Measuring Capital Flight: Estimates and Interpretations
Total Page：16
File Type：pdf, Size：1020Kb
Working Paper 194 Measuring Capital Flight: Estimates and Interpretations Benu Schneider March 2003 Overseas Development Institute 111 Westminster Bridge Road London SE1 7JD UK Acknowledgements This paper is the first part of a project on Capital Flight from Developing Countries. The project is funded by ESCOR, Department for International Development, UK and we gratefully acknowledge their financial support. The UK Department for International Development (DFID) supports policies, programmes and projects to promote international development. DFID provided funds for this study as part of that objective but the views and opinions expressed are those of the author alone. The author would like to thank Mathieu Sampson and Benno Ferrarini for their research assistance. ISBN 0 85003 633 X © Overseas Development Institute 2002 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publishers. ii Contents Acknowledgements ii Tables and Figures v Abstract vi 1 Introduction 1 2 Defining capital flight 3 2.1 Preferred definition 3 2.2 An overview of definitions of capital flight in the literature 3 2.3 The broad definition of capital flight 4 2.4 Capital flight defined as a response to discriminatory treatment of domestic capital 5 2.5 Defining capital flight as an illegal transaction 6 3 Methods to measure capital flight 8 3.1 Broad measure of capital flight 8 3.2 Components of the broad estimate 10 3.3 Data problems 13 3.4 Capital flight: a sub-set of private capital outflows 15 3.5 Critical assessment of measuring procedure 17 3.6 Measuring capital flight as an illegal transaction 19 3.7 Conclusion on estimation procedures 20 4 Profile of estimated capital outflows 21 4.1 Recent trends in resident capital outflows 21 4.2 A profile of capital lost through trade misinvoicing 24 4.3 Hot money flows 25 4.4 Resident capital outflows one-side of two way flows 26 4.5 Capital account liberalisation and estimated outflows 26 4.6 Unrecorded capital flows and crises 30 4.7 Credibility of the reform process and repatriation episodes 31 Conclusion 32 References 35 Annex A Country groups 38 Abbreviations: 39 Annex B Estimated resident capital outflows 1 East Asian Pacific (1971–98) 40 2 Europe & Central Asia (1972–98) 43 3 Latin America & the Caribbean (1970–98) 46 4 North Africa & Middle-East (1972–98) 52 5 South Asia (1975–98) 54 6 Sub-Saharan Africa (1974–98) 55 Annex C Resident capital outflows as a ratio to GDP classified according to World Bank classification on income & indebtedness 65 Annex D Estimates of misinvoicing 1 East Asian Pacific (1971–98) 71 2 Central Asia And Europe (1971–98) 73 3 Latin America & The Caribbean (1971–84) 74 4 North Africa and the Middle-East (1971–98) 78 iii 5 South Asia (1971–98) 79 6 Sub-Saharan Africa (1971–98) 80 Annex E Hot money flows 1 East Asian Pacific (1971–99) 87 2 Central Asia and Europe (1971–99) 87 3 Latin America and the Caribbean (1970–99) 88 4 North Africa & the Middle-East (1972–99) 89 5 South Asia (1971–99) 89 6 Sub-Saharan Africa (1972–98) 90 Annex F Resident flow estimates from the UK and Finland 92 iv Tables and figures Tables Table 1 An overview of definitions of capital flight in the literature 3 Table 2 Summary presentation of measuring procedure (broad and hot money measure) 9 Table 3 India: Estimates of resident capital flows based on a data set in 1995 and 2000 (in millions of US dollars) 15 Table 4 Summary presentation of measuring procedure (capital flight: a sub-set of total claims) 17 Table 5 Regional estimates of resident capital flow 22 Table 6 Does liberalisation of the current and capital account affect estimates of resident, hot money flows and misinvoicing? An analysis of selected countries 27 Table 7 Banking assets and resident capital flows for selected countries in millions of US dollars 28 Table 8 Estimated resident capital flows for India, Uganda and Chile 31 Data sources 38 Figures Figure 1 Broad measure of capital flight: sectoral coverage of foreign assets 11 Figure 2 World Bank residual and Dooley measure of capital flight (annual flows) 19 Figure 3 Regional estimates of capital outflows (including China) 23 Figure 4 Estimated resident capital outflows as a ratio to GDP – by region 23 Figure 5 Estimated resident capital outflows (adjusted for debt forgiveness and cross currency valuation) by region 24 Figure 6 Estimates of misinvoicing by region 198198 25 Figure 7 Regional estimates of hot money flows 26 v Abstract Capital flight is a phenomenon which, though unobservable, is assumed to be widely prevalent in developing countries. While this is probably true the approaches proposed by economists to quantify it may not necessarily capture what we seek to measure: namely the flight of capital as a response to economic and political instability. Each method