in the Aegean Sea. In response, Turkey developed an arms import substitution program with the intention of using it as a springboard to a generalized industrialization of the requisite underlying support industries. But, as would be true for other countries, Turkey's program was quickly broadened, as regards arms, to the desire to secure multiplesource weapon-supply lines, whether produced indigenously or not, irrespective of whether or not generalized industrialization would follow on the heels of indigenous arms production. Thus, in addition to building requisite industrial infrastructure to support its indigenous efforts, Turkey began a variety of co-production and licensing ventures with countries such as Spain, France, Germany, Italy, and eventually even with the US itself. (On Turkey, see Ayres, 1983, and Giinluk-§enesen, 1993; see Nolan, 1986, on South Korea and Taiwan; and see, generally, Sandier and Hartley, 1995, pp. 187-188). In turn, the case for secure, diversified, multiple-source supply lines eventually gave way to the idea that supply diversification by an arms buyer should permit it to emerge as a niche supplier in its own right, i.e., once our country has achieved some degree of military industry, it could subsidize these efforts by exporting weapons. The case especially of Brazil (Franko-Jones, 1992) illustrates this further shift of emphasis and justification. What started as a notion of arms import substitution became arms-production-as-industrial-policy (Pearson, 1994, p. 33), with or
110 Arming the South
without an arms export component attached. If the pure case of arms import substitution industrialization fails, as the broadening of its scope and mandate suggests, why did it do so? In answer, consider this quote from an early and prominent writer on the subject. Nicole Ball (1988, p. 375) writes: Although the establishment of a domestic capacity to produce arms is often justified in Third World countries by the technological spin-offs that can be expected from the military to the civil sector, present experience suggests that primarily the military sector benefits from know-how and other resources already available in civil-sector industry. [My emphasis.]
This point is empirically demonstrated in Brauer (1991a, 2000): at issue is not military-led industrialization but civilian-led military possibilities. Giinluk-§enesen (1993) documents well how purely indigenous Turkish efforts remain at low technical sophistication and limited spin-offs, and how the more advanced arms production efforts are suffused with foreign co-production, licenses, design assistance, and the like. It is not that analysts dismiss, out of hand, the potential for success of an arms import substitution industrialization strategy. It is, rather, primarily a matter of the degree to which technology and scientific and engineering technology knowledge embodied in products and production processes are available to developing nations. Michael Brzoska puts the point nicely. With respect to India, he writes (1989, p. 514): The Indians have repeatedly bought the newest available technology and then tried to advance from there on their own. After a while, finding themselves falling behind, they again bought the newest technology on the international market.
A country, thus, has a choice. To gain some economic benefits from arms import substitution, produce what you can in fact substitute for, but do not overreach. What many countries can, in fact, substitute for, are relatively simple items ranging from uniforms to platforms of various sophistication, i.e., items on the "Krause-table 1." As Brzoska writes (1989, p. 526): "Arms production can provide a net positive contribution to both the enhancement of military capability and the economy if production is focused on easily produced weapons." But exactly what is
The arms industry in developing nations
111
an "easily produced weapon" depends on already established capabilities. Where many countries ambitiously overreach is on items on the "Krause-table 2." Some countries have been more judicious than ambitious in their arms related decision-making and planning. Nolan's (1986) accounts of the South Korean and Taiwanese arms industries, Bitzinger's (1995) evaluation of South Korea's arms industry, and Anderton's (1995,1996) assessment of mainland China's arms industry and conversion efforts, all suggest a more deliberate, more closely integrated and supervised, and therefore potentially more successful interaction between military and civilian production. But even in the South Korean case of a technically fairly advanced arms industry, Bitzinger writes that "the presence of an already well-established, domestic heavy industrial base," in addition to "extensive foreign assistance" was key to propel the nation's indigenous arms production efforts forward (1995, p. 236). And even then, "South Korean arms exports have consisted mostly of low-tech items such as uniforms, nonlethal military equipment, small arms and ammunition, and patrol boats" (1995, p. 243). Why is this so? Bitzinger summarizes the reasons: (a) lack of interest by South Korean industry because private industry could make more profits in other industrial branches (Korea's arms industry, like Turkey's and Brazil's, is heavily privatized); (b) lack of interest by the South Korean armed forces because of the impetus always to obtain the most advanced equipment available from overseas; (c) structural weaknesses in the defense R&D base, i.e., the lack of linkages between the knowledge establishment and the production facilities; and (d) lack of overall, long-term planning that would allow South Korea to steer resources into appropriate basic research. South Korea is bumping against a "technology plateau" (Bitzinger, 1995, p. 246), and it is highly relevant that other authors make similar observations over and over again for other countries, e.g., Gunliik-§enesen (1993) on Turkey and Franko-Jones (1992) on Brazil. In sum, one cannot reach for too much with too little. But this means only that import substitution as an arms production strategy has failed in its pure form. It does not mean that we will not see more and increased, and increasingly sophisticated, arms production efforts by developing
112 Arming the South
nations. It merely means that the form that arms production takes is changing. I will return to this point later on. EXPORT-PROMOTION INDUSTRIALIZATION AND FOREIGN-EXCHANGE EARNINGS. Turkey and South Korea are examples of nations that wished to create military-led industrialization without much thought for foreignexchange earnings. But in some countries, most prominently in Brazil, an arms export promotion strategy and foreign-exchange earnings potential was instrumental in the argument for the creation and expansion of a domestic arms industry. Especially for Brazil, whose domestic arms needs always have been modest because it does not face any substantial external security threat, arms exports would permit the subsidization of domestic production runs, would thereby permit unit-cost reductions, and would even result in foreign-exchange earnings and therefore ease balance-of-payment difficulties. Again, what is the evidence? If it were true that the foreign exchange cost of domestic arms production outweighs arms import costs, then one should observe increased foreign-exchange use as a country increasingly substitutes domestically produced for imported arms. For example, Terhal's 1982 study ("Foreign Exchange Costs of the Indian Military, 1950-1972") is often cited is this regard. But using Terhal's own data, Deger (1986, p. 137) showed that in spite of increasingly substituting domestic for imported weapons, the Indian military's foreign-exchange needs show "almost no increase at all" (see also Brauer, 1989, pp. 257266). This would suggest that there might be something to the foreignexchange argument. But, to be blunt, there does not exist a single sufficiently documented case in which a developing country exporting arms earned net foreignexchange. Part of the problem with the foreign-exchange point is that the foreign-exchange earnings from selling arms, if truthfully reported, are relatively easy to account for. In contrast, the foreign-exchange costs of domestic arms production and exports are difficult to account for (Smith, Humm, Fontanel, 1985, p. 241). For example, costs arise to import needed machinery and raw materials and experts and licenses and components and sub-systems. Had the resources devoted to arms production been poured into another export sector, perhaps larger foreignexchange earnings would have resulted. If so, these foregone earnings
The arms industry in developing nations
113
would have to be netted out against the actual arms export earnings. The foreign-exchange savings that arise by not importing arms have to be taken into account as well. No one has tried to bring all these facets together into a single, comprehensive, empirical study so that the net foreign-exchange benefit or cost could be assessed, although some (e.g., Batchelor, 1995, pp. 19-21) have made a good effort at least to enumerate some of the usually invisible foreign-exchange costs associated with indigenous arms production (for example, export infrastructure, export incentives; counter-trade and offset deals). In any event, it appears that even the evidence in support of appreciable export earnings is lacking. For the single-best case that I am aware of, Franko-Jones writes with respect to a particular year, 1982, of the Brazilian aircraft maker EMBRAER that "for every dollar of imports, the industry generated two dollars of export earnings" (1992, p. 159). That is the very best foot that has been put forward: a single year, a single industry, a single country. Surely, this does not amount to making the case that net foreign-exchange earnings are in fact commonplace in developing nations' arms industry. Moreover, Franko-Jones points out, as have others, that Brazil's arms exports on occasion seem vastly overstated (pp. 140-147), and this before taking implicit foreign-exchange costs into account. The foreign-exchange earnings case is thin indeed. It should be mentioned that parts of the extant literature overstate the case against arms export promotion and foreign-exchange earnings through arms exports. In the literature, it is far more common (e.g., Krause, 1992a, p. 166 as well as note 31 on p. 257) to find grand, unequivocal statements to the effect that the cost of imported components and subsystems is greater than the cost of importing the whole weapons system than to find carefully documented cases. For instance, Raimo Vayrynen (1992, p. 98) claims: "A detailed scrutiny of the export and import patterns of the Brazilian aeronautical industry in the 1970s and the early 1980s suggests that the imports of various components clearly exceeded the value of Embraer's export." But other writers, in contrast, stress Brazil's deliberate and heavy use of domestically produced inputs (preproducts) that went into the Brazilian arms output (Franko-Jones, 1992) so that a writer as unsuspected of arms industry sympathies as Brzoska (1989, p. 516) concludes that "the Brazilian arms industry seems
114 Arming the South
to have been an important foreign exchange earner from the mid-1970s." Thus, despite the assured tone of certain pronouncements, and in spite of spectacular cases such as Israel's Lavi fighter, that would tend to support the critics of the foreign-exchange earnings argument, by far the best evidence that the net foreign exchange effect is negative is not that anybody has conclusively proven anything about the import content and foreign-exchange cost of indigenously produced arms; rather, the best evidence is that no one has made a nearly convincing case that the net earnings are in fact positive. In any event, Brazil's arms exports in particular rode high on a "speculative bubble," namely the Iran-Iraq war throughout the 1980s. Franko-Jones (1992) presents a nuanced analysis of why Brazil's arms industry collapsed in the late 1980s and early 1990s. Based on her research and interviews, she believes that the bursting of the arms export bubble happened to coincide with a fatal misjudgment and with external circumstance: the misjudgment was that exactly at the time when the arms export bubble burst, Brazil also decided to move up the ladder of technological sophistication (up on the "Krause-table 2") so that the cost of imported components rose at a time when export earnings fell, putting many of its largely privately owned military-industrial firms into debt. In addition, the external macroeconomic environment of Brazil at the time - 1,000 percent inflation, several currency reforms, structural adjustment pressures from the IMF, political turmoil - served to divert attention from what was up until then a carefully orchestrated public-private arms industry partnership. In sum, we do not have a definite study that would conclusively settle the matter of the export promotion and foreign-exchange earnings argument. But we do know (a) that foreign-exchange earnings are at times overstated, (b) that foreign-exchange costs are not properly accounted for; and (c) that no one has provided conclusive evidence that the net effect is positive for the arms exporting country. Krause concludes with the suggestion that "arms production may... not provide major spinoff benefits, but may at least ameliorate the negative impact of military spending" (1992a, p. 167), a finding empirically confirmed in Brauer (1993) with respect to military expenditures, arms production, and economic performance in general. Specifically with respect to foreign
The arms industry in developing nations
115
exchange, Brauer (1991b) finds that military investments by developing nations appear to result in domestic absorption rather than foreignexchange leakages, again possibly mitigating the foreign-exchange effect of domestic arms production as compared to the foreign-exchange costs of arms imports. In either case, however, is it not clear that the mitigation effect is due solely to the existence of indigenous arms production. Instead, relative to non-arms producing developing nations, the effect might be due to the fact that arms producing developing nations generally exhibit a stronger economy to begin with so that they could absorb more easily any negative effects stemming from indigenous arms production and military expenditure. An export promotion industrialization strategy via arms exports is a niche market effort at best. Niche markets are vulnerable to rapidly changing market conditions. They can grow "soft," as the collapse of the Brazilian arms industry showed when petroleum prices fell and arms demand from the Middle East for niche market suppliers dried up. An interesting, new development is taking place in South Africa. As Peter Batchelor observes (1995, p. 13), there is some desire by South Africa's neighbors to want to have a South Africa nearby that is arms manufacturing capable and able and willing to refurbish and service weapons - a sentiment that is shared by officials at ARMSCOR, the South African arms marketing agency. In this regard it is noteworthy to recall that before Brazil became a big player in the global arms market during the Iran-Iraq war, it was in fact a big player in the regional arms market in South America, and it is entirely possible that the future will bring more focused efforts by one or the other regional behemoth to take on the role of primary regional arms supplier, at least for certain categories of weaponry. THE DISTRESS ARGUMENT. Assorted other claims about benefits from indigenous arms production have been made. Relatively new is the distress argument. It consists of two components. The first goes as follows: since we already have an arms industry and poured so much money into it, we cannot just abandon it. For if we do, all the cost sunk into the arms industrial projects will be lost for good and be wasted (for examples, see, e.g., South Africa under Mandela in 1994; the CSSR under Havel in 1989; and, generally, Ball, 1988, ch. 9). The second part
116 Arming the South
is this: in the current arms sales market climate - oversupply, relative lack of demand, and therefore fire-sale prices - the prices we obtain do not cover our average costs. To bring down average cost we need to increase the quantity of our arms sales, thus not only avoiding losses but saving jobs. I am not suggesting that expert observers make these arguments; rather I am saying that one hears this line of reasoning expounded in public debates about indigenous arms industries. This line of argument is an example of the fallacy of sunk costs. A simple example illustrates the point. If fixed costs incurred for research and development, set-up of production lines, etc. amount to $1,000 and each unit actually produced amounts to an additional or incremental charge of $100 for materials and direct labor (i.e., a marginal cost of $100 per unit), then the total cost of producing ten units is $1,000 + (10 x $100) or $2,000. Sold at a price of $200 each for a total revenue of $2,000, the profit, from an accountant's point of view, is zero. If market conditions are adverse and the average price that can be obtained falls to $150, an accounting loss of $500 results (total revenue of $1,500 minus total cost of $2,000). In response, an effort might be made to expand production to 20 units. Total costs now run at $1,000 + (20 x $100) or $3,000, for an average unit-cost of $150. Selling each unit at $150 restores the accounting profit back to zero. Thus, in public debates apparently stark choices of "economic reality" are offered to an unsuspecting public: either do not expand arms production and lose money in the short-run by taking losses on each unit sold and also lose money in the long-run by having to close down production lines altogether (loss of unrecovered fixed costs and loss of jobs) or expand arms production, arms sales, and arms exports, even if one might have moral quibbles about doing so. (This argument works special wonders when linked to the notion that it is not arms sellers but arms users who are the morally responsible parties.) From an economist's point of view, the cost argument is different. Once the up-front cost for research and development is committed or sunk into the project, the deed cannot be undone - these costs are "lost" in any event. The only economically relevant point is to make offsetting contributions toward defraying the sunk costs. Thus, if producing one more unit incurs an incremental cost of $100, then any price obtained
The arms industry in developing nations
117
above $ 100 will make a positive contribution to the recovery of the sunk cost. Consequently, following the earlier numeric example, if the price of arms falls from $200 to $ 150, there will still be a positive contribution of $50 toward offsetting the up-front cost (Smith, Humm, Fontanel, 1985, pp. 242-244). Since every experienced business manager, including every arms industry manager, is familiar with this elementary economic principle that decision-making is driven not by average but by marginal costs - my suggestion here is that it is likely that the distress argument is a hoax hoisted upon an unsuspecting public, aimed at evoking nationalistic sympathies toward potentially unemployed workers, even if that implies shunting aside moral scruples in favor of expanded arms production and arms export sales. If the public debate is won by this kind of incorrect argument, arms sellers will be less cautious as to whom they sell weapons and weapons technology to. Assessing the developing world's arms industry Summarizing what has been said so far, developing nations possess legitimate security interests and some of them - for strategic reasons wish to build up their own, indigenous arms industry. But the supposedly positive economic benefits do not exist. Military-led, generalized industrialization, an occasional exception notwithstanding, is virtually always dependent on the prior state of civilian industrial accomplishment, and as regards arms exports, there does not exist a convincing study showing net foreign exchange earnings for any country for any sustained period of time. Where does this state of affairs leave us? I agree with Stephanie Neuman's post-cold war assessment. She writes: "worldwide the military sector is shrinking. Democratic regimes are replacing military rulers. Defense spending is down. Military industries are going out of business. Defense production is declining, military assistance is dwindling, and the arms trade has contracted dramatically." (Quoted in Pearson, 1994, p. 26, based on a conference paper Neuman delivered in 1993.) But, along with a number of experts, I also share misgivings about certain new trends and features of arms production. I take up three themes.
118 Arming the South
LDCs. First, it should be remarkable that former developing nations such as South Korea, Taiwan, Spain, Portugal, and Israel are now "graduated" in a number of arms production capabilities to the ranks of Sweden, Austria, Switzerland, Japan, Canada, and similar countnes, i.e., they have become just another set of industrialized nations. Of all of these countries, none are entirely self-reliant in arms production. Globally, therefore, we note a huge convergence and falling out at the same time. The US and Russia (for the time being, but perhaps not for long) are first-tier producers, who are completely and independently able to design and construct highly sophisticated weaponry across the entire weapons spectrum (stage 11 on the "Krause-table"). Then there is a large cohort of second-tier arms producers including, increasingly, France, Britain, and the aforementioned countries whose primary distinguishing mark is the ever increasing transnationalization of arms design and production - rather like the automobile or other transnational industries. The array of co-production, licensing, joint development, replete with complex counter-trades and offsets generates an ever more dense maze of mind-boggling transnational networks of arms producers. It is very important to differentiate between internationalization and transnationalization of production, /ftternationalization means that a firm operates production facilities in a variety of countries that are relatively independent of one another. In contrast, ^^nationalization means that production and production facilities across countries are interdependent with one another. This is the new technological imperative affecting all globalized production, not just arms production. Instead of producing from the "ground-up," production - including arms production - becomes modular where systems are co-developed across nations and plugged in wherever in the world they are needed. This encourages component buying, modifications, retrofitting, and re-exporting activities in an increasingly Adam Smithian free armaments market. It is co-mingled with the increasingly dual-use nature of components and platforms. Arms production becomes more fluid, more like any other normal industry (see Pearson, 1994, p. 27; Hartung, 1994, ch. 12; Skons and Wulf, 1994; Gold, 1995). It is the era of the two "Krause-tables" that separate capabilities and accomplishments in platform production from module and components production. Some former developing nations, such as the THE MISSING
The arms industry in developing nations
119
aforementioned South Korea, Spain, Portugal, etc., have been able to catch up with the industrialized world, but by far the majority of developing countries are barely able to produce platforms, let alone modern weapons. Following the converging second tier of arms producers, which collectively but not individually, could match the US and Russian arms production capabilities, there is a third tier of arms producers. These are countries that are falling behind technologically. The best of them produce good platforms ("Krause-table 1"), but as regards weapons, control-systems, and sophisticated sub-systems ("Krause-table 2") they are highly import dependent (Brzoska, 1995, 29). This would include countries such as Egypt, Turkey (at the moment), Indonesia, India and Pakistan, Brazil, Argentina, Iraq, Iran, and others. At some point, these nations either will slip up to second tier-status - on account of expanded civilian industrial capacities - or recognize, as some already have, that their only, and crazy, hope lies in building mad-man's weapons of mass destruction. (There is a fourth tier to which I turn in a moment.) ARMS CONTROL. The second major theme concerns arms control. Pearson, for example, notes that "today more countries have more varied and advanced types of equipment from more numerous suppliers" (Pearson, 1994, p. 27). Consequently, writers such as Bitencourt (1995, p. 172-173) suggest that control of conventional arms will become more difficult as there are more suppliers, fewer power-bloc alignments, and more broadly commercial interests by developing and developed countries alike, compounded by increasing dual-use technology usage and developing countries' legitimate interests in high-technology access and use (e.g., in space and therefore missile technology). In this context, I believe that Bitencourt correctly identifies "... growing [LDC] mistrust surrounding the [arms] control regimes. These are frequently viewed as a... disguise used by developed countries because they are seen as having a higher influence on the organizations interested in non-proliferation, and can thus keep their dominance of the market" (Bitencourt, 1995, p. 174). As noted earlier, in the past supply embargoes merely induced recipients to enter the market as producers themselves and, if possible, as exporters. This suggests that, at a minimum, supply embargoes be mixed
120 Arming the South
with an exit strategy, lest the embargoed nation feels compelled to devote scarce resources to armaments production or to develop extensive and costly multiple arms supply lines. But there is a superior approach to arms control, at least in theory. Among others, Bitencourt (1995, p. 175) is entirely correct in pointing out that instead of controlling the supply of arms, what is needed is to encourage regional peace treaties to reduce the demand for arms and arms production (also see Brauer, 1991a, 2000). As in any market, the supply will then take care of itself. Registers of conventional arms and missile control regimes are good ideas and items, but control on the demand side - as the end of the cold war amply demonstrates - works much more effectively to reign into the armaments market. Even if armaments were restrained from the demand side, arms production and war are, despite the current lull (if that's the term to use), not irreversible. The ultimate objective must be to put incentives and institutions into place that make peace achievable and, once achieved, irreversible so that resources saved may be applied to other economic needs. But irreversibility cannot be achieved technologically since the capacity to produce industrial goods implies the capacity to produce arms (Brauer, 1991a, 2000). There is no technology fix to the reversibility problem. The fix must be one of institutions and proper incentives. For example, security between and among nations will be strengthened when regional peace treaties reduce the perception of threat and therefore reduce the demand for weapons (Ball, 1993, p. 342), when the international "self-help" system (Krause, 1992a, 1992b) gives way to cooperative security arrangements that transcend nation-state borders. Yet it must be realized that most conflicts are not of the interstate but of the intrastate variety and that the coming about of any peace treaty itself is subject to incentives and disincentives, and to policing and enforcement activities and costs (on incentives, see, e.g., Cortright, 1997; Garrett, 1997; Brauer and Roux, 2000). SMALL ARMS. Finally, the third major theme in assessing developing nations' arms industry concerns so-called small arms. Obviously, nuclear and major conventional weapons do present substantial dangers, indeed catastrophic dangers. But the fact of the matter is that most wars and other violent conflicts take place in developing nations that are
The arms industry in developing nations
121
characterized by the absence of major weapons: people are primarily killed by dismissively labeled "small arms." Small arms are formidable killers. Ninety percent of all conflict deaths or injuries were, by one accounting, attributable to small arms, and per injured or killed person another twenty or so were displaced or uprooted by the immediate, personal threat these arms pose (Rana, 1995, p. 1). Small arms are particularly prevalent in intrastate rather than interstate conflict, by far the majority of violent conflicts nowadays. Small arms are easily hidden and, therefore, smuggled. At present the world market is flooded with small arms, ammunition, land mines, and the like, making these killers gruesomely cheap. Moreover, they are easy to manufacture even by technologically impaired nations. Since it does not take much by way of civilian industry to produce small arms, we may expect more countries to enter the fray. But that is not the major point here. It is, rather, that if economists wish to account for the cost of indigenous arms production, or any arms production, or any conflict made possible by arms, we must turn away from merely accounting for narrowly construed direct and indirect costs of preparing for conflict. In particular, we finally need to pay much more attention to counting the cost of actual war, i.e., the cost in terms of economic and human development foregone. Initial attempts have been made (see, e.g., Deger and Sen, 1990; Brauer, 1996; and part III in Brauer and Gissy, 1997). Beyond that, it should be economists' task to consider the incentives and institutions that give rise to conflict, war, and peace, and again little work has been done in this regard (Sen, 1992; part IV of Brauer and Gissy, 1997; Brauer and Roux, 2000). Conclusion To summarize the major topics of this wide-ranging chapter, consider the following points: 1
From the early 1980s to the late 1990s, a number of formerly developing nations have "graduated" from relatively low levels and sophistication of arms production to relatively high levels (e.g., South Korea, Taiwan, Singapore, Spain, Portugal, Israel). This coincides
122 Arming the South
with the continued development of their civilian industrial capabilities. Among the remaining developing nations, as of the late 1990s, between 25 and 35 are engaged in some form of arms production and arms (re)exports. 2
"Arms production" is a misleading term. At a minimum, one must distinguish between the production of platforms (air, land, sea) and the weapons themselves. An increasing number of countries are able to build platforms in the course of the development of their civilian transport industries, but are still unable to build weapons, controlunits, and sub-systems. By far, the primary motive for indigenous arms production is acute strategic need. In recent years, a number of countries plead a preemptive strategic need, "just in case" a conflict should emerge.
3
Strategic motivations apart, indigenous arms production efforts have been justified on economic grounds. The two major "economic" arguments are: (1) that building up an indigenous arms industry will spur generalized industrialization by means of spill-over or spin-off effects; and (2) that building up an indigenous arms industry and arms export sales will permit foreign-exchange earnings. On the first point, the evidence is clear: if anything, the development of indigenous arms industries in developing nations depends crucially on already established civilian capacities. As regards foreign-exchange earnings, the evidence is less clear-cut but three points need be kept in mind: (a) foreign-exchange earnings often appear overstated; (b) foreignexchange costs are never fully counted, in part because such costs are socialized rather than borne by the firm or accounting unit that actually produces armaments for export; and (c) no one has ever presented an uncontroversially convincing case that the net foreignexchange effect is in favor of arms exports.
4
The recent emergence of the distress argument relies on generating public sympathies with faulty economics. The relevant costs that flow into decision-making are not average costs, but incremental costs. Low incremental costs, combined with the urge to recover sunk costs,
The arms industiy in developing nations
123
foster domestic and arms export sales at low market prices. This alone makes arms control from the supply side difficult to achieve. 5
Developing nations can graduate to higher levels of arms production sophistication as the underlying civilian capabilities increase and as they become more integrated into the ^/^nationalization of arms production efforts. This also complicates arms control efforts aimed at the supply side. Indeed, supply restrictions tend to drive up prices, thus providing an incentive for new suppliers to enter the market. The real killer, in any event, are "small arms," which are easy to manufacture, transport, smuggle, maintain, use, and pay for.
6
Economists, and others, might therefore consider expending more intellectual resources on understanding the structure of incentives and institutions that give rise to and/or exacerbate violent conflict and war and that prevent the peaceful settlement of conflict. Similarly, we need to study incentives and institutions that would give rise to peace and prevent war, i.e., study the economics of the demand for conflict and the derived demand for armaments. Finally, even economists that are expert in the study of the economics of conflict tend to focus on the economic cost of preparing for war. We would do well also to study the cost of actual war and of its aftermath. For example, the slaughter in Rwanda in 1994 will economically cripple the country for decades to come - surely a cost much greater than the cost of prevention would have been.
References ACDA [US Arms Control and Disarmament Agency]. World Military Expenditures and Arms Transfers, 1996. Washington, DC: ACDA, July 1997. Anderton, Charles. "Assessing the Difficulty of Chinese Defense Enterprise Transition." Peace Economics, Peace Science, and Public Policy Vol. 3, No. 2 (Winter 1996), pp. 13-21. Anderton, Charles. "Defense Conversion in China: From Swords to Plowshares?" Peace Economics, Peace Science, and Public Policy
124 A rm ing the Sou th
Vol. 2, No. 2 (Winter 1995), pp. 19-31. Ayres, Ron. "Arms Production as a Form of Import-Substituting Industrialization: The Turkish Case." World Development Vol. 11, No. 9 (1983), pp. 813-823. Baek, Kwang-Il, Ronald D. McLaurin, and Chung-in Moon (eds.). The Dilemma of Third World Defense Industries: Supplier Control or Recipient Autonomy? Boulder, CO: Westview Press, 1989. Ball, Nicole. "Disarmament and Development in the Third World," pp. 333-343 in Richard D. Burns (ed.), Encyclopedia of Arms Control and Disarmament, Vol. 1. New York: Scribner's Sons, 1993. Ball, Nicole. Security and Economy in the Third World. Princeton, NJ: Princeton University Press, 1988. Batchelor, Peter. "The Economics of South Africa's Arms Trade." Discussion paper #3. University of Cape Town: Centre for Conflict Resolution, August 1995. Bitencourt, Luis. "The Problems of Defence Industrialization for Developing States," pp. 167-175 in Sverre Lodgaard and Robert L. Pfaltzgraff, Jr. (eds.). Arms and Technology Transfers: Security and Economic Considerations Among Importing and Exporting States. New York and Geneva: United Nations Institute for Disarmament Research (UNIDIR), 1995. Bitzinger, Richard. "South Korea's Defense Industry at the Crossroads." The Korean Journal of Defense Analysis Vol. 7, No. 1 (Summer 1995), pp. 233-249. Brauer, Jurgen. "Potential and Actual Arms Production: Implications for the Arms Trade Debate." Defence and Peace Economics Vol. 11 (2000), pp. 461-480. Brauer, Jurgen and Andre Roux. "Peace as an International Public Good: An Application to Southern Africa."Defence and Peace Economics Vol. 11 (2000), pp. 95-111. Brauer,, Jurgen and William Gissy (eds.). Economics of Conflict and Peace. Aldershot, UK: Avebury Press, 1997. Brauer, Jurgen. "Military Expenditures and Human Development Measures." Public Budgeting andFinancialManagement Vol. 8, No. 1 (Spring 1996), pp. 84-104. Brauer, Jurgen. "Defense, Growth, and Arms Production in Developing
The arms industry in developing nations
125
Nations," pp. 229-242 in J. Brauer and M. Chatterji (eds.). Economic Issues of Disarmament. New York: New York University Press, and London: Macmillan Press, 1993. Brauer, Jurgen. "Military Investments and Economic Growth in Developing Nations." Economic Development and Cultural Change Vol. 39, No. 4 (July 1991b), pp. 873-884. Brauer, Jurgen. "Arms Production in Developing Nations: The Relation to Industrial Structure, Industrial Diversification, and Human Capital Formation." Defence Economics Vol. 2, No. 2 (1991a), pp. 165-175. Brauer, Jurgen. "Military Expenditures, Arms Production, and the Economic Performance of Developing Nations." Ph.D. dissertation. University of Notre Dame, 1989. Brzoska, Michael. "Spread of Conventional Weapons Production Technology," pp. 19-47 in Sverre Lodgaard and Robert L. Pfaltzgraff, Jr. (eds.). Arms and Technology Transfers: Security and Economic Considerations Among Importing and Exporting States. New York and Geneva: United Nations Institute for Disarmament Research (UNIDIR), 1995. Brzoska, Michael. "The Impact of Arms Production in the Third World." Armed Forces and Society Vol. 15, No. 4 (Summer 1989), pp. 507530. Cortright, David (ed.). The Price of Peace: Incentives and International Conflict Prevention. Lanham, MD: Rowman & Littlefield, 1997. Deger, Saadet. Military Expenditure in Third World Countries: The Economic Effects. London: Routledge & Kegan Paul, 1986. Deger, Saadet and Somnath Sen. Arms and the Child. Stockholm: SIPRI, 1990. Franko-Jones, Patrice. The Brazilian Defense Industry. Boulder, CO: Westview Press, 1992. Garrett, John. "Policies for Peace: An Analysis of the Causes of Military Expenditures and the Means to Disarmament," pp. 355-375 in J. Brauer and W. Gissy (eds.). Economics of Conflict and Peace. Aldershot, UK: Avebury Press, 1997. Gunliik-§enesen, Giilay. "Turkey: The Arms Industry Modernization Programme," pp. 251-267 in Herbert Wulf (ed.). Arms Industry Limited. Oxford: Oxford University Press, 1993.
126 Arming the South
Gold, David. "The Internationalisation of Military Production," pp. 105-1114 in Manas Chatterji, Jacques Fontanel, and Akira Hattori (eds.). Arms Spending, Development and Security. Ashish Publishing House: New Delhi, 1995. Hartung, William. And Weaponsfor All. New York: HarperCollins, 1994. Krause, Keith. Arms and the State: Patterns of Military Production and Trade. Cambridge: Cambridge University Press, 1992a. Krause, Keith. "Arms Imports, Arms Production, and the Quest for Security in the Third World," pp. 121-142 in Brian L. Job (ed.). The Insecurity Dilemma: National Security of Third World States. Boulder, CO: Lynne Rienner Publ., 1992b. Nolan, Janne. Military Industry in Taiwan and South Korea. London: Macmillan, 1986. Pearson, Frederic. The Global Spread of Arms. Boulder, CO: Westview Press, 1994. Rana, Swadesh. "Small Arms and Intra-State Conflicts." Research paper #34. Geneva: United Nations Institute for Disarmament Research (UNIDIR), March 1995. Sandier, Todd and Keith Hartley. The Economics of Defense. Cambridge, UK: Cambridge University Press, 1995. Sen, Amartya. "Wars and Famines: On Divisions and Incentives," pp. 219-233 in Walter Isard and Charles Anderton (eds.). Economics of Arms Reduction and the Peace Process. Amsterdam and New York: Elsevier, 1992. Skons, Elizabeth and Herbert Wulf. "The Internationalization of the Arms Industry," Annals of the American Academy of Political and Social Science, Vol. 535 (September 1994), pp. 43-57. Smith, Ron, Anthony Humm, and Jacques Fontanel. "The Economics of Exporting Arms." Journal of Peace Research Vol. 2, No. 3 (1985), pp. 239-247. Terhal, Peter. "Foreign Exchange Costs of the Indian Military, 19501972." Journal of Peace Research Vol. 19, No. 3 (1982), pp. 251259 Vayrynen, Raimo. Military Industrialization and Economic Development: Theory and Historical Case Studies. Aldershot, UK: Dartmouth, 1992.
The arms industry in developing nations
127
World Bank. World Development Report 1997. New York: Oxford University Press, 1997. Wulf, Herbert. "Arms Production in Third World Countries, Effects on Industrialization," pp. 357-383 in Christian Schmidt (ed.). The Economics of Military Expenditures: Military Expenditures, Economic Growth, and Fluctuations. New York: St. Martin's Press, 1987.
This page intentionally left blank
6 Domestic Production as an Alternative to Importing Arms Fotis Mouzakis
Introduction For the majority of countries around the world, the provision of arms for the needs of national defense is established as an inflexible and significant component of their national budgets. The amount of information that the public is provided, however, in support of the related policy decisions is minimal, and so is the related political debate. Unless, therefore, the patriotism of the decision-makers is seen as an adequate guarantee for the optimality and correctness of defense policies, transparency and open monitoring could well reveal the contrary in many cases. Economic analysis of defense policies appears to be rather neglected, leaving implementation mainly to practical judgement, instead of resting it on a systematic foundation of facts. In an attempt to improve the availability of analytical tools that can be used in a systematic way this, chapter sets up a neoclassical model of national defense, focusing on countries' decisions to establish an arms industry. If a domestic arms industry is not already in place, the government may consider its establishment, providing they can meet the necessary cost, or avoid this burden and rely on imported arms and labor. In practice, the extent to which a domestic industry covers the domestic needs for arms varies from elementary to fully-grown production of a
130 Arming the South
wide range of major arms. And, since continuous technological innovation is the critical factor that largely determines the efficiency of arms production in practice, the availability of a competitive R&D sector becomes a necessary condition for the maintenance of this industry. The foundations of the modern economic theory on arms races reaches back to Richardson (1960) and Intriligator (1975). Among more recent developments, Garfinkel (1990) examines the allocation of resources to defense and the strategic interaction of a pair of countries involved in an arms race. Levine and Smith (1997) have developed a quasi-general equilibrium model of the world market of arms, including a sophisticated formulation of security and military organization, by separating countries into pure recipients and pure producers of arms. A well-documented contemporary review of empirical work, closely related to our area of interest is provided in Smith (1995) who discusses theoretical and empirical issues related to the study of the demand for military expenditure and the total amount of resources dedicated to defense industry. Smith, Humm, and Fontanel (1987) take a closer look at the issue of the theoretical foundation of an econometric study of the defense industry. They also examine issues of substitution among inputs and base their model on a constant elasticity of substitution (CES) production function. They also outline the importance of the provision of an appropriate modeling foundation to empirical research. In recent empirical studies, Levine, Mouzakis, and Smith (1996; 1998) (LMS) have claimed that the demand for arms imports follows rather predictable patterns. Using national data for 38 countries, LMS estimated that the demand for imports responds negatively to the price of arms and positively to the size of the military budget. Moreover, they found evidence of non-proportional changes in this demand, implying relatively faster and slower responses to the price of arms when the scale of military expenditure and level of development change. The explanation proposed by LMS for the changing pattern of the demand for arms is that when military expenditure is low countries do not consider the establishment of a domestic industry. Instead, they mainly rely on imported arms, so that the demand for them appears relatively more elastic to budget increases as no other close substitutes are available. But when the military budget is high countries increasingly
Domestic production as an alternative to importing arms
131
tend to set up an arms industry of their own, partly substituting imported with domestically produced arms, and they possibly start to export arms, too. Reduced in quantity and value, the arms that still have to be imported possess a more complimentary relationship to domestic arms production, making their acquisition less flexible and their demand less elastic. In a study of panel data, Mouzakis (1999a) added further evidence to the LMS results, by looking into (nearly) global time series, dissaggregated at the continental scale level. This study addresses issues of heterogeneity bias in the results of the previous one and investigates the differences in the demand for arms across the globe. The results point out that the demand pattern is associated with the size and the economic development of countries, whereas it does not appear to change in time. The model presented in this chapter examines the optimal allocation of resources in the defense sector, allowing for the option to include a domestic arms industry. The three inputs in the production of military capability - labor, imported arms, and domestic arms - are imperfect substitutes for each other. Some minimal quantities of the first two inputs are necessary, but domestically produced arms are not. A technically demanding aspect of this set-up relates to the transition of countries into arms production and the related discontinuous change in the allocation of resources for defense. It assumes three factors of production used in defense: imported arms, domestic arms, and armed forces. It is a partial equilibrium model, because it ignores the impact of defense allocations on national factor prices, assuming that the defense sector is small enough so that these effects are negligible. It is also a small country model in the sense that it ignores the impact of each country's demand for arms imports on the world market price, which is assumed fixed. This chapter is organized into four sections and two appendices. The first section looks at empirical evidence and outlines the requirements for the analytical set-up. The second section presents the model and the third examines its properties and compares them with previous empirical results. The fourth section presents results of an econometric estimation of the model. Appendix 6A provides the analytical specification of the model used in the econometric study, and appendix 6B provides an analytical examination of the production function employed in the model.
132 Arming the South
The role of domestically produced arms Observation of available data shows that while the large majority of countries import arms (ACDA, 1995), only a few large countries possess a well-developed domestic arms industry. Of the 166 countries included in ACDA tables, there are only seven countries and the ex-Soviet republics with zero figures for arms imports for the time period 19831993. They are Albania, Bhutan, the Central African Republic, Fiji, Malta, Swaziland, and Trinidad and Tobago. All seven are very small both in terms of economic size and, with the exception of Malta, in terms of economic development. There are reasons, however, that suggest that even these seven countries do import some arms and that the zero figures are due to technical reasons. In the case of Albania for instance, the unrest of 1996-97 involved arms of Soviet origin, including tanks, artillery, and light arms. According to news reports, some of these light weapons evidently had previously been imported and eventually ended up on the black markets of neighboring countries. It is unfortunate that systematic data series with comparable measurements for countries' arms production are not generally available. SEPRI, for example, releases reports on countries' arms production in almost every Yearbook but does not include measurements of the value or volume of this output at the national level. SIPRI does, however, provide the output of the world's 100 largest arms producers in its "SEPRI Top 100" table, thus providing an approximation of the required information. Table 6.1 displays the ratio of arms imports to armed forces, using SIPRI and ACDA data. This ratio varies from 23.77 in the capital (arms import) intensive case of Switzerland to 0.13 in the labor (soldiers) intensive case of Morocco. A similar variation is seen with the use of SIPRI volumetric arms imports in constant 1990 million dollars divided by ACDA's measure of armed forces (in thousands). The large difference in the composition of inputs between these two cases should not simply be attributed to the availability of domestic arms, since Switzerland is an established producer of arms, whereas Morocco is not. Rather, the large variability of the "arms per soldier" ratio across countries indicates that in reality defense technology allows for a variety of different mixes of the
Domestic production as an alternative to importing arms
133
Table 6.1: Capital-labor ratio (arms imports per soldier) Country Switzerland Japan Portugal Norway Singapore Australia Greece Bulgaria Algeria Israel Finland Canada Netherlands Egypt Belgium Germany, FR Venezuela Taiwan Thailand
SIPRI 12.57 9.32 6.96 6.82 6.4 5.15 4.41 4.37 3.75 3.69 3.14 2.85 2.44 2.29 2.16 2.08 1.93 1.92 1.89
ACDA 23.77 4.73 5.25 7.32 4.48 4.73 1.17 2.87 1.73 2.86 0.79 0.98 4.13 1.42 2.48 1.41 2.46 1.69 1.32
Country Spain Chile UK Bangladesh India Pakistan Poland Turkey Iran France Indonesia Korea, South Brazil Hungary Italy Syria US Morocco China
SIPRI 1.82 1.56 1.56 1.37 1.22 1.21 1.2 1.12 1.05 0.94 0.78 0.77 0.62 0.35 0.2 0.2 0.15 0.12 0.04
ACDA 1.15 0.75 2.81 0.38 1.04 0.74 0.37 1.38 4.24 0.49 0.52 1.13 0.81 0.00 0.47 1.94 0.85 0.13 0.06
Note: Either ratios of arms imports to arm forces are averages for 1990-91. The ACDA (1995) ratio is given by ACDA imports in constant 1990 million dollars divided by ACDA (1995) armed forces in thousands. The SIPRI ratio uses SIPRI volumetric measurement of arms imports, always measured in constant 1990 million dollars, divided by the ACDA figure for armed forces.
three inputs. In other words, the empirical evidence appears to support the notion that the technology of production of contemporary defense allows for a considerable degree of substitution among inputs. For analytical purposes, the variety of arms is aggregated into two assumed homogenous categories: imported and domestically produced arms. These two categories of arms are assumed to be imperfect substitutes in the production of military capability, since countries can defend themselves using imported arms without having to establish a domestic industry. Moreover, imported arms differ from domestic arms by having the characteristics of "necessity," whereas domestic arms rather have the characteristic of "luxury." This distinction is based on the
134 Ai in ing th e So u th
observation that the smaller is a country's defense budget the more it tends to rely on arms imports, whereas countries with larger military budgets tend to increase their use of domestic arms. It has not been possible to find an analytical framework in the literature that captures the required characteristics of production technology, at least not with relative analytical simplicity. But it appears that the desired features can be provided by an extension of the StoneGeary production function. This is discussed in detail in appendix 6B. In its standard form, the Stone-Geary function (Chung, 1994) allows for some minimal positive quantities of the inputs, called "subsistence quantities," that are absolutely necessary for production. Appendix 6B examines a generalization of the Stone-Geary function in which the subsistence quantity of one of the inputs can be negative, representing a minimum amount of the input always available as a free good. Furthermore, appendix 6B introduces fixed costs to the use of this input, which can be avoided if the input is not used. This modification apparently supplies the desired model characteristics, including the entry of domestic production, when the military expenditure budget rises to a sufficient degree. A simplifying assumption of this work is to consider the defense budget as prefixed, reflecting government's perception of the country's needs for security. Moving first, the central government defines the level of the budget without taking into account any secondary effects on national security, such as the impact on an opponent country's allocations (i.e., an arms race). Then the military sector allocates resources maximizing output. This implies that central government is not informed about the returns to scale of the defense sector, also neglecting any disproportional effects of the budget to the produced output. Instead they believe that every additional unit in the budget buys the same amount of security. Restrictions in the availability of information about the returns of the defense sector, however, appear less unrealistic considering the general state of secrecy that usually surrounds the internal organization of national defense sectors.
Domestic production as an alternative to importing arms
135
The model Assume that the only output of countries' military sector is military capability, K, which creates national security. Also assume that the government of a representative country has fixed preferences in terms of allocating resources between national defense and other expenditures. According to this assumption, central government moves first to define the defense budget which is then fixed. With this budget fixed, the defense planner allocates it to the factors of production with the purpose of maximizing the output of military capability. Inputs to the production are domestic arms, imported arms, and armed forces (labor). We assume a static setting where inputs and output are subject to full depreciation in each time period. The set-up costs for the establishment of such a sector are generally large and mainly consist of expenditure on R&D. Also, the differentiation of domestically produced from imported arms is based on two characteristics. One is that domestically produced arms provide higher security of supply, especially in the case of a war when they are most needed. The other, which is technically related to another simplifying assumption, originates from the aggregation of numerous different kinds of arms into a single homogenous good that in reality is a bundle of a variety of different sorts of arms. If the domestic sector gets established, the quality and composition of the bundle of domestic arms could be adapted to the specific needs of the country. It is therefore possible that the availability of a domestic arms industry can raise the returns to scale of the defense sector. Regarding the properties of substitution among inputs, the two categories of arms are assumed to be imperfect substitutes in this production. Imported arms and labor are imperfect complements and a minimum quantity of them, the subsistence quantity, is necessary for any production to exist. Domestic arms are gross substitutes with imported arms and labor; and they may be fully substituted by the other two inputs. In this set-up, the option to export domestically produced arms is not taken into consideration. Let the production function for military capability be
136 Arming the South
(6.i)
K=(D-rDYn(M-rA.fYM(L-rLY''
;K D + KM + KL = I
where K is military capability, KD9 KM9 and KL are coefficients such that KD, KM, KL e (0, 1). Variable M> yM>0 stands for imported arms and yM is the subsistence quantity of them. Variable L > yL > 0 is armed forces and yL is their subsistence quantity. L excludes labor employed in the domestic arms industry. Domestically produced arms D > 0 have a negative subsistence quantity, yD < 0. The price of labor is w and of imported arms is pM, both assumed fixed. When the domestic production is not established, D is simply set to zero, y KDD becomes a fixed coefficient in (6.1) and, as equation (A6.2) shows, returns to scale are lower. The domestic arms industry follows a Leontief production function with constant returns to scale. If we adjust the measurement units of D to be equal to the average product of labor we may write the production function as (6.2)
D=LD
where LD is the labor employed in this industry. The production of D requires total capital YD which consists of fixed capital F > 0 and variable capital bLD in a constant proportion b with labor (6.3)
YD = bLD + F.
The price of labor is w and the price of capital is normalized to unity. Then, the total cost function of D is CD = wLD + YD = (w + b)D + F. Dividing this expression through by D we obtain (6.4)
pD = (w + b) + F/D
where pD is the average and (w + b) is the marginal cost of domestically produced arms. Also, pD is the break-even price at which the profits of this industry vanish. According to (6.4), when the set-up costs are not zero, the average cost diminishes with output. We also assume that factor prices are fixed, remaining unaffected by the demand for capital and
Domestic production as an alternative to importing arms
137
labor. The two cases of discrete choice examined here are when the domestic arms industry is established and when it is not. Following (A6.2), we may use the binary (dummy) variable E to denote the two cases, that is, set E = 1 when the industry is established and E = 0 when it is not. In terms of notation, we distinguish between the two cases using a tilde (~) to distinguish the variables that refer to the case without domestic production. If G is the total defense budget, G - EF is the disposable budget, net of set-up costs for the domestic industry. Without the domestic arms industry, E = EF = 0, and the disposable budget is G. But when the domestic arms industry is established, EF = F and the disposable budget is G - F. Under these assumptions, the budget constraint for the production of military capability is (6.5)
G > (w + b)ED + pMM + wL + EF.
Of course, E = 0 implies that D = 0, too, in which case (6.5) reduces to G > pJM+wL. Solving the maximization of (6.1) subject to the constraint (6.5) using the Lagrangian method, when the domestic arms industry is established, the demand for imported arms is
(6.6)
M=yM
+ ?JL[G-F-(w+b)yD
- pMyM - wy L].
PM
Similarly, the demand for armed forces is
(6.7)
L = yL + ^-[Gw
F-(w+b)yD
- pMyM
and the demand for domestically produced arms is
- wy L],
138 Arming the South
(6.8)
£>= yD + - ^ - [ G - F - ( w + Z>)rD - p „ r „ - w y , ] W
u
T
When the domestic industry is not established D = 0 and the demands for the two remaining inputs are
(6.9,
M=ru
+
^
G
p
-
«
r
" -
w y
L
and K
(6.10)
L = yD +
L
G
~
\~ KD
PMYM-^YL W
The two systems of demand equations have different characteristics dependent on the establishment of the domestic industry. Apparently, an advantage of this set-up is the algebraic simplicity of the solutions. A standard requirement of the Stone-Geary function is that the budget should necessarily cover the cost of subsistence quantities of the inputs, i.e., G > GL where (6.11) GL = pMyM
+ wrL.
When the domestic industry is established, the non-negativity constraint of input D requires from (6.8) that G> GL where
(6.12) GL = F+pMyM +
wyL-^^-{w+b)yD. K
D
Otherwise, if G < GL , domestic arms production is not established. When the budget exceeds this critical level the planner chooses the kind
Domestic production as an alternative to importing arms
139
of production that maximizes output. It can be proved that in the range [GL9 <*>) there is a critical level of the budget G* such that the two cases give equal output. It can also be proved that G* always exists (see appendix 6B). When the budget is lower than this critical level, producing K without a domestic arms production sector gives higher output than producing it with an established domestic sector. When G > G\ output with the use of domestic arms dominates.1 Summarizing, we have determined three critical levels of the budget. In order of magnitude, they are: (6.13) GL< GL< G*. Comparison of G with these values determines the establishment of the domestic arms industry
(6.14) E
0;
G < G*
1;
G>G*
undefined; G = G* When G = G\ either case - with or without domestic arms production gives equal output, and the planner is indifferent and the solution undefined. This is the reason why we may not write condition (6.14) in a necessary and sufficient form. In inverted form the necessary conditions for the establishment are
(6.i5) ;1
E= 0 ^ G< G E= 1=> G> G
Properties of the demand for imported arms One of the objectives of this chapter is to provide a theoretical explanation for the results of a cross-sectional estimated equation (Mouzakis, 1999b)
140 Aiming the South
InM - -21.64 - 0.561n/?M + 1.521nG - 0.06(lnG)2 + 4.74hw 0.29(lnw)2. Only the price elasticity is estimated as constant, independent of the level of pM. Income elasticity sG decreases when G grows but stays always positive, nearing zero in the case of the largest observation (for the US). The elasticity of imports toper capita income sw changes from positive to negative when w grows and so does the estimated response of demand, with a turning point just below the sample mean of w. Analysis of the impact of measurement errors indicated possible underestimation of eG and ew. From the demand functions for imported arms (6.6) and (6.9) we may calculate the own-price elasticities for the two cases as
(6.16) s = -K M
and sn
G-F-(w+b)yD-wyL
^Q
PMM
-
< 0 pMM
\-KD
Similarly, the income elasticities are
(6.17)
eG = KM
—>0andsG PMM
= —M
77>^G>^
\-KDPMM
and the elasticities of imports with respect to the average national wage are w
(7 n + 7 / )
Domestic production as an alternative to importing arms
141
Since yD < 0 and yM < 0, the elasticity of arms imports to labor costs with the domestic sector established, ew9 can either be negative or positive, depending on the relative magnitude of the two subsistence quantities. When the domestic sector is not established this elasticity is limited to be negative. This might well be the case in the cross-sectional study of LMS who estimated this elasticity to change from negative to positive when w rises. In either case, the own-price elasticity is negative and might either increase, decrease with or remain unaffected by the establishment of the domestic arms production sector, depending on parameter values. Notice that if F > -(w + b)yD the demand for M is always more elastic when the domestic arms production sector is not established. Income elasticity sG is positive, depends on G, and decreases when the domestic arms production sector is established. Since the establishment of a domestic arms production sector depends on G as well, when G increases beyond G* the domestic arms production sector gets established, and imports M and elasticity eG jump discontinuously to lower levels. Without a domestic arms production sector, the elasticity of imported arms to labor costs is negative. When the domestic arms production sector is established the elasticity sw increases in value, and it is positive if, and only if, -yD > yL. The two panels of figure 6.1 illustrate the LMS reaction of countries M (dashed curve), as well as examples of reactions predicted by the model M and M. Panel A presents the demand for arms as a function of the budget G and panel B as a function of labor costs (w). A comparison of the theoretically predicted reaction with the estimated reactions shows that the two are in line or, at least, they are not contradictory. Panel B gives an example of the predicted response to changes of economic development of the country, which mainly include the impact of labor costs. Again, the set of two linear reactions provides a possible explanation to the estimated parabolic response.
142 Arming the South
M
M
Panel B
Panel A
M M M A
M
M
-G G(L)
G*
-w w'
Figure 6.1: Predicted and estimated demand curves for imported arms
Empirical results The model provides analytical means for an empirical examination of the demand for arms imports, provided data are available. We may consider quantities G, w, M, pM,, and L as observable, i.e., government budget, average national wage, military expenditure, price of imported arms, and armed forces, respectively. Data series for the domestic production of arms D are not directly available in the required form. Nevertheless, based on available output measurements of major arms industries (e.g., SIPRI, IISS), construction of such a series could be a possibility. But any attempt to approximate a series of domestic arms production would unavoidably raise questions about the quality and adequacy of the constructed data series. Alternatively, as the following analysis demonstrates, we may avoid the quantitative measurement of countries' domestic production of arms and attempt an estimation of the model
Domestic production as an alternative to importing arms
143
merely by using the existence of a major domestic arms industry, that is an approximation of variable E. Separating observablefromunobservable quantities, we can rearrange the twin demand functions for arms (6.6) and (6.9) as
(6.19)
~KM(byD
PMM=
+ F) + KMG+
-KM(7D
+ A>L)W>
(6.20)
E
=
(1-
KM)yMpM
l
PMM^^^G+^^yMpM--^yLw;E=K, 1
K
D
[
K
D
l
K
D
respectively, where KL + KM + KD=1. Let Ei be the dummy variable for the establishment of a domestic arms sector for country /, which equals 1 when a domestic industry is established and 0 when it is not. Also, let Ax = pM[Mj be the ACDA measurement of the value of arms imports andpMi the implicit price index of imported arms. Then, equations (6.19) and (6.20) can be written jointly with the use of intercept and slope dummy variables as A. = m0El + mxGl + m2pMi + m3wi + /w 4 (£.G,.)
where mj are coefficients and uMi is a random term. Expression (6.21) can be estimated using the LMS cross-sectional data set, which consists of 1990-1991 average annual figures of arms imports and military expenditure by SIPRI, arms imports by ACDA, and per capita GDP from the Penn World Tables. Arms imports and military expenditure are in constant 1990 millions of US dollars and per capita GDP is in constant 1990 US dollars. The variables are transformed to US dollars with the use of average annual exchange rates mdper capita GDP is transformed with the use of purchasing power parities. The implicit price index for imported arms is calculated by dividing the ACDA measurement of the value with SIPRIs volumetric estimate of imported
144 Arming the South
arms. Unlike LMS, however, the estimated equation is not logarithmic; instead the variables are in levels. Systematic measurements of countries' production of arms are not generally available in adequate coverage and quality for the needs of our study. As an approximation we can use available figures on the largest arms producing firms and, also, tables on arms exports, assuming that a permanently high level of arms exports indicates a well developed domestic arms industry in the country. This information is available from two tables: the 1993 SIPRI Yearbook "100 Largest Arms Producing Companies" (pp. 470-474) and "25 Leading Suppliers of Major Conventional Weapons" in the 1995 Yearbook (p. 493). Of all the countries in the sample, 10 countries have entries in both tables. These are Canada, France, Germany, Israel, Italy, the Netherlands, Spain, Switzerland, the UK, and the US. In addition to these, companies from India and Japan were also listed in the table of the 100 largest arms producing companies; and companies from Brazil, China, South Korea, Norway, Pakistan, and Poland comprised additional listings in the table of the 25 leading arms suppliers. Thus, these 18 countries can be characterized as having an established domestic arms production sector leaving the remaining 21 counties in the sample characterized as net importers. Table 6.2 presents the results from the estimation of (6.21) by OLS and the solutions for the coefficients (see appendix 6A), including the standard errors and the p-values of the two-tailed tests of significance from zero. The R2 of the regression is 0.655 and the standard tests of the residuals take relaxing values. In particular, White's F-test for heteroskedasticity gives a p-value of 0.908 for the hypothesis of homoskedasticity, Ramsey's F-test for functional miss-specification gives a p-value of 0.209 for the hypothesis of correct linear specification, and t h e / 2 test for the hypothesis of normal distribution of the residuals has a p-value of 0.209 as well. The coefficients KM, KD>, and KL have values within the acceptable range (0, 1) and their signs are well supported by the t-tests. Since KL = 1 KM - KD, testing for the significance of this coefficient as different from zero is equivalent to the test for significance of the sum of the two other coefficients from unity. Subsistence quantities yM and yL are also positive
Domestic production as an alternative to importing arms
145
Table 6.2: Estimated coefficients
Coefficient Standard error
524.74 179.70
*\f
Coefficient Standard error p-value
0.1347 0.0250
*D
78.027 31.754
KL
-0.1747 0.0102
7A/
-0.1292 0.0251
-110.00 118.53
7L
yD
0.0056 0.0022
0.9585 0.0178
0.0359 0.0159
90.391 35.775
0.1296 0.06371
0.8946 1.9655
0.014
0.000
0.031
0.016
0.050
0.652
0.0117 0.0153
with signs statistically significantly different from zero at the 5 percent level. Note that as a result of the normalization of the units of labor equal to the average output of the arms industry, coefficient yL is measured as a proportion of the domestic procurement of the country. The last estimated parameter, subsistence quantity yD, instead of the expected negative value has a positive estimated value, which implies a contradiction with one of the fundamental assumptions of the model. Comparing the low statistical significance of yD, however, with the high significance of the sign of all other coefficients we may well conclude that the hypothesis of a negative yD is not contradicted by the data. The empirical results so far appear either in line with or at least are not contradictory to the hypotheses of the model. It should be clear that the scope of this econometric examination is not an in-depth empirical study of arms production and trade, but merely a test for the applicability of the theory. The weak aspects of these results include the overidentification of the estimated system, measurement errors in the variables, and limited information about domestic arms procurement. Also, the model limits the establishment of an arms industry to the same fixed cost for all countries. This typical simplifying assumption (e.g., used in Levine and Smith, 1997) contributes to the divergence of the theory from empirical reality, the examination of which probably requires a more flexible specification. This is partly relaxed in Mouzakis (1999b) who assumed different fixed costs for large and smaller producing countries.
146 Arming the South
Conclusions The study presented in this chapter does not claim exhaustive coverage of its topic; the presented analysis is an initiation of an analytical path rather than a stand-alone theory of arms production and trade, attempting an analytical formalization of the emergence of national defense industries. The final form of the model developed in this chapter has been largely based on the technical assumption of modifying a Stone-Geary function (see appendix 6B) in a way that provides the desired properties of substitution. As a newly introduced method little tested yet, however, it should be looked at from a critical perspective. From the results, it appears that the performance of the developed framework corresponds to the initial requirements. In line with empirical evidence, the analysis in this chapter suggests that the establishment of a domestic arms industry may have a considerable effect on the demand of countries for internationally traded arms. With the establishment of a domestic arms production sector, domestic arms substitute for imported arms and the demand for the latter becomes less responsive to the military budget of the country. Also, the effect of economic development on this demand changes from positive to negative. The effect of the price of arms imports appears to remain always negative and unaffected by the existence of a domestic arms industry. These results appear to provide possible answers to the questions brought up by the cross-sectional examination of LMS. The econometric study also reveals supporting evidence about the predictions of this theory and about the applicability of this analysis in empirical studies. The high estimated value of the weight of domestically produced arms indicates that the establishment of a domestic arms industry, and the availability of domestic along with imported arms, has a positive effect on the returns to scale of the defense industry. There is also evidence for considerable subsistence quantities of imported arms and labor but not enough evidence to support the negative sign of the subsistence quantity of domestic arms. A possible way to carry this study forward would include the examination of a wider theoretical framework, introducing a welfare function for the central government. This would allow for a more realistic
Domestic production as an alternative to importing arms
147
distribution of resources between defense and other sectors and undermine the notion that welfare benefits from national security change in proportion to the military budget. Another, more complete approach would be to use the developed formulation of the demand side in a world market model. On the empirical side, the availability of data series of arms production would enable the estimation of a complete version of the model. Note A version of this paper was presented at the conference "The Economics of Military Expenditures in Developing and Emerging Economies," held at Middlesex University, 13-14 March 1998, London. The author wishes to thank the organizers and the ESRC for their support. 1. Setting the indirect utility functions for the two cases equal, with some algebra we obtain the necessary and sufficient condition for indifference to the establishment of domestic industry of arms
-[F+ (w+b)yD + pMy
M
+ wyL]= 0
As a real polynomial of G, the above expression does not take a general algebraic solution. Providing the knowledge that a single real solution does exist and is unique, we can solve for G* with the use of numerical methods. References ACDA. World Military Expenditures and Arms Transfers 1993-1994. Washington, DC: US Arms Control and Disarmament Agency, 1995. Chung, J.W. Utility and Production Functions. Oxford: Blackwell, 1994. Garfinkel, M.R. "Arming as a Strategic Investment in a Cooperative
148 Arming the South
Equilibrium." American Economic Review Vol. 80, No. 1 (March 1990), pp. 50-68. Intriligator, M. "Strategic Considerations in the Richardson Model of Arms Races." Journal of Political Economy Vol. 83 (1975), pp. 4362. Levine, P., F. Mouzakis, and R. Smith. "The Arms Trade: Some Theory and Econometrics." Discussion paper, Department of Economics, University of Surrey, No. 8/96, August 1996. Levine, P., F. Mouzakis, and R. Smith. "Prices and Quantities in the Arms Trade." Defence and Peace Economics Vol. 9, No. 3 (1998), pp.223-236. Levine, P. and R. Smith. "The Arms Trade." Journal of Economic Policy (October 1997), pp. 337-370. Mouzakis, F. "Regional Variations of the Demand for Arms Trade: An Empirical Study of Global Panel Data." Discussion paper in economics, University of Surrey, No. 2/99, April, 1999a. Mouzakis, F. "Theoretical and Empirical Aspects of the International Arms Trade." Doctoral Thesis, University of Surrey, 1999b. Richardson, L.F. Arms and Insecurity: A Mathematical Study of the Causes and Origins of War. Pittsburgh, PA: Homewood, 1960. SIPRI. SIPRI Yearbook. Oxford: Oxford University Press, annual. Smith, R. "The Demand for Military Expenditure," pp. 69-87 in Keith Hartley and Todd Sandier (eds.) Handbook of Defence Economics. Amsterdam: North-Holland, 1995. Smith, R., A. Humm, and J. Fontanel. "The Economics of Exporting Arms." Journal of Peace Research Vol. 2, No. 3 (1987), pp. 239-247.
Domestic production as an alternative to importing arms
149
Appendix 6A The estimated system of equations The reduced form of the system can be calculated with the use of (6.19) and (6.20) as follows. Directly from the definition of the dummy variable we obtain (6A.1) K M = mx + m4 , and from the coefficients of G from (6.19) and (6.20) we obtain (6A.2) KD = and (6A.3) KL - 1 - KD - K M Continuing, we may isolate yM from the coefficient ofpM in (6.20), which after simplifications takes the form m (6A.4) y M = — ^ - . 1 - mx From (6.20) we have (6A.5) yL = - — ml and with some algebra from the coefficient of w in (6.20) we write
(6A.6) y
D
m, m, + m, 6 ^ ^ - - ^ -. ml m] + m4
The coefficients b and F cannot be identified. Their calculation requires
150 Arming the South
additional information to that provided by (6.21). Appendix 6B An extension of the Stone-Geary
function
An important consideration regarding the use of a utility or a production function is that it provides and specifies the type and degree of substitution among inputs. Ranking from one extreme of perfect substitutes to the other extreme of perfect complements, the extent of substitutability among the inputs of a transformation process is a critical concern in the selection of the appropriate analytical tool. Whilst the contemporary literature on transformation functions seems to focus on the development of less restrictive and more powerful analytical frameworks, it is not very informative on the issue of complete substitution of one input by the others. This appendix examines an extension of the StoneGeary function, which appears to provide the desirable properties. It also examines the case that the substitutable input requires a fixed cost if it is used. As a starting point we may briefly look at ways that elasticity of substitution affects the geometry of production functions. In the case of a function with unitary elasticity of substitution (Cobb-Douglas, CD, is a typical example), all isoquants (isoproduct curves, surfaces, volumes, or hyper-planes in the cases of two, three, four, or more inputs, respectively) have the axes of the inputs as asymptotes. A more general function, with parental relationship to CD, is the CES function and the Generalised CES which, however, rule out zero levels of inputs (for a recent review of production and utility functions see Chung, 1994). When inputs are complements, the asymptote of every isoquant is located at some positive quantity of the input, denoting the lowest possible quantity required for any level of output, called subsistence level. When the elasticity of substitution is higher than unity the isoquants are expected to intersect the axes of the inputs, allowing for complete substitution of one input by the others. Of course, the range of the function is set accordingly to rule out negative values of the inputs. The Stone-Geary function (e.g., Chung, 1994, pp. 23-31) has mainly
Domestic production as an alternative to importing arms
151
been used as a utility function in the well-known Linear Expenditure System and, like CES, can be considered as a generalization of CD. In logarithmic form, we can define the Stone-Geary (SG) function as follows (6B.1)
\nU=YJKi\n<
where 0 < KX < 1, S ^ = 1, U is output (utility or product), qt are quantities of the inputs, and yt are subsistence quantities so that q( > yi > 0. It is nonhomothetic but strictly monotonic and convex, thus well defined as a utility or production function. Assuming constant prices the budget constraint is G - ^
PlQl where/?, is the price of input /. Solving for
the standard consumer's (or producer's) optimization problem the demand for the inputs of this process is given by G-YSJPJYJ
(6B.2) qt = n + K, —
0 e j) . Pi
Looking closer into this function, the own-price elasticities of demands are
G-I PJ
J1 J
(6B.3)
£U
=
^LPL=
m
< 0 >
where 5 stands for apartial derivative. Examination of (6B.3) reveals that using SG implies that demand is always inelastic (| su \ < 1). The crossprice elasticities are Sq, P, PJ, J J (6B.4) 8 = -f-^- = -Kt - - - < 0 SPj q, Ptfi
{it
j)
152 Arming the South
which limit inputs to be gross complements. The income elasticity is
(6B.5)
dqi G SG qi
€Gi = —^--=Kl
G ' piqi —
<0
which restricts inferior goods. The Allen-Uzawa cross-partial elasticities of substitution are ^hPhY h (6B.6) (j.. = 1 "™ " < 1 u PiPfltti
{itj\ijeh)
limiting the inputs to gross complements, as the cross-price elasticities also show. From the previous description the SG function does not seem to accomplish the requirements for this study since it would not allow the elasticity of substitution to be higher than unity between any pair of inputs. A modification of SG that provides the desirable properties is to allow subsistence quantity of one and only one of the inputs to be negative. The second modification that needs to be introduced to the standard SG set-up is to allow for fixed costs in the use of the input in question. The impact of this alteration can either be seen as causing a diminishing price of the input or as reducing the budget by a fixed amount (of set-up cost) but leaving the price of that input fixed equal to marginal cost. Assume that all assumptions hold as in the usual SG set-up except one: the subsistence level of only one input, say input/, is negative. The subsistence quantities of all other inputs remain positive as usual, or
(6B.7)
\nU=YJK,Mqi-yi)
whoe 0 4W > Yt*f > ° a n d K/ < ° - 4 / -T^ negative subsistence level gallows input/to be a gross substitute instead
Domestic production as an alternative to importing arms
153
of being limited to a net complement with all other inputs. Now, input/ can be fully substituted by the other inputs, in which case (B6.7) takes the form
(6B.8) l n ^ = ^ / l n ( - x / ) + X I - ^ / l n ( 9 , - - ^ ) • Note that (6B.8) is just a special case of (6B.7) where qfxs set to zero. As a convention, we distinguish the type of transformation that excludes input/from the case that includes it by noting the related variables with a tilde (~). The second alteration we introduce to SG is that only when input qf is excluded is the disposable budget G, whereas when/is included it reduces to G - F, because of fixed costs F in the use of/ The fixed amount of set-up costs F has to be spent when any qf > 0 is used in production. As a result of the restriction qi < yi9 in either type of transformation the necessary condition for any output to exist (U> 0) is that the budget can cover at least the expenditure for subsistence quantities of all other than/inputs
(6B.9) G>GL = Yl*fPiY,. When qf\s included in the transformation, the budget in addition to GL has to cover fixed costs F. Thus, a necessary but not sufficient condition for using qfis (6B.10)
G> GL = GL + F.
Including input/in the process, the demand functions for inputs are essentially as in the general cases (6B.2) with the budget reduced by the fixed costs
(6B.11)
qr,. = y. + x-,
G-F-HJPJYJ l
(iej),qf>0.
P, Under (6B.8), however, the returns to scale are decreasing compared to
154 A rm ing the Sou th
(6B.7) and the demand equations for inputs take the form
(6B.12)
qi = yt + ~ \-Kf
(/ e j\qf
= 0.
p.
Solving the problem of output maximization, we may first look into restrictions (6B.9) and (6B.10). When GL > G > GL the use of input/ in the production is impossible and we can directly detect the demand for other inputs using (6B.12). To determine the optimal allocation when G > GL, where either type of production is possible, we must ensure that both output is maximized and also that the non-negativity of inputs is satisfied. While the non-negativity of other than / inputs is not problematic, following the standard SG case, we must ensure that the demand for qfis positive. Thus, applying the restriction qf> 0, with the use of (6B. 11) we obtain (6B.13)
qf > 0=> G> GH
where GH - Gl
PfY f • Note that under the assumptions that K
f yf<0 and F > 0 we always have GH > GL > GL . Thus, condition (6B.13) fully replaces condition (6B.10). Having detected the necessary conditions and solutions for the allocation of resources in each of the two types of production, we need a sufficient condition that defines the optimum type of production in the general case. We can do this using the indirect utility functions for each case. Substituting from the demands for inputs (6B. 11) and (6B. 12) in the production functions for the two cases, after simplifications the two indirect utility functions become n (K
(6B.14)
\ *>
w = ( G - F - X l P / 7 l ) n ( — J ;?/><>
Domestic production as an alternative to importing arms
(6B.15)
^ - ^ ( G - E ^ n (1-*/)
'*/
;
155
;qf = o.
»*/V/V
From the definition of maximality, the condition for the inclusion of g in the production is (6B.16)
qf > O o
u> u
Substituting from (6B.14) and (6B.15) to the right hand side of (6B.16) and rearranging we obtain (6B.17)
^ > 0 o
z> 0
where
z = z(G)= G-a(G- GLfKf ~ (GL + pffy ,)\ (
(6B.18)
a=
\Ks
PfYf \ Kf J
K
f)
Variable z is proportional to the output surplus from the inclusion of input / Note that it is impossible to obtain a general solution for the inverse function of z(G). Therefore, we can only solve for values of G that correspond to given values of z with the use of numerical methods. Differentiating z with respect to G once we get dz I dG - 1 - a{\- K f)(Gd z I dG
GL)
f
- CIK f~ (1 - K f)(G - G j)
and differentiating twice we get -\-K
J
> 0 , showing convexity
of z to G. Solving from the first-order condition, dzldg = 0, for the global minimum of z we obtain the neat result that z get its minimum at GH. Thus, z(G) is strictly increasing in its range [GH9 °°). The domain of z is bounded downward by the global minimum z(G^) but not upward, since
156 Arming the South
lim G _ z = «>. We can also show that z(GH) is always negative. Substituting for GH in (6B.18) we obtain •M-K,
z(GH)=
PfYf'Kf
-a F-{\-Kf)pfyflK
and after substitution for a and some algebraic manipulation we obtain
z(G / y )<0<=>
F \-K
PfYf f V
K
\~K<
>o,
/ i
QED. Thus, as claimed we have shown that z is never higher than zero. Summarizing, we have shown that in its range z{G) is continuous, strictly increasing, and quasi-convex, with a negative lowest bound and upward unbounded. Since 0 is always included in the domain of z using the mean value theorem we have shown that there is always a value of G, say G*, that satisfies z{G^) = 0. Thus, z has at least one root in its range. Furthermore, since z(G) is a strictly increasing function, using the fixed point theorem we have shown that this root is unique. This proof can be expressed in a more formal way as follows. Let
(qi - y[)K'
U - | j
be a function
that maps a
transformation of(quq2, ...,#„) into R where /7 e N, 0 < /c, < 1, Y*XK1 = 1> ?/>/• > 7/>/ > 0, qf > 0, yf < 0 and pi9 KX and yt are fixed. If >0 +
Z/A^
^;
G - ^
where F > 0, //ze/7 ///^rg emte one
and only one G e ( G ^ o o ) , where \-Kf
GH = GL + F-
K
pfy
and G L = ^
f
^ 7 7 , such that the
f constrained maximization of U under G implies that.
Domestic production as an alternative to importing arms
*i GL
157
G 1
- JPfj
and
G>G
=> qi^yi
+ Ki
G-F-IJPJYJ
;
(iej).
By definition, when G = G*, either previous solution gives equal output. An interpretation of this result can be as follows. At the lowest boundary of the budget both types of transformation are possible (i.e., when G = GH). Transformation with input/excluded always gives equal or higher output than with / included. Thus, at GH the output of the transformation without input / weakly dominates the output of the transformation with input / whereas if F > 0 the latter type strictly dominates the former. When G rises, excluding/keeps dominating until G reaches the critical level G* > GH. When G exceeds G* production with /included dominates in terms of output. When G = G*, the two types of transformation give equal output and the dominant type is undefined. The critical level of transition G* is unique. Panel A of figure 6B.2 plots the indifference map of the two inputs when F > 0. When/> 0 the budget constraint is the straight line with slope -Pf/pq9 but at/= 0 (i.e., at the intersection with the vertical axis, q) it does not include the usual endpoint at g = G/pq. Instead, it includes a point which is detached from the rest of the curve, shifted upward by amount Flp . This shift represents the release of resources from fixed costs F when input/is not used, measured in units ofq. If, for example, the budget is Gx and input/is included in the process, maximum output is uv When i n p u t / i s excluded, output reaches level ux > ux using quantity qx of input q. The case of indifference occurs when the budget is G2 = G* and either allocations (q2, 0) and ( ^ 5 / 2 ) Y^d exactly the same output u2 - u2 . If the budget grows higher than G\ as in the case of G3, output level u3 that requires q3 and / 3 quantities of the inputs
158 Arming the South
B
J.8 S G?t>W
8
f
G2,q2
0
l' 1 /! 1 "3
U
4i
u
th i
u
Mi
in+Z y
i
Ofi h h
f
°GLGH
G
Figure 6B.2: Equilibrium when input/requires fixed costs
dominates output u3. Panel B presents the output and demands for inputs as a function of G, when F > 0. The difference between the previous and this case is that the demand for input q is not continuous anymore but consists of two line fragments detached at G*. Output is still a continuous function of G but not differentiable, changing slope discontinuously at G*. The transition from the type of transformation that includes/to the type that excludes it, and vice versa, depends on the relative prices of inputs, the budget, and the other parameters of the production. Under the new definition, simple observation of the elasticities from (6B.3), (6B.4), (6B.5), and (6B.6) reveals that the properties of SG are affected in a desirable way. The own-price elasticity is not any more limited below unity for all inputs except the input substitute / The cross-price elasticities and cross-partial elasticities of substitution now reflect the substitutability of this input by not being limited below zero and unity, respectively. In particular, (6B.6) shows that input/is a gross substitute with any other input since for any i e n9 aif= or> 1.
Domestic production as an alternative to importing arms 159
The type of transformation that excludes/differs from the case that includes it, in that neither the technological coefficient (~7y) unitary, nor is the sum of K coefficients, i.e., 2^
K f x
is
< 1 • Another
difference between the properties of the two types is that the transformation type that excludes input/has lower returns to scale to the type that includes it. Since SG is not homothetic the returns to scale are not globally equal but depend on the inputs. In order to compare the returns to scale of the two alternative methods of production we can examine the change of the ratio of the two outputs when all inputs except / a r e equal and change by a factor /, for any qf> 0.
^f-Yj-^W^Xlq^Y^ v-^./ / / y x i*j (6B.15) — U ~ (-YfY'U^^-Y.) U
JL
(lqf-r^f •yf
> i
)
The fact that U grows faster than U when inputs grow at the same proportion indicates higher returns to scale for the former function. According to the standard assumptions of the Stone-Geary function and those that originate from the alterations we introduced, we can relate the entry of the input in question to the transformation with the size of the resource constraint. Briefly, this is as follows. There is a minimal required level of the fixed resource (the budget) for any output to exist. This is standard in the Stone-Geary framework: it only examines positive subsistence quantities of the inputs. When the budget grows beyond the subsistence quantities the production first takes place without the substitutable input. When it grows further beyond some other, higher, critical level the input in question is included in the transformation. Confirmation of this analytical result can easily be obtained with the use of numerical methods.
This page intentionally left blank
7 Domestic Procurement, Subsidies, and the Arms Trade Maria del Carmen Garcia-Alonso and Paul Levine
Introduction The end of the cold war led to a precipitous drop in the world arms trade. The real volume of the arms trade reached a peak in 1984, fell for two years, rose in 1987, and then declined by over two-thirds up to 1993. World arms sales agreements also declined substantially from a peak of $69.8 billion in 1985 to $38.0 billion (US ACDA, 1995). However, from 1992 to 1993 arms sales agreements rose in nominal terms. The main reason for the decrease in the real volume of arms trade has been the reduction of superpower competition in various regions which has also led to a decrease of arms aid from the US and Russia to their former allies. Recipients must now pay full price for their weapon systems. However, developed countries continue to play an important role in the international arms market providing more than 90 percent of world arms exports between 1983 and 1993. The reason for this is that decreased domestic demand for weapons among the main arms producers has forced defense contractors to seek export sales to compensate for this decrease or bear the high fixed costs of production. Despite the decrease in the domestic demand for weapons, developed countries are of course still interested in having their own national military industries. Among recipient countries, many nations are involved in arms races. Therefore, the security perception of the producer countries plays an
162 Arming the South
important role in the structure of their military industries. Firms producing weapons in the different supplier countries are sometimes referred to as "national champions" because they enjoy a national monopoly in one or more specific areas in the domestic defense industry. The relation between national champions and their respective governments is very close for without the support of the state, its financing, and access to its defense market, these companies would disappear. Besides, technologies are tightly controlled as strategic company assets and their transfer is closely regulated by governments. This chapter develops a model of the international arms market that captures the dependency of weapon producers on their respective governments' subsidies. Levine, Sen, and Smith (1994) provide a formal model of this trade which allows for competing, forward-looking suppliers whose welfare depends on both the economic benefits from sales and the security repercussions of the recipients' behavior. Their model consists of an arms trade market with similar suppliers who maximize a utility function which depends on both profits obtained from arms trade and security. Arms are sold to a generic recipient characterized by a demand function. Levine and Smith (1995) present a dynamic model of a fully structured arms trade market. They introduce arms races among the arms importers.1 Meanwhile, exporters compete in the international arms market. As in Levine, Sen, and Smith (1994), exporters are also concerned about the consequences of arms exports. Both of these papers have in common the existence of a unique decision-maker in each of the arms exporting countries. Firms and governments are fused into the same agent which has both political and economic concerns. As in Levine and Smith (1995), in this chapter both the governments on the supply and the demand side of the international arms market care about security and private consumption. The international demand for weapons is characterized by an arms race situation in which security depends positively on domestic procurement and negatively on the adversary's procurement, and the security for suppliers is negatively affected by global exports. We consider firms and governments as different agents with different interests: firms are profit maximizers and decide upon the quantity of exports, whereas governments are welfare maximizers and decide the quantity of domestic defense procurement and
Domestic procurement, subsidies, and the arms trade
163
consumption. 2 We assume that weapon producers are the only suppliers for their respective government, i.e., they are "national champions." But in the international market, producers compete for the importers' demand for weapons. As in the two papers mentioned above, importers are nonproducers. We also assume that both the governments on the supply and the demand side of the international arms market care about security and private consumption. But the perception of security is different on each side of the market. The international demand for weapons is characterized by an arms race situation in which security depends positively on domestic procurement and negatively on the adversary's procurement. The suppliers' security is affected by the global amount of exports. The supplier governments subsidize domestic arms production by ensuring non-negative profits to national champions and decide on the quantity of domestic procurement. Doing so ensures the existence of a domestic military industry. The chapter is organized as follows. In the next section we present the model and main assumptions. The section following thereafter presents the results of the comparative statics of our model. The final section states the main conclusions and suggests future lines of research. The model We consider an international arms trade market with ns producers. Each firm is the only supplier of weapons in its own country and competes with the other firms in the export market. On the demand side of the international market are 2nr countries, each involved in an arms race with its neighbor. Recipients have no domestic arms production sector and rely on imports for their military procurement. We assume for simplicity that each isolated seller is too small with respect to the whole export market to consider the effect of the weapons it sells on its own security function. In other words, both ns and nr are so large that sellers and buyers are price takers and there is no strategic interaction between them. However, governments of producer countries do act strategically with respect to their domestic firms.
164 A rm ing the Sou th
The supply side On the supply side we have ns identical countries, each with a national champion producing a homogeneous weapon. The military good is both consumed by the domestic country and exported to 2nr identical recipients. Let us denote P as the price of weapons in the international arms market and;? > P be the price paid by the domestic government per unit of military procurement. Each producer supplies g goods to its own government and exports a quantity of x. Since we assume that those countries which have a national champion do not participate in the import market, g therefore constitutes the total domestic procurement for that country. The cost of producing the amount of weapons required by the domestic government and those to be sold in the international arms market is a quadratic function, C[y] = D + cy +dy\ where y denotes global arms production (y = g + x), c and d are constants, and D represents fixed development costs. As a result firm's profits are given by (7.1)
x=pg + Px-c(g + x)- d(g + x)2 - D.
The weapon producer maximizes profits taking the domestic procurement price and quantity, p and g, as given since these are set by its government. The first-order condition for profit maximization then yields the following export supply function
P-c
The government acts as a leader and exploits the reaction function (7.2) in its choice of the procurement quantity, g, and the arms production subsidy,/? - P. As was said, we assume that exporting firms and recipient governments are price takers, hence the exports price of weapons, P, is taken as given. The government must fulfill its national budget constraint given by
Domestic procurement,
(7.3)
subsidies, and the arms trade
165
YS = CS + GS.
This implies balanced trade (i.e., financial autarchy). Here, Y is the output of the representative seller country which we assume is independent of the size of the military sector. Non-military consumption is denoted Q, and Gs = wsLs +pg is total military expenditure (where Ls is military personnel and ws is the wage rate in the military sector). Military expenditure on personnel and equipment ("input") results in military capability ("output") which is given by the production function &s = Ks [Ls, g]. In general one expects some substitutability between Ls and g but in the interest of tractability we assume fixed factor ratios and set Ks= ag = bLs . By a suitable choice of units for defense goods and military labor we can also set a = b = I. As a result we may rewrite military expenditure in the following way:
(7.4)
Gs = fc+p)g.
The government's objectives in any of the seller countries is defined by a utility function that depends both on consumption and security Us = Us [Cs, Ss]. Let us assume that the utility function has a Cobb-Douglas form:
(7.5) Us = C?Slr* = C? {(asYs + /U) VK])'"P Here, cos is the weight that consumption has in the utility function of the seller countries. The security term consists of two components: domestic security, which is an increasing and linear function of military capability Ks = g, and global security which depends on the aggregate arms trade Xs = ns x. Let us assume that security also follows a Cobb-Douglas form where ju is the weight of domestic security in the security function. In the definition of domestic security, asYs are the fixed benefits of defense and/?5, which can be assumed to be greater than one, captures the variable benefits of defense. Global security,/^], is a function of global
166 Arm ing the Sou th
exports. It would be reasonable to assume that weapon exporters realize the effect that global exports have on the situation of the recipients. A large amount of total weapon exports can create instability in recipient countries and this has a negative effect on sellers' security. But, consistent with the assumption that each arms selling country is small with respect to the rest of the market, the individual exporting country takes global exports,^, as given. Therefore, global security plays no role in the decisions taken by each individual government. Without loss of generality we may therefore let ju = 1 in equation (7.5). Finally, each selling government's optimization problem consists of maximizing welfare - equation (7.5) - subject to the budget constraint given by equation (7.3), where Gs is given by equation (7.4), and knowing that its "national champion" firm behaves according to its export supply function, equation (7.2). The government's choice variable is the quantity of domestic defense procurement, g. This will endogenously determine the domestic price to be paid, /?, which, in turn, ensures the existence of a national military sector. The reason why government wishes to subsidize weapons production by paying a price equal to or higher than the international arms market price,/? > P, is of course the desire to have a secure domestic supply of arms. To achieve the survival of the national military industry at minimum cost there is a participation constraint for firms which must be fulfilled. This participation constraint must ensure non-negative profits for the domestic firm, n > 0. Recalling equation (7.1), the minimum subsidy must be
(7.6)
r 1
+ D- Px
Note that —r = — > 0, i.e.,/? is strictly convex in exports. 6X g Introducing equation (7.6) into equation (7.4) and using equations (7.2) and (7.3), the government's utility function can be expressed in terms of g.
Domestic procurement, subsidies, and the arms trade
167
As an aside, it is interesting to see what happens if government chooses both g and exports x. In this case, the firm is regulated, the government's utility depends on g and x, and the firm's profit maximization equation (7.2) no longer applies. Maximization of log Us with respect to g and x yields the following first-order conditions
Sp g
(7.7)
^ Tx
„ .
Sp
s— = 0, i.e., — = 0 Ys -7= (ws + p)g Sx -G)<
Sp (w ™s s ++ p) P +S~ +
(7.9)
w
Ys ~ { s + P)
i}Z^s)Ps _ Q <xsYs + Psg
[Equation (7.8) is presented later on.] Note that equation (7.7) in fact fulfills the supply function of the firm, equation (7.2). Consequently, we can state the following proposition. Proposition 7.1: Ifns is large, then individual governments do not perceive the need to regulate their national arms producer. That governments do not perceive a need to regulate their national arms producer in equilibrium does not mean that efficiency gains from regulation would not be possible. But in our framework, these can only be realized if countries were to coordinate their choice of regulated exports. We now return to our original problem. The government's decision rule can be obtained from equation (7.9). Let us first note from equations
Sp (7.6) and (7.2) that g—= equation (7.9) yields
P- p. Hence, substituting this into
168 Arming the South
(-6)sasTs(ws
+ p)-
G)s{Jsg(ws
+ P)+
Ys(l-G)s)/3S
Rewriting this expression we obtain the government's demand for defense procurement, g, as a function of both the procurement price,/?, and the international price for weapons, P
(7.10) g=~
7= PsyWs +
jr r-y—. VsP+V-asjP)
Finally, using equation (7.4), we obtain government's total military spending as a proportion of national income
_m£s (7.11) — — Y S
{™s + P)S — Y S
{ws + P){{1 ~ as)fis - 0}sas{ws —
/ /3s[ws+cosP
/ +
+ p
))
v \ (l-cos)p)
To summarize, equations (7.10) and (7.11) give government purchases and total government military spending in terms of the domestic procurement price, p, and the international arms market price, P < p. Equation (7.6) presents the procurement price necessary for the existence of a domestic military producer. Finally, equation (7.2) provides the export supply function of the firm. The demand side and general equilibrium On the demand side of the international weapons market we have 2nr recipients of arms engaged in an arms race with their respective rivals. As for suppliers, we have a representative buyer whose budget constraint is (7.12) Yr = Cr + Gr
Domestic procurement, subsidies, and the arms trade
169
Here, Gr is the recipient's military expenditure. Following the same line of argument as presented above, this military expenditure equals the product of the quantity of weapons imported, m, and the sum of wages paid each soldier plus the international price of weapons (7.13) Gr = ( M T +
p)m.
Let us denote the analogous variables for the rival as
Yr ,C*r,G"r,wr y . As in the sellers' case, the recipients' utility function also depends on consumption and security and follows a Cobb-Douglas form, Ur - C®r Sfc°r . But the recipients' perception of security differs from sellers'. Here, security is an increasing function of imported weapons and a decreasing function of the adversary's imports, (7.14) Sr = arYr +
Prm-m\
Taking the rival's imports, m\ as given, the recipient's government maximizes its utility, Ur - subject to its national budget constraint given by equations (7.12) and (7.13) - by choosing weapon imports, m. We then have the following first-order condition
(715)
%-(wr+P)m arYr + P/n-m~
Rewriting this expression yields the reaction function that characterizes recipients' arms race situation
170 Arming the South
r)p, - wr(wr + P)ar )Yr. Assuming symmetry between rivals we have the following individual demand function for imports
(7.17) m = r a = -
j=
x
j=
x—.
As before, we can express this in term of recipients' military expenditure as a proportion of their national income
(7.18) -L=(w
+P)==±
,jri
r
-^-1
}
—.
This completes the demand side. Finally, world arms trade is given by equating total world supply with total world demand (7.19) Xs =nsx = 2nrm. For convenience, table 7.1 summarizes the model by repeating equations (7.6), (7.10), and (7.17) combined with (7.19). Cooperative solution among sellers In the previous section we saw that since governments are small with respect to world market supply, they do not perceive a need to regulate the arms trade. That is to say, the optimality condition for exports coincides with that of the firms. Let us see what happens if seller governments coordinate the quantity of world arms exports. Recall the sellers' utility function
Domestic procurement, subsidies, and the arms trade
171
Table 7.1: Model summary
r 1 c(g+x)+d(g+x)2 p = p[g,x\ = g ({l-(0s)ps-(0sas{ws
(7.6)
(7.10)
Ps(ws +
(7.17) and (7.19)
+ D- Px + P))Ys
cosP+(l-cos)p)
2nr((l-a>r)fir-a>r{^r+P)ar)Yr
Xs =
firfa+P)-(Drfc+P)
1-//
Us = CssSls"°s
C?((asYs + fisgY(f[xs])
Since Xs is no longer taken as given, global security now does play an important role. In the cooperative case, sellers maximize their joint utility and jointly decide on the quantity of weapons to be procured and exported. Maximizing ^ log Uls with respect to g* andx', and dropping superscripts, the first- order conditions are
(7.7')
•==
•*% . 0 - ^ - . ) ^ •+
Y
w
-CD,
(w + p)+
s ~ ( s + P)
{1.9)
*
+
0.
f[Xs] Sp^
**J + — (i-»M
Ys ~ w{w p) P)g Ss +
a Y
ss
+
Psg
0.
172 Arming the South
Equation (7.7') can be expressed as
Sp_{Ys-fc+p)g)(l-(Qsll-M)6f[Xs] 8x
cosg
f[Xs]
Sp
5x Sf[Xs]
Note that the sign of —— depends on the sign of . It is easy ox ox to see thatp is a convex function of exports. Therefore, if exports have a negative effect on security,
Sf[Xs]
< 0, the optimal amount of arms
OX
exports decided in a cooperative way by seller countries is smaller than the optimal amount as decided by firms in a competitive way. Thus, with cooperation suppliers see the need for the regulation of the arms exports. Comparative statics The structure of the global arms market is summarized in figure 7.1. Each producer of weapons faces two different markets: the export market and the domestic market. In the export market we have 2nr recipients on the demand side and ns firms compete as suppliers. In the domestic market government procures weapons from a single domestic producer. In what follows we first analyze each of the markets separately. We then analyze the comparative statics of the arms market's general equilibrium. The export market Let us recall equation (7.2), the supply curve of arms exports
P-c (7.2)
n - ~ -
g
.
Domestic procurement, subsidies, and the arms trade
173
Figure 7.1: The arms market
As can be seen, the amount of weapons offered by a firm in the international market is an increasing and linear function of their international price, P. Changes in the other parameters will shift the supply curve in the x, P quadrant (see figure 7.1). A decrease in domestic procurement, g, or in the parameters in the cost function, c and d, shifts the supply curve outward. The reason why domestic procurement affects exports in the same way the cost function parameters do is that although domestic procurement is not a decision variable for the firm, the decreasing returns to scale assumption in the variable cost function implies increasing marginal costs of domestic production. The aggregate demand function, which is the demand for weapons each producer faces in the international market, can be derived from equations (7.17) and (7.19).
174 Arming the South
2n\{\-co))pr-cot{^^P)a)Yr (7.1V) x = — - -
7
w=
x
—.
Remark 7.1: x > 0 o ( l - 0 r ) # . - Q)r(wr + P)ar > 0. The aggregate demand for firms' exports, equation (7.17'), is a decreasing and convex function of the export price, P. It can also be proved that the aggregate demand curve shifts outward with an increase in the number of recipients, nr or the national income in the recipient countries, Yn and with a decrease in the number of producers, ns, or in the parameter which measures the importance of the alternative consumption good in the utility function of the importers, cor. However, the effect of an increase in the variable benefits of defense for the recipients, Sx/3j3n on the aggregate demand curve is not clear. Remark 7.2: ——< O o (l-CDr)-(wr+ 0pr
P)ar > 0.
From
Remark 1,since fir > 1, 0 < COr < 1, ( l - (0r)- (wr + P)(Xr > 0 is sufficient for demand for weapons to be positive. Introducing equation (7.17') in equation (7.2) we obtain two equations which express the export price, P9 and the quantity of exports, x, as an implicit function of the aggregate demand function parameters, the cost function parameters, and domestic procurement, g.
x(fir- a>r)(wr + 2d(x+ g)+ c) (7.20a)
-^{(\-co^ r-cor{^r n s = F[g,c,d,a)r,pr].
+
2d{
+g)+c)ar)7r
Domestic procurement, subsidies, and the arms trade
2nX{X-^r)^-cor^^P)ar)Yr
175
P-c
= H[g,c,d,o)r,Pr]. Applying the implicit differentiation rule to equations (7.20a) and (7.20b) we can derive the effects of a variation in the various exogenous parameters on equilibrium exports and the export price. At this stage we treat the domestic market as completely exogenous. That is to say, we do not take into account that a variation in the export market would induce variations in domestic procurement and therefore further variations in the export market equilibrium. The results obtained can be summarized by the following two propositions. Proposition 7.2: The equilibrium export level in the isolated export market decreases with increases in domestic procurement, g, the parameters of the cost function, c and d, and the weight of consumption in the recipients' utility function, cor The effect of an increase in the variable benefits of defense, /?r, depends on the parameter values.
7 7 < 0 o x(Vr + 2d(x +g)+c)~ — ( l - G>r)Yr > 0. This expression can be written as
8x ^<0*(m(wr -m(P-2d(x+
+
p)-(l-cor)Yr)
g)-c)> 0.
An increase in the variable benefits of defense,/?,., allows the recipient country to enjoy the same level of security while buying fewer weapons.
176 Arming the South
The existence of alternative consumption in the utility function of the recipient country allows this country to care about income. Therefore an increase in fir generates a substitution and an income effect. The substitution effect encourages consumption of the good which has become cheaper, security, whereas the income effect encourages an increase in alternative consumption. When the valuation of income, which is measured by con is small, an increase in /?,. will increase the amount of arms imports. But if the valuation of income is sufficiently high - such that the income effect is much greater than the substitution effect - then the recipient will increase expenditure for consumption goods and reduce arms imports. Proposition 7.3: In the isolated export market, an increase in the domestic procurement, g, the cost function parameters, c and d, and a decrease in the weight of consumption for recipients, cor, will increase the equilibrium export price, P. Finally, an increase in/? r decreases the equilibrium export price, P, under the assumption stated in Remark 7.2. The domestic market The relevant supply curve for the domestic market can be derived from equations (7.2) and (7.6). Equation (7.6) can be written as
r
i
D
c(g+x)^d(g+x)\
D-
P(x+g)
g This states that the price to be paid by the government must cover the international arms market price, P, plus the average losses the firm would incur were it to serve both the domestic and the foreign market at the same price. P < p implies a positive assumption on loses. Introducing firms' optimized decisions, equation (7.2), into equation (7.6) we obtain
Domestic procurement, subsidies, and the arms trade
177
if
(P-CY p+-\D--—— g{ 4d
(7.21) p=
This curve is the participation constraint for the domestic firm. In figure 7.1 all points to the right of the curve ensure positive profits for the domestic firms. This constraint is represented in the/?, g plane where/? is a decreasing and convex function of domestic procurement. Increases in the fixed cost D and in the parameters of the cost function, c and d, shift the supply curve outward by increasing its slope. An increase in the international price, P9 shifts the supply curve inward although it shifts the abscise P to the right. The government's decision rule, equation (7.10), constitutes the domestic demand function for weapons
( Q - f l s K - ®saS(ws + p))Ys
This expression is decreasing and convex in the price that the government pays for defense procurement. Moreover, increases in the international price for weapons, P, and the weight of consumption in seller countries, cos or as, shift the decision rule inward, and increased domestic national income, Ys or fis shift the decision rule outward. Introducing equation (7.21) into the government's decision rule we obtain the solution for the level of domestic defense procurement, g
(
\ {-
(P-cY)
(1- as)fis\Ys - D+
4d
/—
y-
j - cosas(ws + P)YS
(7.22a) g =
Psfc+P) and introducing this back to the first equation we have the value for the price of domestic procurement,/?
178 Arming the South
D
(/>-c)n Ad
h{ w^
+ P
(7.22b) p= P +
(••^WV
D
+ ~^~\
~ G>sas(ws + P)YS
Using equations (7.22a) and (7.22b) we can derive the effects of a variation in the various exogenous parameters on the equilibrium level of domestic procurement and its price. Again, we treat the export market as completely exogenous. That is to say, we do not take into account that a variation in the domestic market would induce variations in the export market. The results we obtain are summarized in the following propositions: Proposition 7.4: A decrease in the international price for weapons, P, fixed costs of production, D, the parameters of the cost function, c and d, the weight of consumption in the sellers' utility function, co^ and an increase in domestic income, Ys, or fis will increase the equilibrium procurement, g, in the isolated domestic market. Proposition 7.5: An increase in the fixed costs of production, D, the parameters of the cost function, c and d, the weight of consumption in the domestic utility function, cos,, and a decrease in domestic income, Ys, or /3S will increase the equilibrium domestic price, p, in the isolated domestic market. Finally, the effect of a variation in the international price for weapons, P, on the equilibrium domestic price, /?, depends on the concrete parameter values. The general
equilibrium
So far we have analyzed shifts in the supply and demand curves in the export and domestic market and the comparative statics of the equilibrium of each of these two markets, taking the other as exogenous.
Domestic procurement,
subsidies, and the arms trade
179
But the comparative statics in general equilibrium - i.e., the analysis of the effects of parameter changes on the the equilibrium values of domestic procurement, domestic and international weapon prices, and the quantity of exports - requires the calibration of the model.3 In this calibration, the subsidy, r, is defined as the difference in home and export price as a proportion of the export price
P-P .
T -
P As can be seen in figures 7B.2 and 7B.3 (in appendix 7B), as the weight of consumption in the domestic utility function increases, domestic procurement decreases, exports increase (figure 7B.2), the international price for weapons decreases and the subsidy to the firm increases (figure 7B.3). We can explain the intuition of this result starting from the comparative statics in the isolated markets stated in the previous subsection. In the domestic market, an increase in the weight of consumption in the utility function, that is to say, a decrease in the importance of security in the sellers' utility function, will have a negative impact on the optimal amount of procurement. Since decreased domestic demand could make domestic production unprofitable for the firms, the per unit subsidy to the firms will need to increase. In the export market firms will try to compensate for the decrease in domestic demand by increasing exports, which has a negative effect on the international price. As indicated in figures 7B.4 and 7B.5, as the variable benefit of defense for the recipients, /?,., increases, domestic procurement in seller countries decreases, exports increase (figure 7B.4), the international price for weapons increases, and the subsidy to the firms decrease (figure 7B.5). Again we can explain the intuition of this result by recalling the results in the previous section. As stated in Remark 7.2, the effect of an increase in Pr on the equilibrium in exports and export price in the isolated export market could be either positive or negative, depending on the importance of security in the recipients' utility function. We see that for the calibration we performed, the parameter values are such that the final effect of an increase of/?,, over the equilibrium amount of exports
180 Arming the South
and export price is positive. The expected effect of an increase in the export price on the domestic market is a decrease in the subsidy and an decrease in the amount of procurement. Figures 7B.6 and 7B.7 show that as the variable benefit of defense for sellers, 0S, increases, domestic procurement increases, exports decrease (figure 7B.6), the international price for weapons increases, and the subsidy to firms decreases (figure 7B.7). An increase in fis will increase the equilibrium procurement, g, in the isolated domestic market and will decrease the equilibrium domestic price, /?, in the isolated domestic market. The increase in the domestic procurement, g, will have a positive effect on the equilibrium export price, P, and a negative effect on the equilibrium export level. In figures 7B.8 and 7B.9 we show that, when there is a general increase in the weight of consumption in the utility function of the countries involved in arms trade, both domestic procurement and exports decrease (figure 7B.8), the international price for weapons decreases, and the subsidy to firms increase (figure 7B.9). The general decrease in importance of security discourages consumption of weapons for both sides of the arms market. In the domestic market the decrease in the importance of security in the utility function of the seller will have a negative impact on the optimal amount of procurement. In the export market firms will try to compensate for the decrease in domestic demand by increasing exports, which has a negative effect on the international price. However, the decrease of the importance of security for the importers of weapons discourages arms imports. As a consequence, subsidies will have to increase dramatically for the sellers to remain selfsufficient in military procurement. Conclusion This chapter presents a model which captures some of the characteristics of the structure of the current international arms market. It considers both the domestic and the international market for weapons. This allows for the analysis of the interaction between the two markets and the effect that a variation in the relevant parameters has on the optimal decisions of the agents involved.
Domestic procurement,
subsidies, and the arms trade
181
The model highlights the role of export markets as a main determinant of defense procurement policies in the producer countries. Compared with previous arms trade models, ours describes the military sector and the governments of weapon exporters as two different sets of decision-makers with different decision variables and objectives. Firms decide on the quantity of weapons that maximize their profits. Governments decide on the quantity of defense procurement that maximizes their utility. However, there is a close relationship between governments and firms: the national champions are guaranteed positive profits by the choice of the procurement price. Finally, exports affect the level of security of the seller countries and this creates further interdependence among all participants. In this chapter, we obtain the main properties of the equations that characterize the behavior of importer countries, firms, and governments of exporter countries. The aggregate demand for firms' exports is a decreasing and convex function of the export price. The arms export supply curve is an increasing function of the export price. The domestic procurement price that ensures the existence of the firm is a decreasing and convex function of domestic procurement. Finally, the domestic demand function for weapons is decreasing and convex in the price that the government pays for defense procurement. As a result of the calibration of the model, we derive the effect of a variation of the weight of consumption in the utility function and the variable benefits of defense on the general equilibrium in the arms market. The comparative statics results give a picture of the main characteristics of the amis market. Among other results, we find that a decrease in the weight of security in the utility function of exporter countries decreases the amount of domestic defense procurement, increases the subsidy to the firms and the quantity of exports, and decreases the export price. However, a general decrease in the importance given to security in the utility function decreases both domestic procurement and exports. In future research, we intend to include imperfect competition among weapon producers in the international market. With a small number of producers and governments, each government would recognize the effect its actions have on security. Governments would then act strategically
182 Arming the South
with respect to each other and our model would acquire some of the characteristics of those in the strategic trade theory literature. Notes Work on this chapter was made in the context of the authors' participation in a group working on arms trade, financially supported by the ESRC under grant R000235685. We wish to acknowledge the members of the group for their valuable comments. For more information on the Arms Trade Group see its web site at http://mubsr.mdx.ac.uk/ArmsTrade. This chapter was produced during a stay of the first author at the University of Surrey. 1. For a review of the arms race literature see, e.g., Brito and Intriligator, 1995. 2. It is standard in the new trade theory to analyze the relation between government and firms and its effects on international trade. For a review of this literature see, e.g., Helpman and Krugman, 1989. 3. Most of the assumptions over the value of the parameters used in the calibration of the model are based on data from SIPRI and ACDA, the two main sources of military information. References Brito, Dagobert L. and Michael D. Intriligator. "Arms Races and Proliferation," pp. 109-163 in Keith Hartley and Todd Sandier (eds.) The Handbook of Defense Economics. Amsterdam: North-Holland, 1995. Helpman, Elhanan, and Paul Krugman. Trade Policy and Market Structure. Cambridge, MA: MIT Press, 1989. Levine, Paul, Somnath Sen, and Ron Smith. "A Model of the International Arms Market." Defence and Peace Economics Vol. 5 (1994), pp. 1-18. Levine, Paul and Ron Smith. "The Arms Trade and Arms
Domestic procurement, subsidies, and the arms trade
183
Control."Economic Journal Vol. 105 (1995), pp. 471-484. United States Arms Control and Disarmament Agency (ACDA). World Military Expenditures and Arms Transfers 1993-1994. Washington, DC: ACDA, 1995.
184 Arming the South
Appendix 7A: Calibration of the general equilibrium First, consider the calibration of a,-, /?,-, coi9 wi9 where i = s, r. We choose values that replicate observed values of G, / Yt and the ratios of military expenditure on personnel and equipment. Regarding the latter let Rs = (ws L^ /pg = ws //? be the personnel-to-equipment expenditure ratio for the suppliers, and Rr = wr/P bethe corresponding ratio for the recipients. Then, we set ws = Rsp, wr = RJP, and/? = (1+T)P. We normalize the export price, P = 1, and we impose /?, and ar = as = a and cor = cos = co. Introducing this in equations (7.11) and (7.18) we have two equations for co and a si
(7A.1) (\-6))/l-6)(Rr
+ l)a =
(p-6>)-=r=fc-faG>,
where
(1-<»)/?-
+ 0) + ( l - 0 ) / ? ) G F
K + i)/> Gs
5
*
=
^i -
^
PT Gs \—=~. i)p Ys
where 91 S, = P6 ^=^; f2^ = 7 Ys' (RS +
Dividing equation (7A.1) and (7A.2) and isolating co we get
t5(p-h)-p o) = 7
+ 43 -.
u
x
Rr + l
rr, where L = —(
[h-P-hih-fi))
r.
M^+i)
Domestic procurement, subsidies, and the arms trade
185
Now, from equation (7A.2) ( l - 0))P-
a=
3 + >4Q)
0)1 >{Rr +
l)
Using equations (7.17) and (7.10):
((\-o)p-G)(l+Rr)a) Px (7A.3)
1 In m
rel
(\+ R,)(j3- a V
pg
1+T ns g
A
\+T ({\-cQ)lJ-o){pRs
+ \)ay
p(pRs + a> + (l-a))p)
2nJr
where rel = — ! = ^ measures the relative size of recipients in economic n Y ss terms. The above expression can also be written as Px/pg =f\x\. Now consider the calibration of c, d, and D. Let 11 =
{p-cf
: — = K + D be the profit flow. Then we can obtain an
4d
expression for c (7A.4) c = 1 - 2y[dYl. Equation (7.2) gives
fn (7A.5)
x+g
Using equations (7A.5) and (7.19) we have
186 Arm in g th e So u th
ft O
I l vX/
~ g = 2nrm.
[\d
n
nr
Vd
rio
Hence, J — = 2 — m + g. This gives the following expression for d
n^ (7A.6) d =
—{
_
m
K nr Y
_
2 •>
^
m
where m - = a « J a = = .
As before, let 11 =
(i>-c)2
Ad
= n + D be the profit flow. Then, using
D-U equations (7.6) and (7.2), we get that r =
. Let £> = (1 + 0 ) 77, g (6 > 0) and 77 = r(g/d). Then we obtain military profits as a percentage of GDP
n (7A.7) = =
Ys
T((\-o))p-co(pRs
9
+ \)a)
pRs + a + ( l - co)p)
T
9
We can use data on military profits as a percentage of GDP in order to calibrate 6 or r. Finally, we normalize the GDP of sellers, Ys = 1, to obtain rel. In the calibration we set 6 and x. This completes the calibration. Table 7A.1 summarizes the results of the calibration for
Domestic procurement, subsidies, and the arms tn
assumed and calibrated parameters. Table 7A.1: The model calibration Gs/Ys Gs/Ys Rs R, 9
=0.03 =0.06 = 0.50 =0.25 =3
d x rel
=0.86 =0.2 = 0.4
ar = a s =0.14 cor = cos = 0.9 c = 0.93 D = 0.0044
188 A rming the South
Appendix 7B: Figures 7B.2 to 7B.9
0.022
0.018 2
0.016
•2
0.014
8
0.012
o
0.008 0.006 0.004 0.002 + 0.98 omegas
Figure 7B.2: Government purchases and exports against sellers' weight on consumption (see narrative for explanation)
1.2 1.1
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
0.9
0.91
0.92
0.93
0.94 omegas
0.95
0.96
0.97
Figure 7B.3: Export price and subsidy against sellers' weight on consumption (see narrative for explanation)
0.98
Domestic procurement, subsidies, and the arms trade
0.03 0.025 o
0.02 % 0.015 S3 |
0.01 0.005
1.2
1.4
1.6
2.2
1.8
2.4
2.6
2.8
betar
Figure 7B.4: Government purchases and exports against importers' variable benefit of defense (see narrative for explanation)
1 0.9 0.8 •K 0.7 3
•S 0.6 8 0.5 0.4 0.3 ]
0.2 <1.2
1.4
1.6
2
2.2
2.4
2.6
betar Figure 7B.5: Export price and subsidy against importers' variable benefit of defense (see narrative for explanation)
2.8
189
190 A rm ing th e So u th
0.02 0.018 a 0.016 x
1
0.014
§ 0.012 ^
0.01 0.008 0.006 2.65
Figure 7B.6: Government purchases and exports against exporters' variable benefit of defense (see narrative for explanation)
1 0.9 0.8 .-2 0.7 0.6 •a 0.5
0.4 0.3 0.2 2.65
Figure 7B.7: Export price and subsidy against exporters' variable benefit of defense (see narrative for explanation)
Domestic procurement, subsidies, and the arms trade
191
0.02 0.018 GO
o X (I)
0.016 0.014 0.012
GO CO
03
x: o u,
0.01 0.008
> 0.006 O
0.004 0.002
0.9
0.92
0.94
0.96
0.98
omega Figure 7B.8: Government purchases and exports against the weight of consumption (see narrative for explanation)
1.4 n
1.2 |
CO
-°3
i
CO
I
•o 0.8 H c cd
f
06
!
j
0.4 j 0.2 I 0.9
0.91
0.92
0.93
0.94
0.95
0.96
0.97
omega Figure 7B.9: Export price and subsidy against the weight of consumption (see narrative for explanation)
0.98
This page intentionally left blank
PART III: COUNTRY STUDIES
This page intentionally left blank
8 Saudi Arabia: Defense Offsets and Development Ron Matthews
Introduction At the beginning of the 21 st century technology offsets have become commonplace for countries undertaking major overseas arms procurement. The purpose of offsets is to ensure that a proportion of funds spent are re-invested to achieve economic development goals. The development significance attached to offsets is such that after import substitution and export promotion strategies they are often viewed as a "third-way" for economic development of underdeveloped countries. Although this may be hyperbole, arms purchasing countries wield substantial market power. Market power has resided in the hands of arms importing developing countries for almost two decades now, coinciding with the draw-down in global defense expenditure and the consequent decline in defense procurement (reduced by 34 percent between 1987 and 1997).1 As the arms trade takes place within a buyers' market, it follows that arms importing countries are in a dominant position to extract compensatory investment. Such "offsetting" technology transfer has grown apace with the tightening of the global arms market. Evidence of the growing importance of offsets can be found by referencing to the 1970s when only a handful of countries possessed countertrade and offset guidelines; there are now upwards of 150 countries with published countertrade requirements.
196 Anning the South
The majority of defense industrialists, Ministry of Defense civil servants, and informed observers understand and are comfortable with offset terminology, but the current trend is to drop the term offsets in favor of a more positive rubric such as industrial "cooperation" or "partnership." These terms connote long-term, mutually beneficial relationships that are arguably more acceptable than "offsets;" the latter being viewed as short-term and negative, i.e., the price to be paid for winning a contract.2 Purchasing countries and vendor companies alike, are now synchronized to the view that governmental and corporate stakeholders, representing both sides of the contractual relationship, see offset programs as an opportunity not a threat. Moreover, although vendor companies have traditionally regarded offsets as an incumbrance, distracting them from their main activity of selling arms, most firms now interpret offsets as a strategic marketing vehicle for locking-in overseas customers into existing and future product portfolios. From a scholarly and commercial perspective, these issues are challenging, but in practical terms they are sterile unless offsets make a positive contribution to local economic development. The offset impact can be felt from either the defense or civil standpoint. Indeed, given the increasing policy emphasis on dual-use, civil-military development and production activity, offsets may straddle the civil-military divide. Whatever the perspective, the bottom line is that there must be a development impact; if this is absent, then there is really no point in pursuing offsets. To achieve this impact, the authorities must develop a vision, translating offsets into a vehicle for the long-term, dynamic, and, importantly, sustainable infusion of "technology" into the local economy. The aim should be to create domestic jobs; but of equal significance, the jobs should be highly skilled, maximizing added value. The ultimate aim must be to develop indigenous, innovative, and appropriate output, but this is not a simple task. Whether offsets are destined for the defense or civil sector, the problem for developing countries is that they will always be playing technological "catch-up" with the established, advancedcountry companies. These foreign enterprises will normally have the benefit of a long-standing international reputation, as well as overseas marketing and distribution networks, and these advantages will be in addition to all the familiar Western corporate attributes of skill, scale,
Saudi Arabia: defense offsets and development
197
scope, learning economies, and capital availability. The question that needs to be put, therefore, is whether offsets actually work. This is a deceptively simple question, but one that carries great policy significance. Without trying to pre-empt this paper's conclusions, there is clearly merit in the view that offsets are a mode of technology transfer that have the potential to facilitate local economic and technological development. However, whilst offsets may act as a stimulus to the promotion of indigenous production, by themselves they are not sufficient. Also important is the local capacity to absorb the technology transferred. Technology transfer will not be viable or sustainable in the absence of adequate domestic technological absorptive capacity. Hence, the easily defendable assertion that offsets will likely work, and indeed have done so in advanced, knowledge-intensive, and technologically sophisticated economies, such as those of Singapore and Japan, but are unlikely to be successful in impoverished, less industrially mature environments. The Philippines provides a classic example of the problems that the less industrially advanced countries face in the implementation of sustainable offset programs. In the early 1990s, the Philippine government procured from the British company, GKN, 150 "Simba" armored personnel carriers at a cost of $56m (Villanon, 1998, p. 4.11). A local company (the Philippine Asian Armoured Vehicle Technologies Inc - Philcorp) was created for the in-country assembly of 142 Simba vehicles. Britain's GKN fulfilled its commitment and Philcorp's licensed production of the Simba was completed on schedule.3 There was, though, only limited technology transfer, and on contract completion, Philippine employment at the plant fell from 62 to 29 workers, with no fresh orders or export possibilities in sight (Villanon, 1998, 4.12). The purpose of this chapter, then, is to explore the development relationship between primarily defense-related offsets and local industrial and technological development. The key issue to be addressed is whether offsets contribute to development. The presumption is that they may work for Singapore and Japan, but do they work for developing countries such as arms-importing Middle Eastern countries, e.g., Kuwait, the United Arab Emirates, and Saudi Arabia?4 To tackle this question, it is necessary to lay out a framework to define and scope the constituent elements of
198 Arming the South
offsets, integrating this framework into the broader fabric of a countertrade typology. Once these building blocks are in place, it is helpful to study the impact of defense offsets on development by reference to the case of Saudi Arabia. Saudi Arabia is relevant because it is a good example of a country which since 1985 has sought to exploit the potential that offsets are believed to offer. The Kingdom is an industrializing country, highly dependent on one resource, oil, for the majority of its income; yet at the same time has been the world's first or second biggest importer of arms for the last 15 years.5 The focus on Saudi Arabia begins by evaluating, chronologically, the development of the Kingdom's various offset programs and then appraising their developmental impact, or lack of it. The chapter closes by speculating on the appropriate policies for strengthening the contribution that offsets can make to Saudi development. Defense offset typology Defense offsets should be examined as part of a much bigger picture of countertrading and reciprocal purchasing activity. Figure 8.1 illustrates the various elements of this framework. Barter, counterpurchase, and technology transfer represent the principal non-monetized and compensatory aspects of countertrade. These trading mechanisms diverge from the conventional cash transactions model and instead concentrate on commodity-based exchange and trade reciprocity. Barter is perhaps the best-known form of non-monetized trade, which in its simplest form refers to the one-off exchange between countries of, say, one ship-load of coal for another of iron ore. Alternatively, a contractual relationship may be extended over time (so-called clearing arrangements) to incorporate higher value transactions, e.g., arms deliveries, where there would have to be substantial and ongoing deliveries of perhaps oil or agricultural commodities to pay for such shipments. A further derivation of barter is called "switch-trade." This brings several countries worldwide into the contract, enabling deals to be struck where the purchaser's currency is neither convertible nor acceptable to
Saudi Arabia: defense offsets and development
r
j CounterJ trade . T i ! 1 ! Baxter |
| Counteri purchase
Offsets (technology transfers) Primary defense contract Defense offset
Direct
Civil offset
Primary civil contract Civil offsets
Indirect
Figure 8.1: Countertrade and offsets: a typology
Purchaser
Russia
Primary contract Payment in Indian commodities/manufactures Vendor...