differs conceptually as to what defines capital flight based on distinctions between ‘normal flows’ and ‘capital flight’, ‘short term’ and ‘long term’ and between ‘legal’ and ‘illegal transactions’. Consequently estimates of capital flight vary. This paper makes a critical evaluation of estimation methods used in the literature and applies them to a sample of 116 countries for the period 1971–98. The paper makes the case that the most widely prevalent measure of capital flight can at best be treated as a resident capital flow, which captures not only capital flight, but other influences as well. The estimates can be used in conjunction with explanatory variables and country-specific information to determine the type of capital flow one seeks to explain. The estimates reveal that irrespective of definition and the corresponding estimation method, estimated resident capital flows based on these methods are high for some countries and regions even in the 1990s and that it occurs in economies both with and without wholly or partially liberalised capital controls. The estimated capital flow is unevenly distributed across countries, both in absolute numbers and as a percentage of GDP. In recent years the outflows have been high from East Asia, Europe, Central Asia and Latin America. Although subject to data problems and corresponding errors in magnitudes. the Mexican and the East Asian crisis and the impact of other events are discernible from the estimated resident capital outflows. Contrary to a widely held belief that outflows from sub-Saharan Africa are high, the estimates show that since 1992 resident capital outflows from the region were repatriated with a small outflow in 1995. East Asia and Latin America show a rising trend in the 1990s compared to the 1970 and 1980s. Loss of capital through misinvoicing of trade documents is high for some countries. As a region, they are the highest for East Asia and the Pacific especially during the crisis period. In Latin America, the Mexican crisis episode is clearly seen in this estimated series. Individual country experiences vary, because besides the flight motive, capital movements through this channel can also reflect tariff-jumping, subsidies to exports or even smuggling. Hot money flows reflect the smallest measure of capital movement in developing countries. In spite of the data problems and problems of interpretation, estimates for East Asia do capture the crisis period. The study provides a database on estimated resident outflows of capital for further research on capital flight, capital movements and financial integration. vi 1 Introduction Since the debt crisis in the early 1980s a growing strand of literature has focused attention on the outflows of resident capital in response to the distortionary impact of domestic policies and political instability. The phenomenon was termed ‘capital flight’ and became a heated issue in the 1980s. It came to be recognized as a reasonably good indicator to the investment climate in a country and the confidence domestic residents have in their own economic system. It thus gained increasing prominence as an indicator of credibility in the wake of the debt crisis in the 1980s when many of the countries that had problems in servicing their external debt were also experiencing capital flight. It was argued that the occurrence of capital flight severely constrains the development of economies that are already burdened by debt and poor economic performance. Capital flight measures are thus viewed as important indicators of a nation’s predicament in financing international debt repayments and a warning to the international bank community as to the risk of further lending to these countries. Capital flight is a phenomenon, which though unobservable, is still assumed to be widely prevalent in developing countries. This paper reviews the methods adopted in the literature to estimate capital flight and provides corresponding estimates for a large data set of 116 countries for the period 1971–1998. This exercise was motivated by the fact that that we had no record of recent experience and that the compilation procedures for Balance of Payments at the IMF – on which these estimates are based – has changed since 1995. Therefore a consistent database was needed to review historical experience.1 A large data bank was chosen to identify the occurrence and scale of the problem across countries and regions. This paper examines the evidence to see if these estimates show the movement of unrecorded capital in the recent period in the developing world and finds that estimated resident capital flows are large even today in many countries and regions in the developing world. The sample chosen covers countries that still maintain controls on the movement of capital as well as those that have partially or wholly liberalised their capital accounts.