Rolls Royce pic
Figure 8.2: The process of switch-trade
199
200 Arming the South
the vendor in the initial contract. An example may best explain how switch-trade works in practice. Figure 8.2 illustrates a hypothetical, but not unrealistic, case where the Russian government seeks to develop its local aerospace industry through the purchase of British Rolls Royce (RR) aeroengines. The British government and RR are happy for the deal to be closed, but there is one difficulty: RR is unwilling to accept payment in Russian roubles. Nor is the British company prepared to barter its aeroengines for unattractive Russian civil goods, or even vodka.6 There is an alternative possibility, however. Russia has built up a considerable credit balance with India from the 1980s barter (clearing arrangements) deals of T-72 Main Battle Tanks for Indian rice, tea, and textiles. Russia may offer to "switch" these trade credits with India to Britain as payment for the engines to be purchased from RR. Russia is happy (it receives the engines), RR is content (being paid in products eminently more saleable than, say, Russian black and white television sets), and India is delighted (because it gets to market its commodities, and even manufactured goods, in Europe). Money is not used at any stage in barter transactions, representing simply a mutual exchange of wants as per classical economic scriptures. By contrast, counterpurchase - the second major element of countertrade - does use money. It is employed to lubricate both the primary defense transaction and an additional reciprocal or "counter"-purchase obligation the nature of which is determined at the time the primary defense contract is signed. Counterpurchase deals normally require that the arms vendor purchases specified commodities from the purchaser's host economy, up to an agreed percentage value of the primary defense contract.7 Technology transfer is the third and final form of countertrade mechanism. As can be observed from figure 8.1, it is more complex than the barter and counterpurchase arrangements. The aim is clear, however: to create a win-win situation whereby the arms purchaser is able to extract inward technology transfer from the vendor; the latter benefitting not only from the current sale, but also from anticipated future sales on the back of the goodwill generated. This is the theory, at least. In reality, of course, much depends on the ability of the vendor company to facilitate reciprocal investment flows.8 It goes without saying, that there is a real and obvious cost of establishing offset companies in overseas
Saudi Arabia: defense offsets and development
201
countries,9 and it is unrealistic to believe that a high proportion of the offset credits earned will be sourced from the vendor company directly.10 Offsets will in any case be difficult to achieve, and may be slow in forthcoming. Exporters, for instance, may be content to continue with direct sales - an easy form of market access. Acting rationally, they will be loathe to expose themselves to unnecessary market-risk by transplanting manufacturing enterprise in difficult, uncertain, and possibly dangerous foreign environments. Such risk aversion will make it difficult to encourage businesses to enter into joint venture deals. If anticipated offset targets are not met, then future arms sales will simply be "will-o'-the-wisp" contemplations. Technology transfer can be tied to either defense or civil contracts. The opportunity to obtain offsets from major defense procurement is an option available to all countries, industrialized or developing. By contrast, offsets linked to high-value civil contracts is a policy route available only to developing countries. In January 1996, the World Trade Organization (WTO) introduced the Government Procurement Agreement (GPA), whereby, under Article XXIII, member states agreed not to seek offsets when purchasing goods and services (US Department of Commerce, 1998, p. 48). The only permitted exception to Article XXIII are developing country purchasers.11 For example, Indonesia's early 1990s purchase of Airbus and Boeing airliners for its national airline, Garuda, was tied to in-country licensed production of components and subassemblies for these aircraft. Offset options stemming from the procurement of defense equipment are more varied. Most major purchases of military equipment will attract an offset requirement. The offset can link directly into the primary defense contract (e.g., licensed production of the weapon systems procured), or, alternatively, maybe funneled into other defense work that is incidental or "indirect" to the primary defense contract. Defense-based offset investment is the approach favored by the UK Ministry of Defence (Matthews and Williams, 2000). Here, the aim is to compensate for placement of the defense contract with an offshore supplier by encouraging inward technology transfer into the local defense industrial base, to a similar value and quality of work that has leaked abroad. Direct defense offsets only make sense, in the present tight international
202 Arming the South
economic climate, if the purchasing country possesses a robust defense industrial base: it is a rational development approach for the advanced (and, possibly, industrializing) countries which have high levels of defense expenditure and scale in procurement and production, but it is hardly justified for other countries with minimal defense budgets and low levels of procurement. Finally, offsets tied to major defense purchases can be directed to the civil sector. This has been the trend in recent years amongst the developing countries. The majority of major defense procurements are linked to some form of countertrade requirement. For instance, Saudi Arabia pays for British Tornados, Hawks and other defense equipment in oil (a barter deal, under the Al Yamamah contract), and, linked to the defense contract, there is a requirement for reciprocal offset investment into Saudi Arabia. The factors determining the nature of countertrade deals depends, first, on whether the objective is short-run economic stimulus or long-run technological dynamism, and, second, whether the bias is toward civil or defense-industrial development. As a generalization, barter is employed to facilitate trade, which otherwise might not have taken place. Financial illiquidity is normally the conditioning factor in this regard. Counterpurchase, in contrast, is viewed as a short-term mechanism to strengthen demand in the local economy, through the circular-flow impact of economic multipliers. Additionally, there is a marketing aspect to counterpurchase. For example, in the mooted sale of (BAE Systems) Hawk trainers to India, a counterpurchase contract will likely be required. In all probability, this will involve the marketing of Indian manufactured goods in Europe. A counterpurchase deal 10 years ago would have included Indian rice, tea and textiles, but not now. The final dimension of the countertrade typology is technology transfer. This is viewed by the arms purchasing countries as harboring the greatest potential impact for long-term, dynamic, and sustainable economic and technological development. It is the recognition that offsets may act as the investment conduit for local technological development, which brings this introductory discussion to a close. Having put in place the contextual backdrop, and additionally defined and scoped the principal elements of countertrade, particularly offsets, the stage is now set to explore the impact of offset programs, with particular reference to
Saudi Arabia: defense offsets and development
203
the case of Saudi Arabia. Chronology of Saudi offset programs The objectives of the Kingdom's offset programs were specified in a 1983 Saudi Economic Offset Committee (SEOC) document (Matthews, 1996, p. 239 and Al-Ghrair and Hooper, 1996). Principally, offsets were viewed as a mechanism to raise the rate of economic diversification by stimulating viable and sustainable commercial activities. Hence, projects directly related to upstream oil and gas investments are excluded. Other primary objectives include: 1
broadening the Kingdom's economic base and increasing import substitution and exports so as to contribute to the growth of national income;
2
facilitating the inward transfer of state-of-the-art technology in order to provide increased opportunities for Saudi Arabian managerial and technical personnel in the private sector to develop their expertise and gain access to modern technology;
3
providing more investment opportunities for Saudi investors so that they can share more fully in the Kingdom's economic prosperity.
These offset policy objectives have shaped the process of implementation. However, whilst there has been an emphasis on economic development and diversification, there is considerable variety in the contractual conditions among Saudi Arabia's six offset programs. The remainder of this section lists and explains the principal features of these programs. Peace Shield I The origins of the Saudi offset programs date back to 1985. Peace Shield I (see table 8.1) was the first to be implemented and was linked to the US$3.8 billion air defense contract signed the year before. The prime
204 Arming the South
Table 8.1: Saudi Arabia: prime contracts and associated offset programs Prime contractor
Contract description
Year signed
Value (US$ million)
Offset program
Boeing
AWACS Platform Command, Control, and Communications Systems for the RSAF
1984
3,800
Peace Shield
British Aerospace (now BAE Systems)
Tactical fighter aircraft, associated equipment and services, and airbase construction for the RSAF
1988
7,600
Al-Yamamah
Hughes Aircraft
Extension of the Peace Shield Program
1991
837
Peace Shield II
General Dynamics
Supply of M1A2 Abrams Mam Battle Tanks and associated equipment and systems for the Royal Saudi Land Forces (RSLF)
1992
n/a
General Dynamics Economic Balance Program
McDonnell Douglas
Supply of F-15 fighter aircraft and associated equipment and systems for the RSAF
1993
n/a
McDonnell Douglas Peace Sun IX
ThomsonCSF
Supply of frigates and associated weapons systems for the Royal Saudi Naval Forces
1994
3,500
Al-Sawari
AT&T International (now Lucent Technologies)
Sixth Telecommunication Expansion Project (TEP-6) for 1.5 million new telephone lines and 200,000 GSM lines
1994
6,000
AT&T Offset
Sources: Adapted from Riyadh-based Economic Bureau (Spring 1998), p. 51, and Matthews (1996), pp. 233-251.
contractor for the contract was Boeing, and under its management five offset companies were established: two are aircraft and aircraft equipment repair and maintenance companies - Alsalam Aircraft Co Ltd and Aircraft Accessories and Components Co (AACC) - one is engaged in
Saudi Arabia: defense offsets and development
205
the manufacture and repair of telecommunication and electronic equipment - Advanced Electronics Co (AEC). Another is a small computing systems enterprise - International Systems Engineering (ISE) - and the fifth specializes in the maintenance and repair of gas turbine engines and components - Middle East Propulsion Company (MEPC). Peace Shield I was a defense offset program built around an obligatory 35 percent offset target, as measured against the technical value of the primary defense contract. This was the Kingdom's first and last offset arrangement involving compulsion. By 1993, the offset target had been achieved, indeed surpassed: the offsets totaling at that point 37.5 percent of the primary defense contract value (Matthews, 1996, p. 248; original source, interview with SEOC secretary, January 1994). Al Yamamah Although there have been two major British-Saudi government-togovernment defense contracts (1985 and 1992), only one Al Yamamah offset program has been implemented. This was initiated in 1988, three years after the signing of the first primary defense contract involving the Saudi purchase of 72 Tornado fighter bombers, 30 Hawk trainers, and diverse numbers of other defense-related projects, including airfield construction (Matthews, 1996, p. 235). The Al Yamamah offset program has been dogged by difficulties, including: delay, caused by the Gulf War; problematic international economic conditions; and the risk-averse nature of British corporate investors, themselves affected by the early 1990s domestic recession. However, during the latter part of the 1990s, there have been a number of high profile offset ventures implemented, notably a sugar refinery and a pharmaceutical plant. The number of Al Yamamah offset companies now amounts to seven, comprising the United Sugar Company Investment; Glaxo Saudi Arabia pharmaceutical products manufacture; Saudi Development and Training Co (with BAE Systems); a Cyclar Project (SABIC and BP); Dharan Harco Chemical Industries; Rezayat Flover Co Ltd, involving the repair of instrumentation and calibration equipment; and finally Unilube, an engine oil recycling operation. The Al Yamamah offset program is intended to capture both civil and
206 Arming the South
military offset investments. In practice, however, all the offset projects have been located in the commercial field, save for a recent Rolls Royce offset investment, cross-threading into the Peace Shield offset company, MEPC.12 The Al Yamamah offset program is based on "best endeavors," with a £1 billion offset target; that is, 25 percent of the technical value of the primary defense contract, to be achieved across an illustrative 10 year time-frame. The offset deadline was passed in 1998, without the achievement of this target. Peace Shield II Hughes Aircraft Co.13 was awarded an extension to the Peace Shield defense contract in 1991, and this was further extended in 1997. The Hughes Aircraft company's commitment was carried out by General Motors (GM), which was pursuing an ambitious strategy to globally diversify automobile and parts production. Thus, in 1998, GM established a joint venture, The Middle Eastern Battery Co., to produce and recycle automotive batteries. Based at Damman, the enterprise began production with an annual capacity of around 500,000 batteries. US "economic balance" and the F-l 5 fighter offset program In 1992, Saudi Arabia purchased considerable numbers of Ml A2 Abram Main Battle Tanks from the US company General Dynamics (GD). As part of a reciprocal investment arrangement, GD agreed to participate in what was termed an "economic balance" program. No offset target was set, and GD agreed to make a "good faith" commitment to place work into the Saudi economy. As of March 2000, GD has implemented one project in the environmental sector and has awarded several other contracts for electrical assembly work to the Peace Shield company, AEC. A similar "good faith" agreement was entered into in 1993 by McDonnell Douglas, having signed a deal to supply 72 F-15 Eagle fighter aircraft to the Royal Saudi Air Force. The offset arrangement that was agreed between the US company and the Saudi Economic Offset Committee committed the former to implementing specific work
Saudi Arabia: defense offsets and development
207
packages with the five Peace Shield companies, and, in particular, AEC; the latter benefitting from substantial avionics and electronic sub-system contracts. Al Sawari In 1981, the French company Thomson CSF sold four naval vessels and two replenishment tankers to the Saudi Navy. Almost a decade later, in 1990, Thomson CSF established an offset program in anticipation of a further primary defense contract. This second major contract was achieved in 1994, when the Saudi government purchased two stealth La Fayette frigates (a further frigate was purchased in 1997). The offset arrangements were based on achieving 35 percent of the primary defense contract's technical value in any business sector, except oil. Three offset ventures have been initiated, a gold scrap reclamation plant, Dahab Co. Ltd., an oil catalyst venture, Al Bilad, and an electric meter manufacturing establishment. AT&T offset program The AT&T contract is Saudi Arabia's only offset program deriving from a major civil contract. The primary contract is the government's (August 1994) sixth Telephone Extension Project (TEP). The contract is worth a huge $6 billion, involving the supply to the Kingdom of a fully digital communications network. Under the offset agreement, Lucent Technologies (the US company which subsequently acquired AT&T) has committed itself to placing USS300-400 million worth of contracts in the Peace Shield company AEC for the production of printed circuit boards, assembly work, and systems integration. A joint venture company, International Network Engineering (INE), has also been formed in partnership with ISE to undertake focused software development for the Lucent Technologies contract. Development impact Surveying the impact of offset programs on the Saudi economy is not as
208 Arming the South
Table 8.2: Saudi Arabian offset companies Activities
Partners
1. Al Salam
Aircraft airframe overhaul & maintenance
Saudia and other local investors; Boeing
2. AEC
Design and development of civil-military electronic systems
Saudia and other local investors; Boeing
3. AACC
Aircraft accessories and components, overhaul and repair
Saudia and other local investors; Boeing
4. ISE
Design and integration of information systems
Saudi investment; Boeing
5. MEPC
Engine overhaul for civil-military aircraft (Riyadh)
Saudia and other Saudi investors; GE, P&W, RR
6. Middle East power
Assembly and testing of gas turbines (Dammam)
GE; Saudi investment
Production of car batteries
GM/Hughes; Saudi investment
8. Glaxo Saudi Arabia
Production of pharmaceuticals
Glaxo Wellcome; Saudi investment
9. United Sugar Refinery
Sugar refining
Tate & Lyie; Saudi investment
10. Dharam Harco Petrochemicals
Paint and adhesive products
Saudi investment; Harco (UK)
11. Saudi Development and Training
Education programs
Saudi investment; BAE Systems
Program/ Companies Peace Shield I
Peace Shield II 7. Middle East Batteries & Co Al Yamamah
Saudi Arabia: defense offsets and development
209
Table 8.2 (continued) 12. Rezayat Flover
Repair and maintenance of instrumentation and calibration (oil sector)
Rezayat Flover
13. Unilube
Recycling of used engine oil
Unilube (US); Saudi investment
14. Dahab
Gold refining
Thomson CSF; Saudi investment
15. AlBilad
Regeneration of catalysts for oil refineries
Eurcat; Saudi investment
16. Arabian Meter Co.
Production of electric meters
Schlumberger; Saudi investment
Specialized software for communications
ISE; Lucent
Al Sawari
Lucent (AT&T) 17. INE
Source: Data supplied by Saudi Economic Offset Committee, Riyadh, Saudi Arabia (March 2000).
straightforward as it might appear. At first sight, the 17 offset companies created (see table 8.2) through Saudi purchases of up to $100 billion of US, British, and French military and commercial telecommunications equipment, is a small development return for this mammoth defense and civil procurement.14 The American Peace Shield program achieved its 35 percent offset target and also the deadline, and five offset companies emerged from the placement work. Four big US companies (Boeing, Hughes, General Electric, and General Motors) are equity partners in these undertakings. Significantly, all the joint ventures were created primarily to service military requirements. With the passage of time, however, most if not all of these firms have diversified away from near-total dependence on defense work to a greater emphasis on civil commercial activity. The Lucent Technologies (formerly AT&T) contract has played an important role in this respect. AEC, in particular, has been a major beneficiary of the US$300-400 million telecommunications switching equipment
210 Arming the South
subcontracts placed in the Kingdom. Additionally, the firm has won commercial export orders on merit. These include telecommunications deals with Chile and Kuwait. Military work has not declined, but simply become a smaller proportion of an increasing civil-military portfolio of production activity. AEC has also been awarded major electronics subcontract work for the Lockheed Martin F-16 fighter. Additionally, via non-Peace Shield work-placements, linked to Saudi Arabia's defense purchases of General Dynamics Ml A2 Abram main battle tanks and McDonnell Douglas (Fl 5 fighters), there has been a regular flow of work to the new AEC facility, located at King Khaled Airport, Riyadh. The Middle Eastern Battery Company, sponsored by General Motors under Peace Shield II, has expanded production to 700,000 units per year, with 30 percent of this output destined for regional export markets (interview with SEOC secretary, H. Sugair, February 2000). The Middle Eastern Propulsion Company, moreover, has expanded its technical and market capacity through formal incorporation of Rolls Royce, alongside the existing US aerospace companies of General Electric and Pratt and Whitney. Similar progress has been made with Britain's Al Yamamah offset program. Two major joint-venture companies (Tate & Lyle, and Glaxo) commenced operations at the latter end of the 1990s, and two further offset projects (Rezayat Flover and Unilube) were implemented at the turn of the millennium. All the British-sponsored offset companies are located in the commercial sector, and employment of local Saudi labor has been emphasized. The United (Tate & Lyle) sugar refinery and the Glaxo- Wellcome pharmaceutical plant have raised the numbers of Saudi nationals employed in Britain's joint venture companies to between 400 and 500 (British offset officials, Riyadh, February 2000). This represents a sizeable increase in local employment over the early 1990s aggregate figure (approximately 300 Saudi workers) for all the Kingdom's offset joint ventures (Matthews, 1996, p. 248; original source, British offset officials, London, October 1995). In 1995 the value of Al Yamamah offset credits had reached over £300 million, or around 35 percent of the £1 billion target (Matthews, 1996, p. 248). In 1998, this figure had reportedly risen to £450-500 million (Hansard, 21 April 1998, col. 557558). The latest estimate suggests that by 2002, the Al Yamamah offset
Saudi Arabia: defense offsets and development
211
target will have been achieved (British offset officials, London, March 2000). Indeed, with a further four offset ventures due to be implemented shortly, the probability is that the £1 billion offset value will be exceeded. Whilst this milestone will have been delayed by four years, the anticipated achievement of the offset target nevertheless represents an impressive feat. This is particularly the case given that this program attracted considerable Saudi criticism (as recent as the late 1990s) over its slow progress (Economic Bureau, 1998, p. 100). Promotional efforts by the British offsets office will likely continue beyond 2002. Negotiations to encourage investment into the Kingdom have been taking place between British offset officials and representatives from Japan, Australia, and other countries, and are ongoing. The UK is committed to securing offset credits, both for current obligations and to bank for possible future offset programs. The policy thrust initiated by the French of pursuing offsets in anticipation of a primary defense contract now concentrates the minds of all the Kingdom's principal arms supplying nations. Notwithstanding France's well-known proactiveness and professionalism in the offsets field, the French Al Sawari effort looks sickly. Only three small offset ventures have commenced, and one of these is struggling. In addition, the Dahab gold scrap recovery plant has suffered recently from declines in the international gold price. Employment of Saudi labor in the three joint ventures is well below 100 workers (French offset officials, Riyadh, February 2000). Finally, there is the Lucent Technologies (AT&T) telecommunications contract, which has led directly to the creation of a small (Saudi Riyals 17 million) joint venture company, INE. Ownership is shared between Lucent and the US Peace Shield offset company, ISE. INE is a dedicated software facility focused on the telecommunications field. The creation of INE, along with the substantial switching subcontract work that Lucent Technologies is providing AEC, arose because of Saudi Arabia's market muscle in high value civil overseas procurements. The Saudi government recognizes its leverage in this regard and this is reflected in the 1997 Council of Ministers resolution, requiring that all future major offset development programs, both military and civil, must accommodate the Kingdom's Saudi-ization and high
212 Arming the South
technology ambitions. The Council formally designated that strategic sectors of the economy would have offset requirements tied to major overseas purchases. These strategic sectors are listed as: telecommunications; electronics; aerospace; defense; medical fields; transportation; power generation; desalination; and downstream petrochemicals. Whilst no offset percentage target threshold would be imposed, a case-by-case evaluation will be undertaken to determine the nature of the offset agreement. Policy implications Offset programs should represent a win-win relationship between military and civil equipment vendors and the purchasing country. The focus of this chapter has been on Saudi Arabia as a purchaser of major civilmilitary equipment, and the consequent impact of tied offset programs on Saudi development. Saudi Arabia measures the value of offset arrangements in terms of the long-term partnerships that need to be constructed. It is a perspective stemming from the recognition that offsets have the potential to act as focused drivers of indigenous industrial and technological development. Saudi offset programs differ from the norm in that they primarily emphasize a "best endeavors" approach, offer only limited multipliers,15 and refrain from imposing a penalty for failure to achieve the offset requirement by the agreed deadline.J 6 Additionally, Saudi offset programs do not follow the conventional pattern since they reflect neither the dictat of government nor provide business with the prospect of guaranteed government purchases.17 Rather, the thrust of the Kingdom's offset policy is to foster economic diversification by nurturing commercial jointventure projects in both the civil and defense sectors. There is a belief that over the long-term this policy will promote industrial linkages, leading to further investment outside the formal offset structure. There is no doubting the size and success of Saudi Arabia's broadbased joint-venture strategy, which by March 1999 had supported the creation of 1,581 enterprises in the Kingdom at an investment value of over $39 billion (Arab Daily, 21 February 2000). In the more narrowly focused area of joint venture offset companies, the value of committed
Saudi Arabia: defense offsets and development
213
investment is also impressive, amounting in 1997 to over $4 billion for US, British, and French companies (Saudi Offset Limited Partnership, n/d, p. 3). Enhanced Saudi government emphasis on compliance with offset commitments is accelerating the pace of joint venture development, with greater flexibility, innovation, and commercialization characterizing the offset process (Saudi Offset Limited Partnership, n/d, p. 5). Implicit in Saudi Arabia's joint venture offset company strategy is the priority given to Saudi-ization. The creation of a highly skilled Saudi workforce is a constant theme underlying all Saudi development initiatives. Accordingly, twice the value of the costs incurred in training Saudi nationals for management or engineering positions in joint-venture companies can be counted toward the offset target. The Saudi authorities' position on the employment of Saudi nationals in joint-venture companies has hardened. In the early days of industrialization, no compulsion was placed on joint-venture companies to achieve Saudi employment targets. Today, however, a target of five percent local labor is rigidly enforced.18 At the commencement of the Kingdom's first offset program, the 1985 Peace Shield I, the Saudi authorities planned that offsets projects would generate 75,000 Saudi jobs (Arab News, 23 March 1992). Reality, however, has fallen well short of this expectation. Currently, the total employment generated by the various offset and economic balance programs is estimated to be around 2,000.19 Whilst this employment generation remains small, it is a healthy increase on the 1,400 employment figure achieved in the early 1990s (Matthews, 1996, p. 248). These figures provide no indication, however, as to the quantity and quality of the Saudi jobs created. Local Saudi employment in offset companies is estimated to exceed 1,000 workers, or more than treble the 300 worker level achieved in 1994, and there is a sense that they are no longer predominantly unskilled in nature (p. 248). Certainly, this is the case at AEC. The company's trend of performance is impressive: in 1991, Saudi workers accounted for only 25 percent of the workforce of 80 people, and they were mostly employed as clerical staff (p. 241); by 1994, Saudis accounted for 53 percent of the workforce, which had expanded to 276 people, mostly employed as technical staff (p. 241); finally, by February 2000, the workforce had grown still further to 376 employees, of which 71 percent are Saudi, and 86 percent are located in technical and
214 Arming the South
engineering areas (Arab News, 14 February 2000 and AEC Information Pack, February 2000). AEC's employment strategy emphasizes Saudiization through labor growth and company expansion rather than through the replacement of expatriate workers. The increasing proportion of AEC's Saudi workers possessing higher technology skill levels is likely to be reflective of the rising skill levels across the broad array of offset ventures, but this is a generalization rather than an empirically established fact. For Saudi Arabia, as with other industrializing countries, offsets reflect market leverage and an opportunity to stimulate economic development through the promotion of inward technology transfer. Development, diversification, and Saudi-ization are the principal development goals of the Saudi government. These are rightly seen as critical issues in a country highly dependent on oil for income generation and foreign resources for technological development. Through joint foreign and local ownership of offset companies, there is an expectation that in the long-run the dynamic technological benefits of local innovation, promotion of indigenous specialist sub-contractors, and the possibility of wealth and job-creation through domestic and foreign market expansion, can be maximized. This will require that consideration be given to the implementation of an appropriate Saudi science and technology strategy, highlighting the development of local universityindustry-government relationships, promotion of indigenous research and development expenditure, institutional support for small and medium-size specialist enterprises (servicing the high value-added inputs of the major joint-venture manufacturing plants transplanted into the Kingdom), and finally the fostering of appropriate marketable skills through university and vocational industrial education and training to ensure that the technologies transferred can be effectively absorbed and disseminated throughout the Kingdom. Conclusion Defense and civil offset programs will benefit from Saudi industrial development initiatives, but offsets can also make a contribution to the practical implementation of these initiatives. Defense offsets and
Saudi Arabia: defense offsets and development
215
development can thus be a two-way street. For offsets to work, however, technology transfer is not enough, there must also be a high degree of local technological absorptive capacity. Hence the need for a governmentsponsored, civil-military, science and technology strategy. Notes I acknowledge financial support from the Department of Defence Management and Security Analysis, Cranfield University, to conduct empirical research in Saudi Arabia. 1. Defense expenditure declined from US$ 1,086 bn in 1987 to US$ 704 bn in 1997 (calculated in constant prices). See SIPRI Yearbook 1998, p. 214. 2. There is a growing change of attitude reflecting defense, global market characteristics, and, in particular, a recognition that offset partnerships or "package" deals should be entered into for defense contracts. GEC, for instance, views offset partnerships as "opportunities," while offsets in the past were seen as a problem. The UK Defence Export Services Organisation (DESO) argues that such package deals, which include offsets and technology transfers, are one of the major planks of strategic planning in the approach to today's market. See P. Pohling-Brown (1997). However, also note the dangers; see Scott, 1996. 3. GKN formally fulfilled its offset commitment of 15 percent offset and 100 percent counterpurchase in coconut oil, copper cathodes, garments, cordless telephones, fresh frozen shrimps, and shipping services (Villanon, 1998, p. 4.11). 4. Arguably, two of the more successful countries in regard to effective absorption of technology transfer are Singapore and Japan, and neither have formal published offset guidelines. By contrast, Kuwait and UAE publish detailed offset guidelines, but the offset performance for these countries is questionable. See Matthews (1999) and Matthews and Chinworth(1996).
216 Arming the South
5. Saudi Arabia is ranked as the world's leading recipient of major conventional weapons across 1993-97. See SIPRI Yearbook 1998 (1998, table 8.2, p. 300). 6. Note, however, that barter deals with Russia are not uncommon. Around 70 percent of Rank Xerox's contracts with Russia and the CIS come in the form of barter payments, covering animal feed, electric motors and even camel saddles. Similarly, Pepsi Cola's $3 billion contract to supply Pepsi to Russia initially involved no cash payment at all. Instead, the US company received tankers and freighters as well as huge consignments of Stolichnaya vodka. See Matthews, 1992. 7. The early 1990s sale of BAe Hawk aircraft to Malaysia included a 4050 percent counterpurchase obligation. One counterpurchase deal was linked to the sale of GEC-Marconi radars. The deal required the purchase of tea and palm oil, and possibly local manufactures, to the value of the $350 million target. See Matthews, 1991. 8. BAe (now BAE Systems) was appointed as the prime contractor in the Al Yamamah offset agreement, and thus has the responsibility of encouraging inward investment into Saudi Arabia. See Cooper (1997, chapter 6) for discussion on the impact of the Al Yamamah program on UK export performance. 9. In those cases where the countertrade deal is negotiated at the same time as the primary defense contract there is always the possibility that the vendor company will load into the price of the defense equipment the costs of servicing the counterpurchase and offset arrangements. This view is supported by the results of recent research. A questionnaire survey supplied to all 49 members (11 firms provided data) of the British Defence Manufacturers Offset Group (BDMOG) in 1993 discovered that the cost of offsets varies between 0 to 60 percent. However, the percentage of cost premium added to the contract price became 100 percent; that is, the contract price with offsets is double the price without offsets. See Martin and Hartley (1996, p. 353).
Saudi Arabia: defense offsets and development
217
10. For example, Westland Helicopters outsources around 90 percent of the value of output, and thus the achievement of, say, 100 percent offsets through its own resources would be impossible. The only avenue out of this impasse is for Westland to cascade down some of its offset obligations to its principal suppliers. 11. The requirement to ban offsets tied to commercial contracts had earlier been institutionalized in the 1977 General Agreements on Tariffs and Trade (GATT). 12. BAE Systems benefits from this offset credit, but cannot claim MEPC as a British offset company as it is already characterized as such under the Peace Shield offset program. 13. Hughes took over from Boeing as the prime contractor for the Peace Shield installation and associated offset program in 1991. Hughes then became responsible for the additional offsets under Peace Shield II. 14. This is an informed estimate. In 1994 Saudi Arabia's arms deals since the Gulf War were valued at $80 billion. Aggregating this with the French and substantial US arms sales since would bring the total value of sales close to the $100 billion mark. See Matthews, 1996, p. 233. See also the figures contained in Hirst (2000). 15. Investment into particular, pre-determined fields of activity, e.g., training of local labor, may attract higher multiplied credits in some countries' offset guidelines. 16. The UAE, for instance, imposes a penalty of 8.5 percent of the difference between the agreed offset target value against a pre-determined time schedule and the offset value actually achieved. For a review of the performance of the UAE offsets program, see Matthews, 1998, p. 12. 17. This was the original intention of the SEOC, but in practice the offset companies would probably not have survived without Saudi government sole-sourced contracts.
218 A rm ing the Sou th
18. Discussions with SEOC officials and A. Al-Ankari, Head-Marketing, S ABIC, Riyadh (February, 2000). This requirement was introduced in the late 1990s, and applies to companies employing more than 20 workers. 19. Discussion with SEOC officials (February 2000). The Saudi Offset homepage estimated employment at offset companies to be "more than 1500" in 1996 (http://www.saudioffset.com/saudihyp.htm, p. 2). The additional Al Yamamah and Al Sawari offset investment companies since that time must certainly have pushed employment to at least 2000. References Al-Ghrair, A andN. Hooper. "Saudi Arabia and Offsets "pp. 219-244 in S. Martin (ed.). The Economics of Offsets: Defence Procurement and Countertrade. Amsterdam: Harwood Academic Publishers, 1996. Arab Daily, London, 23 March 1992,14 February 2000, and 21 February 2000. Cooper, N. The Business of Death - Britain's Arms Trade at Home and Abroad. London: Tauris Academic Studies, 1997. Economic Bureau. "The Saudi Economic Offset Program." Journal of the Japanese Institute of Middle Eastern Economies Vol. 11, No.39 (Spring 1998), pp. 49-102. Hansard, HM Government, 21 April 1998. Hirst, C. The Arabian Connection - the UK Arms Trade to Saudi Arabia. London: Campaign Against the Arms Trade (CAAT), May 2000. Martin, S. and K. Hartley. "The UK Experience with Offsets," pp. 337355 in S. Martin (ed.). The Economics of Offsets: Defence Procurement and Countertrade. Amsterdam: Harwood Academic Publishers, 1996. Matthews, R. "Singapore Buys Longbows and its Defence Industry Grows." Asia Pacific Defence Reporter (December 1999), pp. 20-21. Matthews, R. "Offsets to Grow 9 " Pointer (Jane's Information Group, December 1998), p. 12. Matthews, R. "Saudi Arabia's Defence Offset Programmes: Progress, Policy and Performance." Defence and Peace Economics Vol. 7 (1996), pp. 233-251.
Saudi Arabia: defense offsets and development
219
Matthews, R. "Butter For Guns: the Growth of Under-the-Countertrade." The World Today (May 1992), pp. 87-92. Matthews, R. "Offset to Decline For Britain's Defence Industrial Base?" RUSI Journal (Winter 1991), pp. 58-63. Matthews, R. and M. W. Chinworth. "Defence Industrialisation Through Offsets: The Case of Japan," pp. 177-218 in S. Martin (ed.). The Economics of Offsets: Defence Procurement and Countertrade. Amsterdam: Harwood Academic Publishers, 1996. Matthews, R. and R. Williams. "Technology Transfer: Examining Britain's Defence Industrial Participation Policy." R USI Journal Vol. 145, No.2 (April 2000), pp. 26-31. Pohling-Brown, P. "Package Deals Required To Develop Emerging Markets." Jane fs Defence Contracts (October 1997). Scott, Rt. Hon., Sir Richard. Report of the Enquiry into the Export of Defence Equipment and Dual Use Goods to Iraq. House of Commons Paper (Session 1995/6) HC 115, HMSO, London (1996). Saudi Offset Limited Partnership. Saudi Economic Offset Program. (Not dated.) SIPRI. Yearbook 1998. Stockholm: Oxford University Press, 1998. US Department of Commerce. Offsets in the Defence Trade: Third Annual Report To Congress. Washington, DC: US Department of Commerce (August 1998). Villanon, A. Philippine Defence Industrial Development and Offsets. Unpublished MDA thesis. Cranfield University (July 1998).
This page intentionally left blank
9 South Africa: An Econometric Analysis of Military Spending and Economic Growth Alvin Birdi and J. Paul Dunne
Introduction There is considerable debate in the literature over the effects of military spending on economic growth in developing economies. Following the early cross country studies, using average values over time and simple correlation techniques, the application of econometric models provided a wide variety of studies, but no clear consensus over the results. What started to become clear was that to understand the dynamics of the relation between military spending and growth it was necessary to focus on relatively homogeneous groups of countries as well as undertaking case studies of individual countries (Dunne, 1996). This chapter provides a contribution to the corpus of case studies by providing an analysis of South Africa, a particularly interesting focus of study because of the nature of its military industrial complex, the characteristics of the economy, and the fact that it has undergone considerable change. In addition, the country has relatively high quality data for a developing economy. While there have been some studies of the relation between military expenditure and economic growth in South Africa, noticeably a recent
222 Arming the South
special issue of the journal Defence and Peace Economics [Vol. 11, No. 6 (2000)], they are by no means exhaustive of the possibilities. This chapter provides a critical review of the approaches taken in econometric analyses of the defense-growth nexus in South Africa and then develops the analysis to overcome perceived deficiencies. New results are then provided using a cointegrating vector autoregressive (VAR) approach. Historical background Military spending has played an important role in South Africa. During the apartheid period high military burden supported the system internally and externally as part of a military-energy industrial complex (Fine and Rustomjee, 1996). In anticipation of the 1977 UN arms embargo a state arms producer, Armscor, was formed to develop domestic arms production capability and in 1982 entered the arms export market, having a policy of selling to all comers. This represented a huge change as prior to 1963 South Africa spent 70 percent of its defense budget overseas whereas by 1984 almost 100 percent was procured domestically. With the end of the apartheid system and of the cold war there have been significant cuts in military spending. The decline started in 1989-90 with the start of the abolition of apartheid, the establishment of civilian control of the military, and the termination of South Africa's nuclear weapons program. But the 1989-93 period also showed one of the worst economic recessions in South African history. Still, even though manufacturing declined, the country's external position improved with the end of economic sanctions that had been imposed on South Africa during the apartheid years. Also, important strategic industries, such as Sasol, Armscor, and Mosgas, were commercialized and privatized. With the new ANC government came not only the so-called Cameron Commission report into the country's arms exports practices, but also the voices of a number of "hawks," the smell of political compromise, and a defensive fight by vested interests in the military industrial complex. This has seen pressures for increases in military spending. The present situation sees an improving economic situation, considerable debate over the level of military spending, and an end to its decline. A recent large procurement order with offsets has highlighted the debates (Batchelor and
South Africa: an econometric analysis
223
Dunne, 1998; 2000). One of the arguments in the policy debates concerned the possible negative impact of military spending cuts on economic growth. This has made research on the role and impact of such expenditures an important target for research. Before moving on to review existing applied studies of South Africa, the next section considers the more general theoretical and empirical work. Economic effects of military spending Any evaluation of the impact of military spending on growth is contingent on the theoretical perspective used. Neoclassical models are generally supply side models and focus on the tradeoff between "guns and butter." Keynesian models see military spending simply as one component of aggregate national spending and therefore focus on the economy's demand side (although, when formulated as an aggregate production function, the econometric work does give the Keynesian models neoclassical flavor). A group of institutional economists focuses on the damaging impact of the military industrial complex on the economy, and Marxist views varyfromthose stressing positive economic effects, by preventing underconsumption or realization crises, to those highlighting the possible negative impact of military spending on the profit rate (Dunne, 1990). When we move to empirical analyses it is necessary to determine the level of abstraction at which the analysis is to be presented and to operationalize the theory to form an applied model. This leads to a variety of empirical work from applied econometric to more focused institutional case study approaches. When statistical analysis is used it is generally within the neoclassical and Keynesian approaches, as these are most amenable to the creation of formal models. Some work adopts an ad hoc approach. Studies differ in terms of the country coverage, whether time series or cross section are used, the time period covered, and the empirical methods used (Dunne, 1996). In general, the empirical analyses have identified a number of channels by which military spending can influence the economy and they can be positive or negative. On the one hand, military spending can take skilled labor away from civilian
224 Arming the South
production, but on the other hand can train workers, particularly in developing economies where the military may provide valuable skills. It can take the best capital equipment from civilian industry to produce a high technology enclave, but there may be positive externalities of the development of the military sector on the civilian sector. It can lead to damaging wars, but may maintain peace and lead to economic benefits from more prosperous allies. It can stimulate demand in a stagnant economy and lead to growth, but may create bottlenecks in a constrained economy. Finally, it may slow down development through the fostering of a militaristic ideology, but nationalist attitudes may increase effort and output and the military and ideology may be used to control the workforce. Clearly whether the overall effect is positive or negative is an empirical question and is likely to differ across countries. Following the ad hoc approach of Benoit's original study (Benoit, 1978), which found a positive effect of military spending on growth in developing countries, an impressive literature has been built up. This has used econometric analysis of single equation reduced form equations and simultaneous equation models that model both direct and indirect effects (Dunne, 1996). In addition, macroeconometric models have been used to simulate the likely impact of changes in military spending at the country and international level (Gleditsch et al., 1996). Overall, the results tend to show an insignificant or negative impact of military spending on economic growth in developing countries and a clearer negative impact in developed economies, as military spending siphons off investment rather than consumption spending. This does, however, hide the diversity of the literature. Much of the earlier crosssection analyses found sample selection to be important and this led to calls for more case studies. Time series analyses of individual economies and relatively homogenous groups of economies have improved the understanding of the economic processes at work but have also produced a variety of results. What is clear is that while cross-country studies can provide valuable general information it is important to build up more comprehensive case study analyses. The next section considers the analyses that have been undertaken for South Africa.
South Africa: an econometric analysis
225
Analyzing military expenditure and growth in South Africa Econometric analyses of the relation between military spending and growth in South Africa have followed the different approaches in the literature. There are the neoclassical growth models, with the Feder-Ram model being estimated by McMillan (1996) and by Batchelor, Dunne, and Saal (2000). A Keynesian simultaneous equation model with an aggregate production function has been estimated by Roux (1996) and by Dunne, Nikolaidou, and Roux (2000). Finally, Dunne and Vougas (1999) analyzed the relation between military spending and growth using Granger causality and cointegration techniques. There have also been attempts to use industrial level panel data, e.g., by Birdi, Dunne, and Saal (2000). The simple Feder-Ram model has held something of a fascination for defense economists, mainly because of its ability to explicitly treat externality effects of the military on the non-military sector. For the basic model, assume two distinct sectors, military (M) and non-military (C), that labor L and capital K are the divisible inputs, and that the military sector has an externality effect on the rest of the economy. (9.1)
M = M(LM,K
(9.2)
C = C(LC,KC,M)
M
)
with (9.3) (9.4) (9.5)
Q=M+C K = KM + K C L = LM + L C
Allow input productivities to differ such that the ratios of the marginal productivities for the sectors are: (9.6)
M'K/C'K = M'L/C'L=l+5
Military spending can then have a productivity differential effect, 8, and an externality effect, 5C / 5M > 0. Reformulating in terms of aggregate
226 Arming the South
inputs, taking the total derivative of Q, and then substituting and manipulating gives: (9.7)
YD = p LD + a (I / Y) + ((5 / 1 + 5) - CM ) MD (M / Y)
where YD = dY/Y; LD = dL/L ; MD = dM/M. The coefficient on the last term is the sum of the externality and factor productivity differential effects of military spending. Following Biswas and Ram (1986), and assuming that the externality parameter is not CM but CM (M/C) and is denoted 9, allows us to write: (9.8)
YD = (3 LD + a (I / Y) + ((5 / 1 + 5) - 9) MD (M / Y) + 9 MD
Separate estimates of 9 and 5 can be obtained. To operationalize the model for empirical application the instantaneous rates of change of the variables are replaced by their discrete equivalents giving: (9.9)
AY/Y^ - a0 + a, AL/L^ + a 2 1/Y M + a3 AM/M,., (M/Yt.,) + a4 AM/Mt_,
Initially, these models were used on cross sections, but increasingly have been applied to time series for individual countries. When this model was estimated using South African data for the period 1964-1995 in Batchelor, Dunne, and Saal (2000) it gave the results displayed in table 9.1. These results suggest that economic growth in South Africa can be only partly explained by the model, although the specification seems to be reasonable according to the different statistical tests. The R2 suggests that the equation only explains 59 percent of the variation in the dependent variable, which in a time series regression is relatively poor. In addition, only the employment variable is statistically significant at the 5 percent significance level. This variable is the growth in nonagricultural labor which is used to approximate the labor force. Surprisingly, the investment term is insignificant. The military spending coefficient estimates suggest a positive externality effect, but a negative size effect, but these are only significant at 10 percent.
South Africa: an econometric analysis
227
Table 9.1: Aggregate estimation results, South Africa Coefficient
Variable Constant I/Y„ ALt/L„ AMt/M,, (Mt/Yt.,) AM/M,,
0.03 -0.08 0.78 -1.58 0.07
t-ratio 1.4 0.8 4.4 -0.4 1.1
Using AK/K^ instead of I/Yt. t-ratio Coefficient 0.01 0.02 0.72 -1.24 0.05
1.9 0.1 3.4 -0.3 0.4
0 = 0.07 and 5 = --0.60
0 = 0.05 and 5 = •-0.54
R2 = 0.587; DW == 1.68 Serial corre lation -0.69 Functional Form = 3.86 Normality == 0.11 Heteroskedasticity = 0.44
R2 = 0.52; DW = 1.36 Serial correlation = 2.76 Functional Form = 1.81 Normality = 0.05 Heteroskedasticity = 0.23
Joint LLR test for zero restrictions on military expenditure terms chi-squared 2: 0.48 (0.79)
3.65 (0.86)
All chi-squared with 1 degree of freedom except normality test which is 2.
There are of course problems of multicollinearity with an equation of this form particularly between the two military spending terms. This will mean that although the estimates are unbiased they are imprecise and unstable. One would expect a high F statistic, but low individual significance, which is what we observe. However, the joint test of zero restrictions on the military variables' coefficients cannot be rejected, suggesting that there is no significant impact of military spending on growth. The problems with these results led to considerations of how they might be improved. There are a number of options that have been taken by researchers. They have used a more detailed model that identifies more sectors. One study of particular interest is McMillan (1992) who estimates a variation of the model for South Africa for 1950-1985. An extended model developed along these lines, in chapter 13 of this volume, finds improved result for Greece. Another alternative is to consider the
228 Arming the South
impact of military expenditure on the manufacturing sector alone, rather than the whole economy. When Batchelor, Dunne, and Saal (2000) also did this they found that the results for the manufacturing sector in South Africa were much better than for the aggregate data. Finally, the dynamics of the applied model can be considered. The move to a discrete model in the theory is somewhat ad hoc and it is possible that the processes may have a longer memory than allowed for in the derived model Batchelor, Dunne, and Saal (2000) take the simple model for South Africa and use an ARDL estimation procedure to model the shortrun dynamics. This does improve the performance of the model, but the strange nature of the composite variables gives some concern. This leads us to suggest that it is necessary to further investigate the dynamics of the relation between military spending and growth and that it is also worthwhile considering an analysis at the level of manufacturing as well as the aggregate economy. Opting for an approach that uses a structural model, we have seen the problems with using the Feder-Ram model, while the nature of the commonly used Keynesian model limits the scope for developing the dynamic specification. This suggests the need to consider a different approach and to use a different model. An obvious alternative is to use an aggregate production function model and this is developed in the next section. Developing the analysis Taking the simple Cobb-Douglas model (9.10) Q = AK a L p , which in log form is (9.11) q = a + a k + pl, we can simply introduce military spending to this equation, (9.12) q - a + a k + p l + ym
South Africa: an econometric analysis
229
where m is the log of military expenditure. In addition we add in a dummy variable to take account of the effect of sanctions which basic data analysis suggests was an important factor. To develop the dynamics of the model we use a cointegrating vector autoregressive (VAR) approach to estimate it. This allows us to focus explicitly on determining the short-run dynamics and long-run properties of our structural model. Treating this within a VAR estimation framework within Microfit 4.0 (Pesaran and Pesaran, 1997) and starting from an order 4 VAR we get a VAR (2) as the optimal lag length. Using unrestricted intercepts and no trends gives one cointegrating vector, (9.13) z = 0.82 q 4- 1.86 k -7.05 1 + 1.23 m or (9.14) q =-2.27 k + 8.63 1-1.51 m (13.9) (41.3) (7.4) where the asymptotic standard errors are in brackets. The coefficient on military spending is negative, but the results are strange, with all coefficients insignificant and a negative sign on capital. Its persistence profile converges quickly, within 9 years, but the underlying error correction model is a very poor specification for a growth equation, though again there is a negative but insignificant effect of m. (9.15) Aq t = 2.3 + 0.22 Aqt_, + 0.11 Ak M - 0.10 A l t l - 0.03 A m t l (1.6) (0.9) (0.4) (0.3) (1.0) - 0.03 ECM t j - 0.02 DS (1.6) (2.2) where ECM is the error correction term, DS the sanctions dummy, and the values of the t ratios are in brackets. The rather large coefficient on the labor term is worrying and may suggest that we have the wrong dependent variable and may be estimating a labor demand equation. Rather than simply move to another form of
230 Arming the South
model we follow Batchelor, Dunne, and Saal (2000) and estimate the model on the manufacturing data. The order of the VAR is found to be 2 and unrestricted intercepts and no trends gives one cointegrating vector, (9.16) q m = 1.32 k - 1 . 5 3 1 +0.50 m (0.7) (2.1) (0.5) This gives a more sensible specification with the capital coefficient positive and significant at 10 percent. Military spending is now positive but insignificant. The underlying ECM model is (9.17) Aqm t = 1.96 + 0.55 Aqm t x + 1.23 Ak t x - 0.84 Al t , - 0.08 Am t x (1.7) (3.6) (2.0) (1.6) (1.3) + 0.16 ECM t ,-0.04 0 8 (1.6) (2.3) In this case military spending has a negative short-run effect on growth. Given the similarity of the normalized values of the coefficient on k and 1 a test on the cointegrating vector of a = - p in the original CobbDouglas equation seemed worthwhile. This restriction makes output a function of the capital-labor ratio and military expenditure and was accepted with %2 (2) = 0.03 giving (9.18) q m = 1 . 2 3 ( k - l ) + 0.44 m (0.19) (0.09) with asymptotic standard errors in brackets. In fact the coefficient on the capital-labor ratio is not significantly different to one, % 2 (2) = 0.98, giving (9.19) qm = ( k - l ) + 0.51m (0.08) and an error correction equation for growth,
South Africa: an econometric analysis
231
(9.20) Aqm t = 1.31 + 0.55 Aqmt ^ 1 . 0 2 AkM - 0.89 A1M - 0.07 AmM (1.7) (2.6) (1.9) (1.7) (1.3) + 0.15 ECM M - 0.05 DS (1.7) (2.3) These results show a positive long-run relation between military spending and manufacturing output, but a negative short-run effect (significant at 10 percent) of the growth of military spending on the growth of manufacturing output. The composite effect of the short-run coefficient on military spending and the error correction term suggests that the shortrun impact of cuts in military expenditure will at worst not be significant. In this they tend to support the results of the previous studies. Conclusions South Africa clearly provides an interesting case study for the analysis of the impact of military spending on growth. This paper has provided some new results and has tried to deal with some of the perceived deficiencies of previous work. Concerns about the Feder-Ram model, in terms of its specification and dynamics, led to the use of an aggregate production function, estimated using cointegrating VAR methods. Concerns with the problem of undertaking the analysis at an aggregate level in a country which, while a developing economy, has an advanced military industrial sector, led to a focus on manufacturing output. When the model was estimated using GDP, the results were rather disappointing but suggested a negative though insignificant effect of military expenditure on growth. When estimated at the level of manufacturing the results show a positive long-run relation between military spending and growth, but a negative short-run effect (significant at 10 percent). The composite effect of the short-run coefficient on military spending and the error correction term suggests that the short-run impact of cuts in military expenditure will at worst not be significant. As we have seen, most of the empirical evidence tends to suggest a negative or insignificant effect of military spending on growth in South Africa. The results of this study, while providing some advance in the
232 Arming the South
econometric analysis, do appear to be consistent with the previous findings. Note This paper is based on research undertaken as part of a project on Defence Industrial Restructuring, Conversion and Economic Growth in South Africa, funded by the Leverhulme Trust whose support is gratefully acknowledged. It was presented at the 1999 Econometrics for Africa Conference in Johannesburg, and we are grateful to the participants for comments. References Batchelor, Peter and Paul Dunne. "Industrial Participation, Investment and Growth: The Case of South Africa's Defence Related Industry." Development Southern Africa Vol. 17, No. 3 (September 2000), pp. 417-435. Batchelor, Peter and Paul Dunne. "The Restructuring of South Africa's Defence Industry." African Security Review Vol. 7, No. 6 (1998). Batchelor, Peter, Paul Dunne, and David Saal. "Military Spending and Economic Growth in South Africa." Defence and Peace Economics Vol. 11, No. 6 (2000), pp. 553-571. Benoit, E. "Growth and Defence in LDCs." Economic Development and Cultural Change Vol. 26, No. 2 (1978), pp. 271-280. Birdi, Alvin, Paul Dunne, and David Saal. "The Impact of Arms Production on the South African Manufacturing Industry." Defence and Peace Economics Vol. 11, No. 6 (2000), pp. 597-613. Biswas, B. and R. Ram. "Military Expenditures and Economic Growth in Less Developed Countries: An Augmented Model and Further Evidence." Economic Development and Cultural Change Vol. 34, No. 2 (1986), pp. 361-372. Dunne, P. "Economic Effects of Military Expenditure in Developing Countries: A Survey," pp. 439-464 in N.P. Gleditsch et al (eds.), The Peace Dividend. Amsterdam: Elsevier, 1996. Dunne, Paul. "The Political Economy of Military Expenditure: An
South Africa: an econometric analysis
233
Introduction." Cambridge Journal of Economics Vol. 14, No. 4 (December 1990), pp. 395-404. Dunne, Paul, Efi Nikolaidou, and Andre Roux. "Military Spending and Economic Growth in South Africa: A Supply and Demand Model." Defence and Peace Economics Vol. 11, No. 6 (2000), pp 573-585. Dunne, Paul and Dimitrios Vougas. "Military Spending and Economic Growth in South Africa: A Causal Analysis." Journal of Conflict Resolution Vol. 43, No. 4 (August 1999), pp 521-537. Fine, B. and Z. Rustomjee. The Political Economy of South Africa: From Minerals-Energy Complex to Industrialisation. London: Hurst and Company, 1996. Gleditsch, N.P. et al. (eds.) The Peace Dividend. Amsterdam: Elsevier, 1996. McMillan, S. "Economic Growth and Military Spending in South Africa." International Interactions Vol. 18, No.l (1992), pp. 35-50. Pesaran, M and B. Pesaran. Microfit 4.0. Oxford: Oxford University Press, 1997. Roux, A. "Defence Expenditures and Economic Growth in South Africa."Journalfor Studies in Economics and Econometrics Vol. 20, No. 1 (1996), pp. 19-34.
This page intentionally left blank
10 Ludwig Erhard in Africa: War Finance and Post-War Reconstruction in Germany and Mozambique Tilman Briick
Introduction The German economist Ludwig Erhard wrote a secret working paper on German war finance and war debt in 1943-44 to prepare for the eventual economic reconstruction of a defeated Germany (Erhard, 1977). Many of the ideas expressed in his paper formed the basis for post-war economic policies which he helped to shape as finance minister and, later, as chancellor of West Germany, and which contributed to the Wirtschaftswunder. Erhard's analysis recognized the key role of war finance for shaping post-war fiscal constraints and people's expectations of future growth. Without a careful position on war debt and the currency, the post-war government was going to risk its liquidity and credibility in the market thereby undermining confidence in itself and in the economy as a whole. Erhard's policy analysis of 1943-44 will be assessed critically and
236 A rm ing th e So u th
then applied to the case of Mozambique's internal war of 1979-92 to understand the current role of war finance for post-war reconstruction in a poor developing economy. I find that Mozambique suffered from a disproportionately large war debt while receiving high levels of aid both during and after the war. But donors failed to use their influence to end the war early and post-war foreign debt cancellation was too limited and too slow to create a resumption of exports and tax revenue. In addition, the Mozambican government could have done more to prepare for the end of the war, especially by following a civil investment program designed to maximize confidence and trust in public institutions, to reduce transaction costs, and to distribute the benefits of growth equitably. War finance in Germany Military expenditure may have positive economic effects by increasing the security of the country and thereby reducing uncertainty and transaction costs (Berthelemy et al., 1994; Brauer and Chatterji, 1993; Gleditsch et al, 1996; Hartley, 1997; Klein et al, 1995; Lamb, 1992; Mohammed, 1999). Yet there are decreasing returns to military expenditure as, beyond a certain point, military expenditure may signal the intention to declare war in the future thus starting a regional arms race. In addition, secret military budgets are more likely to lead to the start of armed conflict and to longer conflicts than could be expected with budgets controlled by a democratic process. In Nazi Germany, the military budget remained concealed and there was no democratic debate about its purpose or about alternative ways of funding or conducting the war. Secret financial preparations for the war had been made since at least 1933 (Schmolders, 1977). The increasing and unconstrained expenditure for the German army and military infrastructure were partly made at the expense of other state spending, which in the 1930s accounted for over a third of GDP, and partly through the hidden issue of debt. By 1938, 60 percent of the state budget was funded through taxes and 40 percent through debt (Schmolders, 1977). With the start of the war, debt became the key source of state spending. The high war debt was an indicator, for Erhard, of the intensity of the
Ludwig Erhard in Africa
237
German war effort. However, the debt also served to reduce the war burden on the current generation of German citizens who believed that their war time sacrifices would result in higher post-war consumption. The government in effect promised future dividends for current, badly paid effort, but covered up the extent to which the war debt was going to prevent the payment of such peace dividend (Erhard, 1977, pp. 10-11). High direct taxes also would have indicated the extent of the mobilization of resources at a time when the German government intended to shield the population from the reality of the war as much as possible for reasons of political control. With the strength of the military sector in Germany, the allocation of the remaining budget disadvantaged weaker social groups and caused further, indirect costs of military expenditure (Erhard, 1977, pp. 34, 6264). One of the key effects of war is its negative impact on income distribution, both through the deliberate reduction in social spending programs benefitting poorer households and through the random expropriation of human and capital resources from the private sector through acts of warfare (Stewart, 1993). A fundamental problem of war debt is that there is a time inconsistency problem, which is asymmetric for a dictator and a democracy. A victorious dictatorship can burden the war costs on the losing country while a victorious democracy may prefer to share the war burden in the interests of stability and democracy (that is, it might apply the lessons of the Treaty of Versailles). The time inconsistency arises in that, unlike a democracy, a losing dictatorship will have no value attached to the future of its own people, thus increasing their expected war burden. Hence, citizens of a dictatorship have a lower incentive to lend to their government, as they know that it has no incentive to repay that money in case of defeat. This might either tie people more closely to their government (wishing that they will win to recoup their assets) or it forces the government to hide the extent of its war finance to reduce people's disincentive to lend further. In Germany, both factors probably applied during the war. The policy challenge after the war was to organize debt consolidation while rebuilding trust, social justice, and output. Given that the allies emerged victorious from the second world war, there was some burden sharing
238 Arming the South
between the allies and West Germany. Erhard's recommendations for post-war reconstruction in Germany For the economist and later West German finance minister Ludwig Erhard, the key obstacle to successful post-war reconstruction was the hidden, enforced, and unjust war debt, which accounted for over half of all government expenditure from 1939-45 (Schmolders, 1977). Given this, Erhard emphasized that at least the repayment of the war debt had to be distributed fairly. Social justice was a key characteristic of his design of post-war economic policy. For this reason, he opposed forgiving the war debt, accepting high post-war inflation, or nationalizing remaining private sector assets as ways to repay the German domestic debt. Erhard's second concern was with the pent-up demand and its likely effect on the post-war price level (Erhard, 1977, p. 21). In this, he shared a key concern of Keynes, writing in 1939 about the expected effects of the war on the British economy (Keynes, 1939; 1978). The eventual solution to this dual problem of consolidating the war debt and managing excess demand was to break people's expectations. This was brought about by total military defeat, the huge German population displacement in Central Europe, the allied occupation of Germany (which made people realize that they were not going to recoup their war time debt), and the subsequent monetary refonnof 1949. This currency reform, introducing the DM, created new expectations of incentives, once more holding out the promise of future reward for current sacrifice. But this time, the economic policy was transparent, time consistent, and backed by the allied forces still present in West Germany. This helped reduce uncertainty and thus increased investments (Collier and Gunning, 1995). Writingduring the war, Erhard underestimated some factors that were going to shape the coming West German Wirtschaftswunder he was going to father later. First, Erhard did not anticipate that the war loser could count on support from the victor (Erhard, 1977, p. 58). Post-world war I this may have been a prudent assumption but post-world war II the Marshall Plan helped western Europe rebuild its asset structure
Ludwig Erhard in Africa
239
(Burnham, 1990; Milward, 1970). Second, Erhard placed a premium on the obligation of the war time generation to also fund the war effort. In fact, the bankruptcy of the German state in 1945, which he advocated in his working paper, helped consolidate the war debt, thereby reducing the tax burden on future generations. The total West German public debt hence amounted to only 35 percent of GDP in 1950 (Statistisches Bundesamt, 1952, pp. 374, 456). Third, while Erhard realized that repaying the war debt would incur costs and that those costs had to be distributed fairly by appropriate taxes in the post-war economy, he nonetheless overtly stressed the need not only to avoid placing a strong tax burden on firms but also to compensate them for some of their war time investments as these had been made under the influence of the government (Erhard, 1977, pp. 51-52). Such approach neglects the lack of financial scope any post-war government was going to have for raising sufficient funds to compensate domestic economic losers of the war. This approach also fails to realize the degree of collusion with the government that many German firms practiced during the war, the extent of war profits that these firms obtained, and the moral and political responsibilities these firms hence carried for the actions of their government. Fourth, even with his preferences for a "social" market economy, Erhard's approach indicates relatively stronger concern with the welfare of larger private sector firms at the expense of private households. Erhard's thinking may have been influenced by the economic importance of large enterprises in inter-war Germany. Private households, while also bearing a significant cost of the war, would benefit from Erhard's postwar policies only indirectly, on the supposition that very high growth rates would raise disposable incomes significantly. Fifth, Erhard placed low priority on the export sector, thus failing to predict the dramatic increase in West German exports in the post-war period. In addition to the funds of the Marshall Plan, exports were going to finance the fast and strong West German recovery, which eventually permitted a sustainable rise in West German living standards in the 1960s. Finally, Erhard placed strong emphasis on the stabilization of war
240 Arming the South
debt, government finances, and the price level (Schmolders, 1977, pp. xxx-xxxi). Erhard complemented these prescriptions with a strong urge to build confidence, "justice," and social institutions, especially given the post-war setting (North, 1990; Schmolders, 1977, p. xxxiii). However, his view of a social market economy as exhibited in his 1943-44 paper is not a social-democratic or even a "third way" view of the state supporting or enabling people, but one of the state enabling the free market to function properly. The nature of the wars in Germany and Mozambique One of the strengths of Erhard's assessment of the economic effects of the war and his prescriptions for post-war reconstruction is that they can be applied, subject to modifications, to other conflicts as well. This does not imply that most wars are directly comparable to world war II. Rather, this section draws out the differences between world war II and another long-lasting and severe conflict, the internal war in Mozambique, 19791992. While the military and political dimensions of world war II are quite well known, this section also summarizes the key features of the conflict in Mozambique, thus permitting a better understanding of its economic implications. Both conflicts were essentially about the desire to extend political power, thus subjugating economic motives to overtly political ones (see table 10.1). Both wars were also extremely intense, as measured by the relative physical destruction caused, the population movements they induced, and their costs to the respective governments. Both wars lasted many years although world war II had a more clearly shifting frontline, which suggested that the war was likely to end with one side's defeat eventually. In Mozambique, by contrast, the government and the rebels rarely fought traditional battles along clear battle lines thus creating a stalemate in which the rebels could continue to terrorize the rural population quite unhindered (Finnegan, 1992; Geffray, 1991; Vines, 1996). Finally, both wars were ended through the occupation by foreign troops, though in Mozambique, the government maintained formal control during the UN presence (called ONUMOZ) even though it could not match its economic and administrative resources.
Ludwig Erhard in Africa
241
Table 10.1: The nature of the wars in Germany and in Mozambique Aspects of conflicts
Germany, 1939-1945
Mozambique, 1979-1992
international political control long, but dynamic huge population displacement/poverty allied occupation
domestic political control high quasi-permanent huge population displacement/ poverty quasi-UN occupation
world-wide stability and growth strong and war-oriented sovereign states well endowed high tech none
national economic instability weak domestic warlords poorly endowed/looting low tech very high
SIMILARITIES
Cause of conflict Conflict intensity Conflict duration Social effects of conflict Post-war outlook
high
DIFFERENCES
Spread and scale Pre-war situation Industrial base Type of opponent Opponent's strength Weapons used Aid dependency
The key differences between the two wars, apart from their geographical spread and scale, are that Germany had and maintained a strong economy before and during the war while Mozambique entered its internal war having only recently emerged from an independence war and a chaotic independence process. The industrial base and the weapons used also differed in both conflicts. Germany produced its own, capitalintensive weapons while in Mozambique the government imported its armaments and all other equipment whereas the rebels used very simple and cheap weapons to terrorize the population and carry out their attacks. This diminished their need for external, foreign-exchange based war finance significantly. In addition, Germany faced a well-organized and endowed allied army which carried forward its skills and resources into the period of occupation. The Mozambican government, on the other hand, faced a weak, badly endowed and organized rebel movement that sustained its war efforts through looting its resource base. The ceasefire of 1992 did not immediately result in clear post-war political arrangements. While ONUMOZ had a huge budget, its mandate was only agreed at very short notice, its structures were established on the ground with a long delay, and its mandate was limited in time, ending as abruptly as it had started
242 Arming the South
(Barnes, 1998; Synge, 1997). Finally, Germany was not as dependent as Mozambique on either export earnings or foreign aid, using neutral countries, occupied territories, and its domestic economic strength (via domestic debt) to fund its war effort. Mozambique, however, was unable to generate sufficient export earnings to fund its war efforts and consequently depended on very high levels of foreign aid and an increasing foreign debt to fund its budget. This analysis of both conflicts indicates that the severity of the conflict, relative to the level of development, was at least as high if not higher in Mozambique than in Germany. This has important implications for the economic effects of the war in Mozambique and the prospect for post-war reconstruction. The next section addresses these issues for Mozambique, drawing on the issues and lessons identified by Erhard. War finance and post-war reconstruction in Mozambique While Erhard succeeded in creating anew an expectation of future growth, in Mozambique the end of the war did not lead to the expectation, in 1992, of a future boom in output. Factors contributing to this view included the initial post-war insecurity and political instability, the high debt and high budget deficits, the severe destruction of rural infrastructure and assets, people's high discount rates preventing longterm savings or investments, and the continuing obstacles to private sector activities after years of bureaucratic socialism (Briick, 2000). However, the end of the conflict and the successful completion of the peace process delivered immediate and long-term peace dividends in some key social, legal, and political areas. The success of the peace process was based on economic incentives offered to every group of war participants thus encouraging them to end their violent activities. In effect, the peace settlement was Pareto-optimal across all political groups in Mozambique, leaving no group worse off than had the war continued (Berdal and Keen, 1997). In Germany, however, people were not persuaded by economic incentives to end the war but the defeat was imposed militarily. This may have had the advantage that there was less political fragility in the immediate post-war period even though it
Ludwig Erhard in Africa
243
represented a loss of political sovereignty. The peace process benefitted millions of international and domestic refugees who could return to their villages thus re-starting the smallholder, agricultural economy in which 80 percent of all Mozambicans are employed (Briick, 2000, table 4). The cease fire thus suddenly reduced the risk of lethal ambush and allowed people to resume their normal agricultural work without fear of attack or looting. In addition, the peace process gradually increased the protection of property rights and helped to enforce contractual obligations, to facilitate Mozambique's move toward democracy, and to increase the respect for human rights on both sides of the political divide. This may not represent a typical peace dividend derived from cutting military expenditure but it is an immediate peace benefit and a necessary condition for the realization of long-term growth. Furthermore, these benefits helped to strengthen the peace process thereby reducing the risk of renewed warfare even after the departure of ONUMOZ in late 1994. The endogenous peace process in Mozambique resulted in significant economic costs for the government, especially with the demobilization of soldiers and the organization of multi-party elections (Lewis et al, 1999; Mazula, 1996; Vines, 1998). The economic legacy of the war was a huge war debt, which amounted to three and a half times the value of national output at the end of the war (see table 10.2). The high debt had been fueled by high military spending during the war, which had accounted for over one-fifth of total government spending in the last six years of the conflict. Average military spending by all developing countries in 1986-1990, by contrast, accounted for only 9.5 percent of total government spending (Mohammed and Thisen, 1996, p. 378). The crippling level of war finance eventually brought about an endogenous end to the war in Mozambique. Both the looting rebels and the indebted government had exhausted all resources available and necessary for fighting this internal war. Social spending (defined as health and education spending in table 10.2) was less than half of military spending in Mozambique in 19871992 and about half of all developing countries' social spending in 19861990 (Mohammed and Thisen, 1996, p. 378). Social spending in Mozambique fell further during the ONUMOZ period as the economic
244 Arming the South
Table 10.2: Economic indicators for Mozambique, 1987-1997 In percent (unless otherwise indicated)
J 987-92 (late war)
344.1 Foreign debt/GDP 20.5 Military/total expenditure 9.2 Social/total expenditure 5.8 External interest/total expenditure 47.0 Total expenditure/GDP Annual real GDP per capita growt:h 4.2 Food production per capita index 78.6 (1980=100) 5.7 Annual exports per capita (1980 US$) Trade deficit/GDP 51.1 74.4 Foreign aid/GDP
1993-94 (ONUMOZ)
1995-97 (peace)
361.8 18.3 8.3 5.8 44.8 6.8 70.2
320.8 10.0 11.7 6.0 36.7 6.0 87.3
5.4
7.2
59.6 84.9
31.4 55.2
Sources: Briick et al. (2000) and author's calculations.
cost of peace required high government spending on other sectors. Even though social spending increased relatively speaking, its share in GDP remained constant throughout 1987-1997 as total government spending, as a share of GDP, declined by ten percentage points in that period. The negative effects of military expenditure on social justice were thus reinforced in Mozambique due to the negative effect of the ongoing structural adjustment program. Another constraint on social spending was the continuing high and rising levels of external debt service imposed by the extraordinarily high foreign debt. A reduction of such payments by 50 percent, for example, would permit an increase of Mozambican social spending by 25 percent. The population in the previously rebel-held areas in Mozambique thus ended up paying twice for the war. During the war, it had to fund the rebels who took by force whatever resources they needed while in the post-war period the rural areas reaped the smallest benefits from the peace. The high levels of real per capita growth which Mozambique obtained in this period were not driven by rural agricultural production or export earnings, which were both well below pre-war levels in 1997, but by the high levels of foreign aid (table 10.2). In fact, exports were so low
Ludwig Erhard in Africa
245
that the trade deficit in the late war and early post-war period amounted to over half of the country's GDP. The aid mainly benefitted urban service sectors but not rural small-scale producers thus delaying the arrival of a peace dividend or of poverty alleviation in the most waraffected areas and creating a potentially very unequal pattern of post-war growth (Briick et al, 2000). A lesson of the Marshall Plan which was not applied in Mozambique was that of commitment and consistency. Donor funds to Mozambique are supply driven. Depending on political events in the donor countries the amount, the timing, and the conditionality of aid will vary, thus placing a huge coordination cost on the already weak Mozambican state (ECON Centre for Economic Analysis, 1998). Mozambique would benefit from a lower level of aid granted for longer periods and with less administrative conditionality but more concise and relevant economic conditionality attached. In addition, it would have been closer to the spirit of Erhard's recommendations for the post-war German economy had the developed countries (ironically including Germany) been more generous sooner with the implementation of the highly indebted, poor countries (HIPC) initiative, especially for countries emerging from violent conflict. Erhard did not agree with a German post-war government canceling its own domestic war debt, as this would have reduced trust in such government, created unfairness, and reduced the scope for future borrowing (Erhard, 1977, pp. 60-70). Yet these arguments do not apply to foreign debt forgiveness under the HIPC initiative if accompanied by a careful package of forward-looking conditionality. Thus, more current resources could be used to increase current growth while targeting more government resources to the social sectors which in turn would strengthen the peace processes of post-war HIPCs. Erhard's concern for larger enterprises can also be detected in Mozambique where large foreign investments and the role of large businesses often overshadow the economic welfare of the majority of the population who live and work in remote rural areas as subsistence farmers. The large trade deficit is one of the key justifications for the prourban, pro-industrial growth policies pursued by the government since the end of the war. However, such policies neglect the export and welfare
246 Arming the South
potential of the rural smallholder sector. Real exports per capita (which include many agro-exports) were 314 percent higher in 1980 than in 1997 (Briick et al, 2000) suggesting the scale of the unrealized export potential deriving from rural smallholder investments. Given limited government capacity, there is a danger that such policy bias toward large projects could further deepen regional and social inequalities in Mozambique even with strong growth. Erhard's vaguely activist view on the role of the state in a waraffected economy has not been taken up by foreign donors as quickly as his preference for supporting larger firms (International Monetary Fund, 1996; World Bank, 1998). Yet poor war-affected economies will only show a supply response to market liberalization when key market institutions have been build and are allowed to work freely. These include institutions to enable and enforce contracts, to manage information and finances, to increase economic and social mobility, and to influence and control political leaders democratically. Given the serious war destruction, historic lack of capacity, and the absence of long-term experience with a free market in Mozambique, there is still a great need to promote civil and market institutions to enable all citizens to participate in the post-war market economy. Conclusions and policy recommendations Erhard's perceptive analysis of the challenges of German post-war reconstruction supports the view that secrecy in the budgetary process raises the average level of military spending, increases insecurity, and dampens the prospects for fast and equitable post-war growth. Yet the German war debt to GDP ratio proved small in comparison to the relative size of the Mozambican war debt thus facilitating a German peace dividend in the shape of high post-war growth. Rather than the German post-war government canceling its domestic war loans, Erhard recommended that the government should repay its war debt and thus re-establish credibility, trust, and a degree of social justice. These are all informal social institutions, which matter hugely for the restoration of certainty and the lowering of transaction costs, and which in turn increase post-war investment and growth.
Ludwig Erhard in Africa
247
If the war debt is with foreign lenders, however, these lenders should consider forgiving the debt to accelerate post-war economic growth, poverty alleviation, and political stability. This lesson of world war I, though applied to West Germany after 1945, was not practiced as part of the HIPC initiative with the relatively more war-damaged and poorer Mozambique. Some uncertainties as to the sustainability of the peace process and the equity of the post-war growth thus remain in Mozambique. In addition, war time donors should thus use their influence over governments involved in conflict to have them agree to a cease fire. It appears as if international donors failed to attempt such peace-oriented conditionality in the case of Mozambique. These lessons suggest that war time economic reforms in the areas of war debt, military expenditure, and economic institutions are crucial to maintain macroeconomic stability and household entitlements during the war and as a foundation for equitable post-war economic growth, especially in economies with long-lasting or devastating conflicts. Such institutional policies should include the strengthening of the civil service, of the independent judiciary, and of the freedom of expression, and they should lay the legal foundations for future investments and privatization. In addition, the government should collect and publish data on its security spending to increase the public pressure for social over military spending. In sum, Erhard's views of the secrecy surrounding war debt, of the negative effects of a large war debt, and of public expectations in the post-war period also matter for poor developing countries emerging from violent conflicts. However, since 1945 the importance of exports and aid in financing war and post-war reconstruction has increased thus requiring a modification of Erhard's analysis. In addition, his concern for the welfare of firms and his relative neglect of the social effect of war on private households are contradictory to contemporary development policies seeking economic growth and poverty alleviation in war-affected developing economies. References Barnes, Sam. "Peacekeeping in Mozambique," pp. 159-177 in Oliver Furley and Roy May (eds.). Peacekeeping in Africa. Aldershot, UK:
248 Arming the South
Ashgate, 1998. Berdal, Mats and David Keen. "Violence and Economic Agendas in Civil Wars: Some Policy Implications." Millennium Journal of International Studies Vol. 26, No. 3 (1997), pp. 795-818. Berthelemy, Jean-Claude, Robert S. McNamara, and Somnath Sen. "The Disarmament Dividend: Challenges for Development Policy." OECD Development Centre Policy Brief No. 8, 1994. Brauer, Jurgen and Manas Chatterji (eds.). Economic Issues of Disarmament: Contributions from Peace Economics and Peace Science. New York: New York University Press, 1993. Briick, Tilman. "The Economics of Civil War in Mozambique," pp. 191215 in Jurgen Brauer and Keith Hartley (eds.). The Economics of Regional Security: NATO, the Mediterranean and Southern Africa. Amsterdam: Harwood Academic Publishers, 2000. Briick, Tilman, E. V. K. FitzGerald, and Arturo Grigsby. "Enhancing the Private Sector Contribution to Post-War Recovery in Poor Countries." QEH Working Paper Series, No. 45. Oxford: Queen Elizabeth House, July 2000. http://www.qeh.ox.ac.uk/. Burnham, Peter. The Political Economy of Postwar Reconstruction. New York: St Martin's Press, 1990. Collier, Paul and Jan Willem Gunning. "War, Peace and Private Portfolios." World Development Vol. 23, No. 2 (1995), pp. 233-241. ECON Centre for Economic Analysis. Donor Coordination in Mozambique: Discussion Paper. Maputo, August 1998. Erhard, Ludwig. Kriegsfinanzierung und Schuldenkonsolidierung: Denkschrift von 1943/44. Frankfurt a/M: Propylaen, 1977. Finnegan, William. ,4 Complicated War: The Harrowing of Mozambique. Berkeley, CA: University of California Press, 1992. Geffray, Christian. A Causa das Armas. Porto: Edifoes Afrontamento, 1991. Gleditsch, Nils P., Olav Bjerkholt, Adne Cappelen, Ron P. Smith, and J. Paul Dunne (eds.). The Peace Dividend. Amsterdam: Elsevier, 1996. Hartley, Keith. "The Economics of the Peace Dividend." International Journal of Social Economics Vol. 24, Nos. 1/2/3 (1997), pp. 28-45. International Monetary Fund. Republic of Mozambique: Recent Economic Developments. Washington, DC, 8 November 1996.
Ludwig Erhard in Africa
249
Keynes, John M. "The Income and Fiscal Potential of Great Britain." Economic Journal Vol. XLIX, No. 196 (December 1939), pp. 626639. Keynes, John M. "Paying for the War," pp. 41-51 in Donald Moggridge (ed.). Collected Writings of J.M. Keynes. Vol. 22. London: Macmillan, 1978. Klein, Lawrence R., Fu Chen Lo, and Warwick J. McKibbin (eds.). Arms Reduction: Economic Implications in the Post Cold War Era. Tokyo, New York, and Paris: United Nations University Press, 1995. Lamb, Geoffrey. "Military Expenditure and Economic Development: A Symposium on Research Issues." World Bank Discussion Papers, No. 185. Washington, DC: World Bank, 1992. Lewis, Neryl, Geoff Harris, and Elisa Dos Santos. "The Demobilisation and Re-Integration of Ex-Combatants." In Geoff T. Harris (ed.). Recovery from Armed Conflict in Developing Countries: An Economic and Political Analysis. London: Routledge, 1999. Mazula, Brazao (ed.). Mozambique: Elections, Democracy and Development. Maputo, 1996. Milward, Alan S. The Economic Effects of the Two World Wars on Britain. London: Macmillan, 1970. Mohammed, Nadir A. L. Civil Wars and Military Expenditures: A Note. Washington, DC: World Bank, 1999. http://www.worldbank.org/research/conflict/papers/civil.htm. Mohammed, Nadir A. L. and Jean K. Thisen. "The Economics of Disarmament in Africa," pp. 359-380 in Nils P. Gleditsch, Olav Bjerkholt, Adne Cappelen, Ron P. Smith, and J. Paul Dunne (eds.). The Peace Dividend. Amsterdam: Elsevier, 1996. North, Douglass C. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press, 1990. Schmolders, Giinter. "Erhards Denkschrift im Lichte neuer Dokumente iiber die Kriegsfmanzierung 1933-45," pp. xxiii-xxxiv in Ludwig Erhard (ed.), Kriegsfmanzierung und Schuldenkonsolidierung: Denkschrift von 1943/44. Frankfurt a/M: Propylaen, 1977. Statistisches Bundesamt. Statistisches Jahrbuch fur die Bundesrepublik Deutschland. Wiesbaden: Statistisches Bundesamt, 1952. Stewart, Frances. "War and Underdevelopment: Can Economic Analysis
250 Arming the South
Help Reduce the Costs?" Journal of Economic Development Vol. 5, No, 4 (February 1993), pp. 357-380. Synge, Richard. Mozambique's Peace Process: UN Peacekeeping in Action, 1992-94. Washington, DC: United States Institute of Peace Press, 1997. Vines, Alex. Renamo: From Terrorism to Democracy in Mozambique? London: James Currey, 1996. Vines, Alex. "Disarmament in Mozambique." Journal of Southern African Studies Vol. 24, No. 1 (March 1998), pp. 191-206. World Bank. Report and Recommendation of the President of the International Development Association to the Executive Directors on Assistance to The Republic of Mozambique under the HIPC Debt Initiative. Washington, DC, 31 March 1998.
11 Angola: Civil War and the Manufacturing Industry, 1975-1999 Manuel Ennes Ferreira
Introduction Angola emerged from Portuguese colonial rule and became an independent country in November 1975. Since then civil war has pitted two groups against each other, the ruling MPLA party and a rebel group, UNITA. For my purposes here, it is useful to divide Angola's twenty-four year history into two distinct periods. The first one, between 1975 and 1991, is characterized by the intention of building a socialist economic and political system, based on central planning and on a Marxist-Leninist oriented single-party system (MPLA/PT).1 During this period, Angola's civil war was subject to strong external influence and material support: socialist countries, in particular Cuba and the Soviet Union, helped the MPLA, whereas western countries, chiefly the US, France, and especially South Africa, assisted UNITA. During the second period, 1992 to 1999, the country was marked by a market-oriented economic system and a multiparty political system. Thus, the first presidential and legislative elections took place in September 1992. But civil war soon resumed. On the eve of Angola's independence, its manufacturing industry was the economy's most dynamic sector, supplying more than half of domestic consumption. It accounted for 25 percent of GDP and showed
252 Arming the South
high annual growth rates (6.9 percent in 1972 and 14.3 percent in 1973). Composed of 3,846 enterprises, chiefly of small and medium-size, the sector employed some 200,000 workers transforming domestic raw materials and other inputs for the food and beverages and light industries and, using imported inputs, for the heavy and light industries.2 In light of the economic importance of the manufacturing industry and the military confrontation that gradually engulfed the entire country, it seems appropriate to investigate the relation between the economic performance of Angolan manufacturing industry and the civil war. While there are many studies investigating the relation between military expenditures and economic growth since Benoit's (1973) seminal study appeared, few research war's impact at the level of manufacturing industry, and none have yet done so for the case of Angola. Moreover, the usual interpretations tend to explain the evolution of industrial production exclusively as a function of civil war as if war were the determining cause of manufacturing performance. But we cannot presume this conclusion. Thus, taking Angola's specific economic, political, and civil war situation into account, this chapter identifies some additional, and perhaps more relevant, factors that explain the performance of Angola's manufacturing industry. Manufacturing industry in a centrally planned economy, 1975-1991 Nationalization and the state-owned industrial sector, 1975-1980 According to Angola's constitution (Lei Constitucional, 1975, article 2) "the MPLA... [guides] the political, social and economic direction of the Nation." Having chosen to establish a socialist society, the MPLA soon tried to build, between 1975 and 1980, the foundation for a socialist economic stmcture. This mainly meant to apply a policy of nationalization and confiscation of private enterprises, reaching all sectors of economic activity. Law No. 3/76 of March 1976, bluntly entitled the Law of Nationalization and Confiscation (Lei das Nacionalizaqoes e Confiscos), provided the legal framework and the basis for setting up state-owned enterprises, called Unidades Economicas Estatais (UEE). The procedures for their operation were later expanded
Angola: civil war and the manufacturing
industry, 1975-1999
253
upon in Law No. 77/77 of 15 September 1977 - the Estatutos das UEE. An essential element was the merging of enterprises within the same industrial sub-sector, thus imposing unique centralization and supervision of their activities. IMAVEST, a state-owned conglomerate created in October 1977 is an example: employing 2,400 workers, it was formed from twelve companies in the apparel and textile sector. Another example is ENEPA, created in 1983, which was assembled from five enterprises that operated in the plastic materials sector. Although most of the nationalization acts had been implemented by 1980, nationalization was actively continued until 1991. One of the main reasons for actively pursuing nationalization of industrial assets was that in a considerable number of cases companies had been abandoned by their owners. This only made it easier to justify the gradual encroachment and take-over by the state-owned industrial sector. Still, the survival of a small private industrial sector was accepted but limited in that difficulties were imposed on it. This was underlined at the First Congress of the MPLA in 1977: while recognizing that private property needs to co-exist with nationalized property, "... the state-owned sector should become dominant...; it must create conditions that permit it to gradually replace the private sector ..."3 (see also Kornai, 1992, p. 444). The creation of a state-owned industrial sector - Kornai (1992, p. 83) considers this the main criterium of socialism - was related to a larger objective regarding the country's development: "the People's Republic of Angola considers agriculture as the basis and industry as the decisive factor of its development" (seeLez Constitucional, 1975, article 8). In this context, and following the Soviet industrialization model, it was proposed that heavy industry, in conjunction with light industry, should be the dynamic element of socialist industrialization (see Departamento de Reconstrufao Nacional, 1976, p. 6). Import substitution industrialization and the deterioration of the military situation, 1981-1991 The strategy of nationalizing industrial assets failed, as indicated by the rapidly falling index of production across various industrial sectors (see table 11 A.l in the appendix). This failure, combined with an upsurge in
254 Arming the South
the intensity of the civil war and, as from 1982, falling world oil prices,4 led the MPLA to adopt a new policy: import substitution industrialization (see MPLA/PT, 1980, p. 34). It was thought that this would create improved conditions for progressive domestic industrialization and, in particular, create a better basis for light industry. Reference to domestic industry protection became part of the annual plans for manufacturing industry (see Ministerio da Industria e da Energia, 1980). But for a variety of reasons, the new model also failed: a worsening military situation, growing external indebtedness, obstacles imposed by the bureaucratic and administrative rules of the central planning system, and the general inadequacy of the new strategy and associated economic policy measures resulted in non-fulfillment of industrial production plans. Thus, following the Second Congress of the MPLA/PT in 1985, an economic reform program, the Programa de Saneamento Economico e Financeiro (Program of Economic and Financial Restructuring), was to be implemented as from 1987. At the same time, Angola applied for full membership in the IMF and World Bank. It was thought that this would permit Angola to reschedule its external debt and make it easier to access international financial markets. Short-term commercial bank borrowing had grown from 16 percent of total external debt in 1978 to 39 percent in 1985 (Ferreira, 1999, pp. 72 and 116). Consequently, short-term debt, as a percentage of total external debt, had increased from 11.5 percent in 1982 to 26.7 percent in 1985 (Ferreira, 1999, pp. 72 and 118). The main objective of the reform program was to restructure the stateowned sector, in particular the industrial sector, and to recast Angola's entire economic system. Contradictions and indecisions about the reform program and process arose within a difficult domestic political environment. The internal discussion was intense. In an attempt to reform while defending "scientific socialism as a programmatic goal" (MPLA/PT, 1990, p. 20), the discussions led to various compromises and generated new terminology, such as "mixed economy" (Santos, 1990b), "peace economy" (MPLA/PT, 1990, pp. 52-53), and "regulated market economy" (Santos, 1990 and Santos, 1990a). In the end, it remained unclear whether or not restructuring of the state-owned manufacturing industry should entail its total privatization.
Angola: civil war and the manufacturing industry, 1975-1999
255
An Angolan defense industrial base? Though engaged in a long-lasting and demanding civil war environment, Angola remained a non-arms producing developing country. It does not truly possess a defense industrial base. Of course, "defense industrial base" can be differently defined (see, e.g., Sandier and Hartley, 1995, pp. 182-185), and it is possible to place Angola into some of those definitions. For example, Haglund5 refers to the existence of national economic sectors that can be called upon to generate goods, services, and technology for ultimate consumption by the state's armed forces. For Krause 6 it would be possible to put Angola into one of his eleven stages of arms production, namely the stage of possessing the capability to perform simple tasks on imported weapons or assembly of imported weapons components. This may explain why some industrial enterprises were placed under the tutelage of the Ministry of Defense, as Angola had become one of the largest weapons importers among developing countries. In the 1982-1986 period for instance, Angola was ranked first in sub-Saharan Africa.7 One paradigmatic example was the so-called BCR unit (Empresa de Reparaqao e Recuperagao de Equipamento Militar). Created in 1983, it was to repair and refurbish military equipment. Other units to support Angola's defense and security needs included Aerovia, Bricomil, and Ecomil 1 and 2, all of them belonging to the construction and assembly sector.8 For Angola, the most meaningful relation between manufacturing industry and national defense needs might be the existence of dual-use activities, i.e., production applicable to both military and civilian purposes (see, e.g., Brauer, this volume, p. 105). But surprisingly, in spite of strong military interest to see industrial production plans fulfilled, industries continually missed specific supply targets for the Angolan armed forces. So there are two aspects here: first, what percentage of manufacturing production went to supply the armed forces and, second, what percentage of what the armed forces wanted to have supplied to them did they actually receive? Starting with the first, we must acknowledge that the percentage of domestic manufacturing output supplied to the armed forces was significant for many goods. For instance, in 1982 this amounted to 31 percent of flour, 40 percent of
256 Arming the South
sponge cushions, 36 percent of tobacco, and 10 percent of blankets. Two years later, in 1984, the values were: 15 percent of aluminum dishes, 95 percent of sponge cushions, 36 percent of sleeping berths, 23 percent of tobacco, 21 percent of tires, 17 percent of batteries and 25 percent of dry batteries.9 Second, what percentage of what the armed forces wanted to have supplied to them did they actually receive? In 1984 only shirts, sheets, cooking and kitchen tools, tobacco, and various general paints stayed closed to the desired supplies (Ministerio da Indiistria, 1985). In contrast, the receipts were less than planned for many others goods: canvas shoes (only 26 percent), trousers (33 percent), sponge cushions (30 percent), blankets (10 percent), beer (31 percent), soft drinks (14 percent), sleeping berths (36 percent), metal boards (6 percent), inner tubes (15 percent), batteries (16 percent), specialty paints (17 percent), and zero percent for military belts, handkerchiefs, hand towels, working suits, tooth brushes, cutlery, radios, and so on. So, for example, in 1984, while 95 percent of total sponge cushion production in Angola went to the military sector (and therefore five percent to the civilian sector), this supplied only 30 percent of what the armed forces had ordered for themselves. The manufacturing industry in disrepair Running a centrally planned economy produced disappointing results for Angola's manufacturing industry (see figure 11.1 and, for the underlying data, table 11A.1 in the appendix). Taking 1975 - the year of independence - as the reference year to construct an output index, we see that manufacturing output declined immediately after independence, recovered briefly but not fully, and then fell continuously. By 1991, the average value of industrial production was only one-third of what it had been in 1975.10 All sectors - food, heavy, and light industry - showed periods of recovery (especially for light industry), but eventually the performance of all sectors fell drastically to between one-quarter and onehalf of their respective 1975 levels. Intra-branch analysis is illuminating. The food sector (sometimes confusingly referred to as "food and beverages") consists of four subsectors: food, beverages, instant coffee, and sugar. The instant coffee and
Angola: civil war and the manufacturing industry, 1975-1999
257
175 150 I?**
,i. •..
100
M* <M
^ \
/ m y^v
\—
-•---•.
• •my. • -
75
/•- " *
50 25
\
• -^f. *•-
—
n
•
A*6 t ^
<&
CA9)
-<& - ^ N -<& - < *
Food industry - -. •- - - Light industry
S> nN
• Heavy industry — e - -IVBGE
Figure 11.1: Angola: Index of manufacturing output by branch (1975=100) and of military expenditures/government expenditures (1978=100), 1975-1991 Source: author's calculations based on Ministerio da Industria (1992), Gabinete de Apoio ao Redimensionamento Empresanal (1998), Ferreira and Barros (1998), and Ferreira (1999).
sugar sub-sectors collapsed in 1986-1987 and ceased all activity as from 1990 onward. The food and beverages sub-sectors, subject to decreasing supplies of agricultural raw materials, saw its production levels diminish almost every year. In 1991, these sectors produced at only 27 percent each of their respective 1975 index value. In light industry, the paper and paper-paste sub-sector saw its output completely paralyzed as a consequence of sabotage in 1983. At the same time, and due in great part to the instability generated by the military situation, the production of plywood and furniture fell abruptly and almost ceased after 1987, while apparel and textile production decreased
258 Arming the South
to about 20 percent of its 1975 level. Processed tobacco and glass works were the only sub-sectors with production above 50 percent of the base year. In the heavy industry sector, with the exception of electrical products, which several times produced above the 1975 level, all other sub-sectors (metals, metallurgy, chemicals, and transportation) produced by 1991 only at between 10 and 20 percent of the 1975 level. As a result of this remarkable decline, manufacturing industry's contribution to GDP fell from 11.3 percent in 1982 to 9.6 percent in 1984, rose to 10.8 percent in 1986, and fell again to 8.2 percent in 1988, 5.0 percent in 1990, and only 4.7 percent in 1991 (World Bank, 1991, p. 334; SecretariadeEstadodoPlaneamento, 1993, p. 42; Institute Nacional de Estatistica, undated, p. 18). Concurrently, the number of manufacturing workers fell from over 200,000 on the eve of independence to about 86,000 in 1990 (Institute Nacional de Estatistica, 1991, p. 34). What accounts for this catastrophic decline? The most popular explanation blames Angola's poor industrial performance on the civil war. Doing so puts the emphasis exclusively on the importance of the relationship between military expenditure (as a proxy for civil war) and the government budget, i.e., the defense effort. Numerous scholars have investigated the relation between this variable and countries' economic performance. Indeed, a look at figure 11.1 could lead one to accept this point of view as straightforward and indisputable: the development of the manufacturing sector appears to follow that of the defense effort (ME/GE) in a counter-cyclical way. The latter decreased until 1981 and industrial production recovered somewhat; thereafter, until 1989, defense effort increased while production shrunk. True enough. One cannot in fact deny the obvious influence the military situation had, direct or indirectly, on general economic activity and on industrial output.11 Direct effects occurred for example from acts of sabotage such as the case of South African aerial bombing of the industrial area of Lubango in 1979 or sabotage of the important paper and paper-paste factory in Alto Catumbela in 1982. Indirect effects occur in two ways, both applicable to Angola. First, military activity absorbs human and technical resources as well as financial resources in national currency or foreign exchange that otherwise could have been allocated to
Angola: civil war and the manufacturing industry, 1975-1999
259
the civilian sector. Indeed, from 1983 to 1991 the Angolan militaryrelated external debt each year amounted to about 68 percent of the country's total external debt.12 This meant fewer resources for productive investment and other inputs such as intermediate and machinery imports necessary for the industry. Second, by creating an environment of risk, the civil war affected agricultural production which in turn was to produce inputs for the food and textile industries. The war also made transportation more difficult and thereby reached upstream economic infrastructure, such as energy and water production and distribution, that affected industrial production (see, e.g., Sogge, 1992).13 So, yes, it is impossible to deny that the war affected economic activity. But a more thorough explanation of Angola's poor economic performance must go beyond the war and must challenge the self-serving point of view the MPLA and its government put forth,14 which one of its former officials captured in a single sentence: "so long as there exist armed bandits [i.e., UNITA], political economy in its true sense can hardly be fulfilled in this phase of building socialism."15 Consider, then, the place of the industrialization strategies: for the first five years after independence the focus was on heavy industrialization, thereafter, until 1991, on import substitution industrialization. Only a mechanistic repetition of the classic model of Soviet socialist industrialization can explain the first choice. And the second choice appears senseless in light of the small domestic market, the absence of any indication that it might expand, indeed the reduced and non-sustainable domestic demand, the population's low purchasing power, the reduced availability of domestic inputs to be transformed by manufacturing industry, and - as we will see next - the inadequate way in which the policy was implemented. Angola's import substitution industrialization attempt was characterized by a complete absence of articulate and coherent economic policy measures necessary to recover and to sustain domestic manufacturing. For example, on the doctrine that "the technique of devaluation is definitively not part of the financial character of a socialist society," ] 6 Angola' s central bank kept the national currency - the Kwanza - unchanged to the dollar until March 1991! The national currency became increasingly overvalued. Consequently, industrial production was
260 Arming the South
strongly punished by competition from much cheaper imported goods. This led to ridiculous and paradoxical efforts: on the one hand, enterprises were punished due to the Kwanza's overvaluation, on the other hand, they were compensated through transfers and subsidies from the government budget. Faced with lower and lower productivity,17 industrial enterprises also had to deal with fixed, administered selling prices of its output, prices which were much lower than the average cost of production. To cover costs, firms received subsidies and financial transfers. As Balassa points out, this produces negative incentives to increase production and productivity,18 the exact opposite of what one would want to achieve. At the same time, fixed nominal interest rates, which were negative in real terms on account of high inflation, and easy access to domestic credit to finance firms' deficits had the same sort of effect that subsidies and transfers had. Angola's state-owned industrial enterprises were, in essence, subject to what Kornai (1992) calls a "soft budget constraint." Clearly, these exchange-rate, monetary, fiscal, trade, and price policies undermined any chance for a successful import substitution industrialization. Foreign-investment policy also was muddled. In contradiction to the Law on Foreign Investment, promulgated in 1979, the Department for National Reconstruction wrote a few years earlier that while "we must use help [from capitalist countries], buying equipment and technology directly, [we must] never permit investments from the capitalist countries" (Departamento de Reconstruct Nacional, 1976a, pp. 10-11). Self-evidently, such position, even more so in light of Angola's already high country-risk, did not stimulate private foreign capital flows. As time went by, industrial enterprises were therefore becoming technologically obsolete. Apart from bungled thinking about industrialization and bungled economic policies, there was a third factor: the influence of central planning. This acted at two levels. At one level, the rigid hierarchy and sluggishness of the administrative system led to delayed shipments of raw materials and other supplies, domestically and from abroad, and caused production slowdowns and, sometimes, production stoppages.19 At another level, the bureaucratic complexity so characteristic of all planned
Angola: civil war and the manufacturing industiy, 1975-1999
261
and centralized systems soon gave rise to vested interests. This was furthered by the existence of state monopolies in external trade and domestic marketing. Well-placed people in public administration, stateowned enterprises, and the party had access to privileges. That was clear as from the beginning of the 1980s. The logic of generating rent-seeking situations (Krueger, 1974; Kornai, 1992) put the official dream of protecting and spurring domestic manufacturing industry in second place. When combined, all these factors not only devalued the importance but also diminished the capacity of domestic industry to successfully implement Angola's import substitution industrialization strategy (also see Myint, 1987, pp. 112-113). Clearly, the war was not the lone reason for Angola's abysmal industrial performance between 1975 and 1991. Manufacturing industry in a market economy, 1992-1999 The end of the cold war and the resulting military impasse on the ground suggested that a peace agreement between the MPLA and UNITA might be reached. Indeed, an accord was signed in late May 1991 in Bicesse, Portugal. Anticipating necessary structural changes, the revised constitution (Lei de Revisao Constitucional) of 6 May 1991 stipulated a market economy and a multiparty system. Presidential and legislative elections took place in September 1992 and appeared to complete the political transformation. But war soon resumed, lasting from late 1992 to late 1994, when a new peace agreement - the Lusaka Protocol - was signed. Yet still the war continued, sometimes sleepily and episodically. Neither did the creation of a Government of National Unity and Reconciliation (GURN), in April 1997, put an end to war. The GURN government was composed of a majority of MPLA ministers but included ministers from UNITA and other parties. Finally, in December 1998, the MPLA-dominated government launched a new, large-scale military offensive against UNITA, a situation that continues into the present. From an economic point of view, the transition to a market-based economy came not without troubles. The breakdown of the former economic system and a total lack of sense in conducting economic policy created serious economic distortions and imbalances (see, e.g., IMF 1997; 1999). For example, in Luanda the unemployment rate reached more than
262 Arming the South
Table 11.1: Angola: Macroeconomic indicators, 1992-1999 1992
1993
1994
1995
1996
1997
1998
1999
Inflation rate
495
1,837
971
3,783
1,650
147
134
329
Fiscal balance (%ofGDP)
31
23
20
27
11
17
14
15
Current account (in US$ mn)
-864
-838
-564
-994
-323
-872
-1,857 -1,449
Current account (%ofGDP)
-10.8
-14.4
-13.9
-19.6
-4.9
-11.5
-24.8
-24.7
-6.9
-24.7
2.5
10.7
11.3
7.7
5.5
4.4
Manufacturing (annual growth rate)
-22.3
-5.8
10.1
-11.4
2.6
9.3
4.9
7.1
Share of manufacturing in GDP (in %)
4.0
5.7
4.9
4.0
3.4
4.4
5.8
n/a
GDP (annual growth rate)
Sources: International Monetary Fund (1997; 1999); Banco Nacional de Angola (2000).
60 percent. Some macroeconomic indicators are provided in table 11.1. Inflation reached hyperinflationary levels, in part because of high budget deficits which, though lower than in the beginning of the 1990s, still ran at about 15 percent of GDP in 1999. The external accounts also deteriorated. The country's indebtedness increased; so did arrears. The current account balance showed a worsening trend throughout the 1990s. The ratio of current-account balance to GDP became worse, reaching a deficit of 24.7 percent in 1999, more than double the ratio for 1992. The petroleum sector alone accounted for more than half of GDP, more than 90 percent of export earnings, and for about 80 percent of government revenues. Due to the growth of this sector, positive annual growth rates for overall domestic output are observed. Yet domestic output in 1997 was 18 percent under its 1990 level (International Monetary Fund, 1999, p. 5).
Angola: civil war and the manufacturing industry, 1975-1999
263
Manufacturing industry was incapable of strong and steady recovery. In the second half of the 1990s, though its annual rate of growth had registered slight increases in its contribution to GDP (to 5.8 percent in 1998), it was far - very far - from its possibilities and needs. (Disaggregated data can be found in table 11A.1 in the appendix.) Between 1992 and 1998 for instance the food and beverage sector increased by a factor of 4.5, reaching 79 percent of its 1975 level. One might think that this was due to generalized recovery among its various sub-sectors. Unfortunately this it is not so. Only three products - breadmaking (20 percent), beer (50 percent), and soft-drinks (20 percent) accounted for fully 90 percent of the food and beverage sector's entire output. The same explanation holds for heavy industry. While the entire index of heavy industry production increased strongly, even surpassing the 1975 level in 1996, 1997, and 1998, this was almost entirely due to the production of ammunition in the chemical sub-sector.20 The other heavy industry sub-sectors such as metals and transportation practically ceased all activity. As the civil war continued through the 1990s, is it possible to conclude that the defense effort, as a proxy for civil war, determined Angola's manufacturing performance between 1992 and 1999? Analyzing figure 11.1 we found a certain relation between these two variables while concluding that other, more important factors, namely those related to economic policy, determined manufacturing performance. In figure 11.2 (see also table A l l . l ) the relation between defense effort and manufacturing performance is much weaker than in figure 11.1. Only in three years is manufacturing performance inversely related to the defense effort. Thus, to identify the reasons for the poor manufacturing performance, we once again need to turn our attention to other possible explanations. Manufacturing industry: deadlock again Following Bicesse, the goals for Angolan manufacturing industry, as viewed by government economic programs, emphasized the need for early recovery of the country's existing industrial capacity and to initiate its modernization. This, it was thought, would then address domestic
264 Arming the South
225 200 175 150 125 100 75 50 25 0
! 1992
1993
1994
1995
1996
1997
1998
— • — Total manufacturing - - - - - - - Total manufacturing (without arrrnunitions) —*—h03E
Figure 11.2: Angola: Index of total manufacturing output (1975=100) and of military expenditures/government expenditures (1978=100), 1992-1998 Source author's calculations based on Ministerio da Indiistria (1992), Gabinete de Apoio ao Redimensionamento Empresarial (1998), Ferreira and Barros (1998), and Ferreira (1999).
demand and also help develop the industry via industrial poles (Governo de Angola, 1994, pp. 47-48; 1997a, pp. 68-70; 1997b, pp. 147-148). Two strategic studies were carried out by the Ministry of Industry, the Piano Director de Reindustrializaqao de Angola of 1994 (Masterplan for Angolan Reindustrialization; see Camera de Comereio e Indiistria Portugal-Angola, 1995), and the Programa de Recuperaqao Industrial para 1998-2000 (Program for Industrial Recovery, 1998-2000; see Ministerio da Indiistria, 1999a). However, manufacturing recovery remained far from what was expected. A number of reasons explain why. Begin, again, with the idea that the war climate may have been
Angola: civil war and the manufacturing industry, 1975-1999
265
responsible, as it was thought to be up to 1991. True, the supply of agricultural inputs continued to be limited. True, the general environment of uncertainty and instability kept the country-risk high for foreign investors. But the military confrontations largely took place in towns in the countryside, not in Luanda, the capital. The war in the countryside did lead to the destruction and stoppage of some industrial activity and was strongly felt in Lobito-Benguela, in Huambo, and in Lubango - but not in Luanda, the seat of the country's main industrial area. As before, it is at the economic policy level and with certain characteristics of the newly privatized industrial enterprises that we find the explanation for Angola's failure of industrial recovery. Begin with the latter. As a consequence of the Redimensionamento do Sector Empresarial do Estado program (Re-dimensioning of the State-owned Enterprise Sector), the Angolan government began privatizing small and medium-size enterprises. This privatization program took place without transparency, but with much corruption, and was suspended in 1994. The beneficiaries, the majority of them linked to the political and party power of the MPLA and to military and security forces as well, became owners of de-capitalized and obsolete enterprises (Ferreira, 1995). These new entrepreneurs, without any kind of prior business experience and possessing at best just the barest of management know-how, and lacking their own financial resources, were now submitted to free-market forces. Suffering competition from the remaining state-owned enterprises regarding access to domestic credit and foreign exchange, the newly privatized industries did not live in an easy economic environment. Moreover, economic policy was a total failure. Exchange rate policy in particular was completely mishandled, as it was before 1992. Several devaluations that were undertaken did not compensate for the high inflation rates, resulting in a national currency that was perennially overvalued. Domestic tradeable goods therefore faced cheap competition from abroad and left them non-competitive. Moreover, the allocation of foreign exchange was based on fixed, administered rates. Few private business owners were granted access, competing unequally with stateowned enterprises and with privileged, "private" groups that relied on the MPLAs political complicity (see, e.g., Munslow, 1999). Similarly, restrictive and politicized trade policy, which required slow
266 Arming the South
and bureaucratic permission to authorize external trade operations, strongly penalized trade on imported inputs for the manufacturing industry, for example through import duties, and did not contribute to the development of domestic industry. Monetary policy and access to domestic credit sources also were mismanaged. Money supply was heavily curtailed to deal with the problem of hyperinflation but this meant that private entrepreneurs essentially could not obtain financing from commercial banks. Obsolete, de-capitalized private firms without access to credit could not be expected to do well. It is not surprising therefore that the majority of industrial output still originated in the state-owned sector. Moreover, the policy of administered output prices continued, as did the policy of subsidies and quasi-subsidies from the government budget. The joint effect of all these factors, together with the still-felt consequences of economic distortion and imbalances created by fifteen years of a centrally planned economy, resulted in an environment too inhospitable for manufacturing industry to recover and to modernize and increase industrial output and productivity. Conclusion Since 1975, Angolan manufacturing industry continuously has been subject to two major forces: the civil war and the economic policies the government pursued. The industry's poor performance, demonstrated by rapidly decreasing production levels of the three main manufacturing branches (food and beverages, light, and heavy industries) as compared to 1975, cannot solely be explained by the civil war situation and its economic proxy, the government's defense effort. It is true that the civil war, especially during the central-planning period, 1975-1991, implied some amount of crowding-out, directing resources away from productive activity, but as this chapter shows, it is in the way that economic policy was pursued that we find the primary causes of Angola's manufacturing performance. Despite talk of an import substitution industrialization policy, the policy was not in fact pursued. Moreover, a strong negative effect was exerted by Angola's overwhelmingly important oil sector. It generated substantial amounts of foreign exchange and allowed Angola
Angola: civil war and the manufacturing industry, 1975-1999
267
to borrow in the international financial markets, but also led the Angolan government to abandon any credible effort to pay attention to the serious difficulties the non-oil domestic sectors faced. Thus, the "output in the domestic non-oil economy remain depressed not only because of the warrelated conditions but because of macroeconomic policies ill-suited to the promotion of economic growth" (International Monetary Fund, 1997, p. 5). The civil war worsened the environment in which manufacturing industry operated. While the war certainly was and is a strong conditioning factor, it was not and is not the underlying fundamental reason explaining Angola's poor industrial performance.21 Notes 1. The MPLA (Movimento Popular de Libertagao de Angola) turned toward Marxist-Leninist ideology and changed its name to MPLA/PT (Partido do Trabalho) soon after the First Congress of the MPLA in December 1977. During the Second Extraordinary Congress of the MPLA/PT, in April 1991, the party returned to its old name, the MPLA. 2. Details may be found in Ferreira (1990) and Ferreira (1999). 3. See MPLA/PT (1977, p. 65). In other words: "revolutionary suppression of private enterprise property is a goal defined by the MPLA which will permit gradually to establish new production relations." 4. Oil pnces (in US$/barrel): 34.2 in 1980; 35.6 (1981); 31.7 (1982); 27.8 (1983); 27.3 (1984); 26.0 (1985); 12.6 (1986); 17.5 (1987); 14.1 (1988); 15.5 (1989); 22.1 (1990); and 18.5 (1991); see Ferreira, 1999, pp. 66 and 112. 5. Haglund, 1989, pp.1-2, as cited in Sandier and Hartley, 1995, p. 182. 6. Krause, 1992, p. 171, as cited in Brauer (chapter 5 in this volume). Also see Brauer for a critique of Krause's stages.
268 Arming the South
I. According to Ohlson and Skons (1987, pp. 187 and 201), among developing countries Angola was the sixth-largest recipient of Soviet arms in the 1982-1986 period (4.8 percent of the total) and first among sub-Saharan African nations. In contrast, for 1977-1981, Angola did not appear among the top-ten Soviet clients. 8. Parallel to BCR, the Ministry of Defense created other enterprises to assure better logistic and operational support for its armed forces. Examples are Agromil, in cattle production, Bacomil, in domestic trade, and Enatec in external trade. 9. Author's calculations from Ministerio da Indiistria (1983; 1992) for 1982 and Ministerio da Indiistria (1985; 1992) for 1984. 10. For a more detailed analysis of manufacturing industry performance in Angola during the period 1975-1991, see Ferreira (1999). I I . An econometric approach, using OLS, is in Ferreira (1992). The model relating the annual non-oil GDP real growth rate to ME/GE turned up a negative coefficient and an R2 of 0.68. Running the equation with the oil-export growth rate yielded a positive coefficient with an R2 of 0.74. The period was 1979-1991. 12. See Ferreira (1999, p. 316), author's own calculations. For a detailed description of the war's influence on Angolan economic activity, see chapter IV, pp. 243-321. 13. Further, specific examples of indirect effects are given in Ministerio da Indiistria (1981; 1986) and MPLA/PT (1977; 1985). 14. Official explanations beyond the war included factors such as lack of spare parts; lack of qualified technicians and skilled workers; technologically obsolete equipment; lack of maintenance and technical assistance; lack of managerial capabilities; and workers absenteeism. But none mentioned economic policy. See, e.g., MPLA/PT (1980; 1985a); Ministerio da Indiistria (1985; 1987).
Angola: civil war and the manufacturing industry, 1975-1999
269
15. "Nesta fase de edificac;ao (constru9ao) do socialismo, dificilmente a politica economica podera adquirir o seu verdadeiro sentido enquanto existirem bandidos armados" (Monakapui, 1983). At the time Monakapui was a member of the MPLA/PT Central Committee and the Co-ordinator of the MPLA/PT Provincial Committee in the Province of Bie. 16. See Banco Nacional de Angola (1977). The National Bank of Angola is Angola's central bank. 17. Taking 1985=100, manufacturing productivity fell to 80 in 1986, to 65 in 1987 and in 1988, to 55 in 1990, and to 51 in 1991 (Ministerio da Indiistria, 1994, p. 53). 18. According to Balassa, "in socialist countries (in Central and Eastern Europe), central planning did not permit competition among domestic firms or from imports. Given the sellers' markets, firms had little incentive to improve productivity" (1989, p. 37). 19. This was in fact officially recognized, year after year, as being one of the most important reasons to explain Angola's weak manufacturing performance. See, for instance, Ministerio da Indiistria (1989) and Santos (Onambwe) (1989). The latter was Minister of Industry at the time. 20. In 1991, at the Bicesse meetings, the US, the Soviet Union, and Portugal agreed not to supply weaponry to either side of the conflict, MPLA and UNITA. This is the so-called "Triple Zero" clause. At the 1994 Lusaka meetings, MPLA and UNITA agreed to a "Bilateral CeaseFire Modalities Timetable" (see HRW, 1996). This prohibited purchase of military equipment, lethal and non-lethal, from external markets. To circumvent this, the Angolan government decided to start producing ammunition internally. 21. See the letter of 3 April 2000, sent by Joaquim David, Minister of Finance, and Aguinaldo Aime, Governor of the National Bank of Angola, to Stanley Fischer, then Acting Managing Director of the International Monetary Fund. This letter accompanied the Memorandum of Economic Policies of the government of Angola. "[i]n March 1999, the Angolan
270 Arming the South
government secured approval from the National Assembly for a mediumterm strategy to address the economic and social problems accumulated over the past decades and exacerbated by the civil war" (see Governo de Angola, 2000). References Balassa, Bela. "Inward-Looking Import Substitution," pp. 36-38 in Gerald Meier and William Steel (eds.) Industrial Adjustment in SubSaharan Africa. New York: Oxford University Press/The World Bank, 1989. Banco Nacional de Angola. Estabelecimento da Taxa de Cdmbio do Kwanza Face ao Dolar Norte-Americano. Luanda: 8 March 1977. Banco Nacional de Angola. Angola Indicators. http://www.ebonet.net/bna (2000). Benoit, Emile. Defense and Economic Growth in Developing Countries. Boston: Heath, 1973. Brauer, Jurgen. "The Arms Industry in Developing Nations: History and Post-Cold War Assessment," chapter 5 in Jurgen Brauer and J. Paul Dunne (eds.) Arming the South: Military Expenditure, Arms Production, and Arms Trade in Developing Countries. London: Palgrave, 2002. Camera de Comereio e Indiistria Portugal-Angola. Piano Director de Reindustrializagao de Angola. Cadernos Economicos PortugalAngola, No. 8. Lisbon: Ed. CCIPA, 1995. Departamento de Reconstruct Nacional. Indiistria: Definiqao de urn Modelo. 4° Seminario do DRN. Luanda, May 1976. Departamento de Reconstru?ao Nacional. Indiistria: Objectivos Iniermedios. 4° Seminario do DRN. Luanda, May 1976a. Ferreira, M. Ennes. Angola-Portugal: do Espaqo Economico Portugues cis Relagoes Pos-Coloniais. Lisbon: Ed. Escher, 1990. Ferreira, M. Ennes. "Despesas Miliatres e Ambiente Condicionador na Politica Economica Angolana (1975-1992)." Estudos de Economia [Lisbon: ISEG/UTL] Vol. 12, No. 4 (July-September 1992), pp. 419438. Ferreira, M. Ennes. "Angola, la Reconversion Economique de la
Angola: civil war and the manufacturing industry, 1975-1999
271
Nomenclature J*etvoliere." Politique Africaine [Paris] No. 57 (March 1995), pp.11-26. Ferreira, M. Ennes. A Indiistria em Tempo de Guerra (Angola, 1975-91). Lisbon: Ed. Cosmos, Institute da Defesa Nacional, 1999. Ferreira, M. Ennes and Carlos P. Barros. "From War to Economic Recovery: Peace as a Public Good in Angola." Defence and Peace Economics Vol. 9, No. 3 (1998) pp. 283-297. Gabinete de Apoio ao Redimensionamento Empresarial. Privatizaqoes em Angola: Apresentaqao. Luanda: Ed. Gare, 1998. Governo de Angola. Programa Economico e Social. Luanda: Governo de Angola, 1994. Governo de Angola. Programa dePolitica Economica e Social. Luanda: Governo de Angola, 1997a. Governo de Angola. Programa de Estabilizaqdo e Recuperaqao Economica de Medio Prazo 1998-2000. Luanda: Governo de Angola, 1997b. Governo de Angola. Angola: Memorandum of Economic and Financial Policies of the Government of Angola. Luanda: 3 April 2000. Human Rights Watch (HRW). Angola - Between War and Peace: Arms Trade and Human Rights Abuses since the Lusaka Protocol. New York: Human Rights Watch Arms Project, Vol. 8, No. 1 (A), 1996. International Monetary Fund. Angola - Recent Economic Developments. Washington, DC: IMF Staff Country Report No. 97/112, 1997. International Monetary Fund. Angola - Staff Report for the 1998 Article IV Consultation. Washington, DC: IMF, 1999. Institute Nacional de Estatistica. Perfil Estatistico de Angola, 1987-1990. Projecto ANG/89/009-PNUD/INE. Luanda, August 1991. Instituto Nacional de Estatistica (INE). Perfil Estatistico Economico e Social, Angola (1991-1994). UCP. Luanda (undated). Kornai, Janos. The Socialist Economy: The Political Economy of Communism. Oxford: Oxford University Press, 1992. Krueger, Anne. "The Political Economy of the Rent-Seeking Society." American Economic Review Vol. 64 (1974), pp. 291-303. Ministerio da Indiistria. Conclusoes do Conselho Consultivo do Mind sobre a Actividade da Indiistria no 1°Semestre de 1981. Luanda: 1415 July 1981.
272 Arming the South
Ministerio da Indiistria. Relatorio de Execuqao do Piano de 1982 do Sector da Indiistria. Luanda, 1983. Ministerio da Indiistria. Relatorio de Execuqao do Piano TecnicoEconomico em 1984 do Sector da Indiistria. Luanda, 1985. Ministerio da Indiistria. Balanqo do Cumprimento do Piano TecnicoMaterial de 1986. Luanda: Gabinete do Piano 1987. Ministerio da Indiistria. Direcqoes principals da Industrializaqdo Continua e do Desenvolvimento de uma Indiistria de Grande Capacidade Recorrendo a Todos os Recursos Naturals. Luanda, 1986. Ministerio da Indiistria. Conclusoes e Recomendaqoes do Conselho Consultivo Alargado do MIND sobre o Piano Tecnico-Material para 1989. Janeiro, Benguela, 1989. Ministerio da Indiistria. Evoluqdo daproduqdo do Sector da Indiistria no Periodo 1970-1990. GEP. Luanda, 1992. Ministerio da Indiistria. Piano Director de Reindustrializaqdo de Angola -Relatorio. Versao preliminar. Luanda, 1994. Ministerio da Indiistria. Balanqo da Produqao. GEPE. Luanda, 1996, 1997, 1998, and 1999. Ministerio da Indiistria. Investir na Indiistria, uma Porta Aberta para o Futuro. Luanda: Ed. IDIA, 1999a. Ministerio da Indiistria e da Energia. Conclusoes e Recomendaqoes do Conselho de Direcqdo Alargado do Minden sobre o Piano TecnicoAdministrativo de 1981. Luanda: 2-4 October 1980. Monakapui, Marques. Discurso do Coordenador do Comite Provincial do Bie do MPLA/PT no Encerramento da Reuniao da 7a Sessao Ordindria do Comite. Bie: 27 August 1983. MPLA/PT. Relatorio do Comite Central ao 1° Congresso do MPLA. Luanda: 1977. MPLA/PT. Orientaqoes Fundamentals para o Desenvolvimento Economico e Social (periodo 1980/85). Luanda: Ed. Secretariado do CC do MPLA/PT, 1980. MPLA/PT. Relatorio do Comite Central ao II Congresso do MPLA/PT. Luanda: 1985. MPLA/PT. Tese: Desenvolvimento da Produqao Material. Luanda: 1985a.
Angola: civil war and the manufacturing industiy, 1975-1999
273
MPLA/PT. Projectos de Teses do III Congresso do MPLA/PT. Luanda: Ed. Vanguarda, 1990. Munslow, Barry. "Angola: The Politics of Unsustainable Development." Third World Quarterly Vol. 20, No. 3 (June 1999), pp. 551-568. Myint, Hla. "Neoclassical Development Analysis: Its Strengths and Limitations," pp. 107-150 in Gerald Meier (ed.). Pioneers in Development. Washington, DC: The World Bank, 1987. Ohlson, Thomas and Elisabeth Skons. "The Trade in Major Conventional Weapons," pp. 181-296 in SIPRI Yearbook. Oxford: Oxford University Press, 1987. Sandier, Todd and Keith Hartley. The Economics of Defense. Cambridge: Cambridge University Press, 1995. Santos, J. Eduardo. Discurso na Abertura da VIII Sessao Ordindria da Assembleia do Povo. Luanda: 29 August 1990. Santos, J. Eduardo. Discurso na Abertura da Sessao Extraordindria do Comite Central do MPLA/PT. Luanda: 25 October 1990a. Santos, J. Eduardo. Discurso de Encerramento do III Congresso do MPLA/PT. Luanda: 9 December 1990b. Santos (Onambwe), Henrique. Discurso do Ministro da Indiistria na Sessao de Abertura do Conselho Consultivo Alargado do Mind. Janeiro, Benguela: 1989. Secretaria de Estado do Planeamento. Contas Nacionais de Angola, 1985-1990. Luanda: Ed. INE, November 1993. Sogge, David. Sustainable Peace: Angola's Recovery. Harare, Zimbabwe: Southern African Research and Documentation Centre, 1992. Stockholm International Peace Research Institute (SIPRI). The SIPRI Military Expenditures Database, http ://freja.sipri.se. 2000. World Bank. Angola: An Introductory Economic Review. Washington, DC: The World Bank, 1991.
274 Arming the South
Appendix Table 11A.1: Angola: Index of industrial output value, by branch and total, at 1987 prices (1975=100) and index of military/government expenditures (ME/GE), (1978=100) Food/ Beverages
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
100 82 62 70 73 70 70 66 64 56 60 48 41 41 34 31 22 17 14 23 18 56 64 79
Light Industry
Heavy Industry
100 74 75 78 86 100 111 102 88 90 90 82 51 55 54 53 35 26 14 10 6 14 19 16
100 70 63 68 93 91 90 73 58 53 68 74 50 39 66 80 51 39 8 62 34 100 208 122
Heavy industry (without ammunition)
100
39 8 9 20 65 9 17
Total Industry
100 77 67 73 82 86 91 82 72 69 74 67 47 46 48 49 33 25 13 24 16 46 69 60
Total industry (without (ammunition)
ME/ GE
100
n/a n/a n/a 100 106 88 81 102 138 156 153 152 169 185 196 159 83 85 172 208 120 96 145 171
25 13 15 13 40 36 42
Sources: Author's calculations based on Ministerio da Indiistria (1992; 1996; 1997; 1998; 1999); Gabmete de Apoio ao Redimensionamento Empresarial (1998); Internal tonal Monetary Fund (1997; 1999); Stockholm International Peace Research Institute (2000); Ferreira and Barros (1998); and Ferreira (1999).
12 Military Spending and Economic Development in Sub-Saharan Africa: A Supply-Side Analysis Oyinlola Olaniyi
Introduction Development theory proposes that physical and financial resources as well as organizational capabilities play a significant role in the determination of economic development. In sub-Saharan countries, a substantial proportion of these resources are allocated to the military sector. Apart from high defense budgets, the sector also attracts a high proportion of highly skilled scientific, technological, managerial, and administrative manpower. Since highly skilled manpower is scarce in sub-Saharan Africa, it is a critical factor in the determination of the absorptive capacity of these economies. Although the cost of anarchy and destruction, or even the economic and psychological implications of total subjugation, cannot be overestimated, it could be argued that a more urgent war which must be won in sub-Saharan Africa today is the war against hunger, ignorance, and disease. According to Sivard (1989), an African is 22 times more likely to die of a poverty-related disease than to die of a war-related cause. Yet most sub-Saharan African countries spend more per capita on military expenditure than to prevent hunger or disease.
276 Arming the South
The relationship between military spending and economic growth in sub-Saharan African countries is not clear. With the aid of an augmented economic growth model, this chapter distinguishes between substitution and externality effects of military spending in sub-Saharan Africa. The chapter then proposes a model of strategic complementarity between the military and the civilian sectors of the economy in which the military participates in the civilian sector to promote sustainable economic growth, which is needed to sustain a strong military. Literature review The question of the macroeconomic impact of military spending has led to the emergence of several schools of thought. The major ones are the neoclassical, Keynesian, and institutional schools. The neoclassical doctrine provides a supply-side argument, which is that military spending provides social security which is a pure public good, nonexcludable, nonrival, and with free-rider characteristics, that make its provision through the market system not feasible. Further, a clear trade-off exists between military and civilian spending and expenditure: the former crowds out civilian investment (Adeola, 1996). In the neoclassical approach the state is a rational actor who balances the opportunity cost and security benefits of military spending to maximize a well-defined and rational social welfare function. A prototypical Keynesian analysis sees military spending as an integral part of public spending, which generates aggregate demand effects through the income multiplier. This approach sees an active state that may use military spending as a counter-cyclical instrument. Yet another school is the institutional school which emphasizes the resource allocation process. This school's reasoning is predicated on the existence of a military-industrial complex and focuses on how military spending can lead to industrial inefficiency and to the development of powerful interest groups (Dunne, 1996). It argues that in addition to the military, members of this complex depend on military spending by the state, in industry, and in other parts of society. They therefore constitute a pressure group to increase military spending independent of any security threat. Thus they push resources to the military sector which negatively affects
Sub-Saharan Africa: a supply side analysis
277
itself and the rest of the industrial sector because of the military sector's lack of market discipline and non-competitive market characteristics. Though the military-industrial complex is virtually non-existent in subSaharan Africa, the military-business complex, the community of contractors who supply defense needs, collaborates with the militaryindustrial complex of developed countries to generate a similar effect in sub-Saharan Africa. The existence of multiple paradigms illustrates the lack of theoretical consensus on the impact of military spending on the economic and social sectors. The watershed of empirical studies in this field is Benoit (1973) which contrary to the neoclassical view suggested a positive relationship between military spending and economic growth in developing countries. Benoit identified, though heuristically, channels through which military spending could positively affect economic growth. Some of these channels include resource mobilization effects, modernization effects, and externalities. The debate that followed his work confirms that the guns-versus-butter trade-off1 is not related to substitution effects alone. Interconnections, interdependencies, multiple causation, primary and secondary effects, as well as short and long-run impacts may exist as well. The debate has also shown the need to differentiate not only between developing and developed countries but also within the group of developing countries, considering their respective development and defense-related characteristics. But no general conclusion is available about the nature, size, direction, or even the presence of causal relationships. Empirical results vary across countries and sometimes over time for the same country even when the same methodology is applied (e.g., Chan, 1985; 1987; and Frederiksen and Looney, 1983). For example, Benoit proposed that military spending is growth-enhancing because it mobilizes hitherto unused resources (Benoit, 1973). While it is correct that developing countries are characterized by unemployed resources, there are other obstacles. For example, foreign currency resource gaps and structural obstacles lead to supply rigidities. This limits the absorptive capacity of the economy and its ability to mobilize non-tradable producer goods quickly for the expansion of production when public spending is increased. And where human capital constraints do exist, especially for
278 Arming the South
skilled labor, expansion of military personnel, bases, and industrial systems would worsen resource constraints. Thus, mobilization of scarce resources such as highly skilled administrative, scientific, and technological manpower for defense purposes may compound supply rigidities. It follows that when higher military spending is supplyconstrained, cost-push inflation may result and reduce investment and growth in the long-run. If demand shortages and excess capacity do exist in industries, domestic production of military hardware could conceivably lead to demand and employment creation. But whether military spending does in fact create more resources or substitutes for civilian investments depends on its overall impact on the factor productivity differential between the military and civilian sectors. If factor productivity in the military sector is higher than in the civilian sector, it may be more beneficial to increase military spending; if not, it may be detrimental. The overall resource mobilization effect thus depends on structural and cyclical conditions of the economy. Indeed, for a cross-section of countries, and contrary to Benoit 1 s observations, Deger finds an overall negative impact because military spending crowds out investment in the civilian sector (Deger, 1986). In Nigeria, military spending also does not appear to contribute to industrial development in spite of the huge capacity underutilization in capital and metallurgical industries (Olaniyi, 1998). Any claimed externality effect of military spending is also dependent on military capital and military research and development to serve some civilian purpose. This would depend, however, on the degree of specialization and the level of high-technology involved. For example, "in countries like Sweden and Japan where military technology is close to civilian technology"(Kaldor, 1991, p. 144), the spill-over effect is high. But even for a country like the UK, where less than 20 percent of military spending is likely to have any beneficial effect on the civilian economy (Kaldor, 1991), the spill-over maybe low. The conflicting theoretical conclusions and empirical results suggest that the demand and supply of military spending depend on and generate a complex web of sometimes opposing relations among various economic and non-economic variables within an economy. The direction and magnitude of these relationships depend on divers endogenous and
Sub-Saharan Africa: a supply side analysis
279
exogenous factors that generate primary and secondary effects contingent on the historical realities of each country. Hence Dunne suggests that the economic effects of military spending are an historically specific and contingent rather than a deterministic or contradictory/dialectical process (Dunne, 1997). Model specification To analyze how supply constraints are affected by possibly contrasting impacts of substitution and externality effects of military spending on an economy, a supply-side model is econometrically estimated. The augmented growth model used here is based on a neoclassical production function. Originally developed by Feder (1982) to estimate externalities between the export sector and the rest of an economy, it was later applied by Ram to the military sector (1986). The model decomposes the impact of military spending on the economy into externality and substitution effects. The details of the full model can be found in Ward, Davis, and Chan (1993). It is given as follows:
M M — +9 — Y M
US
To operationalize equation (12.1), it is converted to an annual, discrete model, and intercept and error terms are added. Thus we have: ^Y
/
02.2)
M
a a
( 8
)kM
6
kM
,
r;- ^ r-/^~s- 'Z/ iu'£
To simplify, the variables can be rewritten as: (
(12.3) g= a0 + ai+
s
fil+\—--
)
0\mr + 9m + s ,
280 Arming the South
where g is the growth rate of national income, i is the ratio of investment over national income, / is the growth rate of the labor force, mr is the ratio of military expenditure over national income, and m represents the growth rate of military spending. Since labor force data are unavailable, the population growth rate is used as a proxy for growth in labor supply. The factor productivity differential is represented by 5, and a0, a, (3, and 9 are coefficients to be estimated, while 8 is the error term. Employing this type of model, Biswas and Ram (1986) found significant differences in the experience of low-income LDCs and middle-income LDCs. The model is criticized, among other items, for not considering the direction of causality (Sandier and Hartley, 1995). But since the study presented here is a cross-country analysis, a causality test will not be useful. However, Mohammed's causality studies revealed a unidirectional causation from military spending to economic growth in Africa (Mohammed, 1993). His sample consisted of 13 sub-Saharan African countries almost all of which are included in my study's sample as well. Twenty-five countries are sampled here. They all share similar strategic, economic, and historical experiences. They are all classified as low-income countries on the basis of 1993 and 1994 data, which are used for this study (World Development Report, 1995; 1996). They are largely dependent on subsistence agriculture and primary exports. They are all arms importing. Their industrial sectors are small relative to their economies except for a few that have large mining sectors and a few others with fairly large service sectors. Manufacturing activities are largely agro-processing and manufacturing assemblages; the rates of economic growth are generally low. These countries and the structure of their economies are shown in appendix 12A.1. Theoretical expectations Production function models are usually expected to generate positive coefficients for all variables. However, in a situation of deteriorating output, it is possible to obtain negative coefficients between output and any of the input factors. Empirically, a wide variety of coefficients has indeed been found ranging from strongly positive to strongly negative
Sub-Saharan Africa: a supply side analysis
281
(Ward, Davis, and Chan, 1993). A number of analyses have suggested that defense spending is likely to dampen capital formation, undermine export competitiveness, and hamper the innovation of civilian products (e.g., Rothschild, 1973). But others argue that military infrastructure such as roads, ports, communication gadgets, military schools and hospitals can all benefit the civilian economy. Also, military human capital and the use of military aircraft to carry VIPs may contribute to economic growth (Ward, Davis, and Chan, 1993). Therefore, the theory is neutral in its predictions. However, since military spending is partly spent on the construction of rural roads, hospitals, educational facilities, and army barracks that open up and stimulate rural economies where they are situated (Haruna, 1978) we would expect a positive externality effect, i.e., 0 > 0. But the posited growth benefits of military spending might be provided more efficiently with investment in the civilian economy (Hewitt, 1991). It is therefore expected that the marginal productivity of the military sector will be less than that of the non-military sector, i.e., 5 < 0. And the combined effect, called the substitution or size effect, of military spending is expected to be negative, i.e., [5/(1+5) - 9] < 0. Analysis of findings Data for the 25 countries examined in this study are for 1993 and 1994. This reduces possible policy shift effects on the estimates. Data on gross domestic product (GDP), investment, and population are obtained from the World Development Reports (World Bank, 1995; 1996). Military expenditure data is obtained from the International Institute for Strategic Studies (1995; 1996). The results of a linear OLS regression on equation (12.3) are: coefficients -0.175 + 0.55 i + 0.51 1 - 1.339 mr + 0.266 m t-values (-2.364) (2.118) (0.54) (-0.164) (0.858) p-values 0.028 0.047 0.593 0.872 0.401 R2 = 0.35; F - 2.692; p-value of (F) = 0.061 An R2 of 0.35 shows that 35 percent of the change in the economic
282 Arming the South
growth rate, g, is explained by movements in the independent variables. The calculated F = 2.692 is statistically significant at the 6 percent level. The estimated signs of the coefficients are in line with expectations. Although small, the coefficient of the domestic investment/output ratio, i, is positive and statistically significant at less than the 5 percent level. It is also relatively small. This may be due to the low capital absorption capacity of these economies and to inadequate infrastructure and other factors such as lack of human and social capital, a poor policy environment, and an inclement economic and political environment, particularly in the industrial sector. The coefficient for the population variable (as a proxy of labor input) is not statistically significant. This is not surprising since population has been increasing as incomes declined. Also investment in human capital enhancing factors like education and health in the sample countries is low. For example, adult literacy ranges from 20 to 79 percent and averages 57 percent for these countries while life expectancy ranges from 39 to 61 and averages 49 years (World Development Report, 1995). Also, the higher rate of armed conflict on the continent has led to a high rate of displacement and to refugee problems that have led to a rise in the number of unproductive labor. Another explanation maybe the slow rate of technological adaptation in agriculture (Microsoft, 1997). The factor productivity differential, 5 = -0.52, 2 shows that nonmilitary spending is economically more productive than is military spending. We may conclude that to the extent that military spending crowds out non-military spending, it had a negative impact on these economies. The coefficient of the size effect of military spending, [5/(1+5) - 9] = -1.339, is negative but not statistically significant. This finding is supported by previous findings for countries with similar characteristics (e.g., Frederiksen and Looney, 1983; Lim, 1983; Adeola, 1996; Brzoska and Wulf, 1979; and Olamyi, 1993). As expected - because military spending is partly spent on the construction of rural roads, hospitals, educational facilities, and army barracks that open up and boost rural economies (Haruna, 1978) - the externality effect of military spending, 9 = 0.266, is positive but it is statistically not significant, possibly because of the absence of deliberate efforts to employ military capital and skills in the civilian economic and
Sub-Saharan Africa: a supply side analysis
283
social sectors. It should be noted that the estimated constant, a0 = -0.175, is statistically the most significant factor in this model. If the constant aggregates the effects of entrenched environmental factors on the dependent variable, then these factors must be very important. From the criteria used to select the sample, poverty is a common factor. Even if this points to a vicious cycle of poverty, experience has shown that merely investing in capital would not necessarily help countries break out of such a cycle. The investment must be human and social capital-enhancing to change attitudes, to increase the level of innovation, inventiveness, reception of and adaptation to new ideas and technology. It must also improve health and reduce the level of mortality and morbidity. SubSaharan Africa must invest in her people. Also, some religious and cultural practices in sub-Saharan Africa impact negatively on productivity and hence on economic growth. These must be identified from society to society and modified. Environmental degradation must also be a matter of serious concern to ensure sustainability of growth. Investment in physical and social infrastructure should also be made to enhance the availability of cooperant factors to foster the productivity of capital. To test whether the structure of the economy affects the impact of military spending on economic growth, data for countries with higher contributions from the industrial sector were eliminated to get a second sample. The countries eliminated are Nigeria (43 percent), Lesotho (47 percent), Zambia (36 percent), and Zimbabwe (36 percent). The other 21 countries have a very low industrial contribution raging from 10 to 31 percent of GDP and high contributions from the agricultural sector ranging from 29 to 60 percent of GDP. The result of an OLS regression on the second sample is shown below: coefficients -0.319 + 1.329 i + 0.276 1 - 1.127 mr + 0.339 m t-values (-4.056) (3.950) (0.334) (-0.159) (1.261) p-values 0.001 0.001 0.743 0.872 0.225 2 R = 0.592; F = 5.802; p-value of F= 0.004 The parameter estimates in this case are generally better than those obtained for the larger sample. R2 increased to 59 percent and the
284 Arming the South
calculated F-statistics is significant at the one percent level. This shows a very high explanatory power. The coefficient for investment is again positive, greater than unity, and significant at the one percent level. This shows that capital is more productive in the agricultural than in the industrial sectors of these countries. The negative impact of the environmental factors is more pronounced than before, confirming that environmental factors limiting growth are related to poverty, ignorance, decease, and morbidity as it is these kinds of variables that deviate widely between the agricultural and industrial sectors. The coefficient of labor productivity became even less significant than in the first case when relatively more industrialized countries were included in the sample. Although the coefficient was not statistically significant in either sample run, this may indicate that the productivity of labor is higher in the industrial than in the agricultural sector. Thus, more labor can still be drawn from the agriculture sector of these economies. The substitution or size effect of military spending, [5/(1+5) - 9] = 1.127, remained essentially unchanged (negative and insignificant at approximately the same level). But the externality effect, 9 = 0.339, reveals a somewhat higher, although still statistically insignificant, level than that obtained in the first case. At best this might show that military spending facilitates agricultural production more than industrial output. Actually, this would not be inconsistent with previous findings. Deger and Sen (1983) found that the spin-off effect of military spending on industrialization in the third world is generally insignificant. Finally, 5, the factor productivity differential can be computed to be -0.44 (see endnote 2). Military spending is less productive than civilian spending. Conclusion The objective of this study was to differentiate between the substitution and externality effects of military spending in sub-Saharan Africa. In addition, it differentiated between the effect of military spending in predominantly subsistence-agriculture based countries and industrializing or mineral-exporting ones. In either case, I find that the effect of military spending (as a percentage of national income) on economic growth is
Sub-Saharan Africa: a supply side analysis
285
negative but not statistically significant in sub-Saharan African countries. But while the direct substitution effect may not be statistically significant, the crowding out of resources needed for investment in much needed human and social capital enhancing investment must not be overlooked. The optimal military doctrine for a poor country is one which minimizes the substitution effects, maximizes externality effects, and maintains a detercence equilibrium. This carries significant implications for national strategy and the defense policies of such economies. The findings suggest that investment in human and social capital constitute the most viable option out of poverty and the eventual realization of self-determination in sub-Saharan Africa. While maintaining deterrence, military spending could be further reduced if the countries were to adopt a collective defense strategy. The estimated positive, but statistically insignificant, externality effects from the military to the civilian sector suggest that existing military capital and manpower could be used without loss, indeed with net gains, to be employed in the civilian sector, to rebuild sub-Saharan African economies and relaunch them on the path of economic growth. In peacetime, highly trained military manpower can be used to construct urban and rural roads and railways or to maintain decaying social infrastructure. Military hospitals and schools can be made more available to the civilian population. The military can contribute in many other sectors of the economy, depending on the available facilities in each country and the area of need or effectiveness. To make specific recommendations of course requires country-specific studies. Notes 1. In the African context, we might refer to this as the guns-versus-gari trade-off, gari being a staple food in West Africa. 2. Since [(5/1+5) - 0] = -1.339 and 9 = 0.266, 6 can be computed to be 0.52.
286 Arming the South
References Adeola, O.F. "Military Expenditures, Health and Education: Bedfellows or Antagonists in Development?" Armed Forces and Society Vol. 22, No. 3 (Spring 1996), pp. 441-467. Benoit, Emile. Defense and Economic Development. Lexington, MA: Heath, 1973. ' Biswas, Basudeb. and Rati Ram. "Military Expenditures and Economic Growth in Less Developed Countries: An Augmented Model and Further Evidence." Economic Development and Cultural Change Vol. 34, No. 2 (1986), pp. 361-372. Brzoska, Michael and Herbert Wulf. "Rejoinder to Benoit's Growth and Defence in Developing Countries: Misleading Results and Questionable Methods." Working Paper. Study Group on Armament and Underdevelopment. University of Hamburg, 1979. Chan, Steve. "The Impact of Defence Spending on Economic Performance: A Survey of Evidence and Problems." ORBISVol. 29, No. 1 (1985), pp. 403-434. Chan, Steve. "Military Expenditures and Economic Performance," pp. 29-37 in US Arms Control and Disarmament Agency (US ADC A), World Military Expenditures and Arms Transfer 1986. Washington, DC: US ACDA, 1987. Deger, Saadet. "Economic Development and Defence Expenditure." Economic Development and Cultural Change Vol. 35, No. 1 (1986), pp. 179-195. Deger, Saadet and Somnath Sen. "Military Expenditure, Spin-off and Economic Development." Journal of Development Economics Vol.13, Nos. 1/2 (1983), pp. 67-83. Dunne, Paul. "Economic Effects of Military Expenditure in Development Countries: ASurvey,"pp.440-464inN.P. Gleditschet.al. (eds.). The Peace Dividend. Amsterdam: Elsevier Science, 1996. Dunne, Paul. "Economics of War and Peace." School of Economics Discussion Paper Series. No.34. Middlesex University Business School, London, 1997. Feder, Gershon. "On Export and Economic Growth." Journal of Development Economics Vol. 12 (February-April 1982), pp.59-73.
Sub-Saharan Africa: a supply side analysis
287
Frederiksen, Peter C. and Robert E. Looney. "Defense Expenditures and Economic Growth in Developing Countries." Armed Forces and Society Vol. 9, No. 4 (Summer 1983), pp. 633-645. Haruna, I.B.M. New Nigerian News Paper, 18 March 1978. Hewitt, Daniel F. "Military Expenditures: International Comparison of Trends."IMF WorkingPaper WP/91/54. Washington, DC: IMF,May 1991. International Institute for Strategic Studies. The Military Balance 1995/1996. London: IISS, 1995. International Institute for Strategic Studies. The Military Balance 1996/1997. London: IISS, 1996. Kaldor, Mary. "After the Cold War: Obstacles and Opportunities in Cutting Arms Budgets," pp. 141-156 in World Bank. Proceedings of the World Bank Annual Conference on Development Economics 1991. Washington, DC: World Bank, 1991. Lim, David "Another Look at Growth and Defence in Less Developed Countries." Economic Development and Cultural Change Vol. 31, No. 2 (1983), pp. 377-384. Microsoft. Encarta Encyclopaedia, 1997. Mohammed, A.l. Nadir. "Economic Growth and Defence Spending in Sub-Saharan African: Benoit and Joerding Revisited." Journal of African Economies Vol. 2, No. 3 (1993), pp. 145-156. Olaniyi, Oyinlola. "Nigeria's National Defence and Economic Development: An Impact Analysis." Scandinavian Journal of Development Alternatives Vol. 12, Nos. 2/3 (1993), pp. 241-252. Olaniyi, Oyinlola. "African Economic Integration and Nigeria's Industrialisation Challenges for the Defence Sector." Defence Studies [Journal of the Nigerian Defence Academy] Vol. 8 (1998), pp. 33-42. Ram, Rati. "Government Size and Economic Growth: A New Framework and Some Evidence from Cross-Section and Time-Series Data." American Economic Review Vol. 76, No. 1 (1986), pp. 191-203. Rothschild, K.W. "Military Expenditure, Exports and Growth." Kyklos Vol. 26, No. 4 (1973), pp. 804-814. Sandier, Todd and Keith Hartley. The Economics of Defence. Cambridge, UK: Cambridge University Press, 1995. Sivard, Ruth Ledger. World Military and Social Expenditure.
288 A rm ing the Sou th
Washington, DC: World Priorities, 1989. Ward, Michael D., David R. Davis, and Steve Chan. "Military Spending and Economic Growth in Taiwan." Armed Forces and Society Vol. 19, No. 4 (1993), pp. 533-550. World Bank. World Development Report 1995. New York: Oxford University Press, 1995. World Bank. World Development Report 1996. New York: Oxford University Press, 1996.
Sub-Saharan Africa: a supply side analysis
Appendix Table 12A.1: Sample countries and sectoral GDP distribution
Country
Agriculture
Industry
Services
Mozambique Tanzania Ethiopia Sierra Leone Uganda Malawi Chad Rwanda Madagascar Guinea Bissau Kenya Mali Niger Burkina Faso Nigeria Togo Gambia Zambia Central African Republic Benin Ghana Guinea Zimbabwe Ivory Coast Lesotho
33 56 60 38 53 39 44 41 34 45 29 42 39 n/a 34 49 28 34 50 36 48 24 15 37 10
12 14 10 16 12 18 22 21 14 19 18 15 18 n/a 43* 18 15 36* 14 13 16 31 36* 24 47*
55 30 29 46 35 43 35 38 52 36 54 42 44 n/a 24 35 58 30 36 51 36 45 48 39 43
Note: * Countries not included in the second sample. Source: World Bank (1995).
289
This page intentionally left blank
13 Greece: Military Expenditure, Economic Growth, and the Opportunity Cost of Defense Emmanuel Athanassiou, Christos Kollias, Efi Nikolaidou, and Stavros Zografakis
Introduction Assessing the impact of military expenditure on economic performance is an area that has been growing fast following Benoit's seminal contribution. It provided a strong impetus for research, both theoretical and applied, on the subject for the past quarter century or so. But no consensus appears to exist as to the nature and extent of the economic effects of defense spending (Dunne, 1996; Ram, 1995; Fontanel, 1990; Gleditsch et al, 1996). This chapter addresses the issue of the economic effects of military expenditure for the case of Greece. Greece, a member of NATO and the European Union, is a particularly suitable vehicle for empirical investigation since for many years it has been allocating a relatively high proportion of its national income to defense. Compared to other countries in NATO and the EU its defense burden (military expenditure as a share of GDP) is often twice as high. At the same time, the Greek economy has gone through periods of high economic growth as well as periods of
292 Arming the South
stagnation and fiscal imbalances. Without necessarily implying any correlation or indeed causation, it is interesting to note that in the mid- to late 1990s its growth rates were among the highest in the EU while its defense burden was twice as high as that of the EU average. In the sections that follow we address the impact of defense spending on the Greek economy using two distinctly different methodologies. The aim is twofold. First we want to examine the issue of the impact of military spending on the growth performance of the economy. The results obtained - namely that defense expenditure is largely neutral vis-a-vis growth - contradict earlier findings using the same method which found a negative relationship. Second, given (a) that the end of the cold war brought about significant reductions in the defense budgets of many countnes and (b) that Greek military spending did not exhibit similar downw ard trends, we want to estimate the effects on the economy had the reduction in defense spending been equal to the NATO average. This scenano is used to illustrate that the defense expenditure to growth relation is affected not only by changes in regime, but also by changes in the opportunity cost of defense. The structure of the chapter is as follows. The next section contains an overview of the main issues associated with the Greek economy and its defense spending. The following section estimates the defense-growth relation for 1960-1996 by employing an augmented Feder-type model consisting of civilian, military, government, and export sectors. Another section uses a Computable General Equilibrium (CGE) model to quantify the forgone benefits had Greece's defense spending been equal to NATO average military spending. A final section summarizes and concludes the chapter. Defense and the economy in Greece With per capita GDP in 1998 of $14,143 Greece, in terms of this traditional indicator of development, is placed among the developed nations of western Europe, albeit in the lower half of the group. It is in fact one of the poorest members of NATO and the European Union. The country is situated in a strategically volatile region at the crossroads of three continents. The Balkans have traditionally been an area of friction,
Greece: the opportunity cost of defense
293
tension, and conflict, and with the collapse of US-Soviet bipolarity the region entered a period of protracted instability as events over the past decade indicate (Glenny, 1995; Larrabee, 1992). In its pursuit of national security, Greece has over the years allocated substantial human and material resources to defense. In comparative terms, Greece is the most militarized country in NATO and the EU (Kollias, 1995a). Expressed as a share of GDP, Greek military spending has invariably been higher than the EU and NATO averages. In fact, over the last decade or so, the Greek military expenditure to GDP ratio has been about twice as that of the EU and NATO averages. For example, in the period 1985-1997 Greece on average allocated 5.1 percent of GDP to defense while the conesponding averages for NATO and the EU were 2.9 percent and 2.4 percent respectively (see figure 13.1). Furthermore, in a period when most defense budgets have been shrinking in real terms, Greek defense spending has increased. For example, according to SIPRI data, in the ten-year period 1989-1998 military spending in Greece increased by about 24.2 percent from $5,001million in 1989 to $6,211 million in 1998 (constant prices). By comparison, total NATO spending fell by about 26.3 percent from $601 billion to $443 billion, and total EU defense spending fell from $209 billion to $183 billion, a reduction of about 12.4 percent for the same time period (constant prices). In many respects, when compared to other EU and NATO members, Greek security concerns are unique and are reflected in the level of resources - both human and material - the country yearly allocates to defense. The Greek defense effort that the various indices reflect cannot be explained only in terms of the broader western security priorities as they have evolved during and after the bipolar era. Thus, during the cold war the country's external security concerns were not only the WTO countries but also its neighbor and NATO ally, Turkey. In fact, Turkey has long been regarded as the main and most imminent source of external threat to Greek sovereignty and national interests. The Greek-Turkish conflict is well documented in the international relations literature (Kurop, 1998; Gurel, 1993; Constas, 1991; Lanabee, 1992). It has also attracted considerable attention in the defense economics literature. A number of studies that have estimated demand functions for Greek military expenditure have reported results indicating that the ongoing
294 Arming the South l
I Greece
6
NATO • European Union
5 4 3 2 10-H 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997
Figure 13.1: Military expenditure as a share of GDP in Greece, NATO, and the EU Source: SIPRI Yearbooks.
conflict with Turkey appears to be an important determinant of Greek military expenditure (Avramides, 1997; Kollias, 1996; 1995a; Kapopoulos and Lazaretou, 1993). Using causality analysis, other studies have empirically investigated the hypothesis of a Greek-Turkish arms race. The findings are mixed. Depending on the analytic method, the data, and the time period of the study, unidirectional, bidirectional, and no causality have been reported (Majeski, 1985; Majeski and Jones, 1981; Georgiou, 1990; Stavrinos, 1992; Georgiou et al., 1996; Kollias and Makrydakis, 1997; Dunne et al, 2001; Smith et al, 2000). Given the well-documented ongoing Greek-Turkish conflict as well as the broader security environment of the region, Greece is forced to allocate a substantial part of its national income to defense at a period when the defense budgets of many European countries are shrinking. This undoubtedly hinders Greece's efforts to achieve economic convergence with other EU members. In fact, Greece was the only member of the EU that wanted but was not eligible to join the Euro currency area. It hopes
Greece: the opportunity cost of defense
295
12 10
1S54
1958
1962
1966
1970
1974
1978
1982
1986
1&)
1994
-•-Milex/CBP —•— GDPgcvtfh rates
Figure 13.2: Military expenditure as a share of GDP and GDP growth rates Source: SIPRI Yearbooks and Greek national accounts.
to do so in 2000 when its application will be reassessed. On a broader level, despite impressive rates of growth during the first two and a half post- WWII decades (see figure 13.2), since the mid-1970s the Greek economy has been facing serious structural and fiscal problems (Alogoskoufis, 1995; Jouganatos, 1992). A number of successive stabilization programs have set the country on the road to economic recovery only in the last few years, correcting its fiscal imbalances and reducing the rate of inflation. Currently the economy is exhibiting relatively high growth rates of 3.2 percent in 1996/97 and 3 percent in 1997/98 compared to EU averages of 2.7 percent and 2.9 percent for the same periods. The economic effects of Greece's relatively high defense burden is a subject that has been attracting attention in the relevant literature. The studies that have empirically evaluated and quantified these effects are not conclusive. Again, depending on the methodology used - adopting single-equation or multi-equation models - and the time period covered, the reported findings are mixed, as Brauer (2001) notes in a
296 Arming the South
comprehensive survey of the studies. A number of them have shown that growth has been retarded through various channels while others have reported stimulative effects through aggregate demand generation (Kollias and Makrydakis, 2000; Antonakis, 1997, 1997a, 1999; Chletsos and Kollias, 1995; Kollias, 1994, 1995; Dunne and Nikolaidou, 2001; Dunne et al, 2001; Nikolaidou, 1999). Studies have also estimated the potential benefits from hypothesized defense spending reductions (Balfousias and Stavrinos, 1996; Kollias and Refenes, 1996). The following sections address the impact of defense spending on the Greek economy using two very different methods. First, the defensegrowth relation is estimated for the 1960-1996 period by employing an augmented Feder-type model consisting of civilian, military, government, and export sectors. Following that, a CGE model is used to estimate the forgone benefits to the economy had Greece's defense spending exhibited a downward trend to equal the NATO average. Military spending and growth: a multi-sectoral analysis In this part of the chapter we investigate the defense-growth relation in Greece over the 1960-1996 period by employing a commonly used supply-side model, the so-called Feder-type model. Although the Federtype model can be employed in either the "overall" form or the "augmented" form, in which case externality effects and productivity differences of each sector are separated from the total (overall) effects, this is not the concern of the present chapter.1 Instead, we account for as many economic linkages as possible by decomposing the economy in four sectors - the civilian, the military, the government, and the export sectors - while the more common approach is to decompose the economy merely in two sectors - the civilian and the military. The aim is to examine the sensitivity of the model to the inclusion of the extra sectors. Model specification Supply-side models for the defense-growth relation have developed from Biswas and Ram (1986), who adopted Feder's (1982) model on the role of exports in economic growth, as a neoclassical two-sector framework
Greece: the opportunity cost of defense
297
(military and civilian) to assess the externality effect of the military sector and the factor productivity variation between the two sectors. Since then, many versions of the Feder model have been developed (i.e., assuming different sets of externalities or more sectors) with most of the studies employing cross-sectional methodologies.2 The form of the model used here assumes that the economy consists of four mutually exclusive and exhaustive sectors - the civilian sector (C), the non-military government sector (G), the export sector (X), and the military sector (M) - so that total output of the economy is the sum of the civilian output, the nonmilitary government output, the export output, and the military output. (13.1)
Y = C + G + X + M.
Capital and labor are allocated among the four sectors at each point in time. (13.2a)
K = Kc + KG + Kx +KM and
(13.2b)
L = LC + LG + LX + LM
where uppercase subscripts denote the civilian sector (C), the nonmilitary government sector (G), the export sector (X), and the defense sector (M). Each of the M, G, and X sectors has an externality effect on the civilian (C) sector. For this approach, the production functions for the four sectors are (13.3a) (13.3b) (13.3c) (13.3d)
G = G(KG,LG)
M = M(K M ,L M ) X = X(KX, Lx) C = C ( K C , L C , G , X, M)
where subscripts C, M, G, X denote sectoral inputs. Allowing for relative productivity differences between the "base" sector (civilian) and the other three sectors, the ratios of the marginal productivities for the sectors are
298 Arming the South
(13.4a) (13.4b) (13.4c)
M L /C L = M K /C K = (1+S m ) G L /C L = G K /C K =(l+8 g ) XL/CL = X K /C K = (1+SX)
where the uppercase subscripts on M, G, X, and C denote partial derivatives (or marginal products) of labor and capital (e.g., ML=SM/8Lm and MK = 5M/8KJ. Also, the size of M, G, and X may act as "externality" factors for the civilian sector (C). In other words, the model also identifies marginal externality effects of each of the three sectors (M, G, X) on the civilian sector (C). So, we will have: (13.5a) (13.5b) (13.5c)
'1+5„)C K X K =(1+5 X )C K MK = ( l + 6 m ) C K G*
and and and
GL = ( l + 5 g ) C L X L =(1 +5x)C L ML-(l+5m)CL
where 8{ is the relative factor productivity differential between the "base" sector and the other three sectors, respectively. If for example the productivity index for defense, 5m , is positive then the defense sector is more productive than the civilian sector. A zero value for 5m would indicate the absence of a productivity difference, while a negative value for 8m would indicate that the civilian sector is more productive. Due to the unavailability of sectoral input data, the model is reformulated in terms of aggregate inputs. The equation for this approach that gives the overall effect can be derived by manipulating the production functions (Feder, 1982; Ram, 1986):
(13.6)
Y = a - + PL + Y (
*x ^ v l + SXJ
+ cM M
(—1 V YJ
where a is the marginal product of capital in sector C (civilian), (3 is an elasticity measure equal to CL(L/Y), and [(8/1+8;) +Q] is the sum of the
Greece: the opportunity cost of defense
299
externality and factor productivity differences (the overall effect of sector i on economic performance). The first part by itself, (5/1+5;), is the relative productivity effect of the ith sector on economic growth, while the second part, Ci9 represents the marginal externality effect of the ith sector on the civilian sector (Huang and Mintz, 1991). Adding an intercept and a disturbance term gives the equation to be estimated for Greece over the period 1960-1996 to get the size (total) effect of each of the sectors on economic growth. Note that in order to estimate equation 13.6 we replace variables' instantaneous rate of change with their discrete equivalents, i.e., Y =AY/Y_!. The estimated four-sector Feder-type model for Greece as specified above, defines simple externalities from each of the defense, export, and non-military government sectors only on the base sector (civilian), derives values for non-military government spending by deducting military spending from total government spending, and provides estimates for the two, three, and four-sector models in order to compare the results each time a sector is added. As such, it overcomes a number of shortcomings associated with this type of model. 3 Data and empirical findings The military expenditure data used here is drawn from SEPRI Yearbooks. Usually labor force data is not available, and its growth is proxied by the population growth rate (Ram, 1986; Ward et al, 1991; Alexander, 1990). But for Greece, labor force data was in fact available from 1970 onwards, with labor force growth proxied by population growth prior to 1970. Data for GDP, investment, government expenditure, exports, and labor force were taken from the OECD database. All figures were first deflated in constant 1990 million drachmas and then converted to 1990 million US$ by means of exchange rates.4 To measure non-military government spending military expenditure was subtracted from government expenditure. This overcomes the problem exemplified by Alexander (1990) where government consumption was used as a proxy for nonmilitary government consumption leading to an overvaluing of government consumption by the amount of military expenditure (the variables are described in the appendix 13A.1). Table 13.1 shows the results for the total (overall) effects of each
300 Arming the South
sector (equation 13.6). The first numeric column gives results for the twosector model (military and civilian), the second column for the threesector model (civilian, military, and government), and the third column gives results for the four-sector model (civilian, military, government, and export). The overall performance of the model in terms of explanatory power is not very satisfactory with R2 being 0.41, 0.47, 0.48 for the two, three, and four sector models, respectively. But another consideration when evaluating the overall performance of this model concerns the coefficients on the investment and labor variables. As Mintz and Stevenson (1995) point out: "In general, one would be more confident in the specification of the model if the coefficients on these variables conform to the standard predictions of the economic theory." If for example investment is found significantly negatively related to economic growth, the validity of the model should be questioned. But this is not the case for any of the three specifications of the model for Greece. In all our estimations, investment is positive and highly significant. But labor force growth has an unexpected, statistically significant negative effect which is problematic as this does not conform to standard predictions of economic theory (although the theories underlying the impact of labor on the economy are less conclusive than those of investment). The negative statistical significance of the labor variable might suggest that, in Greece, increases in the workforce do not necessarily imply a more productive workforce. In fact, this is not an unusual finding as Ward et al. (1991) argue. Furthermore, Antonakis (1997) in his two-sector model for Greece found a negative but insignificant effect for the labor force variable5 justifying it on the grounds that in labor-surplus economies like Greece the natural rate of growth is not a binding constraint. Concerning the total effect of the military sector (in the two sector model), it is positive and significant at 10 percent supporting the modernization and spin- off arguments for defense spending. In the second column of table 13.1, the government sector (excluding the military) enters the equation with a positive sign, significant at 10 percent. As for the effect of the military sector, it is still positive but now insignificant. The intercept, investment, and labor force growth continue to have the same signs as before with their significance levels slightly
Greece: the opportunity cost of defense
301
Table 13.1: Estimation results, Greece, 1961-1996 Dependent variable: GDP growth Independent variable
Eq. (13.6)
Eq. (13.6)
Eq. (13.6)
Intercept I/Y, AL/L, AM/M.,(M/Y,) AG/G.^G/Y,) AM/X ,(X/Y.,)
-0.03 (-1.23) 0.33(3.06)*** -0.92 (-2.50)** 1.59(1.84)*
-0.03 (-1.15) 0.27(2.51)** -0.95 (-2.71)*** 0.43(0.42) 1.04 (1.92)*
-0.03 (-1.29) 0.26(2.39)** -0.83 (-2.2V ** 0.37(0.36) 1.14 (2.05)** 0.49 (0.90)
R2 SE DW F-statistic
0.41 0.03 1.66 F(3,32) = 7.3***
0.47 0.02 1.74 F(4,31)=6.85***
0.48 0.03 1.54 F(5,30) = 5.61***
Serial con-elation Functional form Normality Heteroskedasticity
£ y; £ £
x2 ( 0 = 0.03 (0.861) %2 (1) = 0.63 (0.427) f (1) = 0.05 (0.830) £ (1) = 0.05 (0.825) f (2) = 10.94 (0.004) £ (2) = 8.80 (0.012) £ (1) = 0.18 (0.675) £ (1) = 0.45 (0.501)
(1) = 0.06 (0.801) (1) = 0.01 (0.918) (2) = 9.74 (0.008) (1) = 1.08 (0.299)
Notes: The top rows give the coefficient estimates followed by the t-ratios (in parentheses); the bottom rows give the X2 tests for serial correlation, functional form, normality, and heteroskedasticity, followed by the probabilities (in parentheses). ***: significant at the 1 percent level of significance,**: significant at the 5 percent level of significance, *: significant at the 10 percent level of significance. For all estimations Microfit 4.0 was used.
altered. Finally, by adding the export sector (column 3), all of the variables' signs remain the same, the significance of the government sector increases (from 10 percent to 5 percent) and the significance of investment and labor force growth drops slightly to 5 percent. The effect of the export sector is positive but not significant, which is not surprising for a country like Greece that mainly exports agricultural products. As for the constant term, which measures an average trend rate of technological progress, it is insignificant in all three specifications. Non-nested tests indicated that the three and four-sector models are preferced to the two-sector one, while no clear-cut preference could be
302 Arming the South
made between the three and four-sector models (Akaike's information criterion favored the four-sector model, while the Schwarz Bayesian criterion favored of three-sector one). Given this, plus the fact that results are very similar for the three and four-sector models, reliance on either model is acceptable. It would appear that the military sector in Greece does not have a significant impact on economic growth, and that the same applies to the export sector. Only the non-military government sector seems to be growth promoting in Greece. This is in contrast with earlier findings (e.g., Antonakis, 1997, reports a strong negative effect) although this may be due to differences in the time periods used, or the difference in outcome may be due to a change in regime and a resulting change in the opportunity cost of defense (Murdoch et al. 1997). Change of regime and the opportunity cost of defense As pointed out, the results presented thus far conflict with earlier findings using the same methodology. It has been suggested that the relation between public defense, public non-defense, and non-public expenditure on the one hand, and growth on the other, may be affected by changes in regime, that will in turn affect the opportunity cost when undertaking different types of expenditure. In this section we illustrate this possibility through the use of a CGE model, trying to capture what the change of regime represented by the end of the cold war would have had on the Greek economy, had no other threat to security existed. In order to circumvent the numerous problems associated with public good pricing, we make the simplifying assumption that decisions are taken in such a way so that the change in the slope of the production possibilities frontier reflects the relative marginal costs in undertaking a particular pattern of expenditure. Thus a measure of the opportunity cost would be the difference in the growth rate owing to a change in the pattern of exogenous expenditure. In this part of the chapter, we attempt an analysis of the structural incidence of a hypothetical reduction of Greece's current defense spending, as opposed to investment and labor expenditure, to the average NATO level. Furthermore we obtain results for alternative substitute expenditure patterns. Given that there has been practically no reduction in the level of current defense expenditure in
Greece: the opportunity cost of defense
303
Greece since 1988, the calculations consist of producing a set of opportunity costs cum regime change effects, reflecting the particularity of Greek defense needs. The tool of analysis is a Computable General Equilibrium (CGE) (Zografakis, 1997; Sams and Zografakis, 1997; Dervis et al, 1982) which incorporates the national accounting relationships of the 1988 Social Accounting Matrix (SAM) for Greece. The choice of base year for the analysis, although dictated by the availability of data, is fortuitous since it corresponds to the beginning of the dissolution of the command economies of eastern Europe, and the resultant defusion of East-West tensions. The SAM framework permits the inclusion in the analysis of multiplier effects that work through the distribution of income as well as through transfers among institutions. This is particularly important for countries like Greece where the bulk of armaments procurement is of foreign provenance. Thus the direct structural effect of such spending is minimal. The SAM adopted here disaggregates the national accounts into fifteen sectors (see Athanassiou et al, 1998). The labor input is divided into four skill categories: agricultural, unskilled, skilled, and highly skilled. There are two classes of employment, salaried and self-employed. Labor is assumed to be perfectly mobile among sectors of economic activity. Capital is sector-specific and, due to the comparative-static nature of the analysis, is kept fixed at the initial level. The government sector, net of health and education expenditure, was broken into two components - defense and non-defense - in line with our analytic objectives . The model comprises the real sectors of the economy only, excluding the financial sector. Thus interest rates enter exogenously into the analysis. Three types of markets regulate production and exchange for goods, inputs, and the foreign sector - so that prices and quantities are endogenous to the model. Commodities offered in the domestic market are composites of domestically produced and imported commodities (see appendix 13 A.2a). Aggregation is made through use of a CES function, the domestic and imported goods being considered imperfect substitutes according to the Armington specification (Armington, 1969). The same specification is adopted in order to determine the composition of production between exports and domestically offered goods.
304 Arming the South
Domestic prices and wages are determined endogenously by marketclearing equations. The exchange rate is adopted as the numeraire and thus set to unity. Prices of imports are taken as given, following the small country assumption. Labor is offered as a function of the real wage rate, private investment is determined by the return on capital, while public investment is a policy variable. Demand for commodities is formed by private consumption, investment in dwellings, government expenditure and investment, private investment, stock variations, and exports. These demand elements are allocated over the set of commodities. The production technology in the model is of the CES type. A multilayered nested CES format allows the consistent aggregation of the large number of disaggregated inputs by stages, per sector of economic activity, and by type of input (see appendix 13 A.2b). Thus, at the highest level of aggregation the production function accepts two inputs, a quantity of labor and intermediate goods, and a quantity of capital and energy, each being the result of aggregation over different types of sub-inputs. Three closure rules characterize the model and encompass the assumptions about the working of the economy. The first concerns employment and is of the neoclassical variety, thus full employment of labor is assumed. The rule for investment is of the Keynesian variety: the investment rate determines the savings rate. Finally, the external sector is dominated by exogenous transfers from abroad, reflecting the importance of EU and other non-market transfers to the Greek economy. In summary the model exhibits the following characteristics: 1
It explicitly considers market-clearing mechanisms and related price formation in the economy; prices are computed by the model as a result of supply and demand interactions in the markets in which economic agents are price takers.
2
Within markets cleared by prices that achieve global equilibrium, it separately formulates supply or demand behavior of economic agents that individually optimize their objectives.
3
The model exhibits a large degree of disaggregated detail concerning sectors, social groups, structural features, and policy-oriented
Greece: the opportunity cost of defense
305
instruments (e.g., taxation). The figures for defense spending were taken from the 1989 budget, reporting realized expenditures for 1988, and were subtracted from the government consumption figures in the original SAM matrix. Elasticities used in the model are based on estimates undertaken for the HERMES6 econometric model and GEM-E3 (Computable General Equilibrium Model for studying Economy-Energy-Environment Interactions for Europe).7 Simulation results The question we seek to answer is what would be the effects on the Greek economy if the reduction in current defense expenditure (i.e., excluding labor and investment expenditure) were equal to the average decrease for NATO countries over the period 1988-1996. Since current defense expenditure in Greece has remained practically constant for the period in question, the question is what the peace dividend would have been if a considerable reduction in this type of expenditure had taken place. We examined five counterfactual cases. First, current Greek defense expenditure was reduced by about 25 percent, the NATO average for the period under examination. Since the interest rate in our model is exogenously given, the private sector cannot compensate for the reduction in public spending, so this case can be viewed as an initial impact scenario. To better interpret the results of the other scenarios, we use this case as a benchmark to assess the influence of this type of defense spending on the structure of the economy. In the second simulation we consider the effect of a decrease in current defense spending by the same amount as above; this time, however, it is compensated by expenditure on public consumption in such a way as to leave the shares of the various non-military consumption items constant. Non-military public consumption is divided into three categories, education, health, and other government. The remaining three simulations consist of exactly compensating the reduction in military expenditure by increases in expenditure of each of these categories alone. Given that the composition and share in GDP of government current
306 Arming the South
expenditure in Greece has remained practically constant over the period in question, we are in essence considering a two-tier "what-if' situation. First, we estimate the peace dividend, as if Greece faced a level of national security similar to that of the average NATO country. Second, we examine alternative uses of this peace dividend - and hence the opport unity cost of defense spending - while always remaining within the constraint that current public expenditure remains constant. Table 13.2 presents the results. Column A shows the impact of a once-for-all reduction in current Greek defense spending to the NATO average level. GDP would decrease by about half a percentage point. All GDP components would decrease except exports, which would benefit. Consumer prices would decrease by close to 5 percent, a figure that may be compared to the actual increase of the CPI, ranging from 20 percent at the beginning to 10 percent at the end of the period. Real wages would drop by about a quarter of the drop in the CPI. The drop in real wages would be higher the higher the skill level of the work force. Employment would decrease, the lower skill categories decreasing relatively more. Comparing the four alternative cases, we note that if the compensation is such that the shares of the non-military expenditure categories are kept constant at their 1996 levels, the results fall in between the cases where the compensation is imputed to any one category by itself. Due to the homogeneity assumptions made in the construction of the model, this is not surprising. Generally speaking, expenditure on education seems to be the best alternative, followed by expenditure in "other government" expenditure. Expenditure on health is the worst alternative, and in many cases is worse than military expenditure itself. Of course both of these types of expenditure affect welfare by reducing important types of risk, and this is not taken into consideration by the static, non-stochastic framework adopted here. Compensating for the decrease in current Greek defense expenditure in a share-equivalent way would add 0.2 of a percentage point to GDP from 1996 onward. This is about 10 percent above the actual growth rate for that year. GDP growth could have increased by 0.26 percentage points if the compensation were accomplished through increases in education spending alone, or have resulted in a 0.1 percentage point decrease if health alone took up the slack. Expenditure in other government services
Greece: the opportunity cost of defense
307
Table 13.2: Greece: Effects on national accounting aggregates (percent deviation from reference)
GDP f. p. Private investment Total exports Total imports Private consumption Consumer price index Real wage rate Employment ('000s)
A Defense -24.9 percent
B Defense -24.9 percent Gov. cons. (non-def.) +7.16percen1
C D Defense Defense -24.9 percent -24.9 percent Gov. health Gov. educ. +26.4 percent: +40.05 percent
E Defense -24.9 percent Other gov. cons.
-0.45 -1.51
0.19 0.67
0.26 0.84
-0.1 1.48
0.24 0.39
0.81 -3.27 -0.53
-0.4 0.39 0.37
-0.45 0.53 0.50
-0.81 1.75 0.43
-0.29 0.02 0.33
-4.71
2.59
2.87
5.74
1.79
-1.07 -14.67
1.06 11.55
1.17 13.29
1.28 12.73
1.01 11.16
would have resulted in almost as high a gain in GDP growth as education, but with a lower increase in private consumption and private investment. An increase in GDP similar to that of education is achieved by the relatively smaller impact that this combination of expenditure would have had on exports and imports. This scenario is also kindest to price increases but results in the lowest increase in real wages. Expenditure on health comes off badly in growth-enhancing terms due to the increased import requirements and the relatively substantial decrease in exports. Private investment would increase the growth rate the most, while the rise in private consumption would be slightly less than that seen in the case of education. The effect on prices would be considerable, while that on wages would also be greater than that of the case of education. A significant result is that current defense spending seems to be directed more toward the domestic economy than any other type of expenditure. It is the only category of government expense that is positively related to exports and negatively to imports. Since Greece is
308 Arming the South
not a major arms exporter, it would seem that current defense spending is directed to domestic goods whose factor requirements are complementary to exportables. Conclusion The results obtained from the application of the Feder-method to the case of Greece for the period 1960-1996 indicate that the effect of defense spending on growth is neutral. In contrast, non-defense public spending has a positive effect on growth. This is in contrast to earlier findings which showed that defense spending has a negative effect on growth. The change in the estimated coefficients may well be the result of a change in regime (or set of regimes) or be due to the different time-spans covered. While the difference in the results is interesting in itself, the policy implications do not change. A shift of expenditure out of defense and into non-defense public spending would seem to be beneficial for growth. Using a CGE model we examined such a case. In contrast to the two cases mentioned above we used data for defense expenditures excluding expenditures relating to weapon systems procurement. Our base year is 1988 which coincides with the beginning of the change in the security regime as represented by the collapse of the Warsaw Pact threat. The results indicate that there is a positive relationship between transfers of expenditure from defense to non-defense public expenditure and growth. For Greece, this supports the findings of the Feder-type applications. While transfers from defense to non-defense expenditure may be favourable to growth, it is less so than when the transfer is exclusively to general government expenditure, or expenditure on education. But when expenditure on health substitutes for defense spending the effect is negatively related to growth. It therefore follows that policy regimes, as represented by priorities among various types of public expenditure, would also be operative in determining the relation of defense expenditure to economic growth.
Greece: the opportunity cost of defense
309
Notes 1. For an empirical application of the "augmented" Feder-type model for Greece see Antonakis (1997, 1999) and Nikolaidou (1999). 2. For example Ward et al. (1991), Ward and Davis (1992), Huang and Mintz (1991), Mintz and Huang (1990), Atesoglou and Mueller (1990), Mueller and Atesoglou (1993), and Ward, Davis, and Chan (1993). 3. For a discussion of the shortcomings and advantages of the Feder-type model, see Nikolaidou (1999). 4. The conversion to US$ is done for reasons of comparability. Although this raises the issue of introducing into the data series exchange rate conversion problems, this effect, if present, is universal across all the series. 5. Antonakis (1997), as most other studies have done, proxied labor force growth by population growth. This can cause the impact of labor growth to be underestimated, especially in cases where the size of the labor force changes significantly while population remains stable, as is the case for Greece. 6. HERMES: Harmonised Econometric Research for Modelling Economic Systems. Edited by Commission of the European Communities, North-Holland, 1993. 7. The GEM-E 3 model was built under the auspices of the European Commission (DG-XII; co-ordinator P. Valette) by a consortium involving CORE, NTUA, KUL, University of Mannheim, University of Strathclyde, and CEA. References Alogoskoufis, G. 'The Two Faces of Janus: Institutions, Policy Regimes and Macroeconomic Performance in Greece." Economic Policy No. 20 (1995), pp. 149-192.
310 Arming the South
Alexander, R. "The Impact of Defence Spending on Economic Growth." Defence Economics Vol. 2, No. 1 (1990), pp. 39-55. Antonakis, N. "Guns Versus Butter: A Multisectoral Approach to Military Expenditure and Growth with Evidence from Greece, 19601993." Journal of Conflict Resolution Vol. 43, No. 4 (1999), pp. 501520. Antonakis, N. "Defence Spending and Growth in Greece: A Comment and Further Empirical Evidence." Applied Economics Letters Vol. 4 (1997a), pp. 651-655. Antonakis, N. "Military Expenditure and Economic growth in Greece 1960-90." Journal of Peace Research Vol. 34, No. 1 (1997), pp. 89100. Armington, P. "A Theory of Demand for Products Distinguished by Place of Production." IMF Staff Papers 16, No. 1 (1969), pp. 159-78. Atesoglou, H.S. and M.J. Mueller "Defence Spending and Economic Growth." Defence Economics Vol. 2, No.l (1990), pp. 19-28. Athanassiou, E., C. Kollias, and S. Zografakis. "A Computable General Equilibrium Estimation of the Opportunity Cost of Defence in the Case of Greece." Mimeograph, 1998. Avramides, C. "Alternative Models of Greek Defence Expenditure." Defence and Peace Economics Vol 8, No. 2 (1997), pp. 145-187. Balfousias, A. and V. Stavrinos. "The Greek Military Sector and Macroeconomic Effects of Military Spending in Greece," pp. 191 -213 in N.P. Gleditsch et al. (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Biswas, B. and R. Ram. "Military Expenditure and Economic Growth in LDCs: An Augmented Model and Further Evidence." Economic Development and Cultural Change Vol. 34, No. 2 (1986), pp. 361372. Brauer, J. "Turkey and Greece: A Comprehensive Review of the Defense Economics Literature," forthcoming in C. Kollias and G. Giinluk§enesen (eds.), Greece and Turkey in the 21st Century: Conflict or Cooperation? The Political Economy Perspective. Chletsos, M. and C. Kollias. "Defence Spending and Growth in Greece 1974-90: Some Preliminary Econometric Results." Applied Economics Vol. 27, No. 9 (1995), pp. 883-890.
Greece: the opportunity cost of defense
311
Constas, D. (ed.). The Greek-Turkish Conflict in the 1990s. London: Macmillan, 1991. Dervis, K., J. De Melo, and S. Robinson. General Equilibrium Models for Development Policy. Cambridge: Cambridge University Press, 1982. Dunne, P., E. Nikolaidou, and R. Smith. "Arms Race Models and Econometric Applications," forthcoming in P. Levine, S. Sen, and R. Smith (eds.). The Arms Trade, Security and Conflict. Amsterdam: Harwood Academic Publishers, 2001. Dunne, P. and E. Nikolaidou. "Military Spending and Economic Growth in Greece: A Demand and Supply Model, 1960-1996." Defence and Peace Economics, forthcoming 2001. Dunne, P., E. Nikolaidou, and D. Vougas. "Military Spending and Economic Growth in Greece and Turkey: A Causal Analysis, 19601996." Defence and Peace Economics, forthcoming 2001. Dunne, P. "Economic Effects of Military Expenditure in Developing Countries: A Survey," pp. 439-464 in N.P. Gleditsch et al (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Feder, G. "On Exports and Economic Growth." Journal of Development Economics Vol. 12, No. 1 (1982), pp. 59-73. Fontanel, J. "The Economic Effects of Military Expenditure in Third World Countries ."Journal of Peace Research Vol. 27, No. 3 (1990), pp. 461-466. Georgiou, G. "Is There an Arms Race between Greece and Turkey? Some Preliminary Econometric Results." Cyprus Journal of Economics Vol. 3, No. 1 (1990), pp. 58-73. Georgiou, G., P. Kapopoulos, and S. Lazaretou. "Modelling the GreekTurkish Rivalry: An Empirical Investigation of Defence Spending Dynamics." Journal of Peace Research Vol. 33, No. 2 (1996), pp. 229-239. Gleditsch, N.P., O. Bjerkholt, A. Cappelen, R. Smith, and P. Dunne (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Glenny, M. "Heading Off War in the Southern Balkans." Foreign Affairs Vol. 74, No. 3 (1995), pp. 98-108. Gurel, S. "Turkey and Greece: A Difficult Aegean Relationship," pp. 161-190 in A. Balkir and C. Williams (eds.). Turkey and Europe.
312 Arming the South
London: Pinter Publishers, 1993. Huang, C. and A. Mintz. "Defence Expenditures and Economic Growth: The Externality Effect." Defence Economics Vol. 3, No. 1 (1991), pp. 35-40. Jouganatos, G. The Development of the Greek Economy 1950-1991. Westport, CT: Greenwood Press, 1992. Kapopoulos, P. and S. Lazaretou. "Modelling the Demand for Greek Defence Expenditure: An Error Correction Approach." Cyprus Journal of Economics Vol. 6, No 1 (1993), pp. 73-86. Kollias, C. "The Greek-Turkish Conflict and Greek Military Expenditure 1960-92" Journal of Peace ResearchVol. 33, No 2 (1996), pp. 217228. Kollias, C. "Preliminary Findings on the Economic Effects of Greek Military Expenditure." Applied Economic Letters Vol. 2, No. 1 (1995), pp. 16-18. Kollias, C. "Country Survey VII: Military Spending in Greece." Defence and Peace Economics Vol. 6, No. 4 (1995a), pp. 305-319. Kollias, C. "The Economic Effects of Defence Spending in Greece 196390: Some Preliminary Econometric Findings." SPOUDAI Vol. 44, No. 3-4 (1994), pp. 114-130. Kollias, C. and S. Makrydakis. "Is there a Greek-Turkish Arms Race? Evidence from Cointegration and Causality Tests." Defence and Peace Economics Vol. 8, No. 4 (1997), pp. 335-379. Kollias, C. and S. Makrydakis. "A Note on the Causal Relationship between Defence Spending and Growth in Greece: 1955-93."Defence and Peace Economics Vol. 11, No. 2 (2000), pp. 173-184. Kollias, C. and A. Refenes. "Modelling the Effects of Defence Spending Reductions using Neural Networks: Evidence from Greece." Peace Economics, Peace Science and Public Policy Vol. 3, No. 1 (1996), pp 1-12. Kurop, M.C. "Greece and Turkey. Can they Mend Fences?" Foreign Affairs Vol. 77, No. 1 (1998), pp. 7-12. Larrabee, S.F. "Instability and Change in the Balkans." Survival Vol. 34, No. 2 (1992), pp. 31-49. Majeski, S. "Expectations and Arms Races." American Journal of Political Science Vol. 29, No. 2 (1985), pp. 217-245.
Greece: the opportunity cost of defense
313
Majeski, S. and D. Jones. "Arms Race Modeling: Causality Analysis and Model Specification. "Journal of Conflict Resolution Vol. 25, No. 2 (1981), pp. 259-288. Mintz, A. and R. Stevenson. "Defense Expenditures, Economic Growth, and the 'Peace Dividend': A Longitudinal Analysis of 103 Countries." Journal of Conflict Resolution Vol. 39, No. 2 (1995), pp. 283-305. Mintz, A. and C. Huang. "Defense Expenditures, Economic Growth and the Peace Dividend." American Political Science Review Vol. 84 (1990), pp. 1283-1293. Mueller, M.J. and H.S. Atesoglou. "Defence Spending, Technological Change, and Economic Growth in the United States." Defence Economics Vol. 4, No. 3 (1993), pp. 259-269. Murdoch J.C., C.R. Pi, andT. Sandier. "The Impact of Defence and NonDefence Public Spending on Growth in Asia and Latin America." Defence and Peace Economics Vol. 8, No. 2 (1997), pp. 205-225. Nikolaidou, E. "Military Spending and Economic Growth in Greece, A Multi Sector Analysis." Economics Discussion Paper, Middlesex University Business School, No. 72, May 1999. Ram, R. "Defense Expenditure and Economic Growth," pp. 251-273 in K. Hartley and T. Sandier (eds.). Handbook of Defence Economics. Amsterdam: Elsevier, 1995. Ram, R. "Government Size and Economic Growth: A New Framework and Some Evidence from Cross-Section and Time-Series Data." American Economic Review Vol. 76, No. 1 (1986), pp. 191-203. Sarris A.H. and S. Zografakis. "A Computable General Equilibrium Assessment of the Impact of Illegal Immigration on the Greek Economy." Journal of Population Economics Vol. 12, No. 1 (1999), pp. 155-182. Smith, R., M. Sola, and F. Spagnolo. "Regime-Switching and the Prisoners' Dilemma: A Model for the Greek-Turkish Arms Race." Journal of Peace Research, forthcoming, 2000. Stavrinos, V. "Defence Expenditure in Arms Competition: Modelling and Causality Analysis." Greek Economic Review Vol. 14, No. 1 (1992), pp. 115-128. Ward, M., D. Davis, and S. Chan. "Military Spending and Economic
314 Arming the South
Growth in Taiwan." Armed Forces and Society Vol. 19,No. 4(1993), pp. 533-550. Ward, M. and D. Davis. "Sizing up the Peace Dividend: Economic Growth and Military Spending in the US, 1948-1996." American Political Science Review Vol. 3 (1992), pp. 748-755. Ward, M., D. Davis, M. Penubarti, S. Rajmaira, and M. Cochran. "Country Survey I: Military Spending in India." Defence Economics Vol. 3, No. 1 (1991), pp. 41-63. Zografakis, S. "Economic Policy and Impacts on the Evolution of Income Distribution in Greece: An Analytical Approach Based on a Computable General Equilibrium Model." Unpublished Ph.D. Thesis, Department of Economics, University of Athens, 1997.
Greece: the opportunity cost of defense
315
Appendix 13A.1: Description of variables in the Feder-type model Y = Y/Y , = GDP growth I/Y = I/Y ! = Share of investment in GDP L = AL/ L_, = Labor force growth
M\ ~Tf\ = AM/ M.j ( M/Y.j) = Total effect of defense sector
= AG/ G_, ( G/Y^) = Total effect of non-military government sector
#
.
.(x) X\
= AX/ X , ( X/Y ,) = Total effect of export sector
.(c) M\
= AM/ M., (C/Y.,)= Externality effect of military sector
.(c) (j
• = AG/ G , ( C/Y ,) = Externality effect of non-military government sector
V YJ .(C) X\
where
= AX/ X., (C/Y.,) = Externality effect of export sector
C = Y - M in the two-sector model C = Y - M - G in the three-sector model C = Y - M - G - X in the four-sector model 8m= productivity difference of the military sector with respect to the civilian sector 5g= productivity difference of the government sector with respect to the civilian sector 5X= productivity difference of the export sector with respect to the civilian
316 Arming the South
and
sector Y - GDP in constant 1990 mn US$ M = Military expenditure in constant 1990 mn US$ G = Government expenditure (excluding military) in constant 1990 mn US$ X = Exports in constant 1990 mn US$ I = Private investment in 1990 mn US$ L = Labor force in '000s
Appendix 13A.2a: The structure of consumption Disposable Income
Investment in Dwellings
Saving
Consumption
expend, for cars Durable Goods expend, for domestic appliance f:
Non-Durable Goods
3E Food, beverages, tobacco
Products
L
Services
Energy Clothing, footwear
r>T
Medical care and health expenses
H>
Education and Culture
i Recreation, Entertainment ^ 1 Heating Systems
Others goc and servic
~^1 Communication Services ^ 1 Electricity Transport Services (purchased transport) ^
Vehicle (Petrol for cars)
Greece: the opportunity cost of defense
317
Appendix 13A.2b: The structure of production Production function product
Aggregate: Capital - Energy CES function
Aggregate: Labor - Materials CES function
i Capital
, I
Energy
Employed Labor CES aggregate function
skilled labor - Urban
Materials (Factor demand Interm. Consump. of non-Energy Products)
Aggregate Labor
Self Employed CES aggregate function
skilled labor - Urban
J
L-^J semi-skilled labor - Urbat]
"^j
unskilled labor - Urban |
_frj in agricultural sector
[
j
| semi-skilled labor - Urbat]
^ l
unskilled labor - Urban \
J in agricultural sector
This page intentionally left blank
14 A System Estimation of the Defense-Growth Relation in Turkey Jiilide Yildirim and Selami Sezgin
Introduction The effects of military expenditure on economic growth have been examined extensively in recent years. However there is no agreement regarding the question of how and in what respects military spending affects economic growth. Feder-Ram or Deger-type multi-equation models have been estimated for a number of country studies and/or groups of countries, using both time-series and cross-sectional data (for extensive literature reviews, see e.g., Dunne, 1996; Ram, 1995; Deger and Sen, 1995; and Sandier and Hartley, 1995). Empirical studies that employ multi-equation models are comprised of three to four equations such as for growth, the savings or investment ratio, the trade balance ratio, and the defense burden. These models are estimated using twostage least squares or three-stage least squares. Even though there are a variety of studies, the results concerning the military expenditureeconomic growth relationship are mixed. Among other things, the evident inconsistency of results may be due to the varying sample periods of the studies, the underlying economic models, and the econometric techniques employed. Although multi-equation modeling is popular in the defense economics literature, it also attracts a lot of criticism concerning the endogenous-exogenous division of variables and the assumption of zero
320 Arming the South
restrictions, which are imposed in order to achieve the identification of the model. Various shortcomings of multi-equation modeling can be overcome by employing vector autoregressive modeling (VAR). First, there is no a priori endogenous-exogenous division of the variables; second, no zero restrictions are imposed; and third, there is no strict econo mic theory within which the model is grounded. A VAR model also allows us to identify long-run and short-run dynamics of defense expenditure on economic growth. The study of the effect of Turkish defense expenditure on its economic growth has attracted a number of researchers in recent years (see table 14.1).1 This may be due to the fact that Turkey's defense burden is one of the highest in NATO, that it has a large armed force, and that its defense industry has shown important developments over the last decade. The studies used a variety of methods including Feder-type supply models, Deger-type demand and supply-side models, and Grangercausality tests. In this section, studies on the Turkish defense expenditure and economic growth relation are briefly reviewed. Sezgin (1997) analyzed the Turkish defense-growth relationship using a Feder-type model and found that defense spending has a positive impact on Turkish economic growth. However, with a fairly similar model, Ozsoy (2000) found no significant effect on Turkish economic growth. Yet again a positive effect was found by Sezgin in two further studies (1999; 2001). But two Granger-causality studies by Dunne et al and Sezgin (2000) found a negative relationship between defense spending and economic growth. Overall, the findings are inconclusive. The aim of this chapter is to contribute to the pool of empirical evidence by examining the Turkish defense-growth relation using vector autoregressive modeling (VAR) for the period of 1949-1994. This builds directly on Sezgin (2001) where the relationship was investigated with a demand-supply side (Deger-type) model employing 2SLS and 3SLS simultaneous equation methods. The remainder of the chapter is organized as follows. The next section contains a brief overview of VAR modeling. The section thereafter is devoted to the empirical analysis, and the final section concludes the chapter.
A system estimation of the defense-growth relation in Turkey
321
Table 14.1: Empirical studies on the Turkish defense-growth relationship Author
Model and sample
Remarks
Sezgin (1997)
Feder model with human capital, 1949-93
Effect of defense spending on growth is positive
Ozsoy (2000)
Feder model, 1950-1992
No significant effect of defense on economic growth
Sezgin (2000)
Granger causality, 1924-1996
Negative impact of defense expenditure on growth
Sezgin (1999)
Engle-Granger error correction 1975-1996
Positive effect on the long-run and no significant effect on the short-run
Sezgin (2001)
Deger Model 2SLS/3SLS, 1956-1994
Positive effect of defense on growth, no significant effect on savings and balance of trade
D u n n e d al. (2001)
Granger-Causality VAR, Negative impact of defense 1960-1996 expenditure on growth
Vector autoregressive modeling (VAR) The starting point of the modeling is the formulation of a general unrestricted vector autoregressive (VAR) model, which consists of regressing each current variable in the model on all variables lagged a certain number of times. This can be represented as
(14.1) yt = X n y J V y
+
*/ , where ex ~ z./.c/.(0,Q)
7=1
where yt and st are nxl vectors of observations on the current values of all the variables in the model and a vector of random errors, respectively.
322 Arming the South
Generally, equations in a VAR model are completed by an additional set of deterministic components, such as the intercept term, deterministic trends, and seasonal dummy variables. VAR models have recently been used extensively in macroeconomics, as they are assumed to perform better in forecasting than structural multiple equations. By including all variables of interest in the VAR, there is no need to make any assumptions about the values of the exogenous variables in the forecasting period.2 One substantial disadvantage of VAR modeling is the number of coefficients to be estimated and hence the limited degrees of freedom, especially in small samples. However, cointegration analysis allows for the reduction of the number of parameters to be estimated. Cointegration analysis investigates if we can find any stationary, linearly independent underlying relations among the nonstationary variables of the system. It allows precise formulation of testable economic hypotheses about the behavior of variables. In general, cointegration implies certain restrictions on VAR representation and enables the researcher to test economic hypotheses. Furthermore, this formulation provides an explicit classification into nonstationary and stationary components, which could be interpreted in terms of the dynamics of long-run and short-run effects. When the data yt are 1(1), that is integrated of order one, the VAR system can be reformulated as a vector error correction form: s-\
(14.2) A^ = X n > X - y
+
^ ( - , + £,
7=1
where n = 2mmJX n • - / . As the variables are cointegrated, the matrix n is of reduced rank: (14.3) n = a0' where a and p are nxr matrices of rank r, and P yt comprises r cointegrating 1(0) relations inducing the restricted 1(0) representation:
A system estimation of the defense-growth relation in Turkey 323 s-\
(14.4) A^^n^iaj/J'j/J+v Johansen and Juselius (1994) point out that the reduced form, as given by (14.1), (14.2), and (14.4), uniquely defines the probability distribution of the variables and in that sense qualifies as a statistical model for the data. Estimation results In order to investigate the effects of military expenditure on economic growth, a five equation VAR is considered, where the variables are GNP at 1985 prices (Y), real military spending (M), real savings (S), labor force (L), and the real balance of trade (B).3 Annual data is available for the time period 1949-1994. After allowing for lags, estimation is carried out for 1951-1994, using PcFiml version 9.00 (Doornik and Hendry, 1997). Prior to modeling, the relationships among the variables and their univariate time-series properties are established.4 The results of the augmented Dickey-Fuller tests indicate that all variables considered in the study qualify as 1(1). All dummy variables discussed below are included in the short-run dynamics. Furthermore, the trend is restricted to the longrun dynamics as otherwise it would induce a quadratic trend in levels, for which there is no evidence.5 But from an economic point of view the time trend may pick up the effects of other determinants of economic growth that are missing in the model. Additionally, two dummy variables are included in the model: (a) D75, which takes the value of one for 1974 and 1975, is employed to capture the effects of increased military spending due to the Cyprus conflict; (b) D88 takes the value of one for 1988 and is intended to reflect the possible effects of changes in Turkish economic policy in that year. The results of the specification F-tests, given in table 14.2, indicate that the reduction by 32 parameters for eliminating lags 3 and 4 are acceptable. The Schwarz and Hannan-Quinn criteria also indicate the selection of the two-lag system. Hence a two-lag system with two dummy variables, a trend, and a constant is selected as the final model.
324 Arming the South
Table 14.2: Turkey: specification tests Model
Laglength
Schwarz
HannanQuinn
Model Reduction
F-tests
1 2 3
4 3 2
-7.65 -8.21 -8.54
-10.82 -10.75 -1.38
l->2 2->3 1-+3
F(25,53) = 1.21 (0.27) F(25,72)= 1.91** (0.01) F(50,67) = 1.58 (0.03)
Note: p-values are in parentheses; ** denotes statistical significance at the 1 percent level.
The residual cross-correlation and lag-length statistics of this final model are given in table 14.3, where Fj (.,.) denotes an F-test for the hypothesis that all i-lag coefficients are zero. Table 14.3 indicates that there are large unexplained positive residual correlations among military expenditure, savings, labor force, and income; a large negative residual correlation between balance of trade and military expenditure; and a large positive residual correlation between the military expenditure and savings. Furthermore, the lag-length statistics reveal that the first lags of all variables and the second lag of military expenditure variables are significant, suggesting that a two-lag VAR model is appropriate for modeling. Table 14.4 records statistical information about the unrestricted VAR reported by PcFiml. F} (.,.) denotes the F-tests for the hypotheses of no serial correlation against serial autocorrelation up to order 2 (Far), no autoregressive conditional heteroskedasticity against a four lag alternative (Farch), no heteroskedasticity (Fhet), and a chi-square test for normality (x2); analogous vector tests are also given and these are indicated by superscript v. Although there is some indication of a problem of autocorrelation in the balance of trade equation as well as in the vector estimates, they are not significant at the 1 percent level. Furthermore, all other diagnostics are satisfactory. After the VAR model is adequately specified, cointegration in the five equation system is investigated using Johansen's procedure (1988; 1991). Table 14.5 gives the cointegration analysis. X denotes the eigenvalues, Tr and Max denote the associated maximum eigenvalue and trace statistics.
A system estimation of the defense-growth relation in Turkey
325
Table 14.3: Residual correlations and lag-length statistics for VAR (2) Residual Correlations
M S L B
Y 0.675 0.635 0.373 0.011
B
M 0.357 0.115 -0.179
S -0.053 0.143
-0.111
M 4.84** (0.00) 7.17* (0.00)
S 5.86** (0.00) 0.50 (0.76)
L 2.68* (0.04) 0.84 (0.76)
Lag Length and Dynamics
FS=I(5,26) Fs=2(5,26)
Y 2.94* (0.03) 2.35 (0.06)
B 8.96** (0.00) 1.55 (0.20)
Note. * and ** denote statistical significance at the 5 percent and 1 percent levels, respectively, and p-values are in parentheses.
Table 14.4: Goodness of fit and diagnostic test results Far (2,28) F a r e h (l,28) Fhet (22,7)
f(2) FV (50,76) Xvhet(330) XVnd(10)
Y 0.35 (0.70) 0.51 (0.47) 0.48 (0.90) 0.19 (0.90)
M 1.37 (0.26) 0.62 (0.43) 0.41 (0.94) 0.70 (0.70)
S 0.58 (0.56) 1.26 (0.26) 0.94 (0.57) 2.85 (0.24)
L 0.77 (0.46) 0.001 (0.96) 0.14 (0.99) 4.44 (0.10)
B VAR 0.06* (0.03)* 0.67 (0.41) 1.96 (0.18) 5.05 (0.08) 1.60* (0.03) 370.82 (0.06) 15.65 (0.11)
Note: * denotes statistical significance at the 5 percent level; p-values are in parentheses.
326 Arming the South
Table 14.5: Cointegration analysis r X Max Tr
1 2 0.71 0.59 55.14" 39.96** 129.1** 73.98**
3 0.33 17.76 34.02
4 0.22 11.34 16.26
5 0.10 4.91 4.91
Eigenvectors f3
Y
M S L B
Y
M
1.00 14.89 -1.06 0.79 0.12
-23.12 1.00 1.42 13.90 1.93
S
L -0.26 -0.34 -26.88 -4.29 1.00 -1.85 -4.73 1.00 -0.33 -6.41
B -4.63 59.58 -0.92 7.44 1.00
0.34 -0.01 -0.23 -0.12 0.08
0.008 -0.005 0.02 0.008 0.02
Trend 0.22 12.19 0.17 0.08 1.03
Adjustment Coefficients a Y
M S L B
-0.09 0.03 0.01 0.01 0.12
-0.01 -0.001 -0.003 0.0002 0.002
0.04 -0.003 0.07 -0.008 0.006
Table 14.5 formally supports the hypothesis that there are two cointegrating vectors. There are two relatively large and two relatively small eigenvalues. Thus the presence of two cointegrating vectors is assumed throughout the analysis. Thefirstcointegrating vector resembles a Deger-type growth equation with positive effects from military expenditure, labor force, savings, and trend. The second cointegrating vector can be interpreted as a military expenditure equation and indicates that the deviations of military expenditure from trend are positively related to savings and labor force, and negatively related to the balance of trade. Table 14.5 also presents the adjustment coefficients which can be interpreted as the weights with which the cointegration vectors enter the five equation system. The first column of the adjustment matrix a' = (-0.09, -0.01,0.34,0.04,0.008) can be interpreted as the weights with which the excess output enters into each of the four equations of the system.
A system estimation of the defense-growth relation in Turkey
327
In order to obtain unique cointegrating vectors, the identifying restrictions, which are determined by economic theory, should be imposed on the cointegrating vectors.6 First, weak exogeneity of (M, S, L, B) for the parameters of the income equation is considered. This requires that the first cointegrating vector does not appear in the short-run equations of (M, S, L, B) indicating that military expenditure, savings, labor, and the balance of trade do not react to disequilibrium in real income but still react to its lagged changes. This hypothesis is consistent with the data as the first column of the adjustment matrix has values close to zero and wrong signs, except for the adjustment coefficient of the income equation itself Furthermore, it is observed in table 14.5 that the elements of the fourth and the fifth rows of the adjustment matrix are quite small, suggesting the possible weak exogeneity of labor and savings for the long-run parameters in this model. Additionally, the examination of the eigenvectors and adjustment coefficients in the cointegration analysis indicates that the coefficients of savings and labor variables are quite close to each other in the income equation and military expenditure does not depend on income and trend. Accordingly, in addition to the weak exogeneity restrictions, the following identification restrictions are imposed: for the income equation savings and labor force have the same coefficient; for military expenditure equation, no effect from income and trend. These restrictions are accepted when jointly tested, yielding x2 (9) = 19.91 with p > 0.018, where the degrees of freedom equal the number of over-identifying restrictions. As a consequence of the above analysis two cointegrating vectors are determined, indicating that there are two stationary relationships among the five non-stationary variables of the system. The restricted coingtegrating vectors are defined by CI1 = Yt-7.73*Mt - 0.79*S -0.79*Lt -2.39B, -0.325*Trend, CI2 =Mt -0.07*S - 0.09*L + 0.18*Bt. The long-run income is positively related to all variables in the system. This supports the findings of Sezgin (1999; 2001 ), i.e., that military expenditure positively affects income in Turkey. It is obvious that savings
328 Arming the South
(investment) and labor force are positively related to economic growth, which is standard from any basic growth model. The positive sign for defense expenditure suggests that Turkish defense spending helps its economic growth. The findings are consistent with Sezgin (2001) who used a Deger-type model. The positive sign for the balance of trade equation is not expected because a trade balance deficit implies that net capital inflows from abroad might stimulate economic growth. This suggests that capital inflows from abroad did not help Turkish economic growth in the time-period under investigation. The second cointegrating vector measures long-term military expenditure and depends positively on labor and savings but with a negative effect from the trade balance. The findings imply that labor, savings, and the balance of trade should be considered as explanatory variables for determining defense expenditure. The starting point of the second stage of the analysis is to model changes in the variables of the system as a response to departures from the long-run relationships, augmented by the short-run dynamics generated by the current and lagged first differences of the variables included in the model. The 1(0) system determines seven variables (AYt, AMt, AS{, ALt, ABP CIl t , and CI2t). The information set used consists of the first lags of AYt, AMt, ASt, ALt, and ABt, together with CIlt_„ CI2 tl , two dummy variables, and an intercept, entailing a reduction to parsimonious VAR. It is observed that none of the variables in the system contributes to the explanation of the labor force and the balance of trade equations. Hence these variables are assumed to be exogenous and a three equation system is estimated. In order to enhance interpretability and to reduce its sample dependence, the system is modeled as an 1(0) parsimonious vector autoregression (PVAR). The weak exogeneity of Mt, St Lt, and Bt for the parameters of the first cointegration relation is examined first. Then the conditional factorization is performed. The resulting VAR equations are estimated by FML and are given in table 14.6. Since the test of overidentifying restrictions does not reject at the 10 percent level - yjox (21) - 34.34, p > 0.300 - the model parsimoniously encompasses the PVAR. The short-run estimates of the model in table 14.6 indicate that the changes in income are positively affected by changes in military expenditure, labor, and the current value of savings, while the lagged value of savings has a negative impact on
A system estimation of the defense-growth relation in Turkey
329
differenced income. Thus defense spending in Turkey helps economic growth in the short-run as well as in the long-run. Furthermore, the shortrun income function has an adjustment coefficient of 0.36, which indicates that 36 percent of disequilibrium is corrected in each year. The short-run military expenditure in Turkey appears to depend positively on its lagged values as well as on the lagged value of growth of the balance of trade. However, the lagged value of savings has a negative impact on military expenditure in the short-run. Moreover, the adjustment coefficient is quite high with a strong effect. The short-run savings equation, on the other hand, is positively related to lagged values of labor force. Additionally, the model diagnostic statistics, given in table 14.6, are all insignificant at the 1 percent level of significance, matching the valid reduction from the parsimonious VAR. Conclusions Employing a vector autoregressive model (VAR), this chapter has provided an empirical analysis of the defense expenditure-economic growth relationship in Turkey for the period 1949-1994. The VAR estimates of the initial system revealed that there are two cointegrating relationships in the long-run: the first one measures income as a function of military expenditure, savings, labor, and the balance of trade; the second one measures the deviations of military expenditure from its longrun equilibrium level. In the short-run structure, a three equation system is estimated by FIML. The empirical findings indicate that weak exogeneity of income and military expenditure for the long-run parameters is satisfied. Therefore, income and military expenditure can be regarded as being exogeneously given for the long-run structure. Military expenditure appears to enhance economic growth both in the short-run and in the long-run, supporting the findings of Sezgin (2001) who reports a positive effect of defense expenditure on income. The analysis provided an adjustment coefficient of 0.36, indicating that 36 percent of disequilibrium in income is eliminated every year. Furthermore, military expenditure is highly affected by its lagged values and the yearly disequilibrium is corrected quite quickly. Overall, even though there is an argument against high
330 Arming the South
Table 14.6: FIML model estimates AYt = -0.36 + 2.17 AMt + 1.24 ALt + 1.19 ASt -0.41 ASt_, - 0.36 (CI1) t., (-6.48) (2.63) (7.68) (5.10) (-3.81) (-7.67) AMt = -0.80 + 0.53 AM M - 0.08 AS M + 0.11 AB M - 0.87 (CI2) M + 0.20 (D75) (-7.59) (5.28) (-5.16) (4.72) (-7.59) (5.10) AS t = 0.59 AL,, + 1.02(D75) + 1.17(D88) (2.06) (3.06) (2.93) Model statistics ¥\ (18,88) Fvhet (132,71) X2nd(6)
0.93 (0.54) 1.33 (0.08) 3.11* (0.79)
Note: t- ratios are in parentheses.
military expenditures for developing countries as scarce resources are allocated to unproductive projects, it appears that for Turkey this is not the case. Increased military expenditure leads to economic growth in Turkey. This may be due to the fact that Turkey has one of the biggest armies in NATO, and a major part of military spending is allocated to payments of salaries to armed forces. This may induce demand increases, thus causing economic growth. Notes 1. For a detailed literature review on the Turkish defense-growth relation, see Brauer (forthcoming). 2. For a brief discussion see Charemza and Deadman (1997). 3. For the variable sources and descriptions, see Appendix 14A.1. 4. The results of the unit root tests are given in Appendix 14A.2.
A system estimation of the defense-growth relation in Turkey
331
5. See Hendry (1995) for a discussion of inclusion of the trend in the cointegrating space. 6. For a detailed analysis of imposition of restrictions on adjustment coefficients and eigenvectors, see Yildirim (1997). References Brauer, J. "Survey and Review of the Defense Economics Literature on Greece and Turkey: What Have We Learned?" Defence and Peace Economics. Special Issue on Economic Aspects of Defence in Turkey and Greece, (forthcoming). Charemza, W.W. and F.D. Deadman. New Directions in Econometric Practices. Aldershot, UK: Edward Elgar, 1997. Deger, S. and S. Sen. "Military Expenditures and Third World Countries," chapter 11 in Keith Hartley and Todd Sandier (eds.), Handbook of Defence Economics. Amsterdam: Elsevier Science, 1995. Doornik, J.A. and D.F. Hendry. Modeling Dynamic Systems Using PcFiml 9.0 for Windows. London: International Thomson Business Press, 1997. Dunne, J. P. "Economic Effect of Military Expenditure in Developing Countries: A Survey," chapter 23 in N.P. Gleditsch, O. Bjerkholt, A. Cappelan, R.P. Smith, and J.P. Dunne (eds.), The Peace Dividend. Amsterdam: Elsevier Science, 1996. Dunne, J.P., E. Nikolaiodu, and D. Voguas. "Defence Spending and Economic Growth: A Causal Analysis for Greece and Turkey." Defence and Peace Economics. Special Issue on Economic Aspects of Defence in Turkey and Greece, (forthcoming, 2001). Hendry, D.F. Dynamic Econometrics Advanced Texts in Econometrics. Oxford: Oxford University Press. 1995. Johansen, S. "Statistical Analysis of Cointegrating Vectors." Journal of Economic Dynamics and Control Vol. 112 (1988), pp. 231-254. Johansen, S. "Estimation and Hypothesis Testing of Cointegration Vectors in Gausian Vector Autoregressive Models." Econometrica Vol. 59(1991), pp. 1551-1580.
332 A rming the South
Johansen, S. and K. Juselius. "Identification of the Long-Run and ShortRun Structure: An Application to the ISLM Model." Journal of Econometrics Vol. 63 (1994), pp. 7-36. IMF (International Monetary Fund). International Finance Statistics Yearbooks. Washington, DC: IMF (various years). OECD. Labor Force Statistics. Paris: OECD Department of Economics and Statistics (various years). Ozsoy, O. "The Defence Growth Relation: Evidence from Turkey," pp. 139-159 in Jurgen Brauer and Keith Hartley (eds.), The Economics of Regional Security: NATO, the Mediterranean, Southern Africa. Amsterdam: Harwood Academic Publishers, 2000. Ram, R. "Defence Expenditure and Growth," chapter 10 in Keith Hartley and Todd Sandier (eds.), Handbook of Defence Economics. Amsterdam: Elsevier Science, 1995. Sandier, T. and K. Hartley. The Economics of Defense. Cambridge: Cambridge University Press, 1995. Sezgin, S. "Country Survey X: Defence Spending in Turkey." Defence and Peace Economics Vol. 8, No. 4 (1997), pp. 381-409. Sezgin, S. "Defence Expenditure and Economic Growth in Turkey and Greece: A Disaggregated Analysis." Paper presented at "The Arms Trade, Security, and Conflict" conference at Middlesex University Business School, London, June 11-12, 1999. Sezgin, S. "A Causal Analysis of Turkish Defence Growth Relationships: 1924-1996. "Ankara University Journal of Political Sciences Vol. 55, No. 3 (forthcoming, 2000). Sezgin, S. "An Empirical Analysis of Turkey's Defence-Growth Relationships with a Multi-Equation Model (1956-1994)." Defence and Peace Economics. Special Issue on Economic Aspects of Defence in Turkey and Greece (forthcoming, 2001). SIPRI Yearbook. World Armaments and Disarmament. Oxford: Oxford University Press (various years). State Institute of Statistics Turkey. Statistical Indicators 1923-1991. Ankara, Turkey, 1993. State Institute of Statistics Turkey. Turkish Economy, Statistics and Analysis. Ankara, Turkey, 1995. Yildirim, J. Currency Substitution and the Demand for Money in the
A system estimation of the defense-growth
relation in Turkey
333
European Union. Unpublished PhD thesis. University of Manchester, Manchester, UK, 1997.
334 A> in ing th e Sou th
Appendix 14A.1 Data sources and definitions of the variables The data for this study come from several sources (see references for detailed citations). First, defense expenditure data were taken from various SIPRI yearbooks. GNP, saving (calculated from national accounts), balance of trade (exports of goods and services minus imports of goods and services in the national accounts) were taken various IMF/EFS yearbooks. Labor force data were taken from OECD Labor Force Statistics. All financial data were deflated to 1987 million Turkish liras using GDP deflators from the State Institute of Statistics, Turkey. Definitions of the variables: Y S B M L D75 D88
gross national product (1985 constant prices) gross savings (1985 constant prices) balance of trade (1985 constant prices) Turkish defense expenditure (1985 constant prices) employed labor force Cyprus conflict dummy variable taking a value of one for the years 1974, 1975 and zero elsewhere impulse dummy for 1988 to reflect possible effects of the change in the economic policy
A system estimation of the defense-growth relation in Turkey
335
Appendix 14A.2 Stationarity Analysis In this section conventional unit root tests are applied to real income (Y), real balance of trade (B), real military expenditure (M), real savings (S), and labor force (L). Table 14A.1 below presents evidence relating to the augmented Dickey-Fuller tests. The lag length is chosen according to the AKAIKE and SCHWARZ criteria. The second column shows results of the tests of integration of order one, and the third column presents tests of integration of order two. The results of the conventional unit root tests suggest that all variables are integrated of order one. Table 14A.1: Unit root tests VarLAG
A-DF(X)
A-DF(DX)
Y B S M L
-2.91 -2.40 -2.95 -0.66 -1.95
-3.54* -3.36* -4.79** -5.30** -3.40*
3 3 1 1 1
Note : Significant at 1 percent level (**) and 5 percent level (*).
This page intentionally left blank
15 The Military-Civilian Tradeoff in Guatemala: An Econometric Analysis Kanta Marwah, Lawrence R. Klein, and Thomas Scheetz
Introduction There is a general supposition that a high degree of militarization competes with general economic advancement in the civilian sector of an economy. In order to examine this supposition in a quantitative and detailed fashion it is necessary to consider simultaneously the following effects or economic adjustments to the military situation: (a) the demand side, (b) the supply side, (c) the short-run, and (d) the long-run. The idea of simultaneous consideration means that both direct and indirect effects must be taken into account. Military expenditure is customarily (but not unequivocally) included in GDP, and military employment is included as part of total employment. When military expenditure is increased, assuming a balanced increase in armed forces, too, there will usually be a short-run increase in GDP, directly in the military component of final government expenditures, and an induced increase in spending on the total GDP, partly by the enlarged forces, and partly by others, including business firms. The total result is estimated by the familiar multiplier calculation, under the heading of military Keynesianism. This kind of reasoning can serve only as an approximate and rough first step just on the demand side.
338 Arming the South
Large-scale military activity also diverts the economy to inefficient lines of endeavor for producing goods for civilian consumption and investment. Military goods cannot produce future streams of output, while factories, farms, mines, etc. without military equipment and machinery can do just that. That is an essential supply-side difference between military goods and nonmilitaryfixedcapital. Some capital goods do have dual-use capabilities, but when they are producing military goods/services, they are diverted from producing goods/services that could serve the population at large. In many econometric studies of military activity in a general economy, it has been found that in the first instance, military Keynesian effects are at work, but after the passage of time nonmilitary activity, especially the production and consumption of nonmilitary goods/services, begin to contract (expand) when military activity is expanded (reduced). This aspect is apparent, for example, in three studies of the effects of reduced military spending in the global economy.1 In the present investigation we study the economic effects of militarization from similar points of view (demand and supply side, short-run and long-run) in the context of a newly built econometric model for Guatemala. The next section provides an historical perspective on the subj ect. The following sections describe the database, contain the model, and analyze the trade-off results. The final section concludes. Motivation and historical perspective During the 1970s, 1980s, and early 1990s Central America was a battleground with large-scale intervention by the United States. There were exceptions on the peaceful side, as in Costa Rica, but military intrusion and socio-economic turbulence were prevalent in several places. A careful study of a remote, small, poor country like Guatemala during this period is interesting and important in its own right, but it is also of great significance for the entire region of Central America - not to mention South America and the Caribbean. More importantly, Guatemala has just emerged from 33 years (1963-1996) of guerrilla war with an estimated 100,000+ dead. Our study covers almost the entire period of the war, and the results show that the
The military-civilian tradeoff in Guatemala: an econometric analysis
339
military effort, far from having alleviated the underlying social and economic development issues triggering the war, actually made them far worse, initiating a vicious spiral of violence. In that sense, military spending itself can often be said to have been subversive of the established order. From an historical perspective the initial formation of the guerrilla movement in Guatemala may be traced back to the US-organized toppling of the Jacobo Arbenz government in 1954 and the CIA's secret basing of anti-Castro fighters in Guatemala in 1960. But the beginnings of the internal armed conflict can perhaps best be dated from the formation of the Fuerzas Armadas Rebeldes in November of 1963.2 The period from Arbenz through 1970 can be characterized as alternating among weak civilian elected, and military controlled, governments. From 1966 to 1968 the army organized a major counter-insurgency offensive that drove the guerrillas from the northeastern jungle into urban areas. From 1970 through 1985 the military were continually in power, sometimes through fraudulent elections, sometimes by coup d 'etats. The heaviest internal struggle occurred between 1978 and 1983, especially under the government of General Efrain Rios Montt in 1982 - 1983. Finally, beginning in 1986 democracy was restored, albeit with civilian governments again heavily influenced by the generals. Among the other significant security events of the region during the period under study (1968 - 1994), 19 July 1979 marked the Sandinista's overthrow of the Somoza regime in Nicaragua. The following years witnessed struggles and counter-insurgency movements affecting the entire region, particularly in El Salvador and Nicaragua, with US presence heavily felt. Then on 7 August 1987 regional peace accords were signed in Esquipiilas, committing the respective governments to a process of peace negotiations.3 In Guatemala, however, from 1985 onward there began a gradual winding-down of the war: amidst massive human rights violations, almost entirely attributable to the military, the guerrillas largely had been defeated (though not eliminated). Since the level of defense expenditures are determined by internal and external tensions, our series on real military expenditure (MILR) and international arms purchases clearly reflect the pressure from the guerrillas and the general instability in the region as a whole. The trends
340 Arming the South
in real military expenditure and the ratio of military expenditure to GDP (APMIL) are shown, with time on the horizontal axis, in figures 15.1 and 15.2, respectively. Real military expenditure clearly accelerated from 1978 onward and peaked around 1983. Also, other factors such as a severe February 1976 earthquake and worsening economic conditions (reflected in negative or slow growth in 1975, 1981-1984, and 1986) added to the burden of military expenditure and arms purchases. Beside military expenditure, another variable, a more precise index of internal and external tension was developed and employed in the study. While the civil war continued during the entire period under study, not all of that period experienced the same level of violence, nor threat to state security, by the irregular forces. Nor was the external Central American situation always unstable. Our new variable, denoted by TENS, is an index (0 to 1) representing the political perception of events that we hypothesize influenced the Guatemalan budget process.4 The database: a quantitative challenge An econometric study of any developing country is challenging. A major econometric issue is always the availability of the database. The paucity of quantitative data that would be suitable for careful econometric analysi s is indeed a hurdle. Even in the best of circumstances, when there is no particular reason to cover up or mislead outsiders about the magnitude or type of military effort, it is difficult to estimate reliable economic statistics about a comparatively poor developing country. The errors of measurement are much larger and data are reported less frequently than in major industrial countries, where databases are also not precise. More so, in this geographic area, the quantitative record of military trade is a closely guarded secret, particularly in Guatemala, where US (or US proxies') intervention was quite substantial. Having said this, most models for developing countries rely on international sources (principallyfromthe Stockholm International Peace Research Institute, the US Arms Control and Disarmament Agency, the International Institute for Strategic Studies, or the International Monetary Fund's Government Finance Statistics) for their military expenditure
The military-civilian tradeoff in Guatemala: an econometric analysis
69 70
7b
80
85
*0
69 70 — APUIL
75
80
8S
341
90
Figure 15.1: Guatemala: real military Figure 15.2: Ratio of real military expenditure, 1969-1994 ('000s expenditure to real GDP, 1969-1994 of quetzales)
Note: In this chapter, all figures have time on the horizontal axis, and the variable value on the vertical axis.
database. The military expenditure datafromthese sources for most Latin American countries are of very poor quality (Scheetz, 1994). Moreover, these data (in trends and real outlays) vary greatly from source to source. In the case of Guatemala, at the very outset we judged that international data sources would not adequately reveal what was taking place in the economy (see data appendix table 15A.1). Accordingly, Thomas Scheetz and a team of local specialists gathered most of the data - military expenditure and other fiscal accounts, national accounts, sectoral labor force, and prices - within the country. As such there is consistency among data series in national accounts, especially in fiscal accounts and military expenditure data.5
342 Arming the South
For constructing military expenditure data, the basic NATO definition of military expenditure was used throughout the 27 year period under review. Accordingly, the entire national budget was examined by jurisdictional program for each year. Those programs where more than 50 percent was spent for military purposes were classified as military expenditure. This "more than 50 percent" rule is the IMFs criterion for government spending, including military spending. We extended this rule and applied it to the NATO definition for inclusion of other programs such as military education, health, and retirement benefits - as part of military expenditure. An important sub-component of military expenditure, namely arms acquisitions, was taken from the IMFs "Economic Classification of Government Expenditure." The role of this variable appears separately in our model, specifically in determining the level of total imports. Admittedly, the quality of the data on arms acquisitions until 1975 is quite poor. Parenthetically, it is worth noting that the military expenditure data developed for this study were also used by the United Nations (MINUGUA) in their September 1996 negotiations between the united guerrilla groups and the Guatemalan armed forces. The model specification and estimation Our objectives in econometric model building for Guatemala are to try to show the main interactions among markets or sectors of the economy, which are the main issues for studying military-civilian trade-offs, and to be able to determine main trends or tendencies in the economy especially the macroeconomy. A main paradigm, forming the core of a model for a developed industrial economy, is the IS-LM model. By introducing specific sectors such as international trade, wages, prices, exchange rate, interest rates, and public (military plus nonmilitary) and specific private sectors, one can make such systems informative for policy analysis, including forecast extrapolation. The corresponding paradigm for a developing economy must pay attention to both supply and demand sides. This can be done with the core
The military-civilian
tradeoff in Guatemala: an econometric analysis
343
of the Harrod-Domar model, which allows for production technology and its role in the processes of capital formation. The Harrod-Domar core can also be expanded by opening it to international trade, price, interest, exchange rate, and wage rate determination (or influences). To make it as useful as possible, given the data issues discussed earlier, the Guatemala model presented here is an extension of the Harrod-Domar model for studying the military-civilian trade-off as well as for looking at plausible expansion rates for the economy, over the medium-term, covering horizons of 5 to 10 years. In this context, it is important to note that ever since Benoit's (1973) initial single-equation study was published, research on the economic impact of military expenditure has progressed along several cross-sectional and time-series lines. Some studies have focused separately on developed economies and some on developing. The developed economies generally have good data in quality and availability, the greater abundance of fiscal resources, and the presence of a significant military industry that restrains arms imports. But these pertinent factors are absent in developing countries. Over time, Benoit-style single-equation models have largely been set aside, in favor of ever larger models that better describe details of the economy, and are either supply (and/or demand) driven, or they study special aspects of military expenditure effects. Nonetheless, in comparison with the Guatemala model developed in this study, they all remain, by and large, ad hoc in structure.6 Our model incorporates the joint interactions of both demand and supply effects within the HarrodDomar model. The complete model, together with the definitions of the variables, is presented below (table 15.1). The estimated equations are collected the table 15.2. The equations have been estimated by two methods, ordinarysingle-equation-least-squares (OLS) and two-stage-least-squares (2SLS). (AR1 indicates "with correction for first-order serial correlation".) The 2SLS method employs all the exogenous variables as instruments. The numbers within parentheses, printed below the coefficients, are t-ratios. Altogether, the model has 12 equations of which nine are structural and three are identities. The 12 endogenous variables determined within the model are: GDP, CON, GFI, EX, IM, KRM, L, MIL, NED, p, rf, and
344 Arming the South
vel. The remaining variables are exogenous. Let us briefly review the model specification which, although highly constrained by data availability, is put forward as the type of system that can be estimated from currently available data and be useful in the policy analysis of the economic role of the military sector. The strategic entry points of the military sector into the system are provided by four primary components of aggregate spending on the demand side and the production function on the supply side. Specifically, the first four equations covering the private demand side of the macro economy encompass consumer spending (15.1), business spending for gross private capital formation (15.2), imports (15.3), and exports (15.4). These four items (imports being treated negatively) add to an estimate of GDP when combined with total public spending (G) in equation (15.11). Since there also exist some direct estimates of GDP from overall statistics, equation (15.11) provides the value of a statistical discrepancy (SD) between the two sources of GDP measure. The distribution of aggregate demand over each of the five components (the ratio of each component to GDP) and the ratio of military expenditure to GDP are shown in figure 15.3. A major supply-side aspect of the model comes from the production function, relating factor inputs to output (real GDP), as in equation (15.5). A general price equation and an exchange-rate equation, equations (15.6) and (15.9), show market influences. Since military expenditure is to be examined endogenously, in relation to total government expenditure, which is exogenous, it is developed in equation (15.7). Equation (15.8) shows the scope for fiscal policy by relating spendable income (after taxes and transfers) to total GDP. Two more identities define the relation between investment and capital stock (15.10) and monetary velocity (15.12). Since there are too few statistical data points available to develop an informative series on interest rates, we have used velocity as a proxy for interest rate. This is, in fact, a way of interpreting the LM part of the IS-LM core and is often used in model building for developing countries in the absence of business appropriate interest rate statistics. Equation 15.1: The consumption equation shows the overall relation
The military-civilian tradeoff in Guatemala: an econometric analysis
Table 15.1: Guatemala: econometric model Consumption: (15.1) CON/p*N = Fx [NID/p*N, Ml/p*N, MIL/GDP, SD/p*N, D4, u,] Gross fixed investment: (15.2) GFI/pj = F2 [(GDP/p)2, Ml/p, KRM . „ TENS, D4D5, u2] Imports: (15.3) IM/Pm = F3 [GDP/p, pm/p, ARMS/Pm, Dl5 u3] Exports: (15.4) EX/p = F4 [YRUS, MIL/GDP, Pwt/p, (EX/p). „ u4] Production function: (15.5) ln(GDP/p*L) = F5 [ln(KRM/L), [ln(KRM/L)] 2, ln(MIL/GDP), [ln(MIL/GDP)] 2, TIME, \i5] Price formation: (15.6) Ap = F6 [Avel, Apm, vel. „ p m . „ p . „ u6] Military expenditure: (15.7) MIL/p = F7 [G/p, SD/p, (MIL/p) _,, u7] National disposable income and GDP linkage: (15.8) NID/N = F8 [GDP/N, SD/N, u8] Exchange rate: (15.9)
Arf = F9 [Avel, Ap, CUB/GDP, rf .,, \i9]
Identities: (15.10) KRM = GFI/p ; + 0.95 KRM ., (15.11) GDP = CON + GFI + G + E X - I M - S D (15.12) vel = GDP/Ml
345
346 Arming the South
Table 15.1 (cont.)
Note: Definitions of variables ARMS CUB CON EX G GDP GFI IM KRM L Ml MIL N NID SD YRUS P Pi Pm Pus
P rf vel TIME TENS Dl D4 D5 D4D5 A
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error.
Mean = 0.2532E+07; SE = 277998; adj. R2 = 0.80; d = 1.55 Serial independence (BG-LM(1) test) p-value = 0.426; Homoskedasticity (White-het test) p-value = 0.262; Normality (J-B test) p-value 0.678
Mean = 0.47391 E+07; SE = 623664; adj. R: = 0.87; d = 2.06 Serial independence (BG-LM(1) test) p-value = 0.754; Homoskedasticity (White-het test) p-value = 0.120; Normality (J-B test) p-value 0.246
Mean = 0.3044E+07; SE = 349766; adj. R: = 0.77; d = 2.16 Serial independence (BG-LM(l) test) p-value = 0.541; Homoskedasticity (White-het test) p-value = 0.816; Normality (J-B test) p-value 0.203
346 Arming the South
Gross fixed investment: (15.2) GFI/p , = 35508.40 + 0.92E-08 (GDP/p)2 + 0.60 Ml/p- 0.06 KRM ., + 922414.00 TENS - 304315.00 D4D, (0.08) (5.13) (1.90) (-2.92) (2.41) (-1.22)
Table 15.1 (cont.)
(a) Ordinary Least Squares model
Note: Definitions of variables
Consumption: (15.1) CON/p*N = 509.58 + 0.56 NID/p*N + 0.99 Ml/p*N - 4791.46 MIL7GDP + 0.24 SD/p*N + 125.22 D4 (6.80) (12.20) (3.63) (-2.46) (2.43) (5.02)
ARMS CUB CON EX G GDP GFI IM KRM
Pi Pm Pus
Mean = 1 743.70; SE = 29.80; adj. R2 = 0.95; d = 2.36 Serial independence (BG-LM(l) test) p-value = 0.365; Homoskedasticity (White-het test) p-value = 0.141; Normality (J-B test) p-value 0.421
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error. EX/p = - 0.18E+07 + 772.77 YRUS - 0.65E+08 M1LVGDP + 822924.00 p j p + 0.56 (EX/p)., (-1.08) (2.31) (-2.69) (1.51) (4.32) Exports: (15.4)
P rf vel TIME TENS Dl D4 D5 D4D5 A
IM/pm = - 959481.00 + 0.58 GDP/p - 0.37E+07 pn/p + 8.45 ARMS/pm - 0.22E+07D, (-0.88) (10.56) (-4.39) (1.90) (-4.32) Imports: (15.3)
L Ml MIL N NID SD YRUS P
Table 15.2: Estimated models, 1969-1994
National disposable income and GDP linkage: (15.8) N1D/N = - 43.80 + 0.90 GDP/N + 0.10 SD/N AR1 (-3.77) (238.80) (2.41)
Mean = 1715.52; SE = 21.28; adj. R2 = 0.99; d = 1.89; rho = 0.53 Serial independence (BG-LM(l) test) p-value = 0.004; Homoskedasticity (White-het test) p-value = 0.605; Normality (J-B test) p-value 0.206
346 Arming the South
Mean = 249576.00; SE = 29917.50; adj. R2 = 0.88; d = 1.18 Serial independence (BG-LM(l) test) p-value = 0.192; Homoskedasticity (White-het test) p-value = 0.382; Normality (J-B test) p-value 0.888
Table 15.1 (cont.)
Production function: (15.5) ln(GDP/p*L) = 9.87 - 2.02 ln(KRM/L) + 0.14 [ln(KRM/L)]: - 2.83 In(MIIVGDP) - 0.31 [In(MILVGDP)]2 - 0.01 TIME (1.62) (-1.46) (1.73) (-12.46) (-2.33) (-5.97)
Note: Definitions of variables ARMS CUB
Military expenditure: (15.7) MIE/p = -14398.50 + 0.08 G/p - 0.03 SD/p + 0.53 (MIIVp) ., (-0.58) (4.57) (-2.31) (5.03)
o
Mean = 8.94; SE = 0.04; adj. R = 0.80; d = 1.56 Serial independence (BG-LM(l) test) p-value = 0.471; Homoskedasticity (White-het test) p-value = 0.268; Normality (J-B test) p-value 0.652 2
CON EX G GDP GFI IM KRM
Pi
L Ml MIL N NID SD YRUS P Pm Pus
Mean = 0.12; SE = 0.06; adj. R2 = 0.88; d = 2.09 Serial independence (BG-LM(l) test) p-value = 0.564; Homoskedasticity (White-het test) p-value = 0.180; Normality (J-B test) p-value 0.421
Co
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error. (-2.00)
(3.29)
(0.25)
(6.09)
(2.56)
P
Table 15.2 (cont.)
rf vel TIME TENS Dl D4 D5 D4D5 A
Price formation (15.6) Ap = 0.03 Avel + 0.39 Apm + 0.44E-03 vel., + 0.34 p m ., - 0.21 p .,
OO
346 Arming the South
Identities:
Table 15.1 (cont.)
Note: Definitions of variables ARMS CUB
Exchange rate: (15.9) Arf = - 0.18 Avel -* 4.82 Ap - 15.75 CUB/GDP - 0.45 rf., (-1.87) (540) (-3.23) (-4.22)
CON EX G GDP GFI IM KRM
Pi
L Ml MIL N NID SD YRUS P Pm
vel = GDP/Ml
(15.12)
GDP = CON + GFI + G + EX - IM - SD
(15.11)
Pus
Mean = 0.43; SE = 0.34; adj. R2 - 0.69; d = 2.16; sample 1984-1994 Serial independence (BG-LM(1) test) p-value = 0.749; Homoskedasticity (White-het test) p-value = 0.000; Normality (J-B test) p-value 0.701
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error.
Note: The property of serial independence of the residuals is tested by the BreuschGodfry LM(1) (BG-LM) test and the Durbin-Watson d test; homoskedasticity, by the likelihood ratio (LR-het) test; and normality, by the Jarque-Bera (J-B) test. The p-value is to be interpreted as the risk involved in rejecting the stated null hypothesis.
KRM = GFI/p , + 0 95 KRM.,
(15.10)
P rf vel TIME TENS Dl D4 D5 D4D5 A
Table 15.2 (cont.)
346 Arming the South
Table 15.1 (cont.)
(b) Two Stage Least Squares model
Note: Definitions of variables ARMS CUB
Price formation: (15.6) Ap = 0.03 Avel+ 0.41 Apm + 0.61E-03 vel ., + 0.37 p m .,-0.24 p ., (1.72) (6.19) (0.34) (3.40) (-2.18) Mean = 0.12; SE = 0.06; adj. R2 = 0.88; d = 2.08
CON EX G GDP GFI IM KRM
Pi
L Ml MIL N NID SD YRUS P Pm
EX/p = - 0.21 E+07 + 826.15 YRUS - 0.68E+08 MI1JGDP + 912756.00 pu1/p + 0.55 (EX/p)., (-1.20) (2.41) (-2.76) (1.64) (4.22) Mean = 0.30E+-07; SE = 350002.00; adj. R2 = 0.77; d = 2.1 5 Exports: (15.4)
Pus
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error. IM/pm = - 188872.00 + 0.57 GDP/p - 0.46E+07 pm/p + 7.30 ARMS/pm - 0.20E+07D, (-0.16) (10.10) (-4.75) (1.57) (-4.32) Mean = 0.47E+07; SE = 642029.00; adj. R2 = 0.87; d = 2.1 5 Imports: (15.3)
P rf vel TIME TENS Dl D4 D5 D4D5 A
Table 15.2 (cont)
Consumption: (15.1) CON/p*N = 508.34 + 0.56 NID/p*N + 0.97 M l/p*N - 4675.43 MIL/GDP + 026 SD/p*N + 124.01 D4 (6.70) (11.54) (3.22) (-2.35) (2.45) (4.88) Mean = 1743.70; SE = 29.82, adj. R2 = 0.95; d = 2.37
Gross fixed investment: (15.2) GFI/p,= 137253.00 + 0.84E-08 (GDP/p)2 + 0.65 Ml/p-0.05 KRM ., + 728004.00 TENS - 482374.00 D4D5 (0.29) (4.42) (192) (-2.20) (1.74) (-1.63) Mean = 0.25E+-07; SE = 281 846; adj. R2 = 0.80; d = 1.75
Production function: (15.5) ln(GDP/p*L) = 8.39 - 2.17 ln(KRM/L) + 0.14 |ln(KRM/L)]2 - 3.85 ln(MlL/GDP) - 0.43 [ln(MII7GDP)]: - 0.01 TIME (E28) (-1.51) (1.77) (-2.70) (-2.59) (-5.97) Mean = 8.94; SE = 0.04; adj. R2 = 0.80; d = l.58
Table 15.2 (cont.)
346 Arming the South
Table 15.1 (cont.)
o K
ARMS CUB
Pi
Mean = 1772.90; SE = 21.14; adj. R2 = 0.99; d = 2.02; rho = 0.54
CON EX G GDP GFI IM KRM L Ml MIL N NID SD YRUS P
Pm
Pus
vel = G D P / M l
(15.12)
rate: Arf = - 0.18 Avel + 4.82 Ap - 15.75 CUB/GDP - 0.45 rf., (-1.87) (5.40) (-3.23) (-4.22) Exchange (1 5.9)
P
GDP = CON + GFI + G + EX - IM - SD
(15.11)
< •
National disposable income and GDP linkage: (15.8) NID/N = - 49.52 + 0.90 GDP/N +- 0.09 SD/N (-3.73) (224.80) (2.35)
Note: Definitions of variables
^
o
arms imports, thousands of current quetzales (Q). current account balance, thousands of current Q. CUB = EX - IM + net foreign income from abroad. private consumption, thousands of current Q. exports of goods and services, thousands of current Q. central government total expenditure, thousands of current Q. gross domestic product, thousands of current Q. gross fixed investment, thousands of current Q. imports of goods and services, thousands of current Q. real fixed capital stock, thousands of 1987 Q. [KRM = GFI/p , + 0.95 KRM .,] economically active population, thousands, (EAP/1000). money supply, end of the year, thousands of current Q. military expenditure, thousands of current Q. population, thousands, (POP/1000). = national disposable income, thousands of current Q. (CON + GFI + G + EX - IM) - GDP. US GDP in billions of 1987 US$. GDP implicit deflator, 1987 = 1.00. implicit gross fixed investment deflator, 1987 = 1.00. import price, (pus*rf/2.66), 1987= 1.00. US implicit GDP deflator, 1987 = 1.00. weighted WPI of US, Japan, Germany and El Salvador, principal trading partners of Guatemala, 1987 = 1.00. average annual exchange rate, Q per US$. velocity of money, GDP/Ml. trend variable. internal and external security tension constructed variable. dummy variable, 1 for 1982-1994, 1 otherwise. dummy variable, 1 for 1993-1994, 0 otherwise. dummy variable, 1 for 1973, 1983-92, 0 otherwise. dummy variable D4+D5, 1 for 1973 and 1983-94, 0 otherwise. change. stochastic error. KRM = GFI/p , + 0.95 KRM .,
(15.10)
0}
rf vel TIME TENS Dl D4 D5 D4D5 A Identities:
o ^ £
Mean = 249576.00; SE = 29936.10; adj. R = 0.88; d = 2.34 2
a Mean = 0.43; SE = 0.34; adj. R = 0.69; d = 2.16; sample 1984-1994 :
3 expenditure: MIL/p = -15316.30 + 0.08 G/p - 0.03 SD/p + 0.52 (MIL/p) (-0.61) (4.64) (-2.40) (4.84) Military (1 5.7)
352 Arming the South
1.0-
0.8-
APCON
0.6-
0.4
APIM
0.2-
^
- ^ " ' \
\
<,
/ '" .
r
y^
—
—
— — _—--""^
^EX
- ' -
-
N.
^ APMIL
0.0
,
i
6+ 70
i
i
i
— i
i
75
i
i
i
i
i
80
i
i
i
i
i
85
i
1
1
1
1
I
1
90
Figure 15.3: Average proportions of military expenditure (APMIL), consumption (APCON), gross fixed investment (APGFI), imports (APIM), exports (APEX), and government spending (APG) in GDP, 1969-1994
1
The military-civilian
tradeoff in Guatemala: an econometric analysis
353
between total consumer spending and total spendable income (after taxes and transfers), both expressed in veal per capita terms. A liquid balance (wealth) effect, measured by real per capita money supply, is an additional variable that is related to consumer spending. A direct estimate of the effects of military spending on private consumption is shown by the coefficient of the ratio of military outlays to GDP. Interestingly, this effect is negative, implying that rising military claims on output leave less for private consumption - or for exports, since equation (15.4) carries a similar negative effect for sales abroad. The data thus show a crowding-out effect of a large military sector on availability of goods for civilian use and for exports. These are both important coefficients for studying the issues of interest in the trade-off. Lastly, the real per capita statistical discrepancy, SD/p*N, and an allowance for an inexplicable statistical jump in the data series in 199394 by a dummy variable (D4), complete the estimation of the consumption function. Equation 15.2: The gross investment equation, in real but not per capita terms, is expressed as a nonlinear (quadratic) function of GDP, stock of available fixed capital, and real liquid balances. In addition, a measure of tension and an augmented variable (D4D5) to cover data jumps in 1973 and 1983-1994 are included. It is customary to express the consumption function in per capita (or per family) units to take account of the necessity to feed, house, and clothe additional people, as they augment the demographic statistics. In the case of the business sector, it is the drive for profits that motivates spending, rather than the necessity to care, in an economic sense, for each person. The unusual features of the investment equation are that there is no compelling reason for the output effect, or others for that matter, to be linear, and the data support a more rapid nonlinear (quadratic) approach. The variable Ml/p shows some credit or financial effects, while capital stock indicates available capacity already on hand to meet production needs. The unusual variable, TENS, measures societal tension as military forces are built up. A stronger effort in investing is required in order to meet military needs with high priority. That accounts for the positive coefficient of TENS. Equation 15.3: Imports, in real terms, depend on real GDP (the
354 Arming the South
activity effect) and relative prices. These are the main effects suggested by economic analysis. Many imports, however, consist of armaments from abroad, which are treated exogenously in the system, and these constitute another source of import requirements. Finally, heavy imports during the conflict era are specially treated through the use of a shift variable, D,. Equation 15.4: Real exports, like imports, depend on an activity effect and a relative price effect. These are both explicit variables in the equation, with the activity effect being represented by GDP of the United States. There is also an estimate of a direct (aforementioned negative) military effect, as well as a (distributed) lag effect. Equation 15.5: The production function is a logarithmic equation, with output and capital both expressed as ratios to labor input. The capital/labor ratio, in logarithmic terms, is quadratic, as well as the direct effect of the military share of GDP. There is also an additional term. This equation is of extreme importance in the present context for showing how supply conditions come into focus when the military sector expands. The equation is expressed in logarithmic ratio form and is, essentially, an extension of the Cobb-Douglas function.7 It is a generalization since the logarithm of the factor ratio is expressed as a nonlinear effect, with the second-degree term assuming importance as the economy grows. The military shares of GDP enter with negative coefficients to show how production can become disoriented as the military sector grows in significance. In the early stages of a militaryinduced expansion the labor-input effect is positive, but the quadratic term, with labor in the denominator assumes more importance as the economy expands. The military terms are, however, both negative, and tend to restrain output growth. The mean-point partial production elasticities implied by equation 15.5 (2SLS) are 0.489 for capital, 0.511 for labor and -0.166 for the ratio of military expenditure to GDP. Equation 15.6: The equation for price change is made a function of velocity (proxy for interest rate) and import price, with distributed lags. In general, this is a cost-push specification, with capital cost and import cost being the major factors generating inflation. Equation 15.7: Military spending is endogenous in this model and is
The military-civilian tradeoff in Guatemala: an econometric analysis
355
related to total government spending which is exogenous. Also, the statistical discrepancy and a lag distribution are supplementary variables. Equation 15.8: As noted above, this equation shows indirectly how taxes and transfers are related to total activity, by correlating spendable income with GDP and the discrepancy. This is a "legal type" equation and is appropriately put in current price terms, but expressed mper capita values. Equation 15.9: Exchange-rate fluctuations follow interest rate changes (given by the velocity proxy), inflation movements, and current account balance. Since the exchange rate is expressed as quetzales per dollar, it moves in an appreciating direction with respect to improvement in the current balance. Identities 15.11 and 15.12 are simply definitions - of the statistical discrepancy, SD, and of velocity, vel - but the capital accumulation identity 15.10 is based on the accounting rule (capital stock) t = (capital stock) t_x + (GFI/p) t - .05 (capital stock) t ., so that KRM t = (GFI/p) t + 0.95 (KRM) t .,. Apart from the specification and economic interpretation of the estimates, there is the matter of statistical criteria. The equations were estimated by two methods, OLS and 2SLS, and, by and large, the resulting estimates do not differ appreciably. Coefficients and measures of fit are very similar. In all cases, the estimated equations have plausible directions of effects, variable by variable; the correlation coefficients are quite high; and the residuals from each equation neither show significant serial correlation, nor heteroskedasticity. They also are normally distributed. Under the given circumstances of the data problems, the equations turn out to be as good, statistically, as one gets for most developing countries, at the level of aggregation used.
356 Arming the South
System properties: trade-offs While the individual equations appear to be acceptable for further analysis, the system dynamics must be analyzed for what they say about the relationships among the key variables and for what they imply about the trade-offs being studied. Real military spending fluctuated in a narrow range of no more than 100,000 x 103 constant (1987 base year) quetzales per year and then broke out of this pattern between 1980 to 1981 when the total increased by more than 100,000 x 103 quetzales and remained elevated (see figure 15.1 above). Accordingly, from model simulation results, we decided to examine the period 1984-1994 for determining the model's properties with respect to military expenditure. For this period, simulated GDP per capita is correlated with actual GDF per capita with a coefficient of 0.99. Real consumption per capita, which is of unusual significance for our purposes of studying the economic burden, shows a correlation with actual values of 0.94. Gross fixed investment has the same correlation. Simulated exports and imports per head are each strongly correlated with actual values at 0.99 and 0.97, respectively. The closeness of the simulated values to the actual values of the key variables can be assessed from figures 15.4 to 15.11 in which their actual values and the model solutions for 1984-1994 are graphically presented. The simulated values are our baseline solutions and variables are tagged accordingly with suffix SB in comparison to the untagged actual values. Clearly and admittedly, the baseline simulations do contain data-related bias components which, however, can be easily eliminated by a minimal "add factoring".8 Moreover, in the following trade-off analysis, both baseline and controlled solutions contain the same bias components, but these get mutually offset when we measure their deviations. Two scenarios of controlled solutions We ran two sets of simulations, with changes imposed on military expenditure and arms imports, and compare the results to our baseline simulation. In scenario 1, we reduced military expenditure by a constant
The militaiy-civilian tradeoff in Guatemala: an econometric analysis
357
//
-CONP---CONPSB
Figure 15.4: Actual and baseline simulated per capita consumption, 1984-1994 (quetzales per person)
— NIDP
NIDPSB
Figure 15.5: Actual and baseline simulated per capita national disposable income, 1984-1994 (quetzales per person)
GDPP - -GDPPSB
Figure 15.6: Actual and baseline simulated GDP, 1984-1994 (quetzales per person)
Figure 15.7: Actual and baseline simulated gross fixed investment, 19841994 (quetzales per person)
358 Arming the South
-U,-
,jwpsa
Figure 15.8: Actual and baseline simulated imports, 1984-1994 (quetzales per person)
Figure 15.9: Actual and baseline simulated exports, 1984-1994 (quetzales per person)
/
Figure 15.10: Actual and baseline simulated price index, 1984-1994 (actual 1987=1.00)
Figure 15.11: Actual and baseline simulated exchange rate, 1984-1994 (quetzales per US$)
The militaiy-civilian
tradeoff in Guatemala: an econometric analysis
359
Table 15.3: Guatemala, scenario 1 (30 million constant 1987 quetzales annual reduction in military expenditure): percentage changes from baseline simulation Year
GDP
Price index
Exchange rate
Current account balance
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
2.20 2.15 4.35 4.96 5.33 4.14 2.83 3.53 3.71 2.26 2.84
1.00 0.08 -0.07 -0.74 -2.12 -3.78 -4.34 -4.43 -4.64 -4.93 -4.76
-1.98 -1.99 -3.23 -5.52 -8.05 -8.37 -6.00 -5.76 -5.70 -4.64 -3.89
-17.34 -5.54 35.35 -15.77 -4.72 5.99 13.81 11.69 10.31 8.25 10.64
Table 15.4: Guatemala, scenario 2 (30 million constant 1987 quetzales annual reduction in military expenditure and annual 30 million constant 1987 quetzales reduction in arms imports): percentage changes from baseline simulation Year
GDP
Price index
Exchange rate
Current account balance
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
4.39 3.53 5.64 6.16 6.48 4.96 3.69 4.49 4.74 3.01 3.63
2.00 -0.07 -0.74 -1.79 -3.61 -5.56 -5.91 -5.85 -5.91 -6.08 -5.70
4.22 37.08 -2.41 -3.18 -20.44 -8.26 -14.83 -28.21 -13.89 -12.43 -17.51
-112.28 17.94 71.14 -25.86 -6.83 7.23 15.95 13.24 11.51 8.37 11.86
360 Arming the South
CUB per capita
GDP per capita
Price index Exchange rate (Q/USS)
Figure 15.12: Scenario 1: percentage effect, relative to baseline simulation, of yearly 30 million constant 1987 quetzales reduction in military expenditure on per capita GDP, price index, exchange rate, and per capita current account balance, 1984-1994
The military-civilian tradeoff in Guatemala: an econometric analysis
— CONRPSB <- '-CONRPSCI
361
- - CONRPSC2
Figure 15.13: Baseline simulation and Figure 15.14: Baseline simulation and scenanos 1 and 2 effects on per capita scenarios 1 and 2 effects on non-military consumption of annual 30 million constant government expenditure, 1984-1994 quetzales reduction in military expenditures (quetzales p e r person) and annual 30 million constant quetzales reduction in arms imports, 1984-1994
Figure 15.15: Baseline simulation and Figure 15.16: Baseline simulation and scenarios 1 and 2 effects on the scenarios 1 and 2 effects on the exchange price index, 1984-1994 rate (Q/USS), 1984-1994
362 Arming the South
Table 15.5: Per capita real consumption baseline simulation and scenarios land 2 (quetzales, 1987 prices) Year
Baseline simulation
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
1548 1639 1595 1621 1660 1681 1679 1638 1624 1851 2058
Scenario 1 Military spending reduction 1569 1675 1657 1700 1763 1794 1783 1749 1740 1958 2181
Scenario 2 Military spending and arms import reduction 1581 1693 1679 1727 1797 1829 1816 1780 1770 1984 2208
amount in each year (30 million constant 1987 quetzales). In scenario 2, we assumed a further reduction in aggregate military outlays by 30 million constant 1987 quetzales per year in the form of reduced arms imports. If military expenditure is reduced by 30 million constant quetzales, per year (nearly 10 percent, the percentage currently stipulated in the Peace Accords, see note 3), then we find higher values for per capita GDP in all years, 1984-1994 (inclusive). The same is true for real per capita consumption and gross fixed investment. The simulation with military spending restrictions also leads to an appreciating exchange rate. A summary picture is given by a semi-elasticity table (table 15.3) that shows the percentage reductions in four key variables (per capita GDP, price index, exchange rate (Q/US$), and current account balance). These are also graphically presented in figure 15.12. These are indeed substantial and impressive gains that might have been realized, had there been, in fact, no change to the very strong military policy that was in force prior to the build-up. That is, in fact, the trade-off effect. We now consider a larger demilitarization, consisting of both the
The military-civilian tradeoff in Guatemala: an econometric analysis
363
same reduction in military outlays and a reduction of 30 million quetzales in arms imports, scenario 2. This type of shift from military spending can have an added benefit of contributing directly to an improved external balance position. With the reduction in military spending specifically targeted for arms import cut-backs, we see that the net external position (current account balance) ultimately improves more than in the previous case, and the currency exchange position becomes stronger (figure 15.16). GDP growth is better as prices fall more from the baseline simulation values (figure 15.15). A primary motivation in this study is to have a quantitative assessment of "guns-or-butter" for Guatemala. Let us then turn to a comparison of real per capita consumption in each of the two scenarios for arms reduction with the baseline simulation (table 15.5 and figure 15.13). In the mid-1990s, it appears that Guatemala could have enjoyed more than five percent, and possibly as much as seven percent, additional consumption per person with the kind of shift studied here, in the context of a less militaristic economic environment. At the same time, the nonmilitary government expenditure would have gone up without any additional tax burden (figure 15.14). Conclusions Nearly all countries have needs for minimal military or police-guards. In most cases it is the former choice, but Costa Rica does manage to enjoy an improving economic life, spending only for a civil guard. Contrary to much popular thinking along the lines of military Keynesianism, developing countries can prosper with less - rather than more - military spending, if only rational agreement gravitates toward the Cost Rican model. Short of that, the trimming of the military and its demands on scarce resources can result in both short and long-run gains. In the longrun, the gains can be seen in the broadest economic measures, such as GDP, while both short and long-run gains are expected to occur in household consumption. Our findings strongly support President Arias' striking plans for shifts away from large military spending throughout Central America and
364 Arming the South
realization of a better economic life for the population at large. In the context of statistical models like ours, many preferred patterns and paths can be laid out for economies like Guatemala's. To indicate what can be done has been the purpose of this study. Notes This research project was jointly sponsored by Economists Allied for Arms Reduction (ECAAR), New York, and the Arias Foundation for Peace and Human Progress, San Jose, Costa Rica. 1. See the chapters by McKibbin, Li, and Pauly, and by Bayoumi, Hewitt, and Symanski in Gleditsch et al., 1996. 2. Most of the information given here is taken from an extensive unpublished memorandum entitled "Puntos de inflexion, tension y estabilidad militar: 1960 - 1995," written by the Guatemalan news analyst, Edgar Celada. 3. The Esquipiilas II Accords (part of the "Contadora Process") were signed in Guatemala in August 1987 by the Presidents of five Central American states: Costa Rica, El Salvador, Honduras, Guatemala, and Nicaragua. They committed the countries to peace negotiations which would bring an end to the conflicts in Nicaragua (14 February 1989), El Salvador (16 January 1992) and, finally, Guatemala (December 1996). 4. For details on the construction of this variable, see note 2 or contact Thomas Scheetz at [email protected]. 5. The disaggregated fiscal accounts and other data series are available from Scheetz. Also, see the data appendices to this chapter. The data were collected during field work in January and February 1996 in Guatemala. The Guatemalan team included Thomas Scheetz, Edgar Pape, Edgar Celada, and Fernando Solis. 6. Reviews of the earlier literature are found in Chan (1985, 1987), Maizels and Nissanke (1986), Deger (1990), and Dunne (1996).
The military-civilian
tradeoff in Guatemala: an econometric analysis
365
7. It is a CES generalization of the Cobb-Douglas production function. See Intriligator et al. (1996). 8. The systematic deviations between the pairs of curves (actual and simulated) do not appear to be diverging in an explosive way, except for rf, where some interventions would be called for if the spread between the two curves were on the verge of becoming wider and wider. "Addfactoring" is an adjustment of the constant term to reduce the margin of error or forcing the solution to pass through an actual data point. References Bayoumi, Tamim, Daniel Hewitt, and Steven Symanski. "Global Disarmament and Developing Countries: A MULTIMOD Simulation," pp. 491-520 in N.P. Gleditsch et al. (eds.), The Peace Dividend. Amsterdam: North Holland, 1996. Benoit, Emile. Defense and Economic Growth in Developing Countries. Lexington, MA: Lexington Books, 1973. Chan, Steve. "The Impact of Defense Spending on Economic Performance: A Survey of Evidence and Problems." Orbis Vol. 29, No. 2 (Summer 1985), pp. 403-434. Chan, Steve. "Military Expenditures and Economic Performance," pp. 29-37 in US Arms Control and Disarmament Agency. World Military Expenditures and Arms Transfers 1986. Washington, DC: US ACDA, 1987. Deger, Saadet. "Military Expenditure and Economic Development: Issues and Debates," pp. 35-52 in Geoffrey Lamb and Valeriana Kallab (eds.). Military Expenditure and Economic Development: A Symposium on Research Issues. Washington, DC: World Bank Discussion Papers, No. 185, 1990. Dunne, J. Paul. "Economic Effects of Military Expenditure in Developing Countries: A Survey," pp. 439-464 in N.P. Gleditsch et al. (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Intriligator, M.D, R.G. Bodkin, and Cheng Hsiao. Econometric Models, Techniques and Applications. Upper Saddle River, NY: Prentice Hall, 1996.
366 Arming the South
Li, Hung-Yi and Peter Pauly. "Multilateral Disarmament: Project Link Simulations," pp. 521-531 in N.P. Gleditsch et al. (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Maizels, Alfred and Machiko K. Nissanke. "The Determinants of Military Expenditures in Developing Countries." World Development Vol. 14, No. 9 (1986), pp.1125-1140. McKibbin, Warwick. "Military Spending Cuts and the Global Economy," pp. 465-489 in N.P. Gleditsch et al. (eds.). The Peace Dividend. Amsterdam: North Holland, 1996. Scheetz, Thomas. "Gastos Militares en America del Sur," pp. 195-220 in Centro Regional de las Naciones Unidas para la Paz, el Desarme y el Desarrollo en America Latina y el Caribe (ed.). Proliferacion de Armamentos y Medidas de Fomento de la Confianza y la Seguridad en America Latina. Lima, Peru: United Nations, 1994.
The military-civilian tradeoff in Guatemala: an econometric analysis
367
Data appendices Table 15A.1: Guatemala military expenditure. Comparison of international data series (in millions of US$ incurrent prices) Year 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
ACDA 68 74 83 109 116 169 185 132 115 89 138 139 143 145 150 168 166 175
IMF 78 73 85 101 114 149 147 164 180 102 124 129 131 112 132 152 154 175
IISS
65 71 109 91 165 170 180 197 245 106
72 109 123 123
SIPRI
83 103 118 143 161 208 231 270 371 96 124 129 131 112 131 152 154 175
E C A AR/Arias
75 72 90 104 163 177 184 207 224 141 124 146 152 132 119 154 144 153
Sources: Stockholm International Peace Research Institute, SIPRI Yearbook. International Institute for Strategic Studies (IISS), The Military Balance. US Arms Control and Disarmament Agency (ACDA), Military Expenditures and Arms Transfers. International Monetary Fund (IMF), Government Finance Statistics. ECAAR/ARIAS, constructed by Thomas Scheetz and colleagues (see text and note 5). Notes: In all cases the IMFs exchange rate (rf, line 99b) from the International Financial Statistics has been used for conversion to US$. The current US$ figures were given directly by ACDA, but were calculated from all other sources by conversion of local currency with exchange rate (rf, line 99b) from IMF, International Financial Statistics. In all cases we used the most recently published figures accessible to us. ECAAR/Arias data were gathered program by program for the entire Central Government's budget outlays. Comments: The figures vary by as much as 222 percent (SIPRI versus ACDA in 1985). The movements also vary widely. Clearly SIPRI employed IMF data from 19881994. Our locally gathered disaggregated data differ significantly from all other sources in both size and movements.
368 Arming the South
Table 15A.2: Data set for Guatemala study Year
GDP
CON
GFI
EX
IM
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
1390660.0 1453514.0 1610549.0 1715436.0 1904038.0 1984818.0 2101645.0 2569309.0 3161490.0 3645977.0 4365262.0 5480501.0 6070549.0 6903004.0 7879373.0 8607655.0 8717324.0 9049910.0 9470331.0 1.11800D+07 1.58381D+07 1.77110D+07 2.05449D+07 2.36844D+07 3.42903D+07 4.73023D+07 5.39854D+07 6.37336D+07 7.44907D+07
1138787.0 167334.0 263390.0 266135.0 235632.0 283648.0 191739.0 1200819.0 269478.0 296634.0 220852.0 1278609.0 305399.0 298973.0 230979.0 1378846.0 353648.0 338454.0 238562.0 1493313.0 343078.0 371052.0 263568.0 1588013.0 397250.0 389342.0 272530.0 1682221.0 536503.0 519063.0 356577.0 2033633.0 708405.0 811360.0 2469865.0 467890.0 792011.0 857962.0 570759.0 2874856.0 941699.0 1204070.0 900220.0 3395960.0 1340303.0 1438980.0 1038597.0 4126531.0 1303691.0 1655037.0 4674925.0 1217737.0 1473608.0 1784365.0 5433779.0 1286255.0 1747645.4 1963315.1 5880555.0 1295282.0 1471000.4 6469320.0 1443179.0 2031538.5 1288691.0 6588191.0 1629319.2 1309611.0 1175837.2 6920778.0 950247.0 1317050.2 1256198.1 1463574.1 7285509.0 912362.0 2067981.7 8627570.0 1224906.0 2246851.5 2542107.7 1.23027D+07 1593218.0 2311133.7 2807012.5 1.38052D+07 2188317.0 3948468.5 3308543.7 1.60459D+07 2747185.0 4506794.0 4099163.7 1.83287D+07 3254940.0 5322913.0 6775757.5 2.63843D+07 4454972.0 8143114.0 3.64352D+07 5760174.0 8348964.5 1.02162D+07 9482729.0 4.16651D+07 8445259.0 1.47713D+07 5.41645D+07 1.03345D+0'1 1.16128D+07' 1.67651D+07 6.38930D+07 1.057180+0'1 1.31725D+07' 1.85714D+07
Source: Compiled by Thomas Scheetz from domestic sources; see text and note 5.
The military-civilian tradeoff in Guatemala: an econometric analysis
369
Table 15A.2 (continued) Year
G
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
143078.0 156420.0 159271.0 178299.0 196579.0 196687.0 223814.0 254051.0 299786.0 368606.0 507860.0 603258.0 696850.0 795032.0 1095285.0 1355318.0 1212174.0 1043722.0 1004328.0 1015941.0 1427037.0 1871847.0 2230060.0 2624565.0 3255902.0 3737953.0 5103978.0 5896861.0 6368149.0
NID
1364399.0 1446166.0 1639187.0 1673561.0 1770171.0 2196941.0 2709119.0 3109609.0 3367929.0 4663465.0 5157469.0 5945863.0 6817241.0 7500371.0 7551970.0 7925786.0 8296472.0 9674110.0 1.37538D+07 1.52199D+07 1.76208D+07 2.05535D+07 3.02968D+07 4.23402D+07 4.79246D+07 5.73702D+07 6.69304D+07
KRM 1363765.2 2810103.0 4392312.5 5921213.0 7348866.5 8850700.0 1.02672D+07 1.18245D+07 1.32658D+07 1.48277D+07 1.71400D+07 1.96193D+07 2.22210D+07 2.45091D+07 2.63473D+07 2.83309D+07 2.98553D+07 3.04855D+07 3.08930D+07 3.11584D+07 3.14800D+07 3.20943D+07 3.29550D+07 3.39294D+07 3.45860D+07 3.52973D+07 3.67000D+07 3.82513D+07 3.95993D+07
CUB
YRUS
-21424.0 -17298.0 -11086.0 -45938.0 -9521.0 11062.0 -99405.0 -62314.0 -86178.0 -35354.0 -262146.0 -196440.0 -176394.0 -572394.0 -399135.0 -223956.0 -509470.0 -651411.0 -50958.0 -1.17244D+06 -1.09512D+06 -1.04040D+06 -1.24829D+06 -1.06911D+06 -3.66249D+06 -4.19138D+06 -4.01972D+06
2635.0 2693.0 2804.0 2872.0 2878.0 2962.0 3106.0 3264.0 3251.0 3224.0 3377.0 3536.0 3706.0 3795.0 3778.0 3839.0 3758.0 3905.0 4148.0 4276.0 4403.0 4540.0 4719.0 4835.0 4876.0 4822.0 4977.0 5134.0 5341.0
Source: Compiled by Thomas Scheetz mostly from domestic sources; see text and note 5.
o 3 ao 3
o
C/5
H o
a*
I
O o
$
o
Uj
" r
J
~
4
^ ^ H ?
0 U ,
9 i
i
—
: >
r " H !
o
> l r
P M ^ ! ^ P ? ° ^ ^ b b o o b b o o o o o o o o
t O t O ^ ~ ^ ^ M D s ) - - O U i 4 ^ O O v J O \ 0 0 - U i i y i ^ O \ 1 ^ 0 \ t O W ' - N j t \ J U i ^ N ) ( 7 i ^ s l K ) O J U i - C
J
K) O vO ^. b
! O
K) OO MD g 4^
K) OO O ui 4^
N) K) NJ G\ ^ — to 4*. to 4\ w 4^ 4^ ^—
W
to to to 4i w u» ON OO •— 4^ oo 4^ 4^. b —
to to 4^ to 4^
to — •—» C?\
to O Ui Ul
N—
MD MD vj
MD 4^ O 4^ to bo '-^ ^—
to — si OO
O oo O ^—
O to ^ to
MD MD OO ON — Ui MD to OS oo - -
^ b \ b ^ o j b t o b w - b b o u \ b o L n u i c o K ) b si
— ' - > 0 \ 0 \ £ I O O O O O O O C \ I \ 1 - V J N ) N 1 0 \ O C \ C \ C \ 0 \ U 0 0 - S ] | I - ^ 0 \ ^ - O V 1 U I W ^ ^ O O O \ ^ I O O ^ O \
o b b b b
0 0 C 0 s ) 0 N U i 4 ^ 0 0 L K > t O s l - O O ^ K ) 0 0 - O I \ ) O ' ~ -£> " -t^ " O ~ ._ ^ — O ^ - ^ o os u> MD O si O — ^ - - 4^ Ift N! to to C N -fc» MD 4^- O N
t O O 0 \ s
j U C X ) K ,
O O ^ I > J U i V O \ ) K ) ^ W (
O O O O O O O O O O O O O O O O O O O O
l > J
0 0 0 0 0 ^ ^ ^ 0 ' > J ^ ^ - ' 0 ' -
O O ^ X J N J S J ^ ^ M O O M O O N J ^ ^ ^ N I K I ' J J I O
OO^^VOOOOOOOOOOOOOOOOOOOOO^-slNjxl ^ O J K ) - 0 ' O O O N ) a \ U i ^ U i t O H - O l O O O \ l C \ ^ ^
The
military-civi ian tradeoff in Guatemala: an econometric analysis
371
Tablel5A.2(contin ed) Year
P
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
0.16717 0.16784 0.17097 0.17387 0.18256 0.18025 0.17782 0.20358 0.23549 0.26638 0.29699 0.34586 0.36486 0.39623 0.43594 0.47308 0.49674 0.56511 0.58127 0.65453 0.92595 1.00000 1.11655 1.23834 1.73895 2.31411 2.51918 2.86170 3.21638
Source: 5.
PI 0.12270 0.12660 0.12820 0.13210 0.13840 0.14100 0.14660 0.17220 0.23020 0.25650 0.29480 0.31130 0.33990 0.37840 0.42280 0.43720 0.44530 0.44760 0.47230 0.67670 0.84770 1.00000 1.11430 1.24130 1.89330 2.36010 2.66620 3.05180 3.24240
PM 0.10985 0.11368 0.11921 0.12560 0.13199 0.13921 0.14602 0.15541 0.16861 0.18477 0.19669 0.20989 0.22650 0.24650 0.26951 0.29677 0.31508 0.32782 0.46210 0.97646 1.03500 1.00000 0.93699 1.15960 1.93717 2.25333 2.35534 2.61055 2.72240
PWT
RF
0.26050 0.26360 0.26920 0.27760 0.28950 0.29740 0.30880 0.34810 0.42180 0.45410 0.48790 0.52780 0.54690 0.60620 0.69740 0.75760 0.79960 0.83420 0.88090 0.92410 0.95110 1.00000 1.09450 1.23930 1.42370 1.55320 1.70020 1.90380 2.17680
1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.35000 2.75000 2.84000 2.66000 2.40000 2.84000 4.55000 5.05000 5.18000 5.62000 5.74000
Compiled by Thomas Scheetz mostly from domestic sources; see text and note
372 Arming the South
Tablel5A.2 (continued) Year
TENS
Dl
D4
D5
D4D5
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
0.50000 0.50000 0.50000 0.70000 0.70000 0.70000 0.70000 0.70000 0.80000 0.70000 0.70000 0.80000 0.90000 !.00000 1.00000 1.00000 1.00000 0.70000 0.60000 0.60000 0.50000 0.40000 0.40000 0.30000 0.20000 0.20000 0.20000 0.20000
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 1.00000
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 0.00000 0.00000
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
Source:
in the study.
16 The Allocation of Resources to the Armed Forces in Chile: A Case of Limited Transparency Guillermo Pattillo
Introduction The purpose of this chapter is to describe the process by which resources are allocated to the armed forces of Chile. Among the many alternative approaches available, I consider the allocation mechanism only from the point of view of transparency. This means that I analyze processes and procedures more than budget numbers. Also, I advance a diagnosis more than proposals for solution. The common denominator of resource allocation to armed forces in South America is their relative independence from effective social control. Even though civil authorities formally possess the ability to make budget allocations, for a variety of reasons these allocations ultimately occur to a great extent outside their scope of influence, thus lending the armed forces some kind of special status. This special status reduces social welfare and undermines the armed forces' legitimacy in society, promoting conflict and distrust. I addressed transparency for the first time as an important issue in military budgets during a workshop held in Brazil on budgeting and
374 Arming the South
democracy, sponsored by the Lyndon B. Johnson School of Public Affairs of the University of Texas at Austin and the Getulio Vargas Foundation (see Pattillo, 1998). The discussion at that time highlighted the low level of social involvement in the process by which fiscal budgets, and of course military budgets, are allocated in most of the South American countries. Although I survey here only the Chilean case, the issue is relevant because it is widespread, and the truth is that despite the defects of the Chilean situation, we are in this regard in a better position than most other countries in this part of the world. Transparency is an area where much remains to be done, and awareness about where we stand now is the first step to initiate action to improve public decision-making. Transparency always has been a desirable characteristic of public decision- making. But it has been only in the last two decades or so that this term has become widely used in South America, and a growing degree of awareness has arisen among the general public about their right to ask government for better information and to make it accountable for the way in which public policy is defined and conducted. The huge changes experienced during the 1980s and 1990s, with global interdependence becoming a reality of everyday life, have brought about significant restructuring of economic paradigms and of the perceived economic role of government. Certainly, these changes have had an important impact on the way that people today understand the functions of government and what they can and should demand of it. The old approach to government as an agent somehow endowed with special knowledge and only pursuing general social well-being is no longer valid. Government and its agencies are formed by people who respond to incentives, and they maximize a utility function whose arguments surely contains private objectives also. Fiscal transparency has thus become a leading concept and one that is highly correlated with the social image of valid decision-making. There is no universal consensus about the meaning of transparency and about the conditions that must be met before an action can be called transparent; nevertheless, in an effort to provide a set of basic guidelines for common understanding of transparency in public policy, the IMF published a Code of Good Practices on Fiscal Transparency (IMF, 1999). This code consists of four general principles (as quoted from the
Chile: a case of limited transparency
375
IMF http://www.imf.org/external/np/fad/trans): 1
The first general principle - Clarity of Roles and Responsibilities - reflects the importance of clear boundaries within government between fiscal, monetary, and public enterprise activities, and between the public and private sectors. However, the Code does not advocate any particular allocation of responsibility among government agencies.
2
The second general principle - Public Availability of Information - is concerned with the need for both comprehensive fiscal information and for governments to commit themselves to publish fiscal information at clearly specified times. The concept of comprehensiveness goes beyond that typically reflected in government budget and accounts reports. In particular, the Code emphasizes the need to report on any quasi fiscal activities that have been assigned to or otherwise undertaken by agencies outside general government.
3
The third general principle - Open Budget Preparation, Execution, and Reporting - encompasses traditional standards relating to the coverage, accessibility, and integrity of fiscal information. Considerable emphasis is placed on the development and harmonization of international statistical and accounting standards for government reporting.
4
The fourth general principle - Independent Assurances of Integrity - emphasizes the traditional means of providing such assurances through external audit and statistical independence, but then goes beyond this and calls for openness by governments to allow independent scrutiny.
The IMF defines fiscal transparency ... as openness toward the public at large about government structure and functions, fiscal policy intentions, public sector accounts, and projections. It involves ready access to reliable, comprehensive, timely, understandable, and internationally comparable information on government activities - including those activities undertaken outside the government sector - so that the electorate and financial markets can accurately assess the government's present and future financial position. [IMF, World Economic Outlook, Annex I, 1998.]
But transparency, from my perspective, is more than information: it is involvement; so I would state the meaning of transparency in government actions as a way of management, in which decisions are made that clarify to the public the objectives, means, and expected results of fiscal
376 Arming the South
initiatives, and in which there is at least indirect public involvement to generate the final decision. In this sense, the conclusion of this chapter is that the allocation of resources to the Chilean armed forces is not a completely transparent process, especially so in the case of arms acquisitions. The reasons for this, explored in the following sections, are related to at least one fundamental factor: the inability of political parties during the 50 years preceding the establishment of military government in 1973 to fulfill their role as a key component of the set of institutions that shape national defense policy. This gave rise to an institutional structure that does not promote transparency. Contributing to this outcome is the very slow introduction of modern management techniques in the public sector as a whole and the Ministry of Defense in particular. The first of the factors that helps us to understand the present situation is closely tied to the severe undervaluation by political elites of the functions performed by the armed forces and of their social value that used to exist up until the early 1970s. This translated into a sharp division between civil society and the military and resulted in a deteriorated social image of the military career. The consequences of this short-sighted approach were various: one prominent one was the exclusion in the early 1940s of the legislature from analysis and decision-making in the area of weapons procurement and, much later, the limitation of the scope of their ability to modify the armed services' annual budget. Another factor is the prevalence of non-optimal managerial processes. Management techniques used by public agencies, of which the armed forces are but one example, and particularly the budget process, have not been instruments for greater efficiency and transparency of public actions, but instruments for bureaucratic control. The chapter is organized as follows. The next section provides a short description of the organizational structure of the Ministry of Defense. The section thereafter describes the way in which the Chilean armed forces are funded. The final section summarizes the most relevant conclusions. The structure of the Ministry of Defense The Chilean Ministry of Defense (MoD) today is the result of a variety of
Chile: a case of limited transparency
311
factors that shaped the organizational history of the armed forces in Chile. Legal and external influences are predominant. From a structural point of view, the MoD is a kind of multidivisional structure1 that evolved from a group of formerly independent organizations to a unified structure that could better coordinate the services.2 As we will see in the following section, service unification has been only partially achieved. Even if the services had been fully brought under a single umbrella, the organizational structure was not designed to produce proper incentives for joint decision-making. Indeed, the first set of regulations about how the new MoD would function (October 1947) kept most of the relevant decision-making within the armed forces themselves, leaving the Minister of Defense without independent technical assistance (Diaz, 1996). The Ministry is composed of a group of autonomous "core divisions" (the armed forces and the police)3, as well as of the Directorate General of National Mobilization, the National Academy of Political and Strategic Studies, and a set of staff units of different levels and relevance. These staff units are: the Commanders in Chief Committee, the Superior Council of National Defense, the General Staff of National Defense, and the Undersecretariats for the Army, Navy, and Air Force.4 The Commanders in Chief Committee (JCJ) is the Minister's highest level advisor in matters that involve the armed forces. The committee was created in 1960, and its permanent working organ is the General Staff of National Defense (EMDN). The latter is the permanent working and advising organ to the Minister of Defense in aspects related to national defense and the preparedness and uses of the armed forces. It is a joint entity established in July 1942, and one of its main tasks is to put together an overall strategic plan to guide the services' war planning. The Superior Council of National Defense (CONSUDENA) is a consultant organism of the Minister of Defense. Today, it has the task of being the comptroller of financial flows associated with weapon acquisitions funded by the Copper Law (see the subsection on the Copper Law later on). CONSUDENA was created in 1942 to effectively coordinate decision-making, but there is no evidence that it has ever accomplished this missions. Finally, on the staff side, the Undersecretariats are administrative assistants to the Minister, their heads being directly appointed by the
378 Arming the South
Joint Chiefs ofStaff CONSUDENA
-^Ministry of^-Defense
Armedforces
General Staff of National Defense Undersecretariats
Police forces
Figure 16.1: Chile: the Ministry of Defense structure
President of Chile and of his exclusive confidence. Also under the Minister's authority are the Directorate General of National Mobilization and the National Academy of Political and Strategic Studies (ANEPE). The former is the agency mainly in charge of general mobilization planning and conducting the annual military draft (i.e., the recruitment of those required to present themselves for obligatory military service).5 ANEPE is a postgraduate center attended by people from various departments and government agencies to obtain formal training in security and defense issues. The academy was founded in 1982, being the successor institution to the Superior Academy of National Security (1974) and, even earlier, the National Defense Academy (1947). Figure 16.1 summarizes this overall structure. As can be seen from figure 16.1, there is no defense agency in charge of common supply and support. This is a distinctive characteristic of ministries of defense in South America. In the Chilean case, the structure of MoD has remained unchanged since the early 1960s, when there was still a strong influence by the United States.6 Similarly, the structure of the Ministry of Defense does not have in the line of authority any joint military organization that could effectively coordinate armed forces' planning and operation. In theory this is the role
Chile: a case of limited transparency
379
of the General Staff of National Defense, but as a staff unit its capacity to promote jointness is severely limited. To understand that, two points must be taken into account. First, the Chief of Staff (a position that is rotated among the services every two years) is an armed forces staff specialist flag officer with the rank of Major General, Vice Admiral, or Air Force General who directly depends on the Minister and whose authority is in direct proportion to the Minister's willingness to be involved in substantive military matters (something that has been more the exception than the rule). Secondly, the Chief of Staff in some way must report to his service and to his Commander in Chief because after two years in office he will resume his professional career there. In practice, then, key aspects of the services' management and development strategies are conducted by them without much interaction amongst them, and the decision system implied by the MoD structure does not provide the correct incentives to a joint approach to defense issues. The result is a partially integrated defense organization where the different components operate more as independent entities than as parts of a closely interrelated system whose product is not simply the sum of individual contributions but a complex function of them. Because jointness is not a true characteristic of defense management, the potentially synergistic effects of the mixture of capabilities and unity of effort are absent. The evident weakness that the systemic approach to defense issues in Chile has implies that the efficiency of resource allocation is sub-optimal, but this is not all. It has also an effect on transparency because it motivates a less open decision-making process. If an improvement is to be made in the quality of the latter, the organizational structure of the Ministry of Defense must be changed significantly; a matter that is beyond the scope of this chapter. Funding the armed forces Two main sources fund the armed forces (Army, Navy, and Air Force) in Chile. They are (a) the fiscal contribution through the annual budget law that is approved by Congress and (b) a secret law that imposes a tax on the state-owned copper production company Corporacion del Cobre
380 Arming the South
(Codelco). The latter is known as the Copper Law.7 The funds provided by this law are never consolidated with annual budget data so that a comprehensive picture of the annual funds available to the armed forces does not exist. The fiscal contribution included in the annual budget law considers resources in domestic currency, and in US dollars, for the services' operations but not for investment in capital equipment, especially not for arms procurement. What Congress approves each year therefore are funds to cover the set of items that finance the planned uses of existing equipment. The most important of those items is, of course, personnel. The income generated by the Copper Law tax is used to acquire weapon systems and their spare parts and may not be used for anything else. Of course, in any particular year the value of arms bought and imported by the services can be much greater than the proceeds of the copper tax. In this case, the services habitually borrow from suppliers to finance their purchases. But there is no public record of that debt, neither of the actual level of indebtedness, nor of its structure, nor of the conditions under which it was contracted. My conjecture (but I cannot prove it) is that the services are close to the maximum debt they can service insofar as resources for amortization and interest payments also must come from copper-tax income alone.8 Because there is an upper bound to the absolute debt level, credit does not constitute an additional source of financing in the long-run. In the last decade (the 1990s), total funds allocated to the military from both sources, i.e., the annual budget law and the copper law tax, have been declining when measured against GDP. They fell from almost three percent of GDP in 1989 to around 1.8 percent of GDP in 1999 (Ministry of Defense, 2000; also see figure 16.2). The next subsection summarizes the mechanism by which these two types of resources are allocated to the services. The budget cycle The budget cycle for the public sector (see figure 16.3) begins in April and ends with congressional approval of the budget law on the last working day of November of the year preceding the fiscal year. (In
Chile: a case of limited transparency
381
% 3.50 -I
3.00 2.50 2.00 -
1.50 1.00 \
• , , , r-, , r— , -, r 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Figure 16.2: Global fiscal income to the armed forces, 1989-1999 (as percentage of GDP) Source: Ministry of Defense, 2000.
Chile the fiscal year coincides with the calendar year.) Of the four major budgeting stages - executive preparation, legislative consideration, execution, and control - the focus here is on the first two. The Budget Department of the Ministry of Finance centralizes the preparation of the budget (Balbontin and Chamaca, 1999). This process takes place between April and June and ends with two products, (a) the projected fiscal position for the next three years and (b) a projected budget by ministries and public agencies for the following year. The latter is constructed by each government agency according to financial guidelines provided by the Budget Department. Because the projected budget is normally less than what agencies requested, they can at this stage present what is called an "expanded budget". This has, however, a crucial limitation: any request for additional monies must not only be extensively justified but must have its own financing. This implies that additional expenditures must be financed with agencies' own resources, e.g., with the sale of its services or assets, an alternative that is generally not available to the armed forces. After consolidation of all requests, there is a period of administrative hearings during which every budget request is analyzed. This phase,
382 Arming the South
Ministry of Finance " ^ Budget Department
Financial guidelines
H_ Financial requests
Government agencies and departments
Budget Department Proposal (BDP) \ Analysis of BDP by Cabinet \ Proposed budget law Finance Committee of the House of Representatives Specialized subcommittee hearings Bicameral Finance Committee consolidation House of Representatives approval Senate approval ^
Executive promulgation of the Budget Law
Figure 16.3: Chile: the budget cycle
during which specialists from the Budget Department interact with the agency under review, produces the Budget Department Proposal. A final revision of this Proposal takes place at a Cabinet meeting. Once the President approves, a proposed budget law is sent to Congress at the beginning of October. After the executive has submitted the budget proposal to the
Chile: a case of limited transparency
383
legislature, the Finance Committee of the House of Representatives, which receives it, splits the budget into as many parts as agencies will appropriate resources and submits those parts to subcommittees that are formed from members of the Senate and the House. These subcommittees initiate a period of hearings with government departments and agencies. Each subcommittee produces a budget document submitted to the Bicameral Finance Committee, which consolidates the information and, in turn, submits the final document for approval to the House of Representatives. After the Lower House has approved the appropriations, the proposed budget law is sent to the Senate. This has to give its formal consent on or before the last working day of November because the President must sign the budget law on the first working day of December. So far this budgetary process looks very standard. Nevertheless, and even at this global level, there are two important differences to the way things are done in other countries. In Chile, the legislature does not have the ability to reallocate resources within an agency's budget, nor can it increase the level of spending the executive has proposed for any agency. What then is the legislative analysis of the budget all about? It is basically to study the consistency between agencies' stated objectives and requested resources, with the possibility to reduce funds only when more than needed has been requested and the agency's level of spending is not determined by another law. The second limitation is that if Congress does not approve the budget within 60 days, the initial executive proposal will automatically become law. Clearly, there is an imbalance between the executive and legislative branches of government. Chile has a presidential system, but in sharp contrast with other countries ruled by the same general system, such as the United States,9 the Chilean constitution gives much power to the President to the detriment of Congress. The reasons for this are twofold: an historical tradition that favors a strong presidency, and the prominence of the figure of the President at the time the current constitution was written (the late 1970s, during the military government). The budgetary process described here applies to every fiscal agency in Chile. But in the armed forces' case, its budget approval process is subject to an additional and unique restriction in the Chilean public sector.
384 Anning the South
On 27 February 1990, two weeks before the military government left office,10 it enacted the Constitutional Organic Law ofthe Armed Forces (law number 18,948) which mandates, among others things, that fiscal contributions in domestic currency to the services should be, from then on, equal to what they had received in 1989, adjusted each year by the increase in the consumer price index (CPI), with the exception of personnel expenses which should be adjusted at least in line with the rest ofthe public sector (normally more than the CPI variation).11 This law, which requires the approval of 4/7 ofthe House of Representatives and the Senate to be modified, was enacted on the expectation that civilian governments would try to severely reduce funding for the armed forces.12 What maneuver room is left to Congress - or, indeed, the executive - in deciding on budget allocations to the armed forces? Not much indeed. The executive can increase, but not reduce funding. Congress cannol increase or reallocate resources and has a floor to how much it could reduce funding. At a global level the funds the armed forces receive are mainly defined, then, by indexation ofthe 1989 budget and salary increases. Although these limitationsper se do not necessarily imply a reduction in transparency, in practice they contribute to it. Because there is a floor to the amount to be allocated, the armed forces do not need to go into an extensive and in-detail justification ofthe budget each year.13 Another factor that makes the budget process opaque is that in Chile the fiscal budgel is an object-of-expenditure budget - i.e., items are budgeted by categories such as personnel, food, and purchases of goods and services - not a performance budget or a program budget.14 It is therefore impossible to tell from the budget what activities will be performed and how much they will cost. In the case ofthe armed forces, activity and cost are not necessarily known even by Congress, much less by the general public The state of limited and fragmentary information could be overcome if the administration were committed to produce the relevant information, but so far this has not happened. Another central question is how the defense budget is divided among the services. Here emerges another rigidity that is particularly binding in the case ofthe armed forces. The funding of each service is calculated using what is euphemistically called "historical criteria". In practice this
Chile: a case of limited transparency
385
Table 16.1: Chile: services' share in fiscal income in selected years (in percentages, Copper Law included) Year
Army
Navy
Air Force
1994 1998 2000*
41.55 42.44 42.06
34.14 33.82 33.87
24.31 23.74 24.07
Note: * projected Source: Ministerio de la Defensa Nacional de Chile 1994-2000, March 2000.
means that the fraction of the total that is allocated to any particular service today is fairly constant over time and represents its share ofthe budget over some undefined period ofthe post-1973 years. It is interesting to note that before 1973 the services' total budget share was very different from today's shares. The turning point took place in 1975, the last year in which the Navy had the biggest share (36.9 percent), followed by the Army (36.4 percent), and the Air Force (26.7 percent). Since 1976 the Army assumed the budgetary forefront and has remained in that position (see table 16.1). The way in which the defense budget is split has an additional and very undesirable property: it freezes the structure ofthe services in the past. This means that the structure of national defense does not respond to contemporary events, but is artificially locked in the past, reducing the social value ofthe product that the armed forces produce. Nevertheless, relative to GDP, the budgetary allocation to the services has been declining throughout the 1990s and today is around 1.4 percent of gross domestic product (see figure 16.4). Although it is well understood by public officials that budget processes should be a central tool to efficiently allocate government resources, control agencies' performance, and improve social welfare, prevailing practices in Chile do not allow, in general, full use of this instrument. This is particularly true in the case of Chile's armed forces. Military budgets should be clear statements of plans, priorities, cost, and performance, but in Chile they are not.
386 Arming the South % 1.90
1.80 — ^ — 1.70
^ ^ * ^
1.40 j
-S^-
1.30 I 1.20 i
^ ^ ~ " ^
""^ —
^
. , • • , , , • , , • 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Figure 16.4: Chile: fiscal budget allocation to the armed forces (as percentage of GDP), 1989-1999 Source: Ministry of Defense, 2000.
The tax on copper sales During most of the 20th century, and without doubt during the period from the early 1920s to the 1970s, weapon acquisition for the armed forces in Chile was, to a great extent, a process characterized by government improvisation and forced by international events that threatened the territorial integrity ofthe country. In general, the political leadership of those days considered investing in the armed forces a low priority, and they invested only when it was clear that there was no alternative. Usually this meant that the services would be reequipped only when some strategic imbalance challenged national security. Because there was no rational plan to maintain the capabilities ofthe armed forces, there was no planned way of equipment financing, and special funds needed to be raised each time a new threat appeared. This was the case with a classified law known as the Cmisers Law of 1938. It was enacted in response to the Argentine naval plan of 1937. This was the case, too, years later, with the Copper Law (see Meneses, Duran, and Neeb, 2000). But the Copper Law has another vital precedent. In 1942, enactment
Chile: a case of limited transparency
387
of law number 7,144 created the Superior Council of National Defense (CONSUDENA) and established new sources of funds for the armed forces, different from the fiscal contributions included in the annual budget law. The new funds had two distinct origins, according to the currency in which they were raised. Money came in US dollars from income fiscal authorities raised from the multiple exchange-rate system that prevailed for copper exports.15 In addition to this, domestic currency funds were raised by means of a tax on tobacco and wine. But more than for the funds themselves, CONSUDENA was important for the authority it received to approve the uses of those funds. This has had a long-lasting influence on the way defense acquisitions are handled in Chile. In 1958, having extinguished the source of extra-budgetary dollars created in the 1940s and partly motivated by an incident with Argentina over the sovereignty ofthe Snipe islet in the Beagle Channel, Congress enacted a law that imposed a tax on profits ofthe copper mining industry and allocated the proceeds exclusively to the acquisitions of weapon systems.16 In those days the tax rate would have been 7.5 percent of profits and the funds were probably allocated not to the armed forces directly but to the Superior Council of National Defense for distribution among the services according to some set of criteria.17 The intention of the authorities in those days (as in the precedent cases already mentioned) was to establish a permanent flow of US dollars to the armed forces, independent ofthe changes in political willingness. Following tradition, the law was kept secret. Nevertheless, as time went by at least part ofthe law became gradually known for a variety of reasons (and by a variety of means), so that today we can be confident that its important aspects are no longer secret. For sure, there are details that remain unknown but the basic stmcture ofthe post-1973 versions ofthe law are now believed to be in the public domain.18 Apparently the law did not contain any instmction for how to transfer the resources to the services, and for the first two years for which published figures exist (1963 and 1964) the fraction going to any particular service was variable, although in a small range.19 But starting in 1965 a new pattern emerged: the proceeds were divided among the services in equal parts, so each one of them received a third ofthe total. Nobody denies that for decades the armed forces made due with
388 Arm ing the Sou th
chronic underinvestment, and the income generated by the Copper Law from 1959 to 1973 did not change that in any substantial way because the funds produced by the law were insufficient to fill the huge gap that already existed. By the time the military government took power in September 1973, the military capabilities ofthe services were therefore severely limited for lack of modem equipment. By the end of 1973 the new, military, authorities introduced significant changes to the Copper Law. Today, the most relevant and best known changes include the increase in the tax rate from 7.5 to 10 percent, the enhancement ofthe tax base from profits to gross yearly sales of stateowned copper mine exports (at that time the whole industry), the establishment of a US$90 million minimum income for the armed forces,20 the explicit division in thirds of the total, and the direct allocation of that amount to each service. In turn, each service would put forth US$ 3.3 millions a year to CONSUDENA to finance joint projects in which more than one service could be interested. The reasons behind these changes were twofold: (a) to increase funding for the armed services and (b) to prevent disputes among them at a time when their Commanders in Chief were, simultaneously, the legislature (the military junta that replaced the civilian Congress in 1973 was formed with the Commanders in Chief of the armed forces and the General Director ofthe police force). Both purposes were achieved. The changeover ofthe law greatly increased the resources that the armed forces received for the purpose of arms acquisition. During the period 1963-1972 the annual average of total Copper Law proceeds was equivalent to 0.32 percent of GDP, with a standard deviation of 0.11; in the 1974-1985 period that average total income was equivalent to 0.78 percent of GDP, with a standard deviation of 0.21 (calculated from Ministry of Defense, 1998, p. 212). Relative to GDP, Copper Law proceeds increased by 144 percent. In nominal terms they reached a maximum of US$ 31,130,012 in 1969 and declined to less than 10 million dollars in 1971.21 With the change in the law at the end of 1973, the services' income jumped to US$ 102 millions in 1974.22 But the final version ofthe law had not yet been written. In October 1985 the law was modified again and at that time, as far as is known (Villar, 1992; O'Reilly, 1997), the main modifications were an increase
Chile: a case of limited transparency
389
% 1.40 -I
1.20
1 .oo 0.80 -j 0.60
^ V V N.
0.20
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Figure 16.5: Chile: armed forces' income generated by the Copper Law (as percentage of GDP), 1989-1999 Source: Ministry of Defense, 2000.
by 100 percent ofthe guaranteed minimum income to the armed forces to US$ 180 millions, the indexation of this base income to the wholesale price index of the United States, and the inclusion of copper's byproducts in the tax base. From 1986 onward therefore the resources provided to the armed forces jumped to a new level, which continues to the present. They represent today around 0.4 percent of GDP (see figure 16.5). This is the story of how the Copper Law evolved and the amounts that it generated for the armed services. But for my purposes it is more important to describe the features the allocation process has adopted. The first characteristic of that has already been mentioned: the money collected by this tax is divided in equals parts among the services and is distributed directly to them with a marginal amount given by the services to CONSUDENA. This approach has, of course, a negative impact on the efficiency of resource allocation. Since the services do not have to compete for resources, there is nothing that can assure that, from the point of view ofthe national defense system, the best projects are undertaken. It istmethat this type of partial analysis has historically been the way
390 Ann ing th e Sou th
resources were allocated to the armed forces in Chile. But the Copper Law institutionalized this way and made it permanent. When the law established that the tax proceeds would be given directly to each service, it created a huge degree of independence among the armed forces regarding the way they purchase equipment and the way in which they make acquisition decisions. Of course, efficiency and transparency are not necessarily correlated. It is possible to have a transparent allocation process that produces an inefficient outcome, and vice versa. What defines the degree of transparency is the way in which different actors are taken into account at the various stages ofthe decision-making process, who they are, and what level of information is available at reasonable cost to the public at large. It is from this point of view that we have to analyze the allocation process ofthe Copper Law. To summarize what follows, for acquisition an appropriation decree must be signed by the Minister of Defense and the President, so the approval of these authorities is essential. But this is a formality; the relevant question is how deeply the armed forces allow civil authorities to be involved in the analysis of weapon acquisition and their real power of veto. The Ministry of Defense produces basic guidelines to armed forces' planning by means of its periodic strategic assessment and projections. This is made explicit in the Political and Strategic Global Appraisal23 which is written by the Ministry of Defense with input from the armed forces, the Ministry of Foreign Affairs, and other institutions. Under this umbrella the services plan (or should plan) their operations and investment. Because the Copper Law produces funding for the latter, the basic question here is by which mechanism an intended purchase of a certain weapon system becomes an actual purchase. What we know ofthe process suggests that the origin of any arms acquisition project is found in the service that will operate it. There is nothing surprising in this, but because the services receive arms funding independently of potential acquisitions, and ofthe requirements of other services, they do not need to coordinate their intended acquisitions with each other. Thus, the first characteristic ofthe decision-making process is that a joint analysis is generally absent. Once the preliminaries are sufficiently developed by the service
Chile: a case of limited transparency
391
involved, and assuming that it can finance the intended acquisition with its expected net income from the Copper Law,24 two scenarios are possible. If the weapon system under review is particularly relevant because of political, strategic, or other considerations, the service will have to discuss it with the Minister of Defense and his closest advisors, mainly to analyze its political viability. It is my impression that not much discussion takes place about the specific military role that the new weapons are supposed to fulfill or what the alternative options might be. But if the purchase is a relatively minor one, it is more likely that the Minister will merely be informed ofthe service's intentions. In any case, there is a flavor of formality to this process that probably dominates its content. We do not know if this is an accurate description of what really happens because we do not have the required information; but it is likely because the services view the Copper Law funds as their "own", so there is not much that others can say about it, and they generally do not acknowledge that civilians may possess the technical ability to question such projects from a substantive point of view. A recent case illustrates the point. In mid-June 2000, a senator denounced in Congress that the Air Force had recently bought a Gulfstream-IV airplane for non-military uses, this being a non-affordable luxury at a time of severe fiscal constraints. Independent ofthe validity ofthe basic argument, the point is that the very existence of this purchase was completely unknown for a non-insider and of course for public opinion. The first notice of it was that speech in the Senate. In reply the Air Force said that the airplane was operating since May 2000, replacing a Gulfstream-III airplane, that they asked the pertinent authorities for permission, that the airplane was bought with services' "own" resources (i.e., with money from the Copper Law), and that the reserve with which this replacement had been treated is only "normal and legally valid in acquisitions for the armed forces" (El Mercurio, 14 June 2000, p. 16). A completely different scenario arises when the projected Copper Law income is insufficient to finance a project and additional funding is needed. In this case, the earliest participation ofthe Minister of Defense and the Ministry of Finance becomes essential.23 But these cases are exceptional and until very recently there was no public record of any example were additional funding was sought.
392 Arming the South
The state of affairs, at least thus far, is one in which the services finance the acquisition of any desired weapon system with Copper Law proceeds and do not receive any additional financial support. In this scenario, the opinion ofthe Ministry of Defense in the early stages ofthe project cycle is much less relevant. The service will evaluate projects from a technical and financial standpoint. This results in a prioritized list of options from which each service's authorities will choose. They will discuss their selection with the Minister and, that done, the selected projects are sent for approval to the Superior Council of National Defense. The Council is, in theory, but not in practice, the most important actor in the decision-making process. CONSUDENA is formed by the Minister of Defense, who is the chairman ofthe Council, the Ministers of Foreign Affairs and Finance, the Army, Navy, and Air Force Commanders in Chief and their Chiefs of Staff, the Chief of the General Staff of National Defense, and the War, Navy, and Air Force Undersecretaries. Among the functions that the law gave to the Council26 there are at least three that are of huge importance for the equipment acquisition decision-making process. These are: 1 2 3
to study and establish the needs ofthe armed forces and propose the acquisitions and innovations necessary to satisfy them, to authorize the referred acquisitions and investments, and to control the acquisition process.
There is no public information about how these functions have been undertaken by CONSUDENA, but case analysis suggests that the role played by the Council has been much more one of formal approval than of analysis ofthe way in which each branch's project would support its mission and about the contribution this would make to national defense. There is strong support for this hypothesis in the Copper Law itself since it gives the funds, in three equal parts, directly to the services. Since each service has then a property right over a certain level of funding, and because nobody else will help finance the investment it wants, nor the operational cost associated with it, there appears not to exist much room to maneuver for the Council to question any service's project. Furthermore, the resources that a service appropriates by means ofthe
Chile: a case of limited transparency
393
Copper Law do not compete with other public uses of the same funds. They are specific to the armed forces,27 so the most probable scenario at the Council is one in which the Ministry of Finance cannot stop an expenditure with considerations about its opportunity cost; and because the funds "belong" to the services, there is no way that the other two services can argue against a desired project of its fellow service if they want their own to be approved in the future. This leads to the hypothesis that the only objection that the Council could possibly raise is a political one. That has happened, but only in major acquisitions and not really by the Council, but by the President himself28 The Copper Law thus provides decision-makers with incentives to look at parts, not at the whole system. If this is the case, and everything points to that, the role of CONSUDENA is essentially a formal one: to produce the administrative act of approval of any project that the services send for its consideration. It is even hard to believe that, because ofthe rank ofthe members ofthe Council, it ever really achieves a high quorum or full attendance. What CONSUDENA surely does is to engage in the financial control of payments ordered by the services (the third function stated above) in fulfillment of contracts from previous acquisitions and to check the financial feasibility of new projects. It is with the Council's secretary where the administrative control of all financial flows generated by the Copper Law is located. After the service has CONSUDENA's formal approval for an acquisition, a decree is sent for signature to the Minister of Defense and the President. When this administrative act is completed, the service can finally sign the contracts with the firm that will provide the equipment and can start using the money it has in secret accounts at the Treasury. The only external scmtiny of the whole process is an ex-post administrative control made by the General Accounting Comptroller (GAC), an independent public institution in charge of reviewing the legal soundness of administrative actions of public employees. Nevertheless, the control exerted by GAC is limited in scope and superficial in nature, and its results are classified information for everyone except the organizations directly involved. The entire decision-making process, from beginning to end, includes
394 Arming the South
then the service that owns the project, the Minister of Defense, with whom some level of negotiation may take place, CONSUDENA, the executive, and the General Accounting Comptroller. Figure 16.6 summarizes this process. Obviously, the missing actor here is Congress. Although it may seem unnatural that the legislature does not have any involvement in the process of deciding which type of weapon system the country should and will have, this is in fact the way things are handled in Chile today. But this is not new; this dates back to 1942. Why was Congress eliminated from decisions that are so cmcial for the country's long-term survival and position in international affairs? The short answer is distmst and lack of interest of political parties in defense matters. For most of the 20th century, the relation between the armed forces and political parties in Congress was complex. The military tended to see politicians as intrinsically opportunistic and focused, essentially, on short-run election cycles. With opportunistic political parties,29 and in the context of voteassociated demands from a variety of social groups, the expected outcome is a very low priority for defense issues and expenditures. This characterizes the situation that the military faced for decades until 1973. They were poorly paid and worse equipped. Surely nobody would deny now that the relation between the military and politicians was inadequate and that the mistake lasted too long. It is fair to say, then, that Chile's history validates the fear ofthe military that Congressional discussion of weapon acquisitions - implying by definition a long-term view of national interests - could be not appropriately treated and could undermine their operative capacity first and their role and very existence, second. Things are changing and civil-military relations have started to be built on different assumptions, but a long time will have to elapse to modify peoples' firmly rooted perceptions. Nevertheless, in the long-mn the only way to socially validate the resources consumed by the armed forces is to give real control over military expenditure to elected representatives ofthe community. To advance in that direction extensive reform is needed in procedures, organizational design, and in the roles different participants play in the
Chile: a case of limited transparency
Navy
•Army
GAC
395
Air Force
" • CONSUDENA
Ministry of Defense
Ministry of Finance
Presidency Figure 16.6: Chile: The decision-making cycle of Copper Law funds
process of determining national defense goals and stmcture. This reform is, probably, still far away because of lack of confidence and political leadership, but surely at some point in time all parties will realize what is now obvious only for a minority: that the reform of decision-making in defense is a Pareto-improving measure. Conclusions The allocation of fiscal resources to the Chilean armed forces is not a transparent process since it occurs almost completely outside the scmtiny of society and is to a great extent independent of social preferences. No prescribed minimum level of discussion exists at the various stages ofthe allocation mechanism to allow a reasonable understanding of the objectives being pursued, the opportunity cost incurred, the benefits accmed, and to identify those who should be accountable. At the most general level the explanation for today's situation lies in the incapacity shown by all political parties, for decades, to fulfill their role as tmstful counterparts in the decision-making about defense issues.
396 Arming the South
The prejudice with which matters of armed forces were handled in Congress, and the significant undervaluation of the social value of the public good that these institutions produce, at last brought into being the self-exclusion ofthe legislature from the arms procurement process in the early 1940s and the limitation of its capabilities to decide upon other military matters later on. In light of this background, the structure ofthe system that evolved is, in the end, a reflection ofthe most-promising looking strategy for each ofthe different groups in conflict. Behind the veil that shrouds defense issues is the prevalence of an institutional stmcture that does not generate the proper incentives from a national point of view, and in particular does not motivate an open discussion of military expenditure. To the contrary, it helps to hide them from the analysis that should take place if maximizing the social value of defense products were the effective ultimate goal. The long history of distmst between the armed forces and political parties was the leitmotif of the relation between the military and politicians during the period of military mle and was the basis for enacting the Constitutional Organic Law ofthe armed forces in 1990 that forbade future reductions in military expenditure (from domestic currency resources). And it was, at least partly, the reason, too, for the design of a less powerful Congress in the constitution of 1980. An additional factor that contributes to limited budgetary transparency is that Chile's fiscal budget is of an object-of-expenditure type so it is impossible to numerically disentangle what activities will be performed and how much they will cost. Two things would make the allocation mechanism more transparent: (a) to strengthen Congress' ability to oversee all military expenditure operational funds and weapon acquisitions - and (b) to produce an understandable, regular, and timely accounting about the use of taxpayers' funds put forth to finance national defense. These look like easy measures, but they are not. In fact they imply a drastic restructuring of Chile's defense organization and procedures and, more importantly, surmounting a history of mistrust and mutual disqualification. Only in the long-mn can we expect to see new processes and defense institutions in Chile, but with a certain dose of political willingness and
Chile: a case of limited transparency
397
determination the authorities in the Ministry of Defense could initiate even now a significant move in the appropriate direction, making public what should have always been public: disaggregated information about defense income, expenditures, and resource stmcture. This is a role for which the Ministry of Defense has no substitute. Notes I thank Roberto Duran, Margaret Daly Hayes, Miguel Navarro, Thomas Scheetz, Paolo Tripodi, and the editors of this book for very helpful comments. Any remaining errors are my own. 1. A multidivisional stmcture (or M-form) is characteristic of firms that decide to organize themselves by product lines, regions, or customer types. In business, it grew up as a response to problems with functional stmcture in large, diversified firms. 2. The Ministry of Defense was created in 1932 subsuming the Secretaries of War, the Navy, and the Air Force. 3. Although police forces are grouped under the Ministry of Defense, their operational coordination lies with the Ministry ofthe Interior. 4. For a complete description of the defense stmcture see Ministry of Defense (1998, part HI, chapter 4). 5. Other functions of this agency are weapon and explosive control, chemical and noxious weapon control, and martial arts control. See Ministry of Defense (1998, part III, chapter 4). 6. For a brief history of that influence, see Donley, 1995. 7. There are two more sources of financing, operational income (referred to by the services as "own resources") and a revolving fund (FORA). Operational income is received for the sale of goods or services to military personnel and their families (e.g., medical services) and to firms (e.g., maritime security). The revolving fund is not included in the budget
398 Arming the South
law and derives from the sale of goods (e.g., maps) or services (e.g., medical assistance) to the general public. Although the relevance of these funds in total financing differs by service, jointly they do not comprise more than 20 percent of total services' income, with FORA making up the smaller part. 8. This conjecture derives mainly from media discussions about the amount of additional money the Navy and the Air Force are asking for to purchase new equipment as compared to the expected value (in statistical terms) of copper-tax income I projected for the same period. 9. I am grateful to Margaret Daly Hayes for emphasizing this point. 10. The military government lasted from 11 September 1973 to 10 March 1990. The democratically elected government took office on 11 March 1990. 11. The law did not index fiscal contributions in US dollars. Today, the dollar income represents almost 10 percent ofthe armed forces' total fiscal income. 12. In fact, the coalition that was to take office in March 1990 was formed by those that had been in opposition to the military. This was a crucial factor for the military's lack of fiscal confidence in civilian government. 13. Of course, this does not apply if they are trying to get an increase over the CPI variation. 14. In object-of-expenditure budgeting, expenditures are organized according to commodities or resources purchased; in performance budgeting, expenditures are classified by work load or activity; and in program budgeting they are related to public goals (see Mikesell, 1991). Chile once had a program budget from 1959 to 1966 on an experimental basis, and extended it to the entire public sector from 1967 to 1970. However, the results did not convince its opponents and the interest in program budgeting was abandoned (see Petrei, 1997).
Chile: a case of limited transparency
399
15. In 1955 the multiple exchange-rate system came to an end and with it the flow of dollars to the services. 16. This was a classified law, published in November 1958 (law number 13,156). 17. Because ofthe classified character ofthe law, we have to speculate here using fragmentary public information, trying to identify what could have been, or are, the central characteristics ofthe Copper Law. 18. To cite some ofthe most useful references: Carlos Villar, Gasto Militar: Tema de Reflexion, published in the Memorial del Ejercito, July 1992; an editorial of El Mercurio, 1 November 1993; an article by Vice Admiral Patricio Carvajal in El Mercurio, 2 February 1994; and a recent article by Juan Pablo Lorca, "Economia de Defensa y de Guerra," in Revista de Marina, January-February 2000. 19. In 1963 the Army got 36 percent of total funds, the Navy 26.9 percent, and the Air Force 37.1 percent. In 1964 those percentages were, respectively, 37.3, 28.1, and 34.6. 20. And if the proceeds of the law are less than US$ 90 millions, the Treasury would have to make up the difference. 21. In 1972 and 1973, the then-ruling socialist government did not transfer the funds to the services. 22. This is based on figures obtained from a paper by Col. Carlos Villar, Gasto Militar: Tema de Reflexion, published in the Memorial del Ejercito in July 1992. Obviously, a more adequate comparison should include figures in real terms, but it is extremely difficult to know which is the proper index to translate the purchasing power of a dollar's worth of guns and bullets in the 1960s to today. In any case, that index is not the US wholesale price index (WPI), nor the CPI, and not even a specific index of military equipment prices because what Chile could buy in the sixties was not only a problem of price but that developed countries wanted to sell some of their near obsolete equipment. Because my intention here is
400 Arming the South
only to give a rough measure ofthe scale-effect ofthe 1973 law, I kept figures in nominal terms so as not to introduce additional distortions. 23. In addition to the obstacles that might be encountered when pursuing national objectives, the Political and Strategic Global Appraisal is a systematic assessment of the country's domestic and international situation. From this assessment stem opportunities for cooperation and conflict. See Ministry of Defense (1998) for a description. 24. Because weapons are purchased on credit, all services incur debt that must be paid with Copper Law proceeds. There is no public information at all about this debt, its level, maturity stmcture, rates charged, and so on. These are entirely in the domain ofthe services and CONSUDENA. 25. This is the case now with the new fighter that the Air Force plans to have for the next decade, and the frigates the Navy plans to build. 26. Law number 7,144 of 1942 created CONSUDENA and its functioning was regulated in March 1958. A complete analysis ofthe structure ofthe whole institutional system of national security in Chile can be found in Guzman, 1986. 27. This is unique in Chilean legislation because by constitutional mandate taxes cannot be raised for specific purposes. 28. In 1 998 the government decided to postpone the acquisition of a new aircraft for the Air Force. The reason given was that because spill-over effects of the Asian financial crisis were contracting the Chilean economy, and thereby public sector resources, the budget could not afford additional expenses. That this argument could be raised at all at the time was because the Air Force did not have enough financing with its own money from the Copper Law and needed additional funding. Had this not been the case, it would have been much more complicated for the President to postpone the purchase ofthe aircraft. 29. Opportunistic as defined in the economics literature on political business cycles (see, for instance, Nordhaus, 1989).
Chile: a case of limited transparency
401
References Balbontin, Marcelo and Alvaro Chamaca. Sistema de Administracion Financiera del Estado, El Subsistema Presupuestario. Unpublished paper for the Defense Economics course, Master of Political Science, Institute of Political Science, Catholic University of Chile, June 1999. Diaz, Jose. "Administracion de la Defensa: la Experiencia Chilena en el periodo 1932-1973." Fuerzas Armadas y Sociedad Vol. 11, No. 4 (October-December 1996), pp. 6-14. Donley, Michael. "Problems in Defense Organization and Management." Joint Force Quarterly No. 8 (Summer 1995), pp. 86-94. El Mercurio. Various articles and editorials during the 1990s. IMF. World Economic Outlook 1998. Washington, DC: IMF, 1998. IMF. Manual on Fiscal Transparency. Washington, DC: IMF, April, 1999. Guzman, Enrique. "Formulacion Constitucional de la Seguridad Nacional." Unpublished thesis at the National Academy of Political and Strategic Studies (ANEPE), 1986. Lorca, Juan. "Economia de la Defensa y de Guerra." Revista de Marina Year CXV, Vol. 117, No. 854 (January-February 2000), pp. 19-29. Meneses, Emilio, Roberto Duran, and Ricardo Neeb. "Factibilidad de una Agencia de Adquisiciones Militares en el Ministerio de Defensa de Chile." Working Paper, ANEPE, April 2000. Mikesell, John L. Fiscal Administration: Analysis and Applications for the Public Sector. Brooks/Cole Publishing, 3rd edition, 1991. Ministry of Defense. Ministerio de Defensa Nacional de Chile 19942000. Santiago: Ministry of Defense, March 2000. Ministry of Defense. The Book of the National Defense of Chile. Santiago: Ministry of Defense, 1998. Nordhaus, William. "Alternative Approaches to the Political Business Cycle." Brookings Papers on Economic Activity No. 2 (1989). O'Reilly, Kevin. "Chile and Argentina: Asymmetrical Air Force Modernization and the Building of Confidence." Thesis at the United States Naval War College, 1997. Pattillo, Guillermo. Proceedings of the Workshop on Budgeting and Democracy. Lyndon B. Johnson School of Public Affairs, University
402 Arming the South
of Texas at Austin and Getulio Vargas Foundation, August 19-22, 1998. Petrei, Humberto. Presupuesto y Control, Pautas de Reforma para America Latina. Washington, DC: BID, 1997. Villar, Carlos. "Gasto Militar: Tema de Reflexion." Memorial del Ejercito (July 1992), pp. 102-122.
PART IV: THE PEACE MOVEMENT
This page intentionally left blank
17 Arms Sales and Development: The Role of the Peace Movement Tony Kempster
I didn't understand; and I didn't see the blood on my own two hands. Jessica Simpson (1995)
Introduction Peace campaigning is essentially about persuading those in positions of power (the decision-makers) to seek an end to existing conflict through justice and reconciliation, and to remove the reasons for potential future conflict. My remit here is to discuss the role ofthe peace movement in persuading decision-makers to ban (or, at least, severely restrict) arm sales to developing countries. I would like to draw particular attention to the information needs of campaigners. At the outset, it is important to emphasize that, for most people in the peace movement, the sale of arms is a moral issue because it causes human suffering and death. Although economic and political considerations may be important, they are not the main concerns. The tragedy of war According to some criteria, the world is now a safer place than it was in the recent past. The number of ongoing conflicts between states can be counted on one hand. The period of nuclear proliferation appears to be
406 Aiming the South
over, and many "proxy conflicts" that were characteristic ofthe cold war era have also ended. Nevertheless for millions of people around the world, violent conflict remains the most direct threat to their life and liberty, and their ability to remain peacefully in their own homes. Some 30 civil wars and many more lower-intensity conflicts are currently being fought, mostly in developing countries (United Nations High Commission for Refugees, 1997). The spectacle of these conflicts reaches us through the media as pictures of the dead and injured, the burnt houses, and the lines of refugees. But it is in the faces of the innocent victims of conflict, particularly those of individual women and children, that we see the tme horror reflected. A poignant image ofthe early 1990s was Kevin Carter's photograph of a vulture waiting for a baby girl to die in an African refugee camp (which inspired the song by Jessica Simpson from which the opening quote is taken). The United Nations estimates that two million children have been killed in armed conflicts in the past decade (most in developing countries) and three times this number seriously injured or permanently disabled. Millions more have had their lives spoiled forever by the effects of war: hunger, disease, uprooting from home, sexual violence and, for some, the trauma of being made to fight (Machel, 1996). Yet, as John Pilger says in his new book Hidden Agendas, such issues are considered by the international media to be "slow news ... used on days when the authorized sources of information are at rest. Nothing happens then, apart from acts of God and disorder in faraway places" (Pilger, 1998). These issues are not prime news and do not create banner headlines. There is no media outcry like the demand for tighter gun controls which followed the Dunblane school shooting tragedy, although the scale of death is grotesque by comparison. This lack of media interest and the shallow reporting of events often prevent the public from understanding what is really going on. This makes it more difficult for the peace movement to mount campaigns which appeal directly to humanitarian concerns. Instead, we have to rely more on economic and political arguments, appealing wherever possible to enlightened self-interest. In anti-arms trade campaigning, these arguments include the risk that we may be supplying arms to potential
Arms sales and development: the role ofthe peace movement
407
future enemies, and the fact that the government provides excessive financial support to arms exporters. Similarly, any attack on government policy which takes humanitarian issues head on is unlikely to be successful unless there are other positive factors in the balance. This is why scholarly conferences on arms trade are so important to us. They can provide new insights into the analysis of arms transfers and valuable information for campaigning. The facility to meet and talk with experts from other countries is especially important because we are tackling a global issue. To understand the views and motivations of decision-makers in other countries, both exporting and importing countries, is vital. Peace campaigning in Britain Despite the campaigning that contributed greatly to the signing of the 1998 Ottawa Convention banning anti-personnel land mines, it is important to recognize that the peace movement is a relatively weak political lobby. This weakness reflects both the nature ofthe movement and the issues that it addresses. The movement is a loose coalition of organizations with different ideologies and agendas. Its roots are both secular, deriving from the socialist antiwar movement established between the two world wars, and religious, through the Quaker movement and the peace organizations within the main Christian churches and those of other faiths. But, whether secular or religious, most peace organizations tend to be marginalized: even the religious peace groups are on the fringes of mainstream church politics. The movement tends also to have a weak image because its aims, particularly those associated with pacifism and unilateral nuclear disarmament, are seen by many as too idealistic in a world where the maintenance of military strength - the big stick of realpolitik - is regarded as essential. Another difficulty, specific to campaigns against the arms trade is that the issues involved are often secret and technically complex. The concern is about weapons (or components of weapons) which may be used many thousands of miles away against people who are rarely able to say what has happened to them. Even if the facts are reported by the media, they
408 Arming the South
can be readily challenged or distorted by vested interests. The whole business of export marketing, government licencing, and the routes for delivering arms can easily become a tangled web of tmths and half tmths, as the Scott Inquiry into the illegal sale of arms to Iraq so clearly showed (Norton-Taylor, 1995). Moreover, such campaigning is normally prophetic, the force of argument depending on the belief that the weapons now being sold will eventually be used for internal repression or aggression against other countries. Critics can always counter that the weapons are being purchased purely for internal security and defense, the legitimate right of every state. Even if the recipient country has a track record of aggression, critics can argue that the political climate has improved. Indeed, they may say that the sales contract allows the British government to restrict the way in which the weapons are used or even influence the human rights policy ofthe recipient country. Such counter-arguments may well have substance in some circumstances, but they can also be used cynically to justify unethical arms sales. Arms supply to developing countries Ironically, the main arms suppliers are four of the five permanent members ofthe UN Security Council, namely, the US, Russia, Britain, and France. The first three now dominate the global market. This role has been established over many years by their national defense and arms manufacturing policies. With the decline in arms demand in the industrialized countries following the end ofthe cold war, arms manufacturers have had to rely more on sales to the developing world and on the opening up of new markets in the emerging economies. Some 80 percent ofthe international trade in conventional arms now goes to such countries compared with just 50 percent in 1990. The regions with increasing arms procurement are the Middle East, South Asia, and East Asia. Notable for the size of their purchases are Saudi Arabia, Turkey, Indonesia, Burma, Pakistan, India, Egypt, Nigeria, Angola, Iran, and Sudan. Sub-Saharan Africa spends an exceptionally high proportion of its GDP on arms (Stockholm International Peace Research Institute, 1997).
Arms sales and development: the role of the peace movement
409
The ending ofthe cold war has not brought the expected reduction in world conflict: what had begun largely as an East-West ideological stmggle continues under different guises, now often related to ethnic interests and resurgent nationalism. Many countries in the developing world are currently expanding their military capability as fresh divisions and tensions emerge. A more detailed discussion on the arms trade and development has been published by CAAT (Campaign Against Arms Trade, 1997). The extent to which the supply of arms from industrialized countries has fueled these conflicts remains a matter of debate. But, it is clear that major arms procurement is an important cause of tension and instability in developing countries through its impact on their economies and political systems. In the first place, many developing countries suffer from a chronic lack of investment in social development programs which is crowded out further by military expenditure. Moreover, social investment programs are often the first to be cut during financial crises whereas military expenditure tends to be the last as has happened in recent years in Egypt, Iraq, and several South Asian countries. Second, high levels of arms expenditure can push the political process into a preoccupation with military matters. Powerful armed forces seldom go hand-in-hand with democracy in developing countries and often lead to dictatorship. This is because military organizations tend to be ineffective without authoritarian command stmctures, and military regimes when threatened are prone to exhibit intolerance of minorities for security reasons. It is, therefore, extremely difficult for tmly democratic political systems to flower in militarized societies (see Dumas, chapter 1 in this volume). Such purchases also contribute to the debts which have become a major problem for many developing countries. SIPRI estimates that some 20 percent of the developing world's debt can be attributed to arms procurement (Stockholm International Peace Research Institute, 1997). It also considers that much of this can be attributed to the ready availability of loans for arms purchases in the late 1970s and throughout the 1980s. In some countries, the Philippines for example, debt servicing and military expenditure together accounted for most government
410 Arming the South
spending, leaving virtually nothing for development obj ectives. Many are indebted as a result of so-called soft loans, especially to the ex-USSR states, which are now calling in these debts to fund their own domestic development programs. But above all, the instability created can easily become a vicious spiral of increasing civil unrest and ever greater military spending to contain it, which has blighted many developing countries and led to gross human rights abuses. As the situation worsens, civil war may follow with the possibility of a spillover into neighboring countries and the outbreak of international conflict. Making the campaign case So it is often the struggling, debt-burdened countries that Britain and the other arms exporting countries are targeting for sales, with potentially devastating effects on their social stmctures and the lives of ordinary people. The issue for organizations like Campaign Against the Arms Trade (CAAT) is how we obtain the relevant facts and assemble them into a convincing case for campaigning. Where should we focus our arguments and how far can they be generalized across different countries? It is relevant to ask how important the direct effects of military conflict on human suffering are when compared with the indirect effects of disease and malnutrition, caused by the lack of social investment due to excessive arms spending. We maybe in danger of focusing too sharply on military conflict per se, when it is the impact of the arms trade on social spending which is more important. Clearly, we need, as far as possible, to consider the impact of arms sales as part of a holistic view of the various factors influencing development and international security a point made forcibly by representatives of NGOs giving evidence to the Strategic Defence Review currently being carried out by the British Government (Committee Office ofthe House of Commons, 1998). Although the situation is obviously complex, differing from one developing country to another, ideally we need an assertion which can be used generically. This would be equivalent, for example, to the statements made by Oxfam on the immorality of forcing children to be soldiers or statements made by Amnesty International on the violation of
Arms sales and development: the role of the peace movement
411
human rights. Having listened to various talks at scholarly military expenditure and arms trade conferences, it seems reasonable to assert that military spending in developing countries has a negative effect on social spending programs and increases human suffering. This requires just two qualifications. First, we are considering military spending beyond what might be regarded as adequate for defense and the maintenance of law and order. Second, we are considering countries where the military force cannot be used to gain resources. In the past, development has often been aided by military force - war (or at least the threat of war) has been the foundation of empires. But in today's world, only a few powerful countries can use force with impunity to achieve their economic objectives. For most developing countries, certainly those with little international weight, the use of military force in this way is not an option because of world reaction. If they act against the interests ofthe West, particularly of the US, retaliation is likely to be brutal, as it has been against Iraq following the invasion of Kuwait. Having established a generic position, one would then go on to consider country-by-country situations. This requires a careful analysis of the situation of individual countries and, wherever possible, information that is 1
accurate and up to date. In this context, I was concerned to hear Professor Scheetz's reservations about the accuracy of some published statistics on military spending for certain countries (Scheetz, chapters 3 and 15 in this volume);
2
damaging to the image of the arms exporting companies or the government departments that support them. CAAT is a campaigning organization, and we obviously need hard-hitting material, just as a journalist requires original and incisive material to write a good story;
3
supported by the underlying facts and, where appropriate, sound argument. This would include historical trends, the reasons for any policy changes, the legal background to the anus transfers, and so on.
412 Arming the South
At present, information is lacking in a number of important areas, and I would call on researchers and academics who sympathize with the cause to assist wherever they can. Information needs include the export strategies adopted by individual arms companies, in particular how they target specific countries and the tactics they use to gain competitive advantage there. Detailed information on sales policy is important in that it may allow us to demonstrate that companies have acted unethically or have encouraged arms purchases well beyond those needed for defense (actions that seem inevitable in some circumstances, when profit is the bottom line and ethical concerns are a nuisance factor in marketing strategies). As Charles Masefield, Head ofthe Defence Export Services Organisation (DESO) at the British Ministry of Defence so clearly puts it: "The objective is to sell ... we must become more pro-active and instead of reacting to market situations ... we must work upstream to influence market movements in our direction" (Masefield, 1995). Knowledge about the evolving structure of the international arms manufacturing industry, in terms of mergers and take-overs, and prediction of likely future developments are also important because these allow us to plan campaigns more effectively particularly those involving collaboration with activists in other countries. We are also interested in the dynamics of arms purchases. What encourages a certain country to buy, especially when the purchase is beyond its needs for defense? This is important because it may be important for us to campaign for policies that reduce demand, as well as for those that restrict supply. As Dr. Neil Cooper concludes in his excellent overview of current arms diffusion and internal conflict: "The problem is most likely to be addressed by a combination of economic strategies designed either to reduce the recipient's ability to pay or to raise the cost of arms, and by lateral strategies that address the problem indirectly via the development of an internal security community" (Cooper," 1998).1 The availability of such information is also likely to create more opportunities for campaign alliances with other NGOs involved for example with human rights, Third World development, and related issues. We might also be able to develop and promote international codes of conduct for arms transfers which effectively cover all the loopholes
Arms sales and development: the role of the peace movement
413
that might be used by unscmpulous governments or companies to avoid the restrictions. Important points include end-user controls and methods of controlling brokerage and licenced production in other countries (Crowley, 1998). Indeed, these coalitions might go beyond other NGOs and include political parties and government departments concerned with development. For example, we have much common ground with the British policy on international development. In her recent white paper, Clare Short, the Secretary of State responsible for Britain's international development policy, has stated that the objective is to halve the proportion ofthe world's population living in extreme poverty by 2015 (Secretary of State for International Development, 1997). The target would be more realistic if the Government not only increased overseas development aid but also acted to limit the detrimental impact of arms sales on Third World development. It is, therefore, essential to point out that arms sales encouraged by the British government contradict the efforts ofthe Overseas Development Agency to alleviate world poverty. This brings me to consider, as a final issue, the political and economic background against which we campaign in Britain. This background informs the way CAATs campaigns are designed and carried out. The campaigning environment The key factors are as follows. 1
The hypocrisy ofthe new Labor Government's foreign policy. The Foreign Secretary has stated that "Labour will not permit the sale of arms to regimes that might use them for internal repression or international aggression" (British Foreign Office, 1997). But the Government has already reneged over arms sales to Indonesia that were in the pipeline at the time of taking office. It will be increasingly open to criticism if sales to repressive regimes continue in future.
2
The need to increase social expenditure in Britain. This is reflected in the growing tension between the desperate need for the government to find money for its social expenditure program and the massive
414 Arming the South
expenditure on the Eurofighter and Trident programs. 3
The restructuring taking place in Europe's defense and aerospace industry (and the globalization ofthe arms industry in general). This will move the policy decisions of British companies into an international arena where the British Government will have less direct influence.
4
The extension of NATO to Eastern European countries and the alienation of Russia. There is a weight of opinion that favors the offering of economic links to these countries, with longer-term entry to the EU, rather than the creation of a new market for arms (and potentially a new cold war). Although not directly related to the subject of this chapter, it is relevant to include it for completeness because the new NATO countries represent a major arms market.
5
The effect of increasing competition in the international arms market. This reflects the declining global market for arms and the competition from the vast US defense industry. Effective British (or even European) competition is likely to demand even higher subsidies and other kinds of Government support.
6
Growing public awareness of the possibility of terrorism in Britain by people fighting for the interests of developing countries or retaliating for grievances suffered. The spread of sophisticated weapons to developing countries increases the danger.
7
Growing public awareness ofthe link between the armed conflict and environmental degradation. This is particularly important in the context of developing countries.
8
Globalization and the effect of the policies of multinational companies on developing countries, particularly debt and the use of military force to achieve commercial objectives. Underneath the surface of this point are several deeper issues associated with modernity, whether development (in the sense of colonization) is
Arms sales and development: the role ofthe peace movement
415
progress, and the psychological attitude ofthe West to Third World countries. Conclusion We should remember that the peace movement does not campaign alone but in brotherhood (sisterhood) with the people in developing countries. The two flames of anger and hope which historically brought social justice and democracy to the Western world are at work today in developing countries. People in these countries are striving to escape poverty, dependency, and insecurity. We are morally obliged to assist them by removing the burden of amis purchases that our society places on them. References British Foreign Office. Press statement. 27 July 1997. Campaign Against Anns Trade. The Arms Trade and Development. London: CAAT, 1997. Cooper, N. "Arms Diffusion and Internal Conflict." Paper presented at arms trade conference, Middlesex University Business School, 1998. Committee Office of the House of Commons. Evidence given by the Department of Peace Studies, Bradford University and Saferworld and Oxfam to the Defence Committee on 25th March 1998. Crowley, M. "Curbing the EU Anns Traffic." Amnesty (1998), pp. 4-5. Dumas, J. "The Role of Demilitarization in Promoting Democracy and Prosperity in Africa," chapter 1 in Jurgen Brauer and J. Paul Dunne (eds.). Arming the South: The Economics of Military Expenditure, Arms Production, and Arms Trade in Developing Countries. London: Palgrave, 2002. Machel, G. Children and War. Geneva: United Nations, 1996. Masefield, C. "Defence Exports: The Challenge Ahead." RU SI Journal (August 1995), pp. 15-19. Norton-Taylor, R. Truth is a Difficult Concept: Inside the Scott Enquiry. London: Guardian Books, 1995. Pilger, J. Hidden Agendas. London: Vintage, 1998.
416 A rm ing the Sou th
Secretary of State for International Development. Eliminating World Poverty: A Challenge for the 21s' Century. Government White Paper. Cm 3789, 1997. Simpson, J. R. Quote from the song "Kevin Carter." Jessica Ruby Simpson © 1995 Hagweed Music. On Red House Records 1996 (RHR CD 96). Scheetz, T. "Military Expenditure and Development in South America," chapter 3, and "The Military-Civilian Tradeoff in Guatemala: An Econometric Analysis" (with Kanta Marway and Lawrence R. Klein), chapter 15 in Jurgen Brauer and J. Paul Dunne (eds.). Arming the South: The Economics of Military Expenditure, Arms Production, and Arms Trade in Developing Countries. London: Palgrave, 2002. Stockholm International Peace Research Institute. SIPRI Yearbook 1997. Oxford: Oxford University Press, 1997. United Nations High Commissioner for Refugees. The State of the World's Refugees: A Humanitarian Agenda. Oxford: Oxford University Press, 1997. Note I would like to thank Martin Broek for his helpful comments on the text. 1. [Editors' note: Dr. Cooper chose to submit a different paper for the present volume (see chapter 2). The paper referenced here may be requested directly from the author at [email protected].]
Index
ACDA, 102, 104, 132, 133, 161 Adeola, O.F., 276, 282 Afghanistan, 40 Africa demilitarization, 19 demobilization, 24 elections, 22 militarization, 16 poverty, 16 violence, 16 weapons, 16 aid agencies, 46 Al-Ghrair, A., 203 Alexander, R., 299 Alogoskoufis, G., 295 Anderton, Charles, 111 Angola, 6, 7, 46 amis imports, 255 civil war, 8 defense industrial base, 255 demobilization, 25 dual-use activities, 255 import substitution industrialization, 254 industrial performance, 7 manufacturing industry, 7, 251 anti-personnel land mines, 407 Antonakis, N., 296, 300, 302 Arbenz, Jacobo, 339
Argentina, 94 military expenditure, 52 armed forces, 132 civilian control, 56 funding, in Chile, 379 operational capacity, lack of, 57 role of, in Chile, 376 Armington, P., 303 arms imports Angola, 255 elasticity of, 130 Guatemala, 339 arms industry in developing nations, 101 arms production, 4, 99 "graduation" to higher capabilities, 118 alternative to arms imports, 5, 129 and arms control, 119 and foreign-exchange earnings, 112 and small arms, 120 arms exports, 5 as industrial policy, 108 defense procurement, 5 economic motives, 4, 107 Greece, 104 model, 6, 129 non-economic motives, 4, 106 Singapore, 104
418 Arming the South South Africa, 222 stages of, 105 subsistence quantity, 135 substitution among inputs, 133 the distress argument, 115 transnationalization of, 118 weapons platforms versus weapons, 105 arms race Asia. 78 balance of power, 80 arms trade, 99 and arms production, 161 and mercenary companies, 43 and post-modern peace, 36 and the role of the peace movement, 405 Asia, 72 defense policy, 408 depreciation of Asian currencies, 74 embargoes, 106 Latin America, 57 model, 5, 162 replacement of obsolete equipment, 81 secret, 407 ASEAN, 78, 80 long history of non-conflict, 75 ASEAN Regional Forum (ARF), 77 Asia arms race?, 78 arms trade, 72 military expenditure, 71 Athanassiou, Emmanuel, 8, 291, 303 Avramides, C , 294 Ayoob, M , 84, 87 Ayres, Ron, 109 Azar, E., 89, 90 Balbontin, Marcelo, 381 Balfousias, A., 296 Ball, Nicole, 16,26, 107, 110, 115, 120 Barnes, Sam, 242 Barros, Carlos Pestana, 264
Batchelor, Peter, 107, 109, 113, 115, 222, 226, 228, 230 Benoit, Emile, 51, 85, 224, 252, 277, 343 Berdal, Mats, 44, 45, 242 Berthelemy, Jean-Claude, 236 Birdi, Alvin, 6, 221, 225 Biswas, Basudeb, 226, 280, 296 Bitencourt, Luis, 119, 120 Bitzinger, Richard, 111 Bosnia, 46 Bradbury, Malcolm, 36 Brauer, Jurgen, 4, 48, 86, 101, 107, 110, 112, 114, 120, 121, 236, 255, 295 Brazil, 106 arms industry in, 101 Briick, Tilman, 7, 235, 242-246 Brunei, 75 Brzoska, Michael, 102, 104, 106, 110, 113, 119,282 Burnham, Peter, 239 Buzan, Barry, 89 Campaign Against the Arms Trade, 10, 410 Chamaca, Alvaro, 381 Chan, Steve, 88, 277, 279, 281 Chatterji, Manas, 236 Chechnya, 37 Chile, 6, 94, 373 allocation of resources to the armed forces, 9 military expenditure, 52 structure ofthe Ministry of Defense, 376 transparency of budget allocation, 9 China, 71,75, 79, 86, 106 Chletsos, M., 296 Chung, J.W., 134, 150 civil war and economic policies, 266 and manufacturing performance, 252 civilian control over armed forces, 56
Index Clapham, Christopher, 45 Coker, Christopher, 35 Collier, Paul, 238 Colombia, 44 Conetta, C , 80, 82,91 conflict changing nature of, 3 possible versus probable, 91 Congo, 44 Constas, D., 293 conversion, 24 and demobilization, 25 Latm America, 57 Cooper, Neil, 3 , 3 5 , 4 1 2 Cooper, Robert, 36 Copper Law, 380 Cortright, David, 120 cost of peace budget allocation process, 243 forward-looking conditionality, 245 institution building, 246 policy bias, 246 Costa Rica, 18,92 elimination of armed forces, 18 countertrade offsets, relation to, 198 presumed benefits, 2 country risk analysis Latin America, 57 country studies, 193 Crowley, M., 413 Czechoslovakia, 102 Davis, David R., 279, 281 de Waal, Alex, 46 defense policy and institutional rationality, 61 arms trade, 408 cost effective, 92 forward offensive defense, 80 incorrectly formulated, 60 institutional structure, in Chile, 376 non-provocative defense, 64 optimal military doctrine, 285
419
political rent-seeking, 61 defense procurement arms exports, 5 defense-growth relationship empirical studies, in Turkey, 321 Deger, Saadet, 53, 85, 112, 121, 278, 284,319 demilitarization, 3 demobilization, 24 Guatemala, 362 prerequisite to democratization, 30 unilateral, 92 vital to development, 30 demobilization a form of conversion, 25 and discharge, 27 and long-term reintegration, 27 and re-orientation, 27 Angola, 25 assembly and encampment, 26 economic incentives, 95 Eritrea, 25 Ethiopia, 24 Liberia, 25 Mozambique, 25, 243 Namibia, 25 recruitment for mercenaries, 45 Sierra Leone, 25 Somalia, 25 Uganda, 25 democracy and civil society, 21 Guatemala, 339 incompatible with militarization, 19 Latin America, 56 long-run value to development, 22 source of security, 28 Denoon, D., 71 Dervis, J., 303 development and democracy, 22 economic investment, 17 more than growth, 16
420 Arming the South source of security, 28 diaspora funding of arms, 45 Diaz, Jose, 377 Dibb, Paul, 76 disarmament a type of investment, 88 Doornik J.A., 323 dual-use technologies, 338 Angola, 255 used by post-modern armies, 37 Duffield, Mark, 41,42 Dumas, Lloyd J., 3, 15, 16, 29, 409 Dunne, J Paul, 6, 85, 87, 221, 223. 224, 226, 228, 230, 276, 279, 291, 294,296,319,320 Duran, Roberto, 386 East Timor, 76 Economists Allied for Arms Reductions, 10 ECOWAS, 29 Egypt, 106 elections strategy for maintaining power, 22 Ellis, Stephen, 40 Erhard, Ludwig, 7, 235, 237-239, 245 Eritrea, 24 demobilization, 25 Esquipiilas peace accords, 339 Ethiopia, 24 demobilization, 25 ethnic violence, 15 built on population displacement, not popular support, 41 economic rationale of, 42 political economy of, 35 European Union, 29 Feder, Gershon, 279, 296, 298 Feder-Ram model, 225 Felice, W., 83, 89, 91 Ferreira, Manuel Ennes, 7, 251, 254, 264, 265 Findlay, T., 81
Finnegan, William, 240 Fontanel, Jacques, 107, 112, 117, 130, 291 Franko-Jones, Patrice, 106, 109, 111, 113, 114 Frederiksen, Peter, 277, 282 Frelimo, 44 Galtung, Johan, 90 Garcia-Alonso, Maria del Carmen, 5, 161 Garfinkel, Michelle, 130 Garrett, John, 120 Geffray, Christian, 240 Georgiou, G., 294 Germany, 7 post-war reconstruction, 235 war debt, 7 Gissy, William, 121 Gleditsch, Nils Petter, 224, 236, 291 Glenny, M.,293 Glitman, Maynard, 43 Gold, David, 118 Gray, Chris Hables, 40 Greece, 6, 291 arms production, 104 economic effects of military expenditure, 295 economic growth, 8 economy, 295 high military expenditure, 291 military expenditure, 8 opportunity cost of defense, 8 security concerns, 293 Guatemala, 6, 337 data, 340 guerrilla war, 338 military expenditure, 52 military-civilian tradeoffs, 9 guerrilla war Guatemala, 338 Gimluk-$enesen, Giilay, 106, 109-111 Gunning, Jan Willem, 238 Gupta, S., 75
Index Gurel, S., 293 Harris, Geoff, 4, 71,86, 92 Harrod-Domar model, 343 Hartley, Keith, 109, 236, 255, 280, 319 Hartung, William, 108, 118 Haruna, I.B.M., 281,282 Hendry, D.F., 323 Hewitt, Daniel, 281 Hong Kong, 71 Hooper, N., 203 Huang, C , 299 Human Poverty Index, 86 humanitarian concerns combined with self-interest, 406 Humm, Anthony, 107, 112, 117, 130 Hutus, 46 ICRC, 46 ideology decline as legitimizing force for conflict, 40 import substitution industrialization Angola, 254 India, 7 1 , 7 5 , 8 6 arms industry in, 101 Indonesia, 74, 76 International Court of Justice, 93 International Institute for Strategic Studies, 142,281 International Monetary Fund, 246, 262, 267 Intriligator, Michael, 88, 130 Islamic fundamentalism, 40 Israel, 4 arms industry in, 101 Japan, 71,74, 75 weak regional role, 76 Johansen, R., 83 Johansen, S., 323 Jones, D., 294 Joon, N.M., 79-81 Jouganatos, G., 295 Judah, Tim, 45 Juselius, K., 323
All
Kaldor, Mary, 41,46, 89,278 Kapopoulos, P., 294 Keen, David, 40, 44, 242 Kempster, Tony, 10, 405 Keynes, John Maynard, 238 Khmer Rouge, 40 Kingma, Kees, 25, 27 Klein, Lawrence R., 9, 52, 53, 236, 337 Knight, C , 80, 82, 91 Kollias, Christos, 8, 291, 293, 294, 296 Kornai, Janos, 253, 260, 261 Kosovo, 45 Kosovo Albanians, 45 Krause, Keith, 104, 105, 107, 113, 114, 120 Krueger, Anne, 261 Kurop, M.C., 293 Lamb, Geoffrey, 236 Laos, 71 Larrabee, S.F., 293 Latin America, 51, 58 and arms trade, 57 conversion of military industry, 57 country risk analysis, 57 defense policy, 60 democracy, 56 military expenditure, 51 military reform, 62 military's mission, relation to, 58 Lazaretou, S., 294 Levine, Paul, 5, 130, 145, 161, 162 Lewis, Neryl, 243 Liberia, 40, 42, 45 Lim, David, 282 logo warriors quasi-uniform of paramilitaries, 46 Looney, Robert, 277, 282 Luckham, Robin, 37 Luttwak, Edward, 36 Machel, G., 406 Majeski, S , 294 Makrydakis, S., 294, 296 Malaysia, 74, 80
422 Arming the South manufacturing industry and civil war, 252 Angola, 251 South Africa, 230 Martin, B., 94 Marwah, Kanta, 9, 52, 53, 337 Masefield, C , 412 Matthews, Ron, 6, 195, 201, 203-205, 210,213 Mazula, Brazao, 243 McMillan, S., 225, 227 Melko, Matthew, 77 Meneses, Emilio, 386 mercenaries, 36, 39 and virtual states, 43 corporate partners of local elites, 43 Croatia, 43 economic incentives, 45 mercenary companies, 42 recruitment from demobilization, 45 Sierra Leone, 43 Mexico, 40 Mikesell, John, 384 militarization, 409 and civil society, 20 and civil war, 410 and economic growth, in Guatemala, 337 authoritarian structure, 21 culture of obedience, 21 Greece, 293 human rights, 91 impediment to democracy, 19 impediment to development, 16 incompatible with democracy, 19 Nazi Germany, 236 search for significant missions, 56 war debt, in Nazi Germany, 236 military expenditure supply-constrained, 278 absorptive capacity, 277 and disequilibrium, 60 and displacement of social
expenditures, 52, 86, 409 and economic growth, in Turkey, 323 and peace dividend, 85 and post-modern peace, 38 and steady-state growth, 59 Argentina, 52 as a public good, 59 Asia, 4, 71 augmented Feder-type model, 292 Bolivia, 54 budget allocation process, 83, 243 Chile, 52 computable general equilibrium model, 292 data, 53, 101 determinants, in Asia, 72, 81, 87 development, 4, 16, 71 development, in Asia, 85 economic growth, in Asia, 85 economic growth, in Greece, 296 economic impact, 2, 51 effects, empirical question, 224 effects, theoretical, 223 externality effects, 276 factor productivity differential, 278 field work, 101 genuine cuts, 1 Greece, 8 growth-supporting role, 60 Guatemala, 52 impact on income distribution, 237 impact on manufacturing sector, 228 institutional models, 223, 276 interviews with decision-makers, 84 Keynesian models, 223, 276 Latin America, 4, 51 Marxist models, 223 military's mission, relation to, 51 neoclassical models, 223, 276 non-optimal managerial processes, 376 noncontributive activity, 16
Index opportunity cost, 17 opportunity cost of defense, 72 opportunity cost of defense, in Greece, 302 overviews, 13 Paraguay, 52 Peru, 52 production function models, 280 simulated reductions, in Greece, 305 South Africa, 6, 221 strategic posture, 63 sub-Saharan Africa, 8, 275 substitution effect, 276 transition to civilian rule, 55 transparency of, 373 military Keynesianism, 337 military reform Latin America, 58, 62 military spending case study, 221 data, 221 econometric analyses, review of, 222 economic growth, 221 military training social and psychological conditioning, 19 Milward, Alan, 239 Mintz, Alex, 299, 300 Modise, Joe, 107 Mohammed, Nadir, 85, 236, 243, 280 Moller, Bjorn, 77, 78, 80, 81 Mongolia, 86 Moon, C-I, 90 Mouzakis, Fotis, 5, 129, 131, 139, 145 Mozambique, 6, 7, 44 demobilization, 25, 243 post-war reconstruction, 235, 242 war debt, 7, 243 Munslow, Barry, 265 Murdoch, J.C, 302 Myanmar, 75, 79, 86 Myint, Hla, 261
423
Nabe, Oumar, 16 Namibia demobilization, 25 Natsios, Andrew, 46 Neeb, Ricardo, 386 Nepal, 71 Neuman, Stephanie, 117 Nikolaidou, Efi, 8, 225, 291, 296 Nolan, Janne, 109, 111 Nordhaus, William, 394 North Korea, 75 North, Douglass, 240 Norton-Taylor, R., 408 O'Reilly, Kevin, 388 Obasanjo, Olusegun, 22 OECD, 16 offset trades, 195 presumed benefits, 2 Saudi Arabia, 6 offsets and development, 197 as mode of technology transfers, 197 cost of establishing offset companies, 200 countertrading, relation to, 198 creation of skilled workforce, 213 development impact, 207 growing importance of, 195 offset credits, 211 policy implications, 212 purpose, 195 straddle the civil-military divide, 196 technological absorptive capacity, 215 the Philippines, 197 typology, 198 Olaniyi, Oyinlola, 8, 275, 278, 282 Ottawa Convention, 407 Ozsoy, Onur, 320 Pakistan, 75, 86 Paraguay military expenditure, 52 Pattillo, Guillermo, 9, 373, 374
424 Arming the South Payne, James E., 16 peace campaigning, 407 campaign alliances, 412 campaigning environment, 413 information needs, 412 peace groups marginalized, 407 peace movement the role of, 10 information needs of, 405 Pearson, Frederic, 106, 107, 109, 118, 119 Peru military expenditure, 52 Pesaran, B., 229 Pesaran, M., 229 Peters, Ralph, 41 Pilger, J.,406 Portugal, 4 post-modern armies use of dual-use technologies, 37 post-modern conflict, 35 acquisition of booty, not territory, 41 military-business complex, 277 post-modern peace, 35 and arms trade, 36 the thesis, 35 post-modern states privatization of state functions, 38 post-modern weapons, 35 from militarism to armament culture, 37 post-war reconstruction Germany, 235 Mozambique, 235 role of war finance in, 236 Prendergast, John, 24 preventive diplomacy, 94 Ram, Rati, 226, 279, 280, 291, 296, 298, 299,319 Rana, Swadesh, 102, 104, 121 Rathmell, Andrew, 37 Refenes, A., 296
Renamo, 44 Reno, William, 39 resource allocation Chile, 373 South America, 373 transparency, 373 Richardson, Lewis, 130 Rothschild, Kurt, 281 Roux, Andre, 120, 121,225 Russett, Bruce, 28 Rustomjee, Z., 222 Rwanda, 45 ethnic genocide, 15 Saal, David, 225, 226, 228, 230 Sahu, Anandi P., 16 Samudavanija, C.A., 84 Sandier, Todd, 109, 255, 280, 319 Santos, J. Eduardo, 254 Sams, A.H., 303 Saudi Arabia, 6,40, 195 defense offsets, 6 offset programs, chronology, 203 Scheetz, Thomas, 4, 9, 51-53, 337, 341, 411 Schmolders, Giinter, 236, 238, 240 security alternative, 89 and conflict management, 94 and conflict resolution, 93 and peace theory, 89 and social defense, 94 arbitrate and adjudicate, 93 changing nature of, 72 changing type and source of threats, 90 democracy, source of, 28 development, source of, 29 forums, 93 nature of armed conflict has changed, 89 non-aggression treaties, 92 non-military options, 92 privatized, 43
Index Sen, Amartya, 121 Sen, Somnath, 121, 162, 284, 319 Sezgin, Selami, 9, 319, 320, 327-329 Sharp, G., 94 Shearer, David, 44 Sierra Leone, 45 demobilization, 25 Simon, S.W., 77-80 Singapore, 4, 71, 74, 75, 79, 80, 86 arms production, 104 SIPRI, 72, 74, 75, 79, 81, 101, 133, 142, 144,408,409 Sivard, Ruth, 275 Skons, Elizabeth, 118 small countries incapable of defending themselves militarily, 63 Smith, Ron, 107, 112, 117, 130, 145, 162,294 Soesastro, H., 71 Sogge, David, 259 Sollenberg, M., 89 Somalia, 37, 46 demobilization, 25 South Africa, 6, 102, 106, 221 arms production, 222 manufacturing industry, 230 military expenditure, 6 nuclear weapons program, abolished, 222 South Korea, 4, 71,74, 75, 81 arms industry in, 101 Spain, 4 Sri Lanka, 75, 79, 86 Stavrinos, V., 294, 296 Stevenson, R., 300 Stewart, Frances, 237 strategic environment, 1 strong states relation to weak states, 39 structural violence, 89 sub-Saharan Africa, 6 human capital, 283
425
Summy, R., 94 superpower confrontation effect on Balkans, 293 removal of, 1 world conflict, no end to, 409 switch-trade, 198 Synge, Richard, 242 Taiwan, 4, 71, 74, 75, 106 arms industry in, 101 Tamil Tigers, 45 Tan, A., 80 technology transfer, 200 and offsets, 197 Terhal, Piet, 112 Thailand, 74 the Philippines, 74, 81 offsets, 197 Thisen, Jean, 243 transparency and arms imports, 376 IMF guidelines, 374 is more than information, 375 of budget allocation process, 374 Turkey, 6, 106,319 defense-growth relation, 9 Uganda demobilization, 25 UNDP, 2, 73, 86, 89-91 UNITA, 40, 42, 44, 46 United Nations High Commission for Refugees, 406 Ury, W., 93 van Creveld, Martin, 41 Vayrynen, Raimo, 113 Vietnam, 79 Villalon, Leonardo, 15, 22 Villanon, A., 197 Villar, Carlos, 388 Vines, Alex, 42, 46, 240, 243 virtual states and mercenary companies, 43 and weak state elites, 39 in the developing world, 38
426 Arming the South m the West, 38 Vougas, Dirmtrios, 225 Wang, S., 79 war debt Mozambique, 243 time inconsistency problem, 237 transparency of, 238 war finance Germany, 235 Ward, Michael, 279, 281, 299, 300 warlords acquisition of booty, not territory, 41 weak state elites and civil companies, 44 and mercenary companies, 43 weak states, 339 interests of rulers, 39 relation to strong states, 39 relation to virtual states, 39
weapons acquisition status symbols, 80 Weisburd, A., 77 Weissman, Fabrice, 42 West, R., 81 Willet, Susan, 16 Williams, R., 201 World Bank, 104, 246, 258, 281 Wulf, Herbert, 107, 118,282 Yildirim, Julide, 9, 319, 327 Zaire, 38, 46 Zapatistas first post-modern revolutionaries, 40 Zografakis, S., 303 Zografakis, Stavros, 8, 291 Zunes, Stephen, 